Suresh Arya Michael O'Neill George Pincus my exeContents 1. Introduction—Fundamentals Structural System of Foundations, 1; Theoretical Approach, 1; Fundamentals of Theory of, Vibrations, 2; Single-Degree-of-Freedom System, 2; Calculation of Parameters for Mathematical Model, 2; Equivalent Mass, m,. 2: Equivalent Spring Constant, ke, 4; Equivalent Forcing Function, F(t), 4; Formulation of Mathematical Model, 4; Transient or Free Vibrations, 6; Steady-State Solution of Forced Vibrations, 8; Dynamic System Sub- jected to Rotating-Mass-Type Excitation, 11; Terminology, 12; Accelerating Bodies (Ac- celeration, Velocity, Displacement), 12; Amplitude (Displacement, Vibration), 13; Analysis. (Computer, Dynamic, Matrix Method, Modal, Static), 13: Balancing (Static, Dynamic), 14; Beat, 14; Conditions (Boundary, Constraint, Initial), 14; Damping (Coetticient or Constant, Critical, Dashpot, Factor or Ratio, Viscous), 15; Coordinates (Cartesian, Generalized, Nor- mal or Principal), 15; Differential Equations (Linear, Simultaneous), 16; Dynamic (Eigen- values, Eigenvectors, Force, Load, Load Factor, System), 16; Excitation (Impulse, Inertial, Harmonic, Sinusoidal, Periodic, Transient), 17; Foundation Structure (Block-Type; Elevated Frame or Table Top, Mat Slab, Overtuned and Undertuned), 18; Frequency (Angular or Circular, Damped Natural or Harmonic, Excitation—Forcing or Operating, Fundamental, Natural, Rayleigh’s), 18; Magnification or Amplification Factor, 20; Mass (Consistent or Continuous, Equivalent Lumped or Lumped), 20; Motion (Equation of Mo- tion, Periodic, Aperiodic, Simple Harmonic.or Sinusoidal, Subharmonic, Superharmonic), 20; ‘Modes (Coupled, Uncoupled, First, Lowest, Fundamental, Normal, Principal— Eigenvector), 21; Modes of Vibrations, 24; Node (Points, Vibrating Systems), 24; Oscilla- tion, 24; Peak-to-Peak (Double Amplitude of Vibration), 24; Period, 24; Phase (Angle), 24; Resonance (Condition, Frequency), 25; Response (Dynamic, Steady Stato—Forced Part, Transient), 26; Shaft (Critical Speed, Flexible, Rigid, Stit), 27; Spring Stiffness (Constant, Equivalent, Linear-Elastic, Nonlinear, Soil), 27; System (Continuous, Dynamic, Free, Idealized or Equivalent, Linear, Nonlinear, Lumped-Mass Spring-Dashpot, Single- Degree-of-Freedom, Multiple-Degree-of-Freedom), 28; Transmissibility Factor, 31; Ref- erences, 31. 32 2. Development of Analytical Models for Dynamic Systems .. Modeling Techniques, 32; The Lumping of Mass, 32; Elastic Spring Constant, 32; Damp- ing Ratio, 32; Forcing Function, 33; Models, 33; Development of Equations of Motion, 33; Model 1—Vibrating Machines Supported by Block-Type Foundation, 34; Model 2— Vibrating Machines Supported by Mat-Type Foundation, 34; Model 3—Machines Sup- Ported on an Inertia Block and Vibration Isolated from the Foundation, 35; Mode! 4— vVibrating Machines Supported by Cantilever, 35; Model 5—Vibrating Machines Sup- ported by Fixed Beam, 35; Model &—Typical Elevated Pedestal Foundation (Table Top), 36; Model A—Single-Lumped Mass (Uncoupled Superstructure and Foundation), 36; Model B—Multi-Lumped Mass (Uncoupled Superstructure and Foundation), 36; Rayleigh’s Frequency, 36; Modal Multidegree Lumped Mass Analysi Lumped Mass with Coupled Soll-Structure Interaction, 37; Mode! D—Mult with Coupled Soil-Structure Interaction, 38; References, 38. Computer Analysis and Applications: Elevated Foundation Example Problem, 114; Example—STRUDL Coding, 118; Computer Printout in ICES— STRUDL, 121; Interpretation of Results, 157; References, 158. ‘Appendix A. Solution of Mul /-Degree-of-Freedom System . 159 Introduction, 189; Dynamic Analysis, 159; Determination of Natural Frequencies and Mode Shapes, 160; Determinant Equation Method, 160; Stodola-Vianello Method, 163; ‘Steady-State Response Analysis, 166. Development of Information, Trial Sizing, and Design Checklist . Machine Properties and Requirements, 48; Soil Parameters, 47; Environmental Condi- tions, 49; Trial Sizing of a Block Foundation, 49; Trial Sizing of Elevated Foundations (Table Tops), 50; Checklist for Design, 51; Design Conditions and Procedures (Static Con- ditions, Limiting Dynamic Conditions, Possible Modes of Vibration, Fatigue Failures, En- vironmental Demands), 52; References, 56. 4. Geotechnical Considerations Appendix B. Summary of ICES-STRUDL Commands . 169 187 Index 87 Notation for Chapter 4, §7; Evaluation of Soil Parameters, 59; Shear Modulus, 62; Calcula- tion of Shear Modulus for Structure-Soil Interaction Analysis, 68; Selection of Shear Strain Magnitude for Computing Approximate Shear Modulus Beneath Footings, 69; Damping Ratio, 70; Selection of Poisson's Ratio and Soil Density, 71; Effect of Footing Embedment, 72; Effect of Stiff Underlying Stratum, 72; Effect of Stratum of Loose Granular Soil, 74; References, 76. 5. Foundations .. .77 Notation for Chapter 5, 77; Modification of Foundation Response, 78; Vertical Spring and Damping Constants for Flexible Mats, 79; Deep Foundations, 80; Vertical Motion, 81; Pile Groups, 82; Horizontal Motion, 86; Uncoupled Rocking Motion, 86; Testing Methods and Empirical Correlations Based on Tests, 88; Comparison of Theory and Measured Behavior, 89; References, 90. 2 6. Design Examples: Block Foundations . 91 Example 1: Foundation Design for Reciprocating Compressor (Footing Embedment Effect Included), 92; A. Introduction, 92; B. Machine Parameters, 93; C. Soll and Foundation Parameters, 98; D. Selection of a Foundation Configuration, 93; E. Dynamic Analysis, 97; F. Check of Design Criteria—Static Conditions, 97; Limiting Dynamic Conditions, 97; En- vironmental Demands, 98; Nomenclature—Example 1, 98; Example 2: Design of a Foun- dation Block for a Centrifugal Machine, 99; A. Machine Parameters, 99; B. Soil and Foun- dation Parameters, 99; C. Selection of a Foundation Configuration, 100; D. Dynamic Analysis, 100; E. Check of Design Criteria, 100; Static Conditions, 100; Limiting Dynamic Conditions, 100; Possible Vibration Modes, 102; Fatigue Failures, 102; Environmental De- mands, 102; Nomenclature—Example 2, 102; Example 3: Foundation Design for Centrifugal Machines with Different Operating Frequencies and Supported on an Inertia Block, 103; A. Machine Parameter, 103; B. Soll and Foundation Parameters, 105; C. Se- lection of a Foundation Configuration, 106; D. Dynamic Analysis, 106; Selection of Springs for Inertia Block, 106; E. Dynamic Analysis as a Multi-Mass System, 107; F. Discussion of Dynamic Analysis, 107; G. Check of Design Criteria, 110; Nomenclature— Example 3, 111; References, 112. vii1|Introduction- Fundamentals ‘The analysis and design of foundations and structures subjected to vibratory loads is considered a very complex problem because of the interaction of structural engi- neering, geotechnical engineering, and the theory of vibration. These foundations may be designed as a simple block, either of plain concrete or of reinforced concrete, not different in resemblance from a footing designed for static loads. ‘The practicing engineer ho is generally not theoretically motivated ordinarily shuns theoretical investigations partly because these in- vestigations on a massive concrete block do not result in any additional reinforcement other than ordinary ‘minimum percentage of reinforcement required by the governing codes. Even when engineering talent is avail- able for a theoretically exact analysis, other factors such as economy, lack of high-speed computers, or design tradition result in an approximate nondynamic design. Thus, it has become imperative to devise practical de- sign procedures which include the various aspects of design and analysis of these foundations in a way that the least effort is involved in the theoretical investiga- tion. ‘The design engineer should recognize that the theoretical dynamic investigation is an integral part of the design effort. Tn this book, an effort has been made to use and simplify the latest theoretical knowledge available in this field (ref. 1). An easy-to-follow step-by-step routine is developed for actual design problems. In addition, at every step of investigation, a brief description is presented explaining the physical meaning of the parameters used and role they play in the design proces. Structural System of Foundations ‘The structural form of machine foundations is gen rally determined by the information provided by the geotechnical consultant and the machine manufacturer. However, during the design phase, it may become neces- sary to adjust the dimensions or shape of the foundation, partly to meet the design criteria or to avoid interference with other fixed objects such as pipelines and building foundations. The broad categories of foundations are (a) shallow foundation (resting on soil) and (b) deep foundation (supported by piles or piers). A further classification involves the structural configuration of the foundation: 1, Block-type foundation, consisting of a thick slab of conerete directly supporting the machine and other fixed auxiliary equipment. 2, Elevated pedestal foundation (table top), consist- ing of a baselab and! vertical columns supporting fa grid of beams at the top on which rests skid- mounted machinery. ‘These types of foundations are illustrated in Figure 1-1 Theoretical Approach Vibrations developed by operating machinery produce several effects which must be considered in the design of their foundations in addition to the usual design static loads. In some cases, if the size of the machine involved is small, it may be appropriate to design the foundation for equivalent static loads instead of strictly applying the vibration design criteria. However, once the design engineer has recognized the need for a vibration analysis, it is necessary that the designer possess a clear understanding of the fundamentals of the theory of vibration (refs. 3, 5, 6 and 7), modeling techniques (refs. 2 and 8), soil dynamics (ref. 1), and in some cases, the application of computer programs (refs. 2 and 3 of Chapter 7). In the step following the selection of the foundation ‘gross geometry, the design engineer is faced with the2 Design of Structures and Foundations for Vibrating Machines (2) Block-type foundation supporting reciprocating compressor (Table Top) ZS (b) Typical pedestal foundation Figure 1-1. Types of foundations for vibrating machines. vibration analysis. The usual procedure is to establish a mathematical model of the real structure which is ‘a necessary prerequisite in order to apply the theory of vibrations. The mathematical representation of a structural system is usually defined in terms of a lumped ‘mass, an elastic spring, and dashpot for each degree of freedom. The terms which are used in the develop- ‘ment of the theory of vibrations are described in the ‘Terminology section provided at the end of this chapter. Fundamentals of Theory of Vibrations The subject of vibration deals with the oscillatory behavior of physical systems. All physical systems built of material possessing mass and elasticity are capable of vibration at their own natural frequency which is known as a dynamic characteristic. Engineering structures sub- jected to vibratory forces experience vibration in differ ing degrees, and their design generally requires deter- mination of their oscillatory behavior. The present “design office” state-of-the-art considers only their linear behavior because of the convenience afforded by apply- ing the principle of superposition, and also because the mathematical techniques available for their treat- ment are well developed. In contrast, nonlinear be- havior of systems is less well known, and the mathema- tical treatment is difficult ¢o apply. However, all struc tures tend to become nonlinear at high amplitude of oscillation, and a nonlinear analysis is required under those conditions ‘Single-Degree-of-Freedom System ‘An engineering structure (a fixed beam) is illustrated in Figure 1-2a, The beam is supporting a machine generating a harmonic centrifugal force. A step-by-step procedure will be described for modeling the actual structure Calculation of Parameters for Mathematical Model Equivalent Mass, m, ‘The beam has distributed mass along its length, and its ends are fixed against rotation. In calculating the ‘mass for the mathematical model, it is necessary to lump the mass only at points where the dynamic force is act- ing, and also at those points where the dynamic response is required. In this example, the dynamic force is acting at the middle and the response is also required at the ‘middle, The technique for obtaining the lumped mass is to equate the kinetic energies of the real and the equiva- lent systems (refs. 2 and 8). First, a deflected shape of the real system is assumed, Figure 1-26 which corre- sponds to the predominant mode. In this example, the beam can have predominant translational modes in the xy plane, the x2 plane, and a rotational mode about the x-axis, Thus, the model has three single-degree-of- freedom systems independent of each other. Considering only the deflected shape in the x~y plane, and assuming the shape is the same as that which would be caused by a concentrated load P applied statically in the middle, Ja = (P48 EL) (3 — 42) co) Jumex = PP/192 Ely (1-2) ‘Assuming the beam's behavior stays within the elastic range and the maximum velocity at any point along the (0) Detlection Curve Introduction-Fundamentals 3 Wy x2 Concentrated load (3 ~ 4x) Unit. dist. toad = I~ xP hy Ke bar w ay (©) Initial {d) Defiected —__(e) Oscillatory Position Position due Condition ‘to Weight W Figure 1-2. SDOF representation of a fixed beam supporting machinery in the middle. ‘beam is proportional to the ordinate of the deflection curve at that point, then the velocities of the beam are Vy = (CPs'/48 El.) (31 4x) and Ves = O(PE/192 EL) , where C= constant relating velocity to deflection. Thus, Vu = (42'/P) (81 — 45) Vous (3) ‘The total kinetic energy of the beam is given by KBy = fms Vids mg/l nal) S(O! + ef — 24is*) dx, a4 where my mass per unit length. Then, Litas! + 163" — 24ie ax = 13/11207 Equation (1-4) after integration reduces to KEy = (13/70) Prous me 5 where m, = my 1, the total mass of the beam. The kinetic energy of the equivalent system is given by KE, = 3m Ve4 Design of Structures and Foundations for Vibrating Machines Since the kinetic energies must be equal in both sys- tems, and since Vax must equal Ve, Kl KE, (18/70) my Vinux = 3 me Ve 0.371 m, = m, (1-5) ‘Therefore, 37.1% of the total distribution mass should be included as the corresponding mass of the mathemati- cal model. However, the mass of the machine is located at the middle; hence, that entire mass should be con- sidered part of the equivalent mass. 0.371 my + tig, (1-6) Similarly, equivalent mass can be calculated in the x2 plane which will also have the same magnitude since the deflection curve of the beam remains the same. Table 1-1 lists equivalent mass factors for beams and slabs with different types of loads and support conditions. Equivalent Spring Constant, k, ‘The spring constant of an equivalent system is obtained by equating the resistances to deformation of the proto- type and the mathematical model, in this ease, the uni- formly distributed loaded beam vs. the modeled middle loaded beam. The resistance offered by the beam per unit Toad is given by the reciprocal of the deflection produced by the same unit static load applied at that point. Therefore, in the example under consideration, the resistance offered due to a unit concentrated load at ‘midspan (ref. 9) is Ry = 1/(F/192 EL) = 192 E1,/t, Co) which when equated to the resistance of the equivalent system gives ke = 192 Eh,/P a8) In Table 1-1, values of spring constants for equivalent systems are presented for different types of loads and support conditions. Equivalent Forcing Function, F(t) The dynamic force may be distributed over a certain length of the element, and in order to obtain its equiva- lent, concentrated load value for application in the single-degree-of-freedom system, the work done by the actual system is equated to that done on the equivalent system. The load factor, fi, with which the distributed dynamic force should be modified to determine the equivalent, concentrated dynamic force is given in Table 1-1. For the model shown in Figure 1-2a, the dynamic force acts at the middle of the beam, thus, a force modification factor is not required. Formulation of Mathematical Model ‘A procedure for obtaining the values of various parameters ina mathematical model which equal those in an actual system is given in the preceding section. An equivalent mathematical mode is shown in Figure 1-2c. ‘The mass and the spring constants are the equivalent parameters corresponding to an actual system. An equiv- alent damping coefficient is not required in this partc- ular model since it is associated with the velocity of the system only, and its effect is implicitly included when the equivalent values of mass and spring constant parame- ters are caleulated. The chosen model has three inde- pendent degrees of displacement and/or rotation, and therefore, there are three individual equivalent models having a single degree of freedom each. The technique ‘of mathematical formulation for each of the three models the same, and therefore, only one single-degree-of- freedom model wil be examined in detail ‘A model, shown in Figure 1-24 with a weight W, is attached to a weightless spring k, and the spring stretches by an amount 8,.=W/k. The system is intially in a state of static equilibrium with the dead weight W” bal anced by the restoring pull of the spring A. Subse- ‘quently, the weight His set into oscillation by the appl cation of some disturbance. Starting at time equal to zero, the system vibrates freely with an amplitude of =y displacement (Figure 1-2) ‘The forces acting on the body are applied against the direction of motion and include: the resistance offered by the spring & (y+ 81x), the resistance Cj offered by the viscous damping C, and inertia force m3, given by Newton's second law of motion. The latter force is equal to the mass of the system multiplied by its negative acceleration (the state of motion of a mass at any instant may be considered as in a state of static equilibrium upon introduction of the inertia force). The force acting in the dzection of motion is the weight W. Summing up the forces, the resulting equation of mot mt Ghoti) = W, which reduces to mit Oth (1-9) ' ' Introduction-Fundamentals 5 Table 1-1 Dynamie Design Factors for Beams and One-Way Siabs Equvalon Parameters of Type ot Type of Singl-Dogree-of Freedom System Equivalent Stractural Dyramic Len Seg] Mathoraticol Member Loading Load actor be Constant Model Load | one, a] uni ba | oe ‘tyrant ngth 4261 el 0.64 = 80 i applets Lo ‘A, Uniform load 3 ae Fp) i 1. simply aor wo | oe ‘supported fel ht led v eo) rT) beam B. One concen- tatod load Ay Fe ate osr | o7e | ose | S24E!) Fey = hx total + | Tg Tai : a | Gita tad IC. Two equal on the span) moontated toad ee M, = Ky X (total ityranit tng mass on te span) ee a Soy | oss | — | ont | 288 ae mxe ‘A. Uniform load u where k = spring Constant ofa eal K« > sytem ? "Equal of th 2. Fixed ends 1, | “eaual pans of the beam ka] | 0 10 | oar | 122941| concentrated Be | masses aro hmped einer at each concen- fussed eed trated dynamic Tivant tnt toad feces 1073.61 oss | — | 04s El et} re pe F | uniter cae et > Fo 3. Simply support ed and fxed son ct Ta! | 10 | so | ome | 708! B. One concen- u trated toad Ao 24 _i2, 1069 e1 ata siqi@ug! | os | osr | oas | 82 c. two equal Concentrated loads6 Design of Structures and Foundations for Vibrating Machines ‘This equation is an ordinary second-order linear dil ferential equation with constant coefficients, also called ‘a homogenous equation. For the case when a forcing function F(t) is also acting, the resulting equation is classified as nonhomogenous (ref. 10) and is written as mi + G+ = FO (1-10) ‘The solution of Equation (1-9) yields the dynamic characteristic of the system such as the natural frequency, the damped natural frequency, the critical damping coefficient, or the transient motion of the system. Each of these terms has a special signi ‘upon the particular problem at hand. ‘Transient or Free Vibrations A solution of the form y = eis assumed for Equa- tion (1-9) where s is a constant to be determined, and t is the independent time variable; then, jm 5 aste aan Upon substitution of y, 5 5) into Equation (1-9), the following expression is obtained: (F + [O/m}s + fem) 0 (12) Since e** must be greater than zero for all values of &, f+ (C/m)s + (k/m) =0 (13) Equation (1-13) is a quadratic equation having wo 1/2n) [—C + VE Fm] 1/2n) [—C — VG? — ain} (4) Several terms, relating. vari tion (1-9), are defined as n= VEfm is called the circular natural frequency of the system in radians/sec; Go=2V Fm is the critical damping of the system in units of foroe/ velocity; D=C/Co=G/2V km is called the damping ratios tof tae em oth daring eluded ‘The complete solution of Equation (1-9) is ym Ae + Be, (1-15) 8 parameters of Equa- where A and B are arbitrary constants which depend upon the initial problem conditions. The motion de- scribed by Equation (1-15) is called transient motion of the system, and the oscillations die out in a short interval of time when significant damping is present. Equations (1-14) and (1-15) show that the nature of ‘oscillation depends upon the value of C. Four possible values of C will be considered here (ref. 6) to illustrate the physical significance of Equation (1-15) Case 1: C =0 (no damping). This case reduces the problem to an undamped system, and the roots obtained from Equation (1-14) are 4,2 Equation (1-15) ‘can be written as, 2 Adee 4 Bett (1-16) Equation (1-16) can be written in three alternate forms by the use of trigonometric identities and complex numbers: y= Ce (1-17) 2 = By cos ost + By sin wnt (18) 2 = Geos (ust ~ 4) (1-19) Equation (1-17) is in terms of phasors, while C, and { are components of a complex number. B, and By are arbitrary constants in Equation (1-18) representing the initial boundary conditions. For example, at time the system has a init and an initial velocity (0) becomes 7 = Ia 08 eat + (0o/eeg) sin ot (1-20) and the velocity function, Sift = —yosin ext + (t/%y) 605 wt (21) Equation (1-19) can be obtained from Equation (1-18) if the following substitutions are made: B, = C, cos $, and By=Cysing. Then tan ¢=B,/B,, and C= Be + BY. By using the trigonometric identities, 9=C, 005 (ont ~ $) (1-22) S/n = —Cx sin (wt — 0) (1-28) In Equation (1-22), C)= Vin? + (Bajos)? is called amplitude of vibration, and g=tan-* (vo/sgyo) is called the phase angle. A graphical representation of Equation (1-22) is given in Figure 1-3a by the projec- tions of a vector Cy rotating about a fixed point O, with ‘constant velocity w», The projection upon the ordinate axis represents the instantaneous displacement y, while’ the projection on the abscissa gives the velocity function }y/om according to Equations (1-22) and (1-23), respec- tively. A displacement time curve based on Equa- tion (1-20) can be obtained fom Figure I-Ba by project- TB, cos wnt a : 3 8 sinagt 2 “ = 5 i oy ing instantaneous values of point P to the right, as shown, in Figure 1-3b. A velocity versus time plot based on Equation (1-21) can be obtained by projecting point P vertically, as shown in Figure 1-3e. The variation of these terms with time is shown for a complete cycle. The time required for one complete cycle is called the period T and equals 2x/oy sec. The corresponding cycle fre- quency is fy = 1/T = o,/2x eps. These cycles are iden- tically repeated since this system is undamped. Gase 2: C* < 4 km, but > 0 (underdamped). In this case, the roots of Equation (1-14) are complex con- jugates, and 5, and 55 become 4 = oy (—D+ivI—D) o, (—D-iVI=D) when the damping ratio D=G/Go= C/2V im: C/2eqm is introduced, Further substitution of Equa- tion (1-24) into Equation (1-15) and conversion to a tuigonometric form with the aid of Euler's formula a8! = cos 0 +t isin 8, gives (28) 2 + Brcose, /T- (1-25) Introduction-Fundamentals 7 c,c08 (wat) Figure 1-3. Undamped free vibrations ini- tial conditions, y(0) = yo, (0) =up Source: ‘AH. Church, Mechanical Vibrations, sec- ‘ond edition, New York: John Wiley, 1963. 9 = OP Y sin (wd + 4) (1-26) Where B,=Yeos$ and B=Ysing and o: ou VIDE The term wg is called the damped natural frequency, and Y and are arbitrary constants to be determined from the initial boundary conditions in a similar way to the procedure used in Case 1. The type of motion desceibed by Equation (1-26) is oscillatory with frequency of wg and is shown in Figure I-ta. The ampli- tude of oscillation Y will diminish with time and is proportional to «~Px* as shown by the dotted lines. Case 3: C: = 4 km (critical damping). ‘The damping corresponding to this case (C= 2V km, C C/Co= D = 1.0) is referred to as critical damping. For this value of ©? = 4 km, Equation (1-13) has two equal roots, 4,2=—C/2m. In this ease, the general solution of the second-order differential equation is ee 2 Substituting the value of C/2m = 2VFmi/2m and applying the initial boundary condition, y (= 0) = yo, and j (¢= 0) = vp in Equation (1-27) gives 2 = by + (oof) + ye] oath & (1-28)8 Design of Structures and Foundations for Vibrating Machines A graphical representation of Equation (1-28) is shown in Figure 1-4b, The motion described by Equa- tion (1-28) is aperiodic. Since ertical damping repre- sents the limit of aperiodic damping, motion is reduced to rest in the shortest posible time with no oscillation Case 4: C* > 4 km (overdamped). Referring to Equation (1-14), the roots of Equation (1-13) are real and unequal. The value of the roots, after substituting the relationship C/m = 2oy D,is given by hg = oD to VDI (1-29) and the resulting solution by using Equation (1-15) is siven by = APN TTT p-e- VP. 90) Since the roots in Equation (1-29) are real and nega- tive for all values of D> 1.0, the value of y in Equa- tion (1-80) will decrease exponentially without a change ay nh. i va Vy YerOunt pe (2) Underdamped periodic oscillation (2 <4kmor D <1.0) ont 4 Aeron at in sign. A graphical representation of Equation (1-30) is shown in Figure 1-4e, which indicates that there are no oscillations, and the system is said to be overdamped. Steady-State Solution of Forced Vibrations ‘The solution of Equation (1-10) includes two parts (ref. 3): (a) transient or free vibrations and (b) steady- state or forced vibrations, Transient motion, which in mathematical terms is called the complementary func- tion, is a solution of the homogenous equation, as previ- ously noted. ‘The particular integral of Equation (1-10) gives the steady-state or forced-vibrations solution. This solution includes the influence of the forcing function. Structures land machines which are subjected to excitation forces which vary with time are susceptible to vibrations. The excitation can be in the form of a pure, simple, harmonic Do (ae-+ Yolae-st % Sor (0) Critical damped aperiodic oscillation (2 = 4kmorD = 1.0) () Overdamped aperiodic oscillation (c2> 4km or D> 1.0) Figure 1-4. Damped free-vibration response of SDOF system. Source: Willlam T. Thompson, Vibration Theory and Applications, © 1965, pp. 39-40. Reprinted by permission of Prentice-Hall, Inc., Englewood Cliffs, NJ. force or displacement, or it may have some other periodic form. These other periodic disturbances can be resolved into a number of harmonic components in the form of Fourier series as illustrated in Table 1-2. A third type of time-dependent excitation is in the form of a series of repeated shocks and impulses, pulse waves, or step func- tions ot force or displacement applied to the mass or to its support "The most common source of excitation in structures supporting machines is the internal excitation caused by an unbalanced condition in the machines or the external excitation produced by a nearby dynamic system. These ‘excitations are generally in the form of harmonics under steady-state conditions and will be further considered here. Introduction-Fundamentals 9 Equation (1-10) with a harmonic force is my + G+ = Fosinat, (31) where o is the frequency of the harmonic excitation. The particular integral solution for this equation is yp = Arsin ot + ds cos of (0-32) with jp = a cos of — ey sin at (1-383) and jp = wy sin wt — uPA c0s ot (1-336) Substitution of Equations (1-32), (1-33a), and (1-33b) in Equation (1-31) and collections of the coeff- cients multiplying the sine and cosine terms yield Table 1-2 Harmonic Components of Periodic Disturbances (Ret. 7) +54 err10 Design of Structures and Foundations for Vibrating Machines [(¢ — ma*)A, — CwAs] sin ot + [Cady + (k — me") Ay] cos wt = Fy sin wt (1-34) and equating the sine and cosine terms on the two sides of the equation yields [= ma") 4, — Cady] sin ot = [GaAs + (b= ma?) As} co wt = 0 Solving of these two simultaneous equations for the two constants 4, and ds and substituting in Equa- tion (1-82) yields Fa (i — mo) sin wt ~ Fy Ca 008 wt ae = mal + Cay "The particular integrals for various forms of the forc- ing function are presented (in Table 1-3) to illustrate the physical feel of resulting oscillation. An alternate form for Equation (1-36) is et ) (1-36) Jn = BoE me} F (Cay sin (ot — 0) = Ysin (ot — 6), (1-37) where ¥ = Fo//(E— ma®) + (Ca) and 1 = tan [@4/(é ~ ma")] ¥ is the amplitude of the steady-state response, and {gx is the “phase lag” of yp (t) avith respect to the forc- ing function Fy sinet. Substitution of the expression for D and oy in the expression for Y and ¢, and replacing sw/en by the frequency ratio r in Equation (1-37) and rearranging Equation (1-37) in nondimensional form gives \ (1-38) y MO” Tian gu= wi’ 2Dr/(1—P) AF + Rpt Table 1-3 Particular integrals (Ref. 7) ro Jor Y» TRE 7 cmd a tesco FF rtCL UL Fi ol cou D1 | eet attgicnsrcont entre where Mis called the dynamic magnification factor, x has been defined earlier. Equation (1-38) shows that the ‘M and , factors are functions of the frequency ratio r and the damping ratio D. These functions are shown in Figure 1-5. These curves indicate that the damping ratio Dis effective in reducing the amplitude and phase angle in the region of resonance, that is when r approaches unity, the particular values of M and gi depend on the damping ratio D. Dynamic System Subjected to Rotating-Mass-Type Excitation In some dynamic systems, the excitation force present arises out of unbalances in the rotating masses. Examples fof such systems are reciprocating and centrifugal ma- chines. The forces generated by a reciprocating machine are of the form (ref. 1) Fe = (tee + typ) 10" 008 oat H+ mee (P/E) oF €08 2a Fy = (mi) 1 sin ot (1-38) (1-396) where Fr and Fy are horizontal and vertical inertia forces, respectively. ‘There are two masses: one moving with the piston at point P in Figure 16a called May (reciprocating) ; and one moving with the crank pin at point C called Mya (rotating). ‘The eranke mechanisin for this type of machine is illustrated in Figure 1-6a, “The force generated by the unbalanced rotating mass of the centrifugal machine shown in Figure 1-6b are given by Fe = mea" cos at f, (1-40a) meu? sin ot (1-406) Equations (1-89) and (1-40) indicate that the magni- tude of the forcing function is proportional to the rotat- ing mass m4, its eccentricity to the true axis e, and the speed w. The rotating mass and its eccentricity remain constant, but the value of w varies from startup of the ‘machine to its stable steady-state condition. Therefore, during that period, the maximum amplitude of the fore- ing function given by Equations (1-39) or (1-40) is directly proportional to the square of the operating speed. The equation of motion for the forcing function in the centrifugal machine is given by a damped single-degree- of-freedom system in the y-direction: mi + G+ by By comparing Equation (1-41) with a constant-force- amplitude expression, Equation (1-31), and substituting ‘meu? = F, in Equation (1-37), the following expression is obtained: (mee) sin at (4) Introduction-Fundamentals 11 © Phase angle (a) Magnification factor (M) 7 @ Frequency ratio (1) Figure 1-5. Magnification factor (M) versus fre- quency ratio (r; (a) and phase angle (¢,) versus frequency ratio (1); (b) for a single-degree-of- freedom system subjected to a constant force amplitude force, F = Fy sin wt. Source: William T. ‘Thompson, Vibration Theory and Applications, © 1965, p. 54. Reprinted by permission of Prentice- Hall, Inc., Englewood Cliffs, NJ. © Figure 1-6. (a) Crank mechanism of a reciprocating machine; (b) Forces from a centrifugal machine (rotating mass excitation). Ie = mes sin(ut—o)// Ema) + (Ga)* (1-42a) Y = mea!//=mal $F (Gof and j= ant ee (1-420) Goma)12 Design of Structures and Foundations for Vibrating Machines Manipulating and rearranging Equation (1-42a) ina similar manner as was done with Equations (1-37) to (1-38) gives YM me) /& oF ¥/(me/m) = Uvll=FF + BD = at OM =, (1-43) where M, is the dynamic magnification factor for the rotating-mass-type excitation case. Figure 1-7 shows the plot of M; in relation to the frequency ratio r for the various values of damping ratio D. Substituting = m4/m in Equation (1-43) yields M, = Y/ue = ()'M (144) ‘The term ye is called the free amplitude. For a given system, the values of y, ¢, D, and w, are constant, so that Figure 1-7 is, in effect, a plot of the amplitude of the mass against the rotating speed of the unbalanced force for various amounts of damping. For a smnall value of +f, or at low rotating speed, the total mass m moves very little; at a speed approaching the natural frequency of the system, r= 1, the amplitude builds up to large values for small amount of damping. Further on, at higher rotating speeds, r > 20, the curves approach the value M, equal to unity since the inertial force of the total mass is then’ approximately 180° out of phase with the unbalanced force. Comparing the curves of Figures 1-5 and 1-7, it may be observed that resonant peaks occur at r-< 1 for the case of a constant force excitation, and r>1 for the case of rotating-mass-type excitation. The various expressions which may be derived from the equations of motion of these two cases are listed in Table 1-4. ‘The combination of transient or free vibration (com- plementary function) and steady-state vibration (partic- ular integral) gives the complete solution for Equa- tion (1-10) Terminology A single-degree-of-freedom lumped mass system is presented above including the derivation of the differ ential equation describing the behavior of the model leading up to the development of formulae for the calcu- lation of the dynamic response. In the area of dynamics of foundations and structures, the investigation may extend to a variety of systems, some having several degrees of freedom, and as a result, the modeling tech- niques and derived formulae are more complex. The fundamental principles of single-degree-of-freedom sys- tems are also applicable to the- multidegree-of-freedom systems; however, additional information from theory of vibrations ie required for consideration of the more com~ plex multidegree-of-freedom systems. A complete intro- (©) Frequency ratio (4) Phase angie (é) Magnification factor (M,) Frequency ratio (") Figure 1-7. Magnification factor (M,) vs. frequency ratio (r): (a) and phase angle (4) versus frequency ratio (r): (b) for a single-degree-of-freedom system subjected to a rotating-type excitation, F = meu sin wt. Source: William T. Thompson, Vibration Theory and Applications, © 1965, p. 60. Reprinted by permission of Prentice-Hall, Inc., Englewood Cliffs, NJ. duction to the theary of vibration is not presented heres however, Chapter 1 lists a number of references on the subject, The following list of terminologies summarizes the most commonly used terms (refs. 4 and 5) in the field of vibrations. An example of each defined term is alo included to provide further information on its application. 1, Accelerating Bodies ‘Acceleration Definition: Newton's Law of Motion—a vector quan- tity when applied to the mass, produces a force in the direction of application. Example: the rate of change of velocity with time. For the #-coordinate it is denoted by dx/dt* or ¥. See Figures 1-8 and 1-9. Velocity Definition: vector quantity which represents time rate change of position for a particle or body. Example: the rate of change of displacement with time. For the x-coordinate it is denoted by dx/dt or &. See Figures 1-8 and 1-9, Displacement Definition: a vector quantity that represents the change of position of a particle or body from a state of equilibrium, ‘Example: a displacement which is a function of time. For the x-coordinate it is denoted by x. See Figures 1-8 and 1-9, Introduction-Fundamentals 13, Table 1-4 Summary of Derived Expressions for a Single-Degree-of-Freedom System Constant Force Excitation Rotating Mass-type Excitation, Expression Fs Constant Foe mew T = ‘Magnication factor 4 - M = ——— VO=PF ODF VO=P¥ + GD Amplitude at frequency f ¥ = M(Fo/t) Y= My (me/m) Resonant frequency Amplitude at resonant frequency fr ‘Transmissiblity factor r vIFEDA VO=P EDP vieaD 2 iF @DF VO=FF ODP ‘where 7 = w/in 14, (Uadamped natural circular frequency) = ¥@7m) ‘D (Damping ratio) = C/C, , (Critical Damping) = 2 T, = Force transmitted/ Fo T, = Force transmitted/s XK Figure 1-8. Rotating vector representation of a har- monic function x = A sin wt. VEL Rewdsin(ot ¥E) : RCCL Keath sintet +77) Figure 1-9. Harmonic motion representation of dis- placement velocity and acceleration. 2. Amplitude Displacement Definition: the maximum ‘change of position of a body ‘or some part of the system from a reference point (generally equilibrium position) at any given time. Example: the maximum displacement of a sinusoidal quantity += A sin et, which is A in this function. Vibration Definition: the time-varying magnitude of peak dis- placement (of a physical body) from a reference point. Example: a time-varying displacement of a rotor shaft of a machine, or a foundation structure from the static equilibrium condition. 3. Analysis Computer Definition: resolution of complex mathematical prob- Jems into simple elements with digital (discrete number operation) or analog (continuous chart operation) com- puters. Example: solution of indeterminate structures or determination of vibration in a dynamic system using computer programs.14 Design of Structures and Foundations for Vibrating Machines Dynamic (Vibration) Definition: a study of motion of a phyial system at a particular time. Example the ealculation of the amplivade of vibra tions in a machine or in a foundation structure. Matrix Method Definition: the study of motion of masies in multi degro-of freedom systems. Example: the solutions of simultaneous equations sing the techniques of matrix algebra the dynamic analysis of a multidegree-of- freedom system, where the responses in the normal modes, (each treated as independent one-degree systems) are determined separately, and then superimposed to pro- vide the total response, ‘Example: vibration analysis of a “Table top”, when a computer program is employed. Static Definition: the investigation of a physical system in equilibrium under the action of a system of stationary forces. Example: dead-load analysis of a structural system. 4, Balancing Static Definition: adjustment of mass distribution of a rotat- ing body such that statically the system is at neutral equilibrium. ‘Example: see Figure 1-10. Dynamic Definition: the adjustment of mass distribution in a rotating body such that the vibrations are controlled. Example: see Figure 1-11 5. Beat ‘Definition: the maximum resulting amplitude of two simple harmonic wave forms of slightly different fre- ‘quencies which are superimposed. Example: sec Figure 1-12 in which beat frequency (fs) = Abs. [o1— on] /2n. Frequency of combined oscil- lation (f) = (+ 42) /4x. Beat period (Ty) = 1/f- Period of resulting oscillation (T') = 1/f5 %mux = Au + Ass ata = Abs. (di ~ As) 6. Conditions Boundary Definition; the known physical relationships at specific points of a structural body, usually at the supports. Example: see Figure 1-13 in which (Boundary Condi- tions) Deflection: y (x =0,L) =0; Slope: El dy/dx (x=L/2)=0; Moment: EI d'y/dat (x= 0, L) =05 Shear: EI d¥y/de* (x= L/2) =0. BALANCING MASS UNBALANCED MASS Figure 1-10, Static balancing. This system under ro- tation produces equal centrifugal forces, but pro- duces unbalanced moments in shaft and pressure ‘on the bearings. UNBALANCED MASS Figure 1-11. Dynamic balancing. For this system during rotation, not only are centrifugal forces balanced but the forces and moments in the shaft (in ‘one revolution) are also balanced. x X= Ay sin (wy t= 4) + Ar sin (wet — da) Figure 1-12, Resulting motion of two simple har- monic wave forms containing a beat. Figure 1-13. Simple beam loaded with uniform load w. Constraint Definition: the imposition of limitations on the be- havior of a physical body. Example: see Figure 1-14. Figure 1-14. Cantilever of uniformly distributed mass. The cantilever is constrained to deform in a Sosa prota ee q k Figure 1-15. Single lumped-mass dashpot system. Figure 1-16. Critical damped oscillation of a single lumped-mass system. dashpot c a Figure 1-17. Symbol used in a lumped-mass system. Figure 1-18. Coordinates of element "in space. Introduction-Fundamentals 18 Initial Definition: the known motion of a physical body at the reference time, often taken as zer0. Example: the application of brakes on a steadily moving vehicle. At the time of application of brakes, velocity = V and deceleration = 0. 7, Damping Coefficient or Constant Definition: a factor used in a dynamic system to account for dissipation of energy. Example: see Figure 1-15, where equation of motion is m+ C+ kx=0. In this equation, the constant C accounts for viscous damping in the system. Critics Definition: a mathematical equality derived in vis cously damped system, such that the free displacement comes to rest without oscillation. Example: see Figure 1-16 in which critical damping 2Vim. schematic representation of viscous damper. Bxample: see Figure 1-17. Factor or Ratio Definition: the ratio of actual resistance in damped hharmonie motion to that necesary to produce critical damping. : Example; D=C/G.= 6/2 Vim Viscous Definition: a type of damping assumed in a dynamics model such that the dissipation of energy during oscilla- tion is linearly proportional to the velocity of the mass. Example: damping force = Ci. 8. Coordinates Cartesian Definition: linear quantities that describe the location of a point in space with respect to a system of threc- dimensional orthogonal axes. Example: see Figure 1-18. Generalized Definition: a specification of a configuration by a set of independent geometric quantities, which may be lengths, angles, or their combinations. Example: a set of n independent geometric coordi- nates which specify the configuration of an n-degree-of- freedom system.46 Design of Structures and Foundations for Vibrating Machines Normal or Principal Definition: a particular set of generalized coordinates which describes equations of motion such that there is neither static nor dynamic coupling among them. Example: a procedure followed in modal analysis in which the general motions of the masses of a multi degree-of-freedom system can be expressed a8 a super~ postion of its principal modes of vibration. ferential Equations (Equations of Motion) an equation relating to two or more vati= ables in terms of derivatives or differentials such that no terms involving the unknown function or its derivatives appear as products or ate raised to a power different from unity, The order of a differential equation is equal to the order of the highest derivative in the equation. When the independent variable is a time function, then it is called an Equation of Motion. Example: see Figure 1-19 for which the Equation of Motion is mi+Gétke=F(t) or m(dz/dt*) + C(dz/dt) + ke= F(t). This is a nonhomogenous ordi nary second-order linear differential equation with con- stant coefficients. In this equation, z and ¢ are variables, where z is the dependent variable, In case z is dependent fon more than one independent variable, then the equa- tion will change from an ordinary to a partial differen- tial equation, If m, C, and & are not constant and are independent of z oF its derivatives, but are dependent of the variable ¢, then the equation is called a differential equation with variable coefficients. If the right-hand term of the equation F(¢) is zero, then the equation is called homogenous. The solution of a homogenous equa- tion is called its complementary function and is given by ee) = ACP sin (ot +) where A and y are constants to be specified by the initial conditions. This solution gives the transient motion of the system. The solution which satisfies the nonhomoge- nous equation is called the particular integral. For F(t) = Fo sin ot it is given by: a5 = Resin (ot ~ 9)/V =m + (Col? ‘This solution gives the steady-state response or steady- state vibration, The complete solution of the equation is the sum of the complementary function z(t) and the particular integral (2) ‘Simultaneous Definition: linear differential equation which contains more than one dependent variable related to a single independent variable . rot te Figure 1-19. Single degree spring-lumped-mass- dashpot system. Figure 1-20. Two-degree-of-freedom system (multi- degree), Example: see Figure 1-20 in which the Equations of Motion are muéit Cy (é— 4) +h (41 — #2) = Fal) and msés + Caz ~ Gy (£1 ~ ta) + fata ~ by (&1 — #2) = F(t). This two-degree-of-freedom system contains two dependent variables, sz, and an independent vari- able, t, The general solution of these two simultaneous differential equations will consist of a complementary function and a particular integral 10. Dynamic Eigenvalues (characteristic values or natural frequencies) Definition: the roots of the characteristic equation which results from the expansion of the determinant of the simultaneous differential equations. (See also definition of normal modes.) Example: when the simultaneous equations are equa- tions of motion of the free undamped multidegree-of- freedom system, then their roots are called eigenvalues ‘which are equal to the squares of the natural frequencies of the modes. Eigenvectors (characteristic vectors ‘or natural modes) Definition: these are the characteristic vectors which are obtained by substituting the characteristic values or { eigenvalues in a set of simultaneous differential equa- tions of a multidegree-of-freedom system. Alternately, eigenvectors are the independent vibrating modes of a multidegree-of-freedom system such that during vibra- tion the ratio of the displacements of any of two masses is constant with time. Example: a multidegree system has exactly the same ‘number of natural modes as degrees of freedom. Associ- ated with each mode is a natural frequency and a char- acteristic shape. Force, load Definition: a foree whose duration and amplitude is a funetion of time. Example: centrifugal force generated by an unbal- anced rotating mass is given by F = m,eut sin of Load Factor Definition: the ratio of the dynamic deflection at any time to the deflection which would have resulted from the static application of the dynamic load. Example: the dynamic load factor caused by the con- stant centrifugal force of rotating mass on undamped one-degree system is given by DLF= 1/{1 ~ (w/+n)*) System Definition: an elastic system which possesses mass and whose parts are capable of relative motion. Example: an engineering structure, machine, or its components, and most physical bodies consisting of matter, 11. Excitation Impulse Definition: the product of force and time while force is acting on the mass. Example: see Figure 1-21. Inertial Definition: excitation generated by the mass in motion. Exampl ire 1-22. Harmonic, Sinusoidal Definition: a pulsating force of the form: Fy sin ot for F605 at Example: see Figure 1-23. Periodic Definition: a time-function excitation which repeats itself identically at regular intervals of time. Introduction-Fundamentals 17 FORCE TeF:Ay toe IMPULSE = FxAy Figure 1-21. Rectangular pulse. 42S - | [reemestsims Figure 1-22. Rotating mass oscillator generated ex- citation: F(t) = 2m, eu? sin at. FORCING FUNCTION = Fosinut Figure 1-23. Harmonic force. The figure shows a centrifugal force of amplitude Fy generated by a rotating machine. F, Figure 1-24. Forcing function generated by the cam of a machine. Example: see Figure 1-24 for which the following ‘equation holds: Fy = 748 Design of Structures and Foundations for Vibrating Machines Transient Definition: a temporary arbitrary excitation which disappears with time. Example: see Figure 1-25. 12, Foundation Structure (for Machines) Block+type Definition: a small area concrete foundation of such thickness so that the structural deformation caused by the superimposed load is negligible Example: see Figure 1-26. Elevated Frame (Table Top) Definition: a three-dimensional elevated reinforced concrete structure consisting of beams framing into columns and supported by a heavy foundation slab. The tops of the columns are connected by a top slab or heavy longitudinal and transverse beams forming a rigid table ‘on which the machinery rests. The foundation structure may be supported by piles or directly on the soil. Example: see Figure 1-27. Mat Slab Definition; a flexible concrete slab which is resting on soil and supports a machine or battery of similar machinery. Example: see Figure 1-28. Overtuned and Undertuned Definition: a machine foundation is said to be over- tuned when the ratio of the speed of mounted machine to the natural frequency of the foundation is less than 1.0 and is called undertuned when that ratio is greater than 1.0. Example: see Figure 1-29. 13, Frequency ‘Angular or Circular Definition: the time rate of change of angular di placement given in units of radians per second. For an oscillating system, it is the number of vibrations in units of radians per second, Example: see Figure 1-30. Damped Natural or Harmonic Definition: the natural frequency of a linear system which includes viscous damping C. Example: see Figures 1-31 and 1-32 for which the following equation holds: Damped Frequency, ag = 0 VT BF Damping Ratio, D = G/2 vim < 1.0 F, t Figure 1-25. An arbitrary transient forcing function. Figure 1-26. Block-type foundation for a recipro- cating machine. MAT FOUNDATION Figure 1-28. Vibrating machine supported by a mat- type foundation. B gee ification factor, M Frequency ratio w/w, Figure 1-29. Magnification factor (M) versus fre- quency ratio for various amounts of damping ratio (0). @ (RAD/SEC) Figure 1-30. Angular or circular frequency w. Figure 1-31. Damped- free linear system. Figure 1-32, Damped-free oscillation. Introduction-Fundamentals 19 Excitation, Forcing or Operating Definition: the number of times a dynamic force achieves an identical amplitude in a time period of one second, and is given in eycles per second (Hertz) Example: see Figure 1-33. Fundamental Definition: the lowest value of all natural frequencies of an oxcillating system. Example: the frequency associated with the first mode of vibration. Natural Definition: the dynamic property of an elastic body or system by which it oscillates repeatedly back and forth from a fixed reference point when the external force application is removed Example: see Figure 1-34 for which the following equation of motion holds: mi-+ ke=0; Natural fre- quency in Hertz (f,) = (1/2) VE7ma Rayleigh's Definition: natural frequency of a system computed by an arbitrary selection of a deflected shape which satis- fies the system boundary condition so that it gives the values of maximum kinetic energy to make the lowest natural frequency a minimum. In a multidegree system, the displacement 5;, 8: of the masses, caused by the masses acting as static loads, PEL = Wis + 3 Wi + ft i m EL + ‘ a Figure 1-35. Weightless cantilever supporting load W at its end. Figure 1-34. Un- damped free sin- gle-degree-of-free- dom system.20 Design of Structures and Foundations for Vibrating Machines 3 Wiis + Wad +. Wide Web + oe & ‘ _ 2b We Example: see Figure 1-35 for which the following holds: Rayleigh’s Frequency (2) = go VEWa7W =v we cr where 8 = acceleration of gravity 14. Magnification or Amplification Factor Definition: in a dynamic system, it is the ratio of steady-state displacement response caused by a dynamic force to the displacement caused by an equivalent static force of a magnitude equal to the amplitude of the dynamic force. Example: see Figure 1-36, The figure gives the re- sponse curves for a damped system subjected to a forcing function, F(t) = Fysin wt. Steady-state Displacement Response Amplitude, x= femal + Gayl Static Displacement x= Fe/k ‘Therefore, Magnification Factor, Wie 15. Mass Consistent or Continuous Definition: a mass function which is distributed at each point of its domain and has infinite possible num- ber of independent degrees of {reedom. Example: see Figure 1-37. 1 Equivalent Lumped or Lumped 1 ‘Definition: a concentrated rigid mass in an idealized i system which 1s obtained by equating the total kinetic nergy of the actual system to that of the equivalent system. Example: see Figure 1-38. SS 2s yure 1-36. Magnification factor (M) versus fre- quency ratio (w/a,) for various amounts of damping ratio (D). EI te Figure 1-37. A fixed ended beam with distributed mass over the span. Cike Figure 1-38. Idealized SDOF system for a fixed beam of Figure 1-37 (see also Table 1-1). Figure 1-39. Free-body diagram of SDOF system of Figure 1-19. 16. Motion Equation of Motion Definition: a differential equation describing the rela- tionship among acceleration, velocity, and displacement of a mass in a dynamic system, Example: see Figure 1-39 for which the dynamic equi- librium condition = equation of motion, mi + C2 + =O. Periodic, Aperiodic Definition: motion of mass which repeats itself at equal intervals of time and can be resolved into har- monics. These harmonics may be of different amplitudes and frequencies, Conversely, when the mass slowly moves back to the equilibrium position, rather than vibrating about it, the motion is said to be aperiodic. Example: see Figure 1-40 and 1-41. Introduction-Fundamentals 24 Simple Harmonic or Sinusoidal Definition: motion of a body or parts of a system described by a trigonometric function, a sine or a cosine which repeats itself in any equal interval of time. Example: see Figure 1-42, xlot) subharmonic 7 Defnion: sinaoidal quan having frog sindettsesnsat—n} gauiy having te 3 Sat are fetal (12,18 In) cra subi of Flgure 1-40, Periodic motion of cam ina machine. the ering freuen) of pete factn tok iterated { Example: see Figure 1-43. Superharmonic Definition: a sinusoidal quantity having frequencies that are multiple (2, 3, n) of the exciting frequency of a periodic function to which it is related. Example: see Figure 1-44. Figure 1-41. Aperiodic motion of damped-tree ‘SDOF system. 4 17. Modes Coupled Definition: modes of vibration of a multidegree sys- tem where the motions are not independent but influ- fence each other because of energy transfer from one mode to the other. Example: see Figure 1-45 with two degrees of free- dom, * and 4, for vertical and pitching oscillations, respectively. Case I: Coupling due to mass (Center of Gravity of ‘mass eccentric but equal strength supporting springs), also called dynamic coupling. Equations of motion Figure 1-42, Harmonic motion A sin af and its vector representation. (A) Asinut (HARMONIC) (8) Bsin 3 (SUBHARMONIC) xlwt) mi + kw — hb (y~ 1) = 0 @) LGthG+L)0@—hi-L)x=0 (b) x(wt)= (A) + (B), (PERIODIC) Figure 1-43. Subharmonic, harmonic, and periodic motions. In these equations, coupling is due to a mass which does not have its center of gravity at the midpoint of the Lz, then Equations (a) and (b) are inde- FAsinal + Bsin2ut (PERIODIC MOTION) Asingt (HARMONIC) Bsin2at (SUPERHARMONIC) CENTER OF GRAVITY Figure 1-45. Simplified two-degree-of-freedom ‘model of an automobile. Figure 1-44. Superharmonic, harmonic, and periodic motions.22 Design of Structures and Foundations for Vibrating Machines Case 11; Coupling due to spring (center of gravity of rmass centric, but unequal strength supporting springs), also called static coupling. See Figure 1-46. Equations of motion: mi + (+k) x—- Lia he (a) B+ Let ho-L&-wx=0 & In these equations, coupling is due to the unequal strength springs fy, ka, If these springs are equal, then both Equations (a) and (b) ate independent and thus, represent uncoupled (independent) modes. Uncoupled ‘Definition: the modes of vibration of a multidegree system where each mode describes the complete motion of a particular type by a single independent coordinate Example: see Figure 147 with two degrees of freedom, y and 8, for vertical and pitching oxcillation Equation of motion: ni + Uy = FO 9+ Qd*) a= MeO Because of symmetry of mass center of gravity and equal values of supporting springs k, the vertical oscillation Gesctibed by y and pitching oscillation described by # are independent of each other. First, Lowest, Fundamental ‘Definition: in a multidegree-of-freedom system, & mode shape which corresponds to the lowest frequency jr ealled fundamental or first mode. The mode shapes fare determined from characteristic equations ‘Example: see Figure 1-48 where the beam with con- tinuously distributed mass has infinite degrees of {ree- ‘dom, The frequencies 1, and mode shapes ¢» are given by = sin nrx/Ly "E-= modulus of elasticity foment of inertia frost-sectional area ‘material density Also, see Figure 1-49 for the various mode shapes. Normal, Principal (Eigenvector) ‘Definition: the independent natural modes which satily the solution of a multidegree-of-freedom system. They have the following characteristics: ‘a. They represent undamped free vibration. . They are harmonic. CENTER OF GRAVITY Figure 1-46. Two-degree-of-freedom system. Figure 1-47. Mass with two independent degrees of freedom. —— Figure 1-48. Hinged-hinged beam. —___ > 0.5L oo 0.3333L, ‘ —$ 6667. e O25L O75L 0.5L. Figure 1-49, Various mode shapes of a hinged: hinged beam. so com rent) Ae of freedom. ° ee described by two independent coordinates x, and x Equations of motion: * sma (hs he) — hots = F(t) t ets = Fa(t) s “ ‘These are linear second-onder diferent equations and cpg beeen rate do spring he ave for free vibration ofthe sytem, te i conditions are: ao maka + (a+ he) 2 — F(t) = Flt) It is assumed that motion of ev 1h au f every point in the system n= Asin lol +9) \ (148) xe= Assin (of +9) Substituting Equation (1-46) in the homogenous part in the equations of motion the following are obtained: (he + ke myet) Ay ~ keds =0 These is equations are satisfied for any value of Ay anc ifthe following determinant is 2702, vans (at ke~ met) (ke) (ky) (at he = mot) Expanding this determinant, (1-48) ‘mam Introduction-Fundamentals 23, Figure 1-50. Normal mode vibration of atwo-dopree- of-freedom system. oe This characteristic equation in quadratic form in o! leads to two roots w* and ,%, which ra af and or wich give the natural frequencies mand oy Ratio of amplitudes: From Equations (1-4 first mode (0* = a:*) ° cee (41/4) = kei + by — mad) Second mode (s* = af): | (1-49) hth «iy Symmetrical Case: ka kesk Subsiaing in the characterise B ining teristic Bunion (148), ome (Pmt) = O which souls in woroets ante = (2k/m) - VA(R/m)* = (kfm)? Seat mi) = BURY vem w= VIE and a corresponding amplitude ratio in Equation (1-49). First Mode », = VET (As[Aa) = ky [2k — m(k/m)] 1 The mass is Phe mastes appear to move as a single mass in ether direction without deflecting the central spring. Second Mode w. (Ay/As) = [2k Vv 3kjm m(3kjm) {k= ~ 124 Design of Structures and Foundations for Vibrating Machines “The two masses move in opposite direction and there is a node at the center of the middle spring, Each half then behaves as a single-degree-of-freedom system. See also Figure 1-51 18. Modes of Vibrations, Definition: a dynamic system which is undergoing free vibration, where the characteristic shape is such that the of every particle 3s a simple harmonic with com- mon frequency. Txample: sec Figure 1-52 with motion described by harmonic displacement 2 and 19, Node Points ‘Definition: fictitious points used in a computer mathe: matical model for the purpose of determining response values ‘usually located where the masses are lumped and /or response is to be determined. Example: see Figure 1-53. Vibrating Systems Definition: a stationary point in a particular mode shape which has a constant zero amplitude from equilib- rium position. Example: see Figure 1-54. 20. Oscillation Definition: in dynamics, itis a displacement of a mass which moves back and forth with respect to time from aa reference point. Example: see Figure 1-55. 21. Peakto-Peak (Double Amplitude of Vibration) Definition: an algebraic difference between opposite extremes of vibration displacement measured in a rotat- ing mass ‘Example: see Figure 1-56. 22. Period Definition: the time duration for a single repetition of a periodic motion. ‘Example: sce Figure 1-57, where period (T) = W/f, and f = number of cycles/sec 23, Phase Angle Definition A: ina dynamic system it is a measure of the time difference betwoen a periodic excitation and 3 k + H Tx, k + . vy £ Figure 1-51. Normal modes of vibration. i a ™ fr GS x2 Von we me L > FIRST SECOND MODE MODE Figure 1-52. Modes of vibration of a two-degree-of- freedom system. Zz NODE POINTS= Y KR Figure 1-53. Node points in a space frame mode! {usually located where masses are lumped and response is determined). Introduction-Fundamentals 25 —— 00 First move = !0 a oo io ‘SECOND MODE the resulting displacement response oscillating at the same frequency. Example: see Figures 1-58 and 1-59. Definition B: alternately, in rotating vector form, it is the angle lag by which the response vectors stays be- hind the excitation vector, Example: see Figure 1-60. nh ee angle in a damped SDOF is given $= tan [Co/(E— ma) ] 00 03s ‘067 LO 78 ta [2 - * sO ae tarr* [2D (o/on) [1 — (a/on)*)] Example: see Figure 1-61 Figure 1-54. Node points in vibrating strings. 24. Resonance Condition Definition: « phenomenon of uncontrolled increase in vibration amplitude exhibited by a physical system when it is subjected to an external vibration force of a fre- quency (0) that approaches the natural fre oscillation frequency (en), i€(w/om) = 10. In a damped system, a resonance condition occurs when the displacement be- comes maximized as w goes from 0 9 oy REFERENCE POINT Cus Figure 1-55. Oscillation of a simple pendulum. : : sd |x B | peak ___ Y 5 FT pcax-10-PeAs 2 / \ nee F()=Eosinwt Asin wt-d) 3 5, DOUBLE, AMPLITUDE (EXCITATION (DISPLACEMENT 5 (OF VIBRATION FORCE) RESPONSE) 5 Neen Figure 1-58. Damped SDOF subject to harmonic force, F(t) = Fysin wt. Figure 1-56, Peak-to-peak (double amplitude) of vibration. sinat xen) ONE CYCLE br : peas Te LAG toa, 1= p/w) DISPLACEMENT X ° jer ‘ Figure 1-59. Response motion lags harmonic load- ing by phase angle 6. Figure 1-57. Period of periodic motion.26 Design of Structures and Foundations for Vibrating Machines Example: see Figures 1-62 and 1-63. Due to the presence of damping in every system, a resonance condi- tion at which the vibration amplitude A will approach infinity isin fact seldom achieved. Frequency ‘Definition: when the frequency of a dynamic system (selated to the undamped natural frequency) equals the frequency of the applied force, resonance condition ‘occurs. In this condition, the response is maximized. ‘Example (formulae for resonance frequency) 1.For maximum amplitude magnification. ‘A.Damped resonance frequency (constant force oscik lator, F = F, sin ot). See Figure 1-64 forashich the following holds: Resonance frequency (fn) = fx WI— 2B wesivd “Magnification factor (M)= 1/(QDV1= D?) = 1/2D B. Damped resonance frequency (rotating mass oxi lator), P= m, eu? sin uf. See Figure 1-65 for whieh the following holds _ Resonance frequency (fur) = fx V1—2D* (D 1/3) Magnification factor (M) = A/me/m =1/(@DVT= DS) = 1/2D 2.For maximum wansmisibilty factor, T,. Damped sermanee frequency (constant force oscillator, F = Fe fWO=1/2D PrfFe = Vos w- DAFT where U = VD? +1 and Fy = transmitted force. 25. Response Dynamic Definition: the time-varying displacement and/or stresses which result when a dynamic force is applied 10 fa physical system. Example: see Figure 1-66 for which the equation of motion is mi + Ci-+ kx = Fo sin at. The complete gen- feral solution x(t) of this equation of motion is called dynamic response. Steady State (forced part) Definition: the sustained periodic motion of a physical system which has the same frequency and duration as the dynamic force. sara kA (SPRING (ACCELERATION) A RESISTANCE) Keon {vetocrTy) o oA DP Fe (oampinc _{p\_ (Exc! TATION) RESISTAN eA maa (DISPLACEMENT) (INERTIA) MOTION VECTORS FORCE VECTORS Figure 1-60, Response vector lags excitation vector by phase angle ¢. Figure 1-61. Phase angle in damped SDOF system rie Bsin(wt-A Feeinet Figure 1-62. Damped SDOF system subjected to F(0) Fsinat. Resonance 7 Sonoirion o 10 207 wn Figure 1-69, Response curve for damped SDOF sys- tem (Figure 1-62). SEP es $e fee fai iT mieatsinot Figure 1-65. SDOF sys tem subjected to F = m,eu? sin wt. Fefysinot Figure 1-64. SDOF sys- tem subjected to F = Fpsin at. re ee ea General Solution x(t) = x9(¢) + xe(#), where x9(¢) = particular integral or steady-state response xo(t) = complementary function or transient solution Transient Definition: a form of free vibration, which quickly vanishes due to the presence of damping, Example: see Figure 1-67. 26. Shaft Critical Speed Definition: the angular speed at which a rotating shaft exhibits dynamic instability with rapid increase in lateral amplitude. This develops when the angular speed js in resonance with the natural frequencies of lateral vibration of the shaft. Example: see Figure 1-68. Flexible Definition: a rotating shaft of a machine which has 2 first lateral natural frequency which is lower than the rotating speed. Example: according to an industry standard for gas turbines, the first lateral frequency of a shaft shall be at least 159% below any operating speed; the second lateral speed must be 20% above the maximum continuous speed Rigid (Stitt) Definition: a rotating shaft of a machine which has f first lateral natural frequency which is greater than the rotating speed. Example: according to industry standards for rigid- shaft compressors, the first lateral frequency of the shaft shall be at least 20% higher than the forcing frequency which may be the rotor speed or some multiple thereof. 27. Spring Stiffness Constant «_Duiton: a constant of proportionality beween the fore and the relative deformation i produces inthe Girton of application in « mass strstr element elastic spring observes Hoooke’s Law, that is, the ‘Ping fore is Uncarly proportional tothe ping de- Example: see Figures 1-69 and 1-70. Introduction-Fundamentals 27 cc k “J | x Fesinwt Figure 1-66. Damped SDOF system subjected to dy- namic force, F(t) = Fosin wt. = Ea sin(wt-g) DISPLACEMENT X(t) X= Xt Xp XeAEPOO! sin(wgt + Y) Figure 1-67. General solution of the equation of motion of Figure 1-66. Fom invoducton to Stue tural Dynamics by John M. Biggs, Copyright 1964, McGraw-Hill Book Co. oe Disk ‘Shaft (ED) Figure 1-68. Rotating shaft with lateral amplitude in x and y directions. xi i | Te F SN W-u) v a Figure 1-69. Linear spring vie constant. Figure 1-70. Torsional spring constant.28 Design of Structures and Foundations for Vibrating Machines Equivalent Defmnition: an assumed theoretical representation of ‘an actual physical system such that force-displacement felationship in the former is equivalent to the Tatter. Example: see Figures 1-71 and 1-72, For Figure 1-71 assume the following: 1. Frame weight is negligible. 2. Ginder is sufficiently rigid to prevent rotation at top of columns. Equivalent spring stiffness: 12@ED) , = GE = 35,250 Ibs./fe Linear (Elastic) Definition: an elastic spring observes Hooke's Law, that is, the spring force is linearly proportional to the spring deformation. Example: see Figure 1-73. Nonlinear Definition: ce spring is not linearly proportional to the displacement Example: see Figure 1-74, for which the following olds: k(t) = Ap(E) /48(6) in a nonlinear spring, the load in the Soil Definition: in a soil dynamics system, a schematic representation of a linear load-leformation relationship of the soil using a linear force displacement spring: Example: see Figures 1-75 and 1-76. 28. System Continuous Definition: a body which has continuously distributed mass density (p) and elasticity (E) in its domain, In a Nibeation analysis, this body has an infinite number of degrees of freedom. ‘Example: see Figure 1-77. Dynamic Definition: a structural body which has mass and elas- ticity and whose parts are capable of relative motion Example: see Figures 1-78 and 1-79. Free Definition: if a dynamic system is set into motion by some disturbance at initial time equal to zero and there- after no force is applied, the resulting oscillations caused in the system are called free vibrations and the system is called a free system. Example: see Figure 1-80. Idealzed or Equivalent a ea idealized sytem a convenient repr seat of a seu sacar such that mathematic seaesion cane pevforned. The parameters of 3 saute ae usualy selected #0 that the defer sa enoeucated mas the ste a8 that OF Sono suicant print on the proteype structure, THE ao yom with the equivalent parameters called ta equivalent se. saat ge Figures 181 and 1-52. From Teble +L, can 3 equivlent parameter vales ae To (hy) 304 BI/L2 = 088 X 304 BI/L? 208.5 1/1? ofl mL Linear Definition: system where the principle of superposi- tion is applicable and where cause and effect are linearly related. ‘Example: the influence of various forces acting on @ mass is algebraically additive, ax in the case of static analysis, Nonlinear Definition: in dynamics, the vibration whose ampli- tude is Targe such as when sin @ cannot be represented by only the first term in its expansion but must include several terms (sin =@—67/3+0°/5— - --)3 OF ‘when the spring-restoring force on the vibrating mass is rot proportional to its displacement. Example: see Figures 1-63 and 1-84. Lumped-Mass Spring-Dashpot Definition: an idealized system in which the parame- ters of a real elastic system have been lumped and where the translational displacements are defined. Example: see Figure 1-85. Single-Degree-of-Freedom (SDOF) Definition: rectilinear or rotational motion described by a single coordinate associated with a mas. Example: see Figures 1-86 and 1-87. Multiple-Degree-of-Freedom (MDOF) ‘Definition: a rigid body in space has six degrees of freedom, namely, three coordinates to define rectilinear postions and three to define the angular positions. If W = 1000 Ibest = 30x 10%psi 56.4 int L = 30-0" Fi. Figure 1-72. Mathematical model of portal frame (Figure 1-71) with equivalent spring stiffness, k,. P (FORCE) 8 DEFORMATION Figure 1-73, Characteristic of a linear (elastic) spring constant p(t) p(t+At) p(t) j oO DEFORMATION Figure 1-74, Characteristi i it Faure 1:74, lic of @ nonlinear spring Introduction-Fundamentals 29 Elastic Fo he a Figure 1-75. A circular * foning wbjeced to oy. GRE namic force F = Fo sinut Spe kez and resing on sdmcin ft ‘cal aed (elas. gue 1-7, Math te naltspecs) atcal model of tre footing wit on L equivalent soil spring stiffness, k, x Figure 1-77. A body of continuous mass in three- dimensional space. fa (ial ftzoad ( u Upper torso Arm- { % ‘Thorax- shoulder (34 28{ abdomen > system oan U) inp elasticity of spinal Hips: 1 Force applied to standing subject Figure 1-78. The hu- man body—a typical dynamic system. Figure 1-79. Rheo- logical model of a human body. Figures 1-78 and 1-79 are from Shock and Vibration Handbook by C.M. Harris and C.E. Crede, © 1976. Used with permission of McGraw-Hill Book Co.30 Design of Structures and Foundations for Vibrating Machines ‘Spring i amper, © Ny Initial Excitation oe disturbance Static eer equili- Y oe mo position Displacement x(0) Figure 1-80. Free oscillation of a simple pendulum — Figure 4-85. Lumped mass, spring and dashpot. system. Fit) = Fosinat Figure 1-86. Single recti- linear motion in x-direc- tion in a cantilever. 4 Figure 1-81. An actual physical structure of a fixed beam of a uniform mass and subjected to a uniform dynamic force. Fit) = Fy sinat. Fao) m, Ke KF a ° Figure 1-87. Single rota- tional motion in a 0 direc- rau 1-22, Equivalent (Gelae) SOOF st —_ . ‘tion in a torsional pen- a Plastic dulum- hinge a a z ol z ez Figure 1-83. Elasto-plastic behavior in a fixed stec! “Vener beam. cr Ss ax Plastic Resistance x Figure 1-88. A mass element with a six-degroe-of- freedom system. ‘ Deformation Yo Bofors Figure 1-84, Bi-linear spring representation of the elasto-plastic system of Figure 1-83. ols lity Factor (T,) Transmissi Frequency Ratio (r) Figure 1-89. Transmissibility factor vs. frequency ratio for various damping factors. KE losk 2 2 y A 2 wo Frequency Ratio (7) Figure 1-90. Transmissibility factor vs. frequency ratio for various damping factors. Introduction-Fundamentals 31 there are m masses in a system with no constraints, then the total degrees of freedom for that system will be 6x n Example: see Figure 1-88. 29, Transmissibility Factor Definition: the ratio of the magnitude of the force transmitted to that of the impressed force. Example: 1, See Figure 1-89. Constant Force Amplitude Excitation F= F) sin ot Transmissbility (T-) = Fr/Fo = vi @DA*/ VGl= PVF (RDAs, where Fy is the force trans- nitted 2, See Figure 1-90, Rotating Mass-Type Excitation, F= mseat sin wt ‘Transmissibility (T) =F @DnF/V) =eT, (ary + DAF References 1. Richart, R. E., Jt, Hall, J. R,, Jr, and Woods, R.D., Vibrations of Soils and Foundations, Englewood Gliffs, NJ: Prentice-Hall, 1970. 2. Biggs, John M,, Introduction to Structural Dynamics, New York: McGraw-Hill, 1964. 3. Tse, Francis S., Morse, Ivan E., and Hinkle, Rol- land T., Mechanical Vibrations, Boston: Allyn and Bacon, 1963, é 4, Lapedes, Daniel N., Dictionary of Scientific and Technical Terms, New York: McGraw-Hill, 1974. 5. Harris, Cyril M,, and Crede, Charles E., Shock and Vibration Handbook, second edition, New York: McGraw-Hill, 1976. 6. Thompson, William T., Vibration Theory and Ap- plications, Englewood Cliffs, NJ: Prentice-Hall, 1965. 7. Church, Austin H., Mechanical Vibrations, second edition, New York: John Wiley, 1963. 8, Norris, Charles H., et al., Structural Design for Dy- namic Loads, New York: McGraw-Hill, 1959. 9. Roark, Raymond J., and Young, Warren C., Form- tulas for Stress and Strain, fifth edition, New York: McGraw-Hill, 1975. 10. Clough, Ray W., and Penzien, Joseph, Dynamics of Structures, New York: McGraw-Hill, 1975,2|Develop Analytic: Dynamic Systems ‘A detailed dynamic analysis of a structural system as it physically appears in real life is rarely attempted. The ‘usual practice is to choose an idealized model consisting fof springs and lumped masses which will closely perform in the same way as the actual structure. It is only neces- tary that a proper selection of the system parameters be ‘made such that equivalence of the idealized spring, damping element, and lumped mass in the model results jn equivalent displacements at analogous points of signifi- cance in the prototype structure. In addition, the {idealized model should behave, time-wise, in exactly the same manner as the actual prototype structure, Modeling Techniques “The ecniges adopted in the modeling of sructres subjected to syne lads ate sil nthe developing aac apmonch ose in the modeling of simple Spc such sr beam sporting a wating ado 8 a esksype foundation supporting, a machine, i Be ufone However, when the souctres involved aeaerey indeterminate pe which zest on silk and are Spjeced to complex dynamic frees, the modeling ee en depending on the anal (rek 2) These rea not necessarily ean dzagreement i the eee lndanntals bat rather relate msty to the ace wae elfcieny achieved in the solution. During the Toeba the inverigaton of structural systems sed in specs cuportin and more recently in srwcares used ee eee power plans and offhore stares has sated in an etablished state-of-art in te id of eal dynamics (es. 3 and 5). Te rigors use of Fate computers and finite-element anajsis techniques Me bas the principal agents the development of the sa peer (ek 6). Therefore ies imperative chat eigaens who wih to solve structural problems should seoeeequa exposure to these analytical tol, Model ment of al Models for ing of any structual system is dictated by the require- rents imposed on the solution. The desired solution may be for one fundamental frequency or for a spectrum of frequencies of all possible modes. In some cases it may bbe necessary to find the vibration response at various points of interest. There are a few practical considera tions which are commonly used in all model representa- tions. These include the following: cone er ae Seite 8 ae a esas con be represented by three linear springs 3. Damping Ratio. The dashpot of the lumped sys- tem (Figure 1-2) represents the damping in a dynamic System. Damping may occur duc to several factors present 32 Development of Analytical Models for Dynamic Systems 33. in the system, for example, the frictional resistance and slippage occurring at the interface of surfaces at the contact joints or the sliding phenomenon in the molec- ular structure of the elastic spring. ‘The damping assumed in the structural system is of the viscous type and includes the following additional assumptions: (a) the internal damping present in con- crete and steel structures is nominal, i.e, the damping ratio varies from 2.0 to 5.0 percent for concrete struc- tures and 1.0 to 7.0 percent for steel structures and nor- mally is neglected; (b) the damping agent associated with the soil is in the form of internal damping and geometric damping. The internal damping value is of small magnitude in all modes of oscillation except in the rocking mode. Geometric damping has considerable effect on the dynamic response of the system and is generally included in the model representation. Chap- ter 4 further describes the nature and evaluation of soil damping. 4, Forcing Function. The forcing funetion is normally treated as an equivalent concentrated force applied at points where masses are lumped. Torques are applied at ‘mass points either in concentrated form or are converted into an equivalent force-couple. The effect of earthquake forces may be obtained by the application of a time- history acceleration at the mass points. A time-history displacement applied at the supports may also be used {for earthquake loads as a type of forcing function (refs. 3, 4 and 5 of this chapter and ref, 4, 5 and 6 of chap- ter 3) Models Givi engineering structures of vatious kinds use dif ferent modeling techniques depending on the type of solution desired (ref. 1). Modeling types are given in Figures 2-1 through 2-11 (page 39) for typical structures having different constraint conditions. Also shown is the ‘mathematical model used to represent each physical sy3- tem and a short description of the model, as well at list- ing the applicable equations of motion. Each model is further deseribed below. Development of Equations of Motion ‘The equations of motion which describe the behavior of the mathematical model are developed using one of the following two methods (ref. 3) (@). Dynamic equilibrium equation. In this method, the condition of equilibrium of a mass at any instant of time under the influence of forces and reactions is con- sidered. In order to account for dynamic equilibrium, ‘he mas inertia force is included, Consider, for example, the vertical excitation for the “Machine supported on inertia-block and vibration isolated from the foundation” shown in Figure 2-3(a). The dynamic equilibrium equa- tions are derived as follows: Assuming that at any instant of time, the masses m, and m, have moved up through displacements Z and Zz, respectively, from their reference position, then, for 1, Resistance of spring kn = ka (Z:— Zs) (acting downward) 2 Inertia force of mass my =m,(d*Z,/dt!) = miZy (acting downward—opposite to the direction of dis placement). 3. Excitation force = F. (t) (acting upward) Since summation of downward forces = upward foroes, - miZat her (Zs~ Za) = Felt) (21) Similarly for mass ms: 1, Resistance of spring: ker = kas ( upward) 2, Resistance of spring kre ward) 3. Resistance of damping Gm = Cre Zz (acting op} site to the direction of movement Z,, thus acting downward) 4. Resistance of inertia force of mass m;=m.Zs (acting opposite to the direction of movement Zs, thus acting downward) Equating the downward resistance to the upward resistance, 2s) (acting Zs (acting down mba Cor Bat hes Za = shes (Zs Za) tise + Cra Ba ken (Za Zs) + hes Ze=0 (2-2) Equations (2-1) and (2-2) are the same set of Equa- tions (a) shown in Figure 2-3 of model. It should be noted that Z and Z stand for the frst and second deriva- tives of the displacement Z with respect to time #, i, de/dt and Z= d'z/dt* (&) Lagrange’s Equation. Lagrange’s equation, in its fundamental form for a conservative system in ‘generalized coordinates qs is given by ess _O(KB.) , (PE) _ 00D.) _ a(W) aq a hUFLhDF where K.E. = kinetic energy of the system, P-E, = potential energy of the system, DE. = dissipation energy of the system, W, = work done by the real external forces on the system,34 Design of Structures and Foundations for Vibrating Machines ‘The use of Lagrange’s equation will directly yield as eae rte freedom of the system, given that basic energy expres- sions of the system are know. "This method is applied to the model dscused eave where the dynamic equilibrium equation method was ‘ied in section (a). In this example, there are two €0 ortiats iat = Zot Ze "The energy expressions in terms of Z:are as follows: Kinetic Energy = KE. = 44 mi (Zs)? + 34 ms (2a) Potential Energy = PE. = % kes (2s)* + % ka (2,23)? , Dissipation Energy = D.E. = (~ Gre 2:)(Z2) Work by external force = W.= Fe (Z:) ‘The dissipation energy due to damping force must be taken as negative, since a positive damping force is lways in a direction opposite to the positive dsplace- tment. The derivatives with respect to Z, are ACE) _ ARE) «9,4 a a a a(KE) _ 9 te, (Zr — Za) aE) 9 az, ow) an, Substitution of the above in Equation (2-3) leads to aw) _ x, az = Fe) im By + hy (Za — 2) = Fall) 4) "The derivatives with respect to Zs are UKE) _ 90KED gyi, © (AED om mds aie ah as a kn a aw) _ ae) on a2, Substitution of the above in Equation (2-3) leads to the ‘equation of motion: 0 @5) mals + Cuda + be, (Zs — 21) + be Model 1—Vibrating Machine Supported by Block-type Foundation (Figure 2-1) ss ype of foundation isa very common frm of a ro ed's anny conidered by des 8 Dor re pene ad mdr pants Thee an ede shaper ae pomsle and sould ro ee Ge 2). Vena snd horizontal snes PSE ay near aifeenal equation and he are esc Dy aru frequencies and. vibson ston or saan the rocking mode, rere ect of the hart ode may be ignored Se ec foundations In tat cae, hy #20, for voy ste png eer pve ard Us Bas sn te gure 21 reduces t Equation (8) Sin vn atten a) of igre 2 abo “wees Me deere the motion in cordinat® ¥ thas follows HEADLEY = OH = To cos ot (26) is equation slong with Equations (a) and (b) of Fee SEE ge tled according tothe procedurs Fl japan [an example sled in Chapter 6 tien mle te required step in the calulation wi en the wbaton reponse of te coupled a cl foe Equations (6) and (8) of Figure ry the sain ean be fund by subating y= Ay sin wt + Ay, 08 ot Qa) ¥ = Ag, sin ot + Aye 608 ot (7) ating. the sn Equations (c) and (4) and then separating a eatas contain either snes or cosines, This pro- cate al rel in four seltancous equations wi fear untnowse, A complete solution of this type of teuaton is given in Appendix A. Model 2—Vibrating Machine Supported by Mat-type Foundation (Figure 2-2) is i ye used for the ‘This type of foundation system may b ; siucnion here several small units are placed side by ide or where a firm soil with a high water table is en- Development of Analytical Models for Dynamic Systems 3 countered at plant grade level. Due to the flexibility of the foundation mat, a high magnitude of damping will be encountered in the rocking and horizontal modes. ‘Therefore, only the frequency and vibration response calculations in the vertical mode are required, ‘A single Iumped mass model may be used when one set of machines is supported by a relatively rigid mat foundation. However, the model is divided into discrete lumped masses when several sets of machines are lo- cated on a flexible mat foundation. In this case, con- straint conditions are applied to the boundaries in the directions of translation for the sake of stability. The spring constants for each element depend on the mat rigidity as described in Chapter 5. Mode! 3—Machine Supported on an Inertia Block and Vibration Isolated from the Foundation (Figure 2-3) In special cases and due to environmental conditions, it may be necessary to limit the vibration amplitude at the foundation base to much lower values than those usually allawed. This requirement may not be practical to achieve even by proper selection of mass or base area of the foundation. In such cases, use of an inertia block and spring absorbers is recommended. In normal behavior, three forms of excitation are ponsible. Excitation in the vertical direction is inde- pendent of the other forms of oscillation. Excitation in the horizontal direction is generally coupled with the rocking mode; however, for a machine which is located at relatively lowe height (his < Yj) then investigation of the horizontal and rocking excitation independent modes is sufficient. ‘The parameters kx, and ka are properties of the spring absorbers. Parameter m, is the combined mass of the machine and the inertia block together. The parameters ke, kr hoy and Cz, Gr, Cox are spring con- stants and damping coefficients, respectively, of the soil in the three modes considered and should be determined using the elastic half-space theory as described in Chapter 4. Parameters mz and J, are the mass and mass ‘moment of inertia, respectively, of the foundation. The solution of the differential Equations (a) and (b) of Figure 2-3 can readily be found for the natural frequencies, mode shapes, transmissibility factors, and the vibration response. Often, the fundamental fre- quency and the transmisibilty factor are the principal results of the analysis. The set of differential equations (6) of Figure 2-3 is in simultaneous form, and a manual solution is tedious to perform. This system of simul taneous equations is rarely solved by hand unless a thorough investigation of the system is required, and then the solution is obtained with the help of a computer program, However, solution for a similar type of equa- tion of motion has been performed in Appendix A. Model 4—Vibrating Machine Supported by a Cantilever (Figure 2-4) It is sometimes required that a vibrating machine be supported on a cantilever. In such instances, a vibra tion analysis is considered necessary. Two modes (verti- cal and rocking) are possible (ref. 3). The calculation of the rocking mode may be ignored if the distance h is found to be small, and the cantilever arm is rigidly secured, The calculation of the vertical mode is generally performed because this provides the fundamental fre~ quency and the largest vibration response. The mass parameter m, is considered lumped at point © and con- sists of the mass of the machine plus an equivalent mass for a portion of the cantilever calculated according to the procedure explained in Chapter 1. The spring stiff- nes parameter fe is the flexural stiffness of the canti- lever at point O. Damping in the system varies from 0.005 to 0.05 of critical, depending on the material ‘The investigation fer the rocking mode is performed on a similar basis as for the vertical mode. The mass moment of inertia parameter I+ is calculated for the equivalent mass m, about the point. O. The rotational spring constant Ay is calculated by applying a moment at point O about the x-axis (the x-axis is perpendicular to the figure). ‘The maximum vibration response calculated for each mode may occur at different times. Therefore, in ob- taining the total response, the maximum of the sum may ‘occur at some specific time within the interval of inter- cst. However, obtaining this maximum value may be difficult. Therefore, a simple summation of the individ ual maxima is generally performed, which results in a conservative estimate of total displacement. The solution of the equations of motion has previously been described in Chapter 1. Model 5—Vibrating Machine Supported by a Fixed Beam (Figure 2-5) ‘Mathematical modeling technique for this physical system is similar to the cantilever system above except that the parameter determination differs, ‘The mass parameter m, is the combined mass of machine and a certain length of beam and is lumped at the intersecting axes of the machine and the beam ‘The spring constant k, is a function of the flexural stff- ness of the beam. Both parameters me and k, can be eval- uated by using the expression given for Case IT in Table I-l, The parameters J+ and ky can be determined by following the procedure described under Model 4 above.36 Design of Structures and Foundations for Vibrating Machines All modes of oscillation which may occur due to the faction of the forcing function generally need investiga- tion. In this case, the vertical mode and the rocking mode (about the x-axis) fall in this category. ‘The vertical mode investigation is necessary because it gives the lower value of natural frequency and the higher level of vibration response. On the other hand, the rock- ing mode investigation provides higher value of natural Frequency and a lower level of vibration response. ‘Therefore, if the machine operating frequency is found to be very close to the rocking mode natural frequency, then the model parameters may need modification in order to avoid possible resonance conditions. Thus, both ‘modes may need to be considered in some machine supports. Model 6—Typical Elevated Pedestal Foundation (Table Top) (Figure 2-6) In this physical model, there is some variation in the use of the modeling technique (ref. 1). In this example, four models are considered, and the merits of each and the effort involved in their solution is discussed. In the modeling procedure the following factors may be used in determining the type of model to be used. Model A: Single-Lumped Mass (Uncoupled ‘Superstructure and Foundation) (Figure 2-7) —A preliminary investigation is required. “Reliable information on the parameters is lacking —The structural framing system is not well defined ‘and preliminary dynamic characteristies are desired. "The beams have much higher stiffness than the columns — Information on vibration response is not required. Because of the lack of interaction between super structure and substructure, this model gives the dynamic solution of individual subsystems and generally results jn a calculated higher fundamental frequency. There fore, the engineer must use conservative criteria to avoid the resonance condition by first making certain whether the system is to be low or high tuned. This condition is generally achieved by varying the mass or stiffness of the system components, The calculation of parameters for the superstructure representation would require a familiarity ‘with structural frame analysis for the calculation of the spring stiffness constants ky and ky. The mass parameters im, is the mass of machine plus the mass of top framing and upper half-length of columns, The substructure investigation during the initial investigation phase of the total foundation is not conducted because its influence fon the solution of the superstructure portion is small. However, ifthe substructure base slab configuration is fhecesary, then information on dynamic soil properties inl be Fequired, and the slab should be rigid enough Such that it ean be represented by a single lumped mast ‘The model parameters for spring stifness fy fay by and damping constant Ce, Cr, Gy in the three modes of tecitation can be determined from the soil properties by the elastic half-space theory as described in Chapter 4 The mass parameter ma is the total mass of machine and of the entite foundation structure, The parameter Iris the mass moment of inertia of the machine and the entire foundation atthe center Hine and at the base of footing. The equations of motion involved in both the subsystems are linear second-order differential equa tions which ean be solved for the natural frequency find vibration response by the procedures desccbed in Chapter 1 Model B: Multi-Lumped Mass (Uncoupled ‘Superstructure and Foundation) (Figure 2-8) ‘This model may be used when the following types of results are desired according to the listed conditions: —Superstructure is well defined, and the fundamental frequency is to be lower than the machine operating frequency (structure to be undertuned). _-Foundation structure is supported either on highly firm soil or rock formation or a rigid deep founda- tion. Accurate determination of vibration response is not a requirement. Foundation structure height is low (less than 20 {t) and foundation is not supporting more than two machines, ‘The dynamic characteristic of the superstructure may be calculated in either of two ways: 1. Rayleigh's Frequency (see definition in Chapter 1). in this method, the weights of structure and of the machines are applied as static forces at a diterete num- ber of points acting in the direction of the deflected shape of the structure (assumed as the fundamental mode). ‘The following formula is used: ae BRA en, @8) where F; are the equivalent lumped forces of distributed weight of structure and machine acting at point i; 8s are displacements of the structure at points # produced by the forces Fi; g is the gravity constants; oy is circular Development of Analytical Models for Dynamic Systems 37 natural frequency in the mode cometponding 10 direction ofthe acting fres a "The nccuracy of the natural frequency calesaton obtained by ths method depends enirly an how cle to realy are the amumptions made in stuming the fected shape of the sructre, However, note that for the appronimate rls obtained by ths meod, the lower frequeney alway gives the bet sppresination, 2. Modal Multidegree Lumped Mass Analysis. The normnal modes (modal analysis and normal modes are defined in Chapter 1) are determined separately and then superimposed to provide the total response. A normal mode (or natural mode) of vibration is associ- ated with each degree of freedom of lumped mass in the system. The property of a normal mode is that the system could, under certain circumstances, vibrate freely in that mode alone, and during such vibration the ratio ‘of the displacements of any two masses is constant with Mme, The rao defines the character shape of the ‘The equations of motion in matrix form for a multi- degree system, but having no external acting force and xno damping, have the following form: mF + kya = 0 9 The natural frequencies of all modes are found by as- suming a harmonic motion for each mode. During vibration in any single mode, the displacements of sev- eral masses is always in the same proportion, i. all possible postions are geometrically similar. Substitution of the assumed mode function in the equations of mo- tion and rearrangment would result in a characteristic value problem. Briefly, solution of these equations con- sists of expansion of their determinant and subsequent solving for the characteristic values (or eigenvalues, natural frequencies) The vibration response of the multidegree system due to applied forces. or initial conditions is obtained by treating each normal mode as an independent one- degree system. The formulation of the problem by this method of analysis requires a thorough background in dynamics analysis and is generally attempted using a compter progr, The fens obtained from this ‘method will be the natural frequencies for each d Of freedom of the mass, mde shapes for each fe quency, and response at the joints in the form of vibra- tion amplitudes, internal forces, and moments. An example problem for three degrees of freedom based fon this method of analysis has been presented in Ap- pendix A, ‘The investigation of the substructure is performed according to the procedures stated for Model A. The vibration response results obtained for both the sub- systems are combined by tome rational procedure such that maximum values are achieved at a particular mo- ‘ment in time. Model C: Two-Lumped Mass with Coupled Soil-Struc- ture Interaction (Figure 2:9) Model includes soil-structure interaction and also includes the true dynamic characteristics of the founda- tion system. This model representation may be employed when the following conditions are satisfied Foundation structure is supporting not more than two machines, ie, the length of the structure should be small and its height is not greater than 20 ft Foundation structure is well defined, and reliable information on the soil is available. Natural frequencies of all subsystems of the model do not fall in the resonance zone (0.5 to 1.5 of the acting frequency of the forcing function) ‘The types of results obtained from this model are: —Natural frequencies for all important modes which are excited by the forcing function. —Vibration response at the axis of rotation of the machines and also at the base of the foundation. ‘The dynamic behavior of the model has been sep- arated into parts a and b. Part a represents the coupled modes of horizontal and rocking oscillations while part bb represents the behavior for vertical oscillation. The equations of motion of these two parts are deseribed in Model C of Figure 2-9. "Dhe natural frequencies of these models can be de- termined from the equations of motion by intially ignoring the terms containing damping and forcing funetions. An exponential type of displacement funetion is substituted for each degree of freedom (mode shape) jn these equations. A determinant is formed from the resulting equations which, when expanded, results in a characteristic or frequency equation, The solution of the frequency equation will give the expression for the natural frequencies. A general solution can be formu- Jated with the use of natural frequency expressions to ‘obtain the mode shapes. The constant of integration in the equations is evaluated by equating the general solution to the initial conditions. The manual calculation of vibration response of coupled mode part a is tedious38 Design of Structures and Foundations for Vibrating Machines and time-consuming and normally is accomplished with the use of computer programs; however, a manual olution is presented in Appendix A. The techniques fused in vibration response analysis follow a modal anal- ysis of the lumped mass mulidegree system which is briefly described above in discussing Model B. The r= sponse analysis of part b is relatively simple and can be obtained manually ina manner similar to part 2 However, if a computer program is used for the solution of part a, then parameters of part b can also be easily combined in order to provide the total results (ref. 2) Jr that case, the rocking spring br of part a should be represented by three equivalent vertical springs of equal Stiffness (see Figure 2-10) with the following conditions ky = ket ke She en VOERE In this case ke = rocking spring constant,fe is the verti- cal spring constant of mi, and ¢ is the distance between to equivalent springs of stiffness k,. The distribution cof damping coefficient Cr related to Cx is rather com- plex; however, a similar form of logic may be followed Es has been done for the spring constant. The damping onfficients associated with the equivalent vertical spring fh, would not be all equal if the same value of ¢ is used: ‘Therefore, an equal value of the damping coefficient Ce has been used for the exterior springs, and a different value of damping coefficient Cw has been used for the iddle spring. ‘The member which connects the three equivalent springs should possess an infinite flexural suf frase but should also maintain the equivalent values of m, and Te. Model D: Multi-Lumped Mass with Coupled Soil- Structure Interaction (Figure 2-11) “Dhis model provides the design engineer with a com- plete insight of not only the dynamic behavior of the Tuperstrueture but also identifies the critical modes in the soil-structure system, In cases where access to & ‘computer program capable of solving dynamics problem fe available, then it is very convenient to resort to this modeling technique. In this investigation, several kinds ‘Of results can be obtained provided that the following parameters are available: see Member sizes of the structure and geometry are avail ‘able and have been proportioned such that: (a) the rigidity center of gravity of the structure in plan Coincides with the center of gravity of the masses of the structure; (b) the flexural displacement of the top of the structure in either direction is uni form across the length when the top mass of the structure and of the machines are made to act a5 horizontal loads; (c) all columns deflect equally under static loads; (d) the center of resistance of the supporting soil coincides with the centroid of all statically imposed loads. Information on founding depth of the structure and ‘on soil, such as shear modulus (G), Poisson's ratio (@), and bearing capacity of the sil is readily avail- able, “The results that can be obtained from the analysis of this model include __The natural frequencies (eigenvalues) ; the number of frequencies will depend on the number of active fonts with humped masses and the degrees of free- dom of each joint considered in the analysis. __"Themode shapes (eigenvectors) for each of the nat ral frequencies. _The vibration response (in terms of displacements, ‘velocities and accelerations) for each mass having six degrees of freedom. _The moments and forces at each mass joint and the reactions at the supports. ‘A dynamic solution of this model using the computer program STRUDL is presented in Chapter 7 (page 114) : References 1. Arya, Suresh C., Drewyer, Roland P., and Pincus, G. Simulation of Structures Supporting Vibrating ‘Machinery, Summer Computer Simulation Gonfer- ‘ence, Washington, D.C., 1976. 2 Richart, R. E, Jr Hall, J. R., Jr, and Woods, RD, Vibrations of Soils and Foundations, Engle- wood Cliffs, NJ: Prentice-Hall, 1970. 3, Biggs, John M., Introduction to Structural Dy- namics, New York: McGraw-Hill, 1964. 4, Harris, Cyril M., and Crede, Charles E., Shock and Vibration Handbook, second edition, New York ‘McGraw-Hill, 1976. 5, Norris, Chatles H., et all, Structural Design for Dynamic Loads, New York: McGraw-Hill, 1959 6. Logeher, Robert D., et al, ICES STRUDL-H, Em- iineering Users Manual, Vol. 2, second edition, School of Engineering, Massachusetts Institute of ‘Technology, Cambridge, MA 02139, June 1971. Figure 2-1 bd Vertical Mode ey fy Horizontal Mode iF Rocking Mode Development of Analytical Models for Dynamic Systems 39 Xo COMBINED CG. =| Vibrating Machine ‘Supported by Block-Type Foundation (Model 1), Vertical made normaly behaves independent of othr modes, The mass (m) Stacie an funate assed ob coated ane vert pring constant of soil (k), damping in sol (C,), inertia of mi andthe forelngtuncton (of he machine have eine faction cone ing withthe vertical axis. Equation of motion: _— mit O2+kz= Fl) @ The representation forthe hovior 3 the hortzontal mode Involves an appr tis made contrary oe vera mode he mmusoc dont eon te same ronal axis, norte ine of action ofthe forces coincides. Ove to these reasons, this mode is normally couplet eneen ally coupled with the rocking mode. Equation my+Oy+ky=F, (b) The model is better representation of the tus dynamic behavior ofthe structure, However, th aa! suo itt atmet due a cou and rocking motions. This coupling effect should be investigated for the case when the machine is| eventos fo th cae when fre machine locale hgh above fond Ii +6,¥ + (y~ Yio) N,C,¥ = Fy cos ut oy YE + (Cy + MEOW + (ky + hee hye ; (a) =NG,¥— hikyy = (YH = Ty) = FH cost Model 1. Vibrating machine supported by block-type foundation.40 Design of Structures and Foundations for Vibrating Machines Development of Analytical Models for Dynamic Systems 41 (EX sero SX) FOUNDATION Machine Supported on Inertia-Block and Vibration Isolated from the Foundation (Model) (Model 2) _ : t F, equivalent This model representation yi the natural fequondies and wration For a mat foundation, which is relativly rigid (corner dis- t eee response inthe vertical drston Since tere are wo masse nthe model, |F For grnont fo 88 x maximum under machine foad atthe my inertia thee are wo degrees of eedom and two natural fequencie, and mode Pasco eM achine end mat slab can be lumped Into a single IL block shapes willbe obiained. Daring In isolation spring is general nenected, m center ig te voricl mode needs investigation. Class ha However, sll dempingis niin and, theatre, included. uations of trees ety mat as "igi is diacussed in Chapter 8. The nt Equivalent este" conetant k, and the damping constant C, are to be ser cetrom elgatle half-space theory of from soll data Equation of motion: ¢ hyp fa Foundation mie ii + Ci hl 2) hake tt) (@) _ mis og 4+ ke=F.) cy ‘The characteristics of this model are similar to the above model except Lumped Mass: ; Vertical Mode beled tz When the mat foundation is @ thin conerete lab, and dy ky + gj — 0) = Fel) Marne analysis is required, a discrete lumped-mass mode) righ + Cag + hy lly 1+ aha = FINITE rae appropriate for analysis. The mat may be divided the + Casket hy (he 00 * Bate © ELEMENTS ts mrercaymmetical arrangement of finite elements (elther MASS mi — Wianguler or rectangular shape) with at feast bending Capabity, The masses at the joints can be lumped by @ cariputer program command. The mass of the machine is Go'be added through the computer program input at the y., appropriate joints, The soll spring constant ‘calculated — 1 eine total mat should be proportioned to the Joint in rela, tin to the peripheral areas for a rigid mat. For flexible ents, the procedures described in Chapter § should be Tund. A computer program Is necessary for frequencies Ualoutations. H vibration response calculations are desired, Gh equivalent damping constant of about 0.1 may be used ws ied ergs mein pany aan sate ec re ct wi Setesoeuncr sey hates Sein Srse Gee ee ee ky + yy (%j — Xp — Veh) Img + Cake + hake ~ has ~~ Vel) Wada + Cate + myn + Kate = FN (D (o) a SPRING CONSTANT 3 Discrete Number of Lumped Masses Coupled Horizontal and Vertical Mode Rocking Modes Figure 2-2, Model 2. Vibrating machine supported by mat-type foundation. i: Figure 2-3,’ Model 3. Machine supported on inertia-block and vibration isolated from the fo. lated from the foundation,42. Design of Structures and Foundations for Vibrating Machines ie jw x —+y Vibrating Machine Supported by a Cantilever (Model 4) fa similar form of mathematics! can bo represented by a a ‘dynamic model parameters will cl san physi oy Bat etna voundarycontions ysl eure meatal {f = La 'e hy model, However, becal Jrange. Two form of modes In th rating Machine ‘Supported by a Fixed Beam (Mode! 5) use of different e direc Rocking Mode (Rotation about x-Axis) Vertical Mode ti mass of machine 4. Vertical Mode—Mass (m,)i the y equivalent mass of support and is lumped at point 0. Spring stiffness (K,) is the deflection stit- Pees at point O for a flexural member. Damping in ind is neglected. Equation cach a eystem is small and is neglected. Ea Damping sm meter Let ke Fah cos ut tee = M,cosut coer : yy a cantilever. Model 5 is vibrating machine Figures 2-4 and 2-5. Model 4 isa vibrating machine supported P ‘supported by a fixed beam. J axis of Rotation Wy 'ot machines oe Cs soll Supported 2. Rocking Mode—Mass (|) tia of mass of the machin ‘about point O. Torsional fotational stiffness at point Damping is small and ) is the moment of iner- 'e “that portion of support ‘spring stiiness (k,) is the {0 for a flexural member. thus is not considered. Equa~ Figure 2-6, Model 6 is a typical elevated pedestal foundation “table top.” Development of Analytical Models for Dynamic Systems 43. te pe aber hd, Pe sa >t ce fH oe omy) ss eK REPRESENTATION Model A This model is composed of two parts: Figure a representing the top floor plus the supporting columns, and Figure b representing the total structure and equipment plus the bearing soll. These two subsystems are ‘assumed to possess independent dynamic characteristics, and there is a lack of interaction behavior between the two. The basic assumption in this model is to consider the column bases as perfectly fixed, The structure is s0 modeled that only its three predominant motions (later the C.G. of masses. Equations of motion: For (al: mig + hyky = Felt) mda kta = Fel) Vertical, and rotational) are predictable at For (Oi mgiy + Gy, + x)= FO m,d, + C8, + hz FD Lib + O06 + kod = Fe(OH) Figure 2-7. Model A. Single-lumped mass model of table top (uncoupled superstructure and foundation). \ Se / = Model B ‘This model representation similar to Model A, but some refinement is added by lumping the masses in the superstructure at points where dynamic response Is important (see Figure a). The basic assumptions in this model are stil the samo, but each individual element acts independently ofall others. This is generally permis- sible when the natural periods of the elements differ by at least a factor of two in any one direction of motion. In Figure b, the representation is similar to Model A. This model is assumed to include the complete dynamic characteristics of individual members of the superstructure, Equations of motion: For (ay: m+ yy These equations arein ‘matrix form and can only be solved with the help For(ok mg, + Ot, + Kx, =F, (0) may + Oo, +2, = FD Ot ker, H of a computer program. Figure 2-8, Model B. Multi-lumped mass model of table top (uncoupled superstructure).44 Design of Structures and Foundations for Vibrating Machines this models an improvement over Model A duo tothe incorporation of interactive capability petwes fy This mete anich was nogloctd in Model A. The model is shown in two pats, Figure sf" Te ‘coupled hori- tare a Mode and Figure b for vertical mode. The model can be analyzed as how! Natal ‘much zontal and rock raoracyr nowever, a coupled mode of both of these modes can also be Susie, without toe oar ately, The method of parameter calculation also does not change. The drawoaes this model aaa grt wen the natural frequency of individual elements is required or Its necessary to calculate the uibration response at some other points In the structure. Equations of motion: For (a For (Dy Ika Hylity — %y ~ YH = Fea IMigiy + ky (@~ 2) = Fall) ily + Gy + KX ~ Malte #1 — YH) = 0 m8, + Cys + hy (@y ~ 29) + hy 2 =O 1b + Oy6 + makaH + keh = Frolt)-H - Model ©. Two-lumped mass model of table top with coupled sollstructure inloraction. @ EQUIVALENT kinky Figure 2-10. Model C-Aiternate form. Coupled lateral, rocking, and vertical oscillation representations of table top for computer program application, Development of Analytical Models for Dynamic Systems lH oer ed COMPLETE STRUCTURE ++ FOUNDATION ‘TOTAL REPRESENTATION Model D & ae a polled in relation to the previous model, not only with respect to degree of reliabilt sein na en rn gn sa ae Pie nen ef ae nh es iat ee ah Se ee ee aa eee ies to ssn het my en Son Figure 2- igure 2-11. Model D. Multi-lumped mass model of table top with coupled soil-structure interaction.3|Development of Information, Trial Sizing, and Design Checklist ear ig eT dauitamente, ‘Therefore, the required design se ee ae et i es i ee te i gr Capen a td Machine Properties and Requirements ‘of many types mee be classified in oe of we in eer, cme ee en Ai include the fol- machine propetties and parameters include th lowing: Cnitine drawing of machine assembly “Functions of machine . Weight of machine andi rotor components TITER of center of gravity both vercally and Teizntaly a ; Spot ranges of machine and components of {e coency of unbalanced primary and secondary fores iegehade tod direction of unbalanced forts both vereally and boriontally and their points of 59> fetion TITS imposed on the foundation with respect to ifeondel defection between point on the plan area of foundation Foundation requtemens he ‘The physical se of the structure depends on requed bass dimensions forthe machine. ORen, ap arenes such ar platforms acd piping supports re aeraaates in be dimensions. The outline fr the orihe base generally species minimum dimensions sree cate specie area that must be et clear for sre ne tachment For example, in tains, certain ac ander and ever the machine must be Tet clear Gems on ROE on eer Chive fone includes information of the overal ase tel mature the machine, Shut the Pate te of aetna portance ae ann then a more conservative design apo ih kemment, for example, shut down of & small duction, However 8 pup any not affect plant pr : Pa ential sompresor i to be shut down, a ul ra dllat operation may be affected. The designer eter the Teel of conservatives balanced agin nile unnecesary expense. 48 Development of Information, Trial Sizing, and Design Checklist 47 ‘The weights of the machine and its components are provided by the manufacturer and serve to give a pre- liminary indication of sol support feasibility. The weight of rotors and speed in centrifugal machines determine the magnitude of possible machine unbalanced forces. The center of gravity location in the horizontal and vertical planes is often provided. When not available, calculations or assumptions may be needed. Basically, the machine is set on the foundation in such a way as to avoid eccentricities between the resultant of all loads and the support center of resistance, that is, the centroid of the pile group if pile supported or the center of resistance of the supporting soil if soil supported. ‘The speed range and frequencies of primary and secondary forces are required in the dynamic analysis in order to check for posible resonance. The designer is generally only interested in the operating frequencies, although in many machines, there will be particular speeds briefly attained during starcup or shut down ‘where the assembly will be in resonance with the ma- chine frequency. A temporary resonance condition may be tolerated in such cases especially when significant damping is available, ‘The magnitude and direction of unbalanced forces are often not available from the machine manufacturer. Some claim that their centrifugal machines are perfectly balanced, a condition that may be approached initially at the manufacturing plant, However, after a few years of use and due to normal wear, some eccentricity will exist regardless of initial machine and installation work- manship. Eccentricity criteria useful in designing struc- tures supporting centrifugal machines are given in Table 3-1 and Figure 3-1. For reciprocating machines, the unbalanced forces, which are generally of consider- able magnitude, are provided by the machine manu- facturer Limits on differential deflection allowed between points of the foundation are set to avoid possible damage to piping and other appurtenances that connect to the machine. In some high pressure (50,000 psi) piping, differential deflection limits are approximately less than 0.0001 in. ‘This is generally the case for machines with very rigid (thick) attached piping. Foundation requirements refer to minimum depth of foundation, as dictated by expansive soils, frost action, fluctuating water table, piping clearance, or paving elevation. The top layer of weathered soil is often not recommended for supporting foundations since firm, Undisturbed soil is required. Also, the bearing strength required for the soil may dictate placing the bottom of the foundation at a deeper level for soil-supported struc- tures. The recommendations of the geotechnical consul- tant must be integrated into the design process (ref. 3) Soil Parameters Knowledge of the soil formation and its representa tive properties is required for static and dynamic analy- sis, In the case of a sand or clay formation, the informa- tion is to be obtained from field borings and laboratory tests. These are usually performed by the geotechnical consultants. Chapter 4 describes procedures for proper evaluation of these parameters and discusses other soil- related problems. The following parameters are generally required: density of soil, y Poisson's ratio, » shear modulus of soil, G, at several levels of strain (or magnitudes of bearing pressure) Table 3-1 Design Eccentricties for Centrifugal Machines For operating speeds up to 3,000 rpm, we have the following: Eccentricity, in. Operating Speed” (Double Amplieude) 70 “O18.082 1,500 0 3,000 02 “Mair, Vibration Anatyss and Design of Foundations Jor ‘Mathines ond Turbines, Akademiat Riad, Budapest, Collet’s Holdings Limited, London, 1962. ‘The following are the modifed API standardst for centrifugal compressors! el) = a VTRDOUTEp < 1.0 (mil), where « = 045 at installation time, alter eoveal years of operation, rating machine speed, rev/min, Yoon int me th ‘American Petroleum Tastiute Standard for Centefugal Com pressors #617, Section 218.4, a= moxifed by the parameter a For gear wits, we have the following:* Double amplitude {including runout, mils ‘Maximum continuous ‘speed, rpm Shop test Shop teat unloaded Toaded ‘Up t0 8000 20 1B £8,000 10 12,000 15 10 yer 13,000 Less than 1.5 Lesa than 1.0 “mei Inte Pteum Stade 3 Units for Refinery Servi ‘gust 1955, igh Speed, Special ‘API Standard 613, (Table 3-1 continued on page 48)48 Design of Structures and Foundations for Vibrating Machines perma ot eae STR ee ze ‘Peak-to-peak displacement eee oa oan i : ood eam i ae) rn 7 ie as : BP ia é co altgnating-current motors, use the highest synchronous NI [KRUGLOV. ae er cilia rene eae § oe atecr simran Eali 2 dike es Boz ee i as a No eee pan Speed, som kpittuder in. oa (BEARING) |" (GHAFT) sees Sa = Sear seuoo somo?” 600 {500-2,999 0.0020 ‘Vibration Frequency, cpm BE aw a cana every of machinery Triomal Blectrieal Manufacturers Assocation Standard, MEI 2o.52, Balance of Machines," July 1900. Form-wound Squirrel-cage Induction Motorst tt to: peak dleplacement Peak peadey in Synchronous speed, rpm Motoron__ Motor on cimietaount gle mount wae 02 0025 £20 4.000 Sots tion ito Sad above ont ot TAnatcan,Peglcum Insite. Standard, “Recommended Hh cace ‘for Form-wound Squirrel-cage Induction Motors, RET Staedard coefficient of subgrade reaction of soil, if the above parameters are not accurately known the foundation depth and the bearing pressure at which the above parameters are applicable other information required for the static design of the footing ‘The soil density 7 and Poisson's ratio » are normally reported by the geotechnical consultant. Chapter ¢ list the typieal range of values for these parameters. Often, ‘mass density p is reported in lieu of soil density y, but if cone is given, the other is known, since pave where g is the gravitational constant. "The soil shear modulus G is not usually reported unless specifically requested. Since this parameter is a controlling vibration. From C.M, Harris, and C.E. Crede, Shock and Vibration Handbook. Used with permission of McGraw-Hill, 1976. factor in the calculation of the half-space spring con- stants it is desirable that the most reliable value be fvailable. This often means that field testing will be Necessary to obtain G since inaccurate values of this parameter will rerult in a worthless analysis. Therefore, Chapter 4 describes methods and procedures which should be followed in the calculation of G. The soil Swain level has an important effect on the value of the shear modulus and should be accounted for in its caleu- Iation as discussed in Chapter 4 “The coefficient of subgrade reaction may be used in lieu of G or directly in a Barkan analysis (ref. 2). How- cever, since the elastic half-space theory (ref. 1) is recom mended here, its availability may be used as a check on the onder of magnitude of the soil moduli as calculated using the half-space theory. “The soil parameters discussed above generally vary swith depth and effective beating pressure. Therefore, the Specific values to be used in design should correspond to the actual bearing pressure and foundation depth used. In particular, the shear modulus is sensitive to strain level and since dynamic loads often produce low strain levels, the shear modulus used in analysis should cor respond to the actual dynamic strain level expected ‘In addition to the information listed above, the design ‘engineer, either by himself or with the help of the geo" teclinical consultant, should establish the layout for the foundation structure. There are two common types of foundations used: concrete block footing placed directly Development of Information, Trial Sizing, and Design Checklist 49 con the soil or rock, and concrete footing supported by piles or piers, The preference of one system over the other should be decided by taking into consideration: relative economy, settlement, bearing capacity of the soil, vibration isolation, and the level of the underground water table. Pile- or pier-supported footings are the exception and are used only sehere poor soil conditions are found, Environmental Conditions ‘There can be several situations during which a machine installation is in the vicinity of vibration sources such as quarry blasting, vehicular traffic, construction pile driv- ing, o the location is in a continental zone where seismic ‘occurrence is possible, The design engineer must then establish the severity of the situation and, if required, should seek the help of a vibration measurement con- sultant, The information requested should include the character of the vibration and the attenuation at the installation site, The effects of seismic forces can be determined through information and procedures de- scribed in References 4, 5, and 6. Trial Sizing of a Block Foundation ‘The design of a block foundation for a centrifugal ‘or reciprocating machine starts with the preliminary sizing of the block. This initial sizing phase is based on a number of guidelines that are partially derived from ‘empirical and practical experience sources, Initial sizing 4s only preliminary; it does not constitute a final design. ‘A block foundation design can only be considered com- plete when a dynamic analysis and check is performed and the foundation is predicted to behave in an accept- able manner as illustrated in Chapter 6. However, the following guidelines for initial trial sizing have been found to result in acceptable configurations: 1. The bottom of the block foundation should be above the water table when possible. It should not be resting on previously backfilled soil nor on a specially sensitive (to vibration) soil. The recom- mendations of the geotechnical consultant are ‘usually followed with respect to depth of structures supporting dynamic or vibratory machines. Some- times, the soil quality is poor, and the geotechnical consultant may recommend using piles or piers. 2. The following items apply to block-type founda- tions resting on soil: a. A tigid block-type foundation resting on soil should have a mass of two to three times the mass of the supported machine for centrifugal machines. However, when the machine is recip- rocating, the mass of the foundation should be three to five times the mass of the machine. b. The top of the block is usually kept 1 ft above the finished floor or pavement elevation to pre- vent damage from surface water runoff. . The vertical thickness of the block should not be less than 2 {t, or as dictated by the length of anchor bolts uted. The vertical thickness may also be governed by the other dimensions of the block in order that the foundation be considered. rigid. The thickness is seldom less than one fifth the least dimension or one tenth the largest dimension. d. The foundation should be wide to increase damping in the rocking mode, The width should be at least 1 to 1.5 times the vertical distance from the base to the machine centerline. . Once the thickness and width have been selected, the length is determined according to (a) above, provided that sufficient plan area is available t0 support the machine phis 1ft clearance from the edge of the machine base to the edge of the block for maintenance purposes. {. The length and width of the foundation are adjusted so that the center of gravity of the ‘machine plus equipment coincides with the cen ter of gravity of the foundation. The combined center of gravity should coincide with the center of resistance of the soi. ‘g For large reciprocating machines, it may be desirable to increase the embedded depth in soil such that 50 to 80% of the depth is soil- embedded. This will increase the lateral restraint and the damping ratios for all modes of vibra- tion. hh, Should the dynamic analysis predict resonance with the acting frequency, the mass of the foun- dation is increased or decreased so that, generally, the modified structure is overtuned or under- tuned for reciprocating and centrifugal machines, respectively ‘The following guidelines only apply to block foun- dations supported on piles: a. The pile cap mass should be 15 to 2.5 times and 2.5 to 4 times the mass of the machine for centrifugal and reciprocating machines, respec tively. . The thickness, width, and length of the block is selected as in 2(b) through 2(f). . The number and size of piles are selected such that no single element carries over one half ofits allowable design load, . The piles are arranged so that the centroid of50 Design of Structures and Foundations for Vibrating Machines the pile group coincides with the center of gravity of the combined structure and machine Toads. fe. Piles are battered away from the pile cap % ‘carry any transverse and longitudinal unbalanced forces. Vertical piles provide small resistance ¢ horizontal loads, and the batter piles are uswally designed to carry all such horizontal forces as axial loads. {, When piers are used, bells may be desirable t0 increase their overall capacity If resonance conditions ate predicted to occur, modifications are necessary as described in 2(h) above. hy, Piles and piers must be properly anchored to the slab for adequate rigidity and for meeting the design conditions assumed during the analysis phase. T Sizing of Elevated Foundations (Table Tops) Preliminary sizing and geometrical member arrange ‘ment constitute the initial design phase for elevated foundations. Although this preliminary phase is often based on the experience of the designer, suggested guide- fines can be useful in arriving at a satisfactory final design. It should be emphasized that the general guide- fines for trial sizing are only useful in the initial phase land are no substitute fora thorough dynamic analysis ghd check as described in Chapter 7. ‘These general guidelines include the following: 1. The designer should carefully analyze equipment size and clearance requirements to assure that fuficient space is allocated to equipment, anchor bolts, piping, and clearance for installation, maintenance and operation, that is, physical space Timits and requirements should be clearly identified and considered. 2, The bottom of the foundation mat should be placed no higher than the minimom founding Septh recommended by the soil consultant. Tl generally includes considering the location of ade uate bearing strata, water table, depth of frost penetration, paving elevation, and special local foil conditions. However, in very poor soils, the geotechnical consultant may recommend the use ff piles. The mat thickness # should not be Tess than += 007 L*" where Lis the average of two adjacent spans between columns. 3 10. iL ‘All columns should be stressed almost equally when subjected to vertical load. Thus, the column ‘areas should be proportional to the load caried by the column, and P;/Ai should be fairly constant for all columns where P; and A, are the axial load ‘and cross-sectional area of any column, The columns should be capable of carrying six times the vertical load, Column spacing should prefer- ably be less than 12 ft. The intermediate columns should be located preferably under the couplings or the gear box. The beam depth should be a minimum of one fifth of the clear span, and the beam width is normally equal to the width of the column con- Sistent with anchor bolt requirements for spacing, embedded depth, and edge distance. The beams Should not defect over 0.02 in. when subjected to static loads, | The flexural stiffness of the beams should be at least twice the flexural stiffness of the columns. The total mass of the structure including the mat should be no less than three times the mass of the supported machine for centrifugal machines and five times the mass of the machine for reciprocat- ing-type machines ‘The mass of the top half of the structure should not be less than the mass of the supported ma- chine. |. The maximum static-bearing pressure for soil: supported foundations should not exceed one half of the allowable soil pressure. For pile-supported. foundations, the heaviest loaded pile should not carry over one half of its allowable load. ‘The center of resistance of the soil should be within 1 ft of all superimposed loads for soil- supported foundations. For pile-supported founda- tions, the centroid of the piles should be within 1 ft of the superimposed loads ‘The center of column resistance should coincide with the center of gravity of the equipment plus the top half of the structure loads in the Tongi- ‘tudinal as well as the transverse directions, that is, the column moments of inertia should be “bal- fanced” about the centroid of the equipment as shown in Figure 3-2 ‘All the columns should deflect equally in the ver- tical, lateral, and longitudinal directions when subjected to equivalent static machine loads acting jn those directions. These equivalent loads are folten assumed to be 0.5, 0.3, and 0.1 of the total Toad for the vertical, transverse, and longitudinal directions, respectively, with the vertical dead load facting in all conditions. Chapter 7 gives a further Development of Information, Trial Sizing, and Design Checklist. 51 + \ xB 28 CENTROIO OF EQUIPMENT LOADS PLU UPPER HALF OF STRUCTURE . © COLUMN RESISTANCE SI Ete Mw Figure 3-2. Center of column resistance. description of these approximate equivalent static Toadings. The maximum horizontal deflections for these equivalent static loadings should be less than 0.02 in. in all cases. 12, The columns and beams should be checked for individual member resonance with the machine- acting frequency. The lowest natural frequency of the columns approximately given by 44800 (Fe) VP tn eee Senate ae ea the 10500 pa gs Eb te clam Hehe inches, and f, is in rpm. 18, Whe’ ies ecommended by the c- eee eras) Smug 3h) gee’ ahowe for eck foundations re a ee ee Checklist for Design Once the proposed structure is mode structure is modeled (10 Chap te), til ses ne seleced ad a analy i efor (et Chapen 6a 7). The pied eae ote epee bed or eonparel ap ra len vomemen The deign eyes ile leet eds eat stale nd ese omnes (com stoning behing nna remit eet, vel an nan 100 00 ———— the overlying soil, is measured to be 125 pef. Piezometer tests in the borehole indicated a piezometric surface (water table) depth of 5 ft. A consolidation test is con- ducted on the sample, which yields an indicated precon- solidation pressure of 4000 psf and an initial void ratio cf 0.60. The plasticity index of the clay is measured to be 30. 'A set of consolidated-undrained triaxial compression tests is then conducted on the soil in which confining pressures are in the order of a, and Zand g are found to be 3 psi and 20 degrees, respectively. What is G at a shear strain amplitude of 5 x 10-*96? y= 15(125) — 10(62.4) = 1251 psf 4000 pat __ 1251 pet ~ 2? 08 (Figure +5) 0.383 a (1+ 2K) = 8. psi = 0.24 (interpolated {rom Table 4-7) 8.69 psi Goax = 1230 {(2.973—0.60)"/1.60] (3.2)°* (8.1)°* = 16,288 psi tans = {(1.9/2) 8.7 sin 20° + 3 c0s 20°)" = (O.1/2) 8.7)'F* = 5.63 psi G is desired ata shear strain level of 5 x 10-*96: “te ~ (Foax/Gase) X 100 = (5.68/16,288) x 100 = 0.0346% 16,288/(1 + 0.005/0.0846) = 16,288/1.14 14,288 psi c- Note that G is about 88% of Gaus, which is a typical reduction for shear strains in the usual range for machine foundations ‘Seed and Idriss (ref. 14) have shown that the shear moduli calculated from the Hardin-Dmevich equations underestimated in the in situ shear moduli by factors of 4 and 14 for two soft normally consolidated marine clays with void ratios exceeding 1.6, This effect has been observed by others as well, leading to the conclusion that the Hardin-Dmevich equations should not be expected to correlate well with in situ moduli for marine soils with high void ratios and for soils of otherwise relatively high sensitivity. For such soils it is advisable that the shear modulus be determined directly from in situ tests. Alternatively, the results of laboratory resonant column tests can be employed for sensitive clays, provided the effect of sampling disturbance can be estimated in a rational way. Anderson and Woods (ref. 1) have shown that all soils experience a slow, time-dependent increase in shear modulus as measured in a resonant column device after an initial disturbance. This increase appar- cently occurs as a result of reestablishment of a stable soil fabric following sampling or other disturbing processes. ‘The indicated shear modulus generally plots in a straight Tine as a function of the logarithm of time after initial confinement in a cell. For cohesive soils, Anderson and Woods have found that (QG/Gren) (Jo) = Zexp (170.2554 +0.37), (4-10) where AG(%) = the increase in shear modulus per log cycle of time in min, Grooo = shear modulus obtained after sample has con- solidated in a cell for 1000 min (approximately the value that would be predicted using the Hardin-Dmevich equations), su = undrained shear strength of the soil in kg/em* oF tons/fi = initial void ratio of the soil. Factors such as magnitude of confining pressure do not play a major role in the rate of time-dependent increase in the value of shear modulus, or “secondary gain.” Examination of Equation (4-10) indicates that the rate of secondary gain increases with increasing void ratio and decreasing shear strength, which is consistent with the observation of Seed and Idriss that the Hardin. Drevich equations, which do not account for sampling disturbance, underpredict the shear modulus in soft clays with high void ratios. Equation (4-10) can be used to convert in an approxi- ‘ate fashion the shear modulus measured by short time- ‘of-consolidation laboratory resonant column tests, and possibly that predicted by the Hardin-Dmevich equa- tion, into an equivalent value representative of the un- disturbed soil. The value of Gass can be obtained from Equation (4-10) as (uy Geotechnical Considerations 67 where N isthe number of log cycles of time in min (beginning at 1000 min) required forthe sil fabric to reediablidh itself as it existed in situ prior to sampling, ting 1000 min as the point of reference. The correction fan also be applied deetly to G with litle error. For soils ofthe type being considered, the appropriate time required for complete fabric rettablishment depends Inainly upon the mineralogical properties of the si Very litle data exist relative to this point. However, if all the fabric restablishment is aauined to be due t0 thixotropy, as suggested by Andenon and Woods, experi rental relationships developed by Skempton and Northey (Gel. 18) can be wed to obtain order of magnitude values for the time in question. For soils whote sen tivity isles than about 8 (moderate sensitivity), Skemp- ton and Northey indicated that disturbed soils appear to regain thet in ity strength (and by implication their invita structure) ina period approximately equal to the geological age of the deposit. Using the above observa- tion as a citerion, Ny can be defined by the folowing equation: Ne=2.72 + logiedy, (4:12) where Ay = age of deposit in years. Equation (4-12) does not apply to heavily overcon- solidated clays, whose structure is not significantly changed by the sampling process, nor does it apply to soils whote sensitivity exceeds about 8. For the latter class of soils, no reliable methods exist to predict Ne: therefore, only in situ tests should be used to obtain the shear modulus. Furthermore, such soils should be avoided altogether as founding strata for structures supporting vibrating equipment whenever possible. Example. A deposit of slightly overconsolidated clay ‘on an industrial site adjacent to an estuary is known to be 10,000 years old. Soil recovered from the deposit from a depth of 30 ft has a shear strength of 500 pst (0.5 kg/em*) as determined by unconfined compression testing. ‘The soil has an in sit void ratio of 1.6 and a sensitivity of 6. Resonant column tests yield a value of G of 2000 psi, What is the corrected in situ shear modulus? “To correct the computed shear modulus, use Equa tions (4-10), (4-12), and (4-11): (AG/ Grou) (9) = Bexp [1.7 — 0.25 (0.25) +037(1.6)} 8.6% 72 +ogio( 10,000) 2t v 68 Design of Structures and Foundations for Vibrating Machines G= [6.72(18.6/100 + 1) 2000 .25(2000) = 4500 psi Seed and Tdriss suggest that shear modulus be cor- related directly to undrained strength in an attempt 0 circumvent the sampling disturbance problem, primarily because strength is affected much less than modulus by sampling, The shear modulus G was correlated to un- drained shear strength 5. for a number of clays by Seed. and Idriss, For low strain levels 5, as obtained by standard static tests, was correlated with G obtained from in situ seismic tests. For high-strain levels 5_ was correlated with G obtained from a variety of laboratory tests, whose results were corrected by multiplying the value of G obtained in the test by a factor of 2.5 to account for the effects of sampling disturbance. The results of the correlations, which are shown in Figure 48, yield a simple procedure for estimating G from standard static laboratory tests. Calculation of Shear Modulus for Structure-Soil Interaction Analysis Since the shear modulus is a function of effective con- fining pressure (%,) for both sands and clays, the value of the shear modulus is influenced by both the geostatic, stress and the net bearing stress produced by the struc- ture. Therefore, the static vertical stress produced in the soil by the structure should be added to the geostatic vertical effective stress when calculating @ and % and subsequently G from Equations (42), (43), (45), (4-6), and (4-9) or when computing confining pressures for resonant column tests. ‘The principal exception to this statement is that only geostatic effective stresses need bbe considered for overconsolidated clays which remain overconsolidated after being loaded by the structure. Furthermore, static machine and foundation weights can often be neglected in practice when evaluating Equation (4-9), since their effects on shear modulus reduction are small For reasonably uniform soils, it is suficient to evalu- fate the shear modulus for purposes of calculating soil spring constants at only one characteristic depth d, be- low the ground surface, as shown in Figure 4-9. For freedraining soils, itis usually adequate to set # in the various equations for shear modulus equal to Fon, also defined in Figure 4-9. Since Ke is needed in the calcula- tion of 3, OGR, which is needed to obtain Ky should be computed by including the effect of the vertical stresses from the structure, that is, by setting @ equal 10 Gomis (Figure 4-9) in the definition of OCR. If the soil is an overconsolidated clay which remains overconsolidated after loading, the imposed stresses have little effect on G. Hence, values of G measured in situ ‘or computed only from geostatic stresses need be used, Uniform deposits of such soils generally possess a reason- ably uniform shear modulus with depth so that it is usually not necessary to evaluate the modulus at a criti- cal depth ‘When the founding soil consists of a normally consoli- dated clay (which is a rare circumstance) or a clay which will become normally consolidated after the load hhas been applied and excess pore pressures dissipate, the shear modulus will increase with time, and two shear 1 oe Sheor Strain in Parent Figure 4-8. In situ shear modulus for saturated clay (after ref. 14), \ Eerecrwe stress neon mete Ree | ‘3. SUMMATION OF 1 AND 2 (@]—————| 9% oc es EL cuney sare Paessune) rten lan, REF 9) IweoseD STRESSES EwexTH rose OF FOOTING Figure 4-9. Determination of Fm: ‘modulus values and consequently two spring constant ‘values should be obtained and used in the analysis: fist, for the unconsolidated-undrained condition which exists immediately after the static load is applied, and, second, for the fully consolidated condition which occurs at some time in the future. The shear modulus for the former condition should be evaluated using only geo- static stress conditions (in which case de corresponds to the depth of the base of the footing) and soil properties that exist just prior to placement of the structure. The shear modulus for the latter condition should be evalu- ated by the method described above or by assessing the new undrained shear strength at depth d- beneath the edge of the footing at the end of consolidation and using Figure 4.8, Example, Determine kz and ky for the footing shown in Figure 4-10. Assuming that the unit weight of the concrete and unit weight of displaced soil are equal, the net static bearing stress at the base of the footing is 25,000/x4# = 497 pst Using the Boussinesq chart in Figure 4-9, the plot of induced vertical stress at the edge of the footing is made (Line A). The geostatic vertical effective stress is plotted as Line B. Finally, the stresses from Lines A and B are ‘added to produce Line C. It is observed that de is at the base of the footing and that Grmin= 490 psf = 340 psi Since the sand is normally consolidated, Ky=04 (Figure 4-5), and = 0.333() + 2a) 0.333 [3.40 + (2) (0.4) (3.40)] 2.04 psi Using the Seed-Idriss correlation, and assuming a shear strain level of 3X 10°%, C= 833K,(%)** -3(52) (2.04) 187 psi Assuming that embedment is effective (footing is cast against undisturbed soil or dense backfill is provided), ke= [4Grof (1 — v) [1 + 0.6(1 — v) (h/t) 4(6187) 4(12) /0.6} [1 + 0.6(0.6) (2/4) ] 34 X 10" Ib/is and ky = [BGr.4/3(1 —v)] [1+ 1.2(1 — »)(B/re) + 0.2(2—») (h/r0}*] [8 (6187)48"/3(0.6) ][1 + 1.2(0.6) (2/4) + 0.2(1.6) (2/4)"] 26 X 10° inelb/rad. ‘Typical ranges of values of low-strain-amplitude shear modulus for several soils are given in Table 4-8 for the purpose of general information. Table 4-8 should never Geotechnical Considerations 69 ask static) ‘NiFoRM ‘PREEDRAINING NORMALLY ‘CONSOLIDATED DENSE SAND VERTICAL STRESS IN PSF Figure 4-10. Footing for example calculations of soil spring constants. Table 4-8 ‘Typical Values for Low-Strain-Amplitude Shear Modulus Soll Type ‘Shear Modulus (psi) Sole Clay 3,000- 6,000 ‘Stim Clay 10,900-20,000 Very Stif to Hard Clay 320,000 ‘Mediums Dense Sand® 5,000-15,000 ‘Dense Sand® 10,000- 20000 ‘Medium Dense Gravel” 16,000-25,000, Dense Gravel® 20,000-40,000 “Por shallow depths. bbe used as a substitute for a' rational determination of ‘modulus values. Selection of Shear Strain Magnitude for Computing ‘Approximate Shear Modulus Beneath Footings Since the shear modulus is a function of shear strain magnitude, it is necessary to obtain an estimate of the appropriate value of shear strain magnitude to use in calculations of soil spring constants. ‘The authors recom- mend the following approximate procedure for vertical loading, which is based on an analogy with static conditions: 1. Select a shear strain amplitude in the range of the crosshatched area of Figures 46 and 48, and compute G. 2.Conduct the structure-sol interaction analysis and determine the transmissibility factor 7, from Table 1-4 for the forcing frequency desired. 3. Multiply the unbalanced vertical force by the trans- rmissibility factor and divide the result by the con- tact area of the footing to obtain the dynamic bearing stress qa.70 Design of Structures and Foundations for Vibrating Machines 4.The approximate average shear strain y ina central block of soil below the footing of dimensions re X 2re (horizontal) X 47 (vertical) is given by Eq, (4-13) %) = 12 4/6 (13) Equation (4-13) presumes ge to induce the same strains in the soil as it would if it were acting as a static stress (ref. 10). Of course, the strains due to a dynamic bear- ing stress emanate as waves, making Equation (4-13) nontigorous. Nonetheless, Equation (413) will yield ‘order-of-magnitude strain levels that are sufficiently accurate for most analyses. Therefore Equation (4-13) ‘may be used to verify the assumed value of G (Step 1). If the assumed and computed shear strains differ sig nificantly, these four steps should be repeated iteratively, using the value of y computed on the preceding trial to obtain G for the present trial, until the strains close to within an acceptable difference. If a particularly precise analysis is warranted, the approach described above should be abandoned in favor ‘of a more comprehensive technique, such as the finite element method, in which complete modeling of a rele- vant volume of the soil and its constitutive relationships 's considered. It should be pointed out here that the shear strain magnitude beneath a footing should be taken as that produced only by the dynamic component of the footing Toad. The static shear strain should be neglected, since it in effect only provides a nonzero strain level about which the dynamic strain is cycled. The small strain shear modulus relative to that nonzero reference is generally about the same as the small strain modulus relative to a reference level of zero strain, Damping Ratio Damping in a soil-foundation system consists of @ geometric component, which is a measure of energy radiated away from the immediate region of the founda~ tion, and material damping within the soil, which is a measure of energy lost as a result of hysteresis effects Geometric damping ratios have been shown at the be- ginning of this chapter to be related to the mass or inertia ratio of the system through the use of clastic halfspace theory. Relationships between mass and inertia ratios and the geometric damping ratio are shown for the four uncoupled modes of motion in Figure 4-11. Material damping is defined in Figure 4-12, Tt is seen to be proportional to the ratio of A,, the area of the soil hysteresis loop in simple shear (energy lost), to Any the crosshatched area (energy input). Material damping ratios can be obtained as a part of resonant column ' os a 02| Siar oil = —— ol YS at rs .02| or Figure 4-11. Geometric damping ratios for four modes of loading. After Richart, Hall, and Woods, ref. 12. Reprinted by permission of Prentice-Hall, Inc., Englewood Cliffs, NJ. n= A. /4n by Figure 4-12. Definition of ma- terial damping (after ref. 7) testing. After the soil has been vibrating in a steady- state condition, the exciter js stopped and the soil vibra- tions are monitored as they decay. The displacement- time relationship is essentially sinusoidal, but with the ‘amplitudes decreasing with time. If two successive amplitudes are 2, and z., then Dy, = (hn (ea/zal) fax" + (ln fes/zel) 14) ‘Additional procedures are described by Richart, Hall and Woods (ref. 12) ‘Seed and Idriss (ref. 14) have shown that material damping in soils is primarily a function of strain ampli- tude and soil type. Figure 4-13 gives typical values for ‘material damping proposed by Seed and Tériss. In order to obtain the total soil-foundation system damping ratio Ds, the geometric and material damping ratios may be added directly. Since the material damp- ing ratio is significant relative to the geometric damping ratio in rotational modes, the total damping ratio, rather than the geometric damping ratio, should be used when analyzing the response of foundations in those LL 1" 10 0" Sheor Strain in percent Figure 4-13. Material damping rs clays (atter ref. 14). 9s for sands and modes. On the other hand, material damping is small compared to geometric damping in the translatory ‘modes and may often be disregarded. The exclusion of ‘material damping from an analysis will result in ampli- tudes that are too high, especially at or near resonance, and resonant frequencies that, for rotating mass excita- tion, are slightly t0 low. Example. The inertia ratio By for a machine-founda- tion system undergoing rocking oscillations and resting fon a saturated clay subgrade is determined to be 1.3. How does material damping influence the deformation at resonance and the resonant frequency? According to Figure 4-11, the geometric damping ratio of the system is 0.05. If the material damping ‘within the clay is neglected, the amplitude magnification factor for rotating mass excitation (Equation 1-43) is ap- proximately 10 at resonance. Further, far = (1/VT= 2D) fa = 1.0025 fa 1f material damping is considered, according to Ease tion 1-43, its value should be approximately 003 in the typical range for machine foundations. Hence, Di=D+ Dw = 005 +003 =008 Using the value thus obtained for total system damping ee ey er er ere ee reduced to approximately 6, Also, using the total system damping factor of 0.08, faur = 1.0065 fa Gonsideration of material damping thus reduces the ational ampli at mionanee by about 40%, bt has only a minor effect upon the resonant frequency. Other effects, disoused in Chapter 5, may also have ‘an influence on fm, and amplitude of motion in the rocking mode. Geotechnical Considerations 74 Whitman (ref. 18) has proposed that the effects of geometric and material damping be combined and that the total damping ratios be computed as follows: Vertical Translation Mode: Dz = 0.49 (M/ordy** (15) Horizontal Translation Mode: D, = 031 (M/ory"* (16) Rocking Mode: Dy = 0.05 +0.1 (Uy/er I UL + Cy /4ore* 4-17) where M= mass of foundation plus mass of structure or machine vibrating in phase with the foundation, effective foundation radius, ‘mas; moment of inertia of foundation plus that part of structure or machine vibrating jn phase with the foundation about a hori zontal axis through the base of the founda- tion perpendicular to the plane of rocking, ‘otal mass density of the soil (unit weight/ acceleration of gravity). ‘When the foundation beng analyz i rgd fotn within a multidegreeoftreedom wucture and to which ‘ibating machinery isnot diecly attached, it cone tervatve to ealeaate Dy frm the above equations a fuming that M is the mas of the foundation alone, Under the same cicumlancs, ‘when the geoterc doping ratios determined foes Fgare 411 ate ad inert alo agumens may be eomputed from the mass opertic ofthe foundation alone Whitman's propoted values for total damping ratio can be son to be somewhat lover than the gicetc ,Cr—— anomaly is due to the fact chat Whitme’s expresions Fepresent lower envelopes to availble text data and Ipls that geometsic tamping in 2 prototype may be Tes than that predicted by hallspace theory, posbly because of layering and boundary effects Selection of Poisson's Ratio and Soil Density Soil-foundation interaction problems are relatively in- sensitive to the values chosen for y and p. Generally, Poisson's ratio can be selected based on the predominant soil type using Table 49. It is also possible to obtain a value for Poisson's ratio by measuring independently the shear modulus (G) in a laboratory torsional resonantSes Sane 72 Design of Structures and Foundations for Vibrating Machines Table 4-9 ‘Typical Values for Poisson's Ratio Soll Type . Saye Saturated Cl 045-0580 Partially Saturated Clay 35-048 Dense Sind or Gravel? os Mein Bese Sandor Grave 03 a4 SS column test and Young’s modulus (E) in a laboratory longitudinal resonant column test. Assuming isotropy, (EPG) -1 (18) Soil mass density values should always be calculated from the total unit weight rather than the buoyant unit ‘weight because the density term in the mass and inertia ratio equations always represents soil undergoing vibra- tion. Total weights are used because both the solid and liquid phases vibrate. Effect of Footing Embedment Analytical expressions have been given in Tables 4-2 and 44 that are to be employed as multipliers to the equivalent spring constant and geometric damping ratio vvalues whenever the footing is embedded, A foundation should be considered as “embedded,” however, only if it is east against undisturbed soil or if it is formed and backfilled carefully with a high compactive effort using soils with low shrink-swell potential. Casual backfilling is ineffective. Stokoe and Richart (ref. 15) studied the response of model circular footings embedded in a dense, dry, poorly ‘graded sand to vertical, horizontal, and rocking excita- tion. ‘Two sets of tests were conducted: one in which the embedded footings were cast against the soil and fone in which a small air gap existed between the soil and the sides of the footing. The results, which were relatively consistent for all modes of loading, are sum- ‘marized in Figures 4-14 and 4-15. Tt is evident that proper embedment had a significant effect on both total damping and resonant frequency, while embedment without adequate lateral support was ‘essentially ineffective. ‘The benefits of proper footing embedment are espe- cially pronounced for the rocking mode, since D; can be increased by several timiés by embedding the footing to.a depth equal to or greater than its equivalent radius. Tis also desirable to embed pile caps in competent soil, 5 T T 5 o Dy/Dyrutece (ALL MODES) ~ Radius Figure 4-14. Effect of embedment on damping (after ref. 15), q 3 8 3 ea 2 w im Z 7S RPO s ayee 7 £ 1 Bys2 =a B15 LE ° Opon 210) 2520) Embedment Radius Figure 4-15. Effect of embedment on resonant fre- quency (after ref. 15). since the response of pile foundations is sharply peaked fat resonance in all modes. Cap embedment in cases ‘where significant lateral loading occuss in the piles is specially important, since resonant displacements can often be excessive without the damping and load trans- fer provided by the cap. This problem will be addressed in greater detail in Chapter 5. Effect of Stiff Underlying Stratum If the subgrade consists of softer soil overlying an appreciably stiffer soil or bedrock within two to three footing diameters of the base of the footing, the response ‘will be altered significantly. A stiff stratum interface will reflect a considerable proportion of the energy that ‘would ordinarily be radiated away from the foundation, in effect producing a lower geometric damping ratio than would occur in a homogeneous subgrade. This effect is offset to a degree by the fact that the effective Jumped spring constant is increased due to the presence of the stiff stratum. Quantitative research in the vertical response of foun- dations on layered media was described by Warburton (ref. 17). Little usable information is available for the rocking and sliding modes at present, making the use of special modeling techniques such as the finite-element ‘method attractive for important problems in rocking and sliding. ‘Torsional response in a layered system is dis- cussed by Richart, Hall and Woods (ref. 12) Table 4-10 presemts approximate relationships for cal- culating &, for various values of H/rq, where H is the distance from the base of the footing to the stiff soil contact. The stiff soil is assumed to be fully rigid. Establishment of a lumped damping ratio in this case is difficult; however, when 5*# = 14sit - a= AGra/(1 — va) = 48000) (14:8) (12) 06 15 108 Ib/in, ee = [(10 X 10) /(w X 148")] he = 1.98 x 10*Ibfin. Using Equation (5-4) t= (100 100/x)°* = 56.4 1 a= (4 X 8000 X 564 X 12) /(1~ 04) = 36 x 10°Ib/in. hee = (36 10°) X (10 X 10) /(100 X 100) = 0.36 x 10*Ib/in. Deep Foundations pie or per foundations (which hereafter wil be eae ily "ple foundations) are often wed sere Ubestry ads when sil conditions at a ste eepete tha hallow foundations wil result unasept wate pemmanent setement, Present understanding of ee denorof ple foundations under wbratory lading sonny poor, bat itis koown eat the wie of pls Sears fee damping increas the resonant tcpeeney Ete foundation, snd infuences deformation feaeeteance. Since in some css, particularly in the reel tend nade, the effect of pes can be adver vee dhol not Be wed without some understanding of their behavior Th tis chapter approximate procedures developed by ata tad fis asocires for analyzing the response Nh tats of angle piles and pile groupe in the humwgled vera, horizontal, and pure rocking modes aren The sluion ate based the assmption Fao, flly embedded vertical piles interacting with hoc sane sil Pathermoe, the pile tip is asemed wpe ined ago motion, except in the case of vertical ‘eon wee the tip ca be fae or “elena” "Fae Instaton of applying Nova’ procedures to zeal geological materials te obvious; however, they have ca remnable predictions forthe response of soall ree ets in vlaively simples profes, and further eae ercay represent the current (1079) stateolsheart inthe prdcGeal analytical weatment of piles i wae, the slant for spring and damping constants Sloped by Novak ae frequency dependent; howeven serreate frequency independent expressions, have serra developed fr both the spring constant and pe pane vatie These wll be deserted herein. Solutions ‘ee he Sften and damping of torsional Toaded in Cowal giles and for coupled rocking and sliding, ob TEES Nowak and his soca, are alto avalable (refs. 10 and 14), but are not included here. Novas various procedures do not permit calculation of the srees induced in the pile materi, although such testes canbe important, particulary during lateral loading. Ghazsaly, Hong and O'Neil (ref.5) describe 4 numeral algoithm for making ste computations ina pile undergoing harmonic lateral lading. A de. tiled description of that algorithm is beyond the spe ob thi tot Both the spring constants and damping ratios should be obtained experimentally from fullsize groupe of tet piles whenever posible expecially if batter ples are con- tained in the foundation or if short fiton ples are to be employed: (Novalis procedures for ple group alysis do not consider batter pls directly in expres sone for overall group siffes and damping; however, Sauls procedure, referenced atthe end of this chapter, among other, des allow consideration of batter pls in tn appronimate fashion.) Furthermore, according to Novak and Grigg (ref. 18), the “apparent” shear meda- Ins to be used for analyzing laterally loaded pls should be that valve backealeulated frm elatc beam on foune ese ee gery ee renee Se Fe The inal slope of the feehead load vers displace iment curve can be od convenient in the ealeulations. Th res fring th apache het te very near the siface contol the lod deformation prop- eee re ea er rene ee ey re erent are rs which is generally appropriate only for a depth greater ere ee eee re tee A, since a gap often for being a ltrally loaded pil the i of hear mr obtained mb ca provides a convenient cmpirical correction ovals elas solution, which snot rigorous for sch 9 case. In a series of large-scale model tests in sand, Novak ad Grigg (rf 18) determined thatthe apparent seat ‘modules fom statis Intra pleteais as about one ee ee a about $0 pile diametety obtained for the str fom tector test Simple methods of interpreting field Jad test data are dissed briefly a the end ofthis chapter, Approximate procedures for estimating damping. rats and. spring Constants based on results of several vibration tes ‘eported in the Iteratare for vertically exited ction pits are alo given ‘Two important rules shoul be followed relative to the xf and contruction of dep foundations, Fis, the factor of safety relative to the ultimate sate axial load on a pile should exceed $ in order to restrict soil Sse to magnitude that will precude the “resonant Aeivng” effet, wherein the soil around the pile resonates with the ple and loes is ability to caxy Foundations 81 sn appeal ond: Ith static od is oo lage (ee tote tun oe hal the tate npc ofthe il) pang ing la shedding cn eal thes eames ith aie Second the pile cap shoud be buried incompetent sil (pretebly dense gramlar sil) whenever posible in onder totale advantage af the damping afforded by the ll cap. ve al Motion ‘The expression for the effective spring constant for a single end-bearing (“‘fixed-tip") pile undergoing vertical motion, k,, given by Novak (ref. 10) is we (EpA fro)fss,3 (55) where Ey is the Young’s modulus of the pile material, A is the cross-sectional area of the pile, r is the equivalent radius of the pile, and fos. is a factor given by Fig- ture 5-la (concrete piles) or Figure 5-1b (timber piles) ‘as a function of ratios of pile penetration 1 to radius 7— and v,/2e (shear wave velocity in soil above tip/com- pression wave velocity in pile). Note that the factors given in Figures 5-la and 5-Ib are for fixed tip piles (end bearing or combined friction and end bearing piles). For friction piles these factors will be in error by ‘a relatively small amount for values I/rs greater than 60 and for values for v4/ve greater than about 0.03 for timber and concrete piles. For friction piles having lower values of I/r and/or vs/ve, the procedure for “relaxed tip” piles (ref. 11) or an enipitical approach, both de- scribed later, should be employed. For steel piles Novak and Grigg (ref. 13) have given a value of fr, = 0.030 where 74/2 = 0.033 (medium stiff soil) and 1/19 > 80. ‘The effective geometric damping constant for vertical motion in a single pile c. is given by Equation (5-6): 2 = (EyA/vs) frs,e (5-6) where vy it the shear wave velocity of the soil through which the pile is driven (VGcg/y.) and fre. is a factor given by Figures 5-la and 5-1b. The damping factors ‘aa are also approximately valid for friction piles in the ranges described previously relative to the stiffness fac- tors fis, although the ezror in the damping factor is somewhat greater. For a steel pile in medium stiff soil (v/v = 0.033) and for I/re > 80, fs,2 ~ 0.085. For relatively short friction piles (that is, piles with a “relaxed tip”), Novak (ref. 10) suggests computing ket and c's follows: Ket = (BpAlro) Prva (55a)Design of Structures and Foundations for Vibrating Machines oro [a STHFPNESS 7, FOANPING a T tT Te oo bene met ae | eet 008 UA 008 So + sor cree ear seed - 2 oe g 008 sa € S 1 904 £ oo | & 3 ro 2 003 = 008 4 8 004 ss = 004 | ———] ee os 02 202. oz Z | i ool ar ao ol 4 7 ° 80, 100 a a a ee (008 -— i SONCRETE or "TiMaeR PILES 3 5 ce € o 2 40 60 20 00 _ ° semicemness Me a PILE SUENOERNESS dir, Figure 5-1. Stitiness and damping factors for vertically excited piles (after refs. 10 and 11). a. Fixed-tip con- crete piles. b. Fixed-tip timber piles, c. Relaxed-tip S-1a and 5-1b reproduced by permission of the Nati Geotechnical Journal, Vol. 11 (1974), p. 586. and c:!= (y/o) an (5-6a) where fis: and f's,2 are stiffness and damping factors, respectively, given in Figures 5-le and 5-1d for concrete fin Gnber piles, The values in Figures Sa, b and @ are most appropriate for use for ay ( = 2x rol VGag/¥=) between 0.1 and 0.8 where f is the machine frequency. Stevens (ref. 20) suggests that they are valid even for {a as low as 0.05, which means that reasonable results can be anticipated for slender piles and low forcing frequencies. ‘The geometric damping ratio for a single pile sup- porting a structure can be computed from the damping constant by using Equation (5-7) : conorete piles, d. Relaxed-tip timber piles. Figures ional Research Council of Canada from the Canadian De = oh/2Vie me (5-7) where’me is the mass ofthe cap plus machinery of por tion of the structure vibrating in phase with the cap. It ay also be appropriate to include part of the tnass in me, but for fully embedded piles, the ratio of that mass to the mass of the supported weight is usually small enough to be neglected Pile Groups Most piles are installed in groups or clusters. The group stiffness will not in general be the simple sum of the stiffnesses of the individual piles. A similar state- ment can be ‘made relative to damping. If several distinct groups are used to support a vibrating super- structure, itis reasonable to obtain an equivalent spring constant and damping ratio for each cluster. These equivalent constants may then be applied to the struc- ture at the appropriate supports for purposes of analysis of structural response. An example of this approach is a frame whose columns are supported by separate pile ‘groups. Often, however, a block or rigid mat foundation will be supported by a single large group of piles. In such case itis often convenient to represent the structure foundation system by Model ©, described in Figure 29, Chapter 2, where the entire group is modeled by a single spring constant and a single damping ratio in each mode of vibration, Whenever piles are used to support a flexible mat, when large corsional moments are applied to the super structure, or when batter piles are present, itis usually necessary to employ a multidegree-ofreedom computer model (Fig. 2-11) and to represent each individual pile in the system by assigning linear support spring con- stants and dashpots to the point in the structure where the pile head is located. ‘These spring constants repre- sent uncoupled vertical stiffness and horizontal stiffness, in two perpendicular directions. When batter piles are present, axial stiffness and damping parameters may be computed from the “vertical” motion equations, and lateral parameters may be computed from the “horizon- tal” motion equations. Input values for stiffness and damping can then be obtained by taking components in the coordinate directions. The horizontal spring con- stants (described later) are strongly dependent on the manner in which the piles are fixed to the structare. For block foundations and table top structures con- taining a thick mat (which also serves as the pile cap), the pile heads are normally assumed to be fixed rigidly to the structure. While frame analysis programs allow the inclusion of uncoupled rotational stiffness terms at the pile head, it usually is necessary to include rotational springs for rocking about transverse pile axes only when large rocking moments are present and when batter piles are not present in the foundation. Tt is rarely necessary to include a torsional stiffness constant for an individual pile, since torsional moments applied to the structure are resisted almost completely by couples pro- duced by lateral reactions at the pile heads. While the piles are represented individually in this type of analysis, it is stl desirable to include the softening effect of svoup action in the model. This can be done by cal Calating the stiffness for the pile group as a whole in the appropriate mode, distributing the stiffness equally among the piles (presuming that all ples have the same Foundations 83 size and penetration), and using the distributed stiff- ness values in the analysis. In other words, it i not appropriate to use the values given by Eqs. 5-5 or 5-54 or the corresponding equations presented in the follow- ing section on single pile horizontal response. ‘The damping ratio to be assigned each pile for each mode of translatory motion in a multi-degree-of-free- dom-foundation representation should include a consid- eration of group action, which will result in a higher damping ratio for each pile than would be obtained using the single pile damping ratio. Typically, each pile in the system would be assigned a damping ratio equal to the group damping ratio in a given mode divided by the square root of the number of piles in the group. In the remainder of this section simple equations are given for the determination of group stiffness ky? and damping DJ for a group of vertical piles oscillating in the vertical mode. Similar expressions will be pre- sented in the next section for horizontal translatory stiffness and damping for vertical pile groups. Note that, in evaluating group effects for the latter mode it is necessary to estimate the direction of primary horizontal ‘motion. Since no general solutions are available for the assessment of group effects in groups containing batter piles, some judgement is necessary to apply these ex- pressions to batter pile foundations. Novak and Grigg (ref. 13) have argued that the deflection factors proposed by Poulos (ref. 15) for groups of statically loaded piles may also be applied to a pile group undergoing steady-state vibration, Hence, ‘Novak and Grigg propose that De Dros where N= number of piles in group a= axial diplacement interaction factor for a typical reference plein the group relative to At and to all other piles in the group, as- fing the elernee ple ad al ober ples carry the same load. (6-8) ‘The factor a can be evaluated using Figure 5-2. The equivalent geometric damping ratio for the group is given by Bra ay Diem Bie 6-9)Design of Structures and Foundations for Vibrating Machines. 10 7p 08 : — g tt oa = Zoe ESS S/2tq 26/8 Figure 5-2. a4 as a function of pile length and spacing (after ref. 15). where me is defined as in Equation (5-7), that is, the total mass of the cap plus machinery. If the pile cap is not in contact with the ground, Equations (5-8) and (5-9) can be used directly in the structural analysis. Embedment of the pile cap, how- ever, has a favorable effect on the response of the group and should be employed whenever possible. It is good practice to assume that embedment is effective only in the development of side friction between the cap and soil and only when dense granular backfill is used, since soil beneath the base of the cap is likely to be of poor quality and can sete away from the cap. Likewise, cohesive backfill can shrink away from the sides and be- come ineffective. Novak and Beredugo (ref. 12) have given expressions for calculating stiffness and geometric damping con- stants for the embedded cap, which are added to the fness and damping values obtained in Equations (5-8) and (5-9) to arrive at total system stiffness and damping for a group of piles. Those expressions are given here as Equations (5-10) (stiffness, ky’) and (5-11) (damping, c) with only the side’ resistance component considered. Expressions for caps or footings in complete contact at the base can be found in Novak and Beredugo (ref. 12) bf = GAS, (6-10) 1) In Equations (5-10) and (5-11), f is the depth of embedment of the cap, ry is the equivalent radius of the cap, Ge and ys are the shear modulus and total unit weight of the backfil, and S, and 5, are constants given in Table 5-1, in which vy is the Poisson's ratio of the backfill sil Table 5-1 Frequency Independent Constants for Embedded Pile Cap with Side Resistance {after refs. 3 and 12)* » SSS. Ba SS 00 2027 «GT 8882 O25 «270 G7) fo oa 37) e741 os SB ‘Values are appropriate for 20 = 2efra/ VGak/%y in the range 0 043. Reproduced by permission of the National Reeearch Council of Canada from the Conadion Geoleciical Journal, Vol. 9 (1972), p. 496. ‘An interesting comparison from a theoretical study of the vertical response of a machine and its foundation ‘when the foundation is embedded and when itis placed fon the surface without embedment was developed by Novak. The foundation consisted of a solid rectangular block of concrete (16 ft long, 10 ft wide, 8 ft high) embedded 2 ft and not embedded. It was supported on eight 35-f¢ long fixed-tip timber piles in a medium stiff clay. The machine supported on the block weighed 10 tons and the operating speed was varied. The re- sponse of the pile foundation system is shown in Figure 5.3. In that figure, dz is the static deflection that is produced by the unbalanced force at a given frequency, Mis the mass of the footing (cap) plus machine, and ‘me and ¢ are the unbalanced mass and its eccentricity within the machine. The effects of the relatively small amount of embedment are evident. ‘Also shown in Figure 5-3 are the response curves for the machine and rectangular block footing without piles ‘The shapes of the various curves show vividly important aspects of pile-supported structures: (1) damping is very low compared to soil-supported footings and (2) the use of piling increases the resonant frequency and, in this case, increases displacement amplitude at resonance. Damping can be increased by embedding the pile cap. Material damping was not considered in this particular analysis Example. Determine the equivalent spring constant and damping ratio for vertical motion for the pile group depicted in Figure 5-4. Assume that Novak's solution for fixed-tip piles is valid. [5000 (144) (82.2)/110]"* 5510 in Exg/a4)”* = {1(8.5 X 104/150] (144) (32.2))* 10,400 ft/sec t4/e = 0.044; me = (70,000-+12% 12X8X150)/386 = 629 (Ib-sec* te = [(20 X 20)/x}"* = 11.3 in, rg = 90 (12)/11.3 = 95.6 fax = 0.083 fing = 0.053, (E,A/n) fusy = [3.5 X 10° (20)*/11.3] (0.033) 4.0 X 10 Ib/in. (EyA/0,) fins = [3.5 X 10° (20)*/5510] (0.053) 1.3 X10" Ib-see/in. i = Gk 5, = 6000 (6 X 12) (2.7) = 1.2 X 10° lb/in. racap) = (12 X 12/x)"* = 6.77 fe = 81.2 in. of = bre Gore/e)”* 3 = (6X12)(61.2){6000(120) /(1728)(386)]**(6.7) 4.1 X 10' Ib-sec/in. Adjust for group effects. Assuming any pile in the group is a reference pile with ry = 11.3 in., $/2re to an ad- jacent comer pile is 72/22.6 = 3.19 and to the di- agonally opposite pile is 1.414 (3.19) = 4.50. Using Figure 5-2: 90(12)/22.6 = 48 0.5 (nearest to actual value of 0.4) ag = 1 (reference pile) a4 = 0.58 (adjacent corner piles) (2 piles) ‘44 = 0.5 (opposite corner pile) Tiae = 142058) +05 Deed / Dh ag = AOKI _ 6.0 x 10 yin (combined stiffness of the piles) 2.66 Foundations 85 ‘cRouP OF FRED - Tip PILES — ‘CAP NOT EMBEDDED (GROUP OF FIXED- TIP PILES—CAP EMBEDDED AMPLITUDE. MAGNIFICATION FACTOR, M, 025-4060 do 001840 BO FREQUENCY (RAD/SEC) Figure §-3. Effect of type of support on vertical response of a foundation (after ref. 10). Reproduced by permission of the National Research Council of Canada from the Canadian Geotechnical Journal, Vol. 11 (1974), p. 589. yet TfoloHt 4 o|O}- 1 [a venricat nsf tao iL? ee MATERIAL PROPERTIES LEAN CLAY: G, = 5000p8i, », #0.4 lOpef BACKFILL, 6000p8i, 74 = I20pet RIGID CAP: x, = I5Opet Pites. Ep = 3.5x10%psi 72 = |50pet (ASSUME 357 OF THE STATIC COLUMN LOAD VIBRATES WITH THE CAP IN VERTICAL MODE) Figure 5-4. Pile group for example problem.Ce 86 Design of Structures and Foundations for Vibrating Machines Adding in the stiffness of cap due to side resistance: is > Hi = 60 x 10° + 1.2 x 10° = 7.2 X 10° ib/in. ea Bie/(Ga) be 4(13 X 104) 2.66)" = (8.2441) X10 = 7.3 X 104Ib-sec/in, +41 x 10 a /2Hm J") (23 X 10')/2(7.2 X 10* x 620)" = 0.54 Note that over half of the damping is produced by the embedded cap. Material damping is not included. Horizontal Motion ‘The solutions for pure horizontal motion of vertical piles (ref. 10) follow the same logic as those for vertical motion. A notable exception is that a reduced value for G, should be used to represent the action of soil against pile as described earlier. For a single pile, kd = (Bylf/re) faa Ure > 28) (5-12) et = (Epl/n 0) fas Ure > 28) 6-13) where J is the moment of inertia of the pile cross- section about a centroidal axis perpendicular to the direction of translation. The subscript x denotes horizon= tal motion, and furs and fu,2 are factors for fixed-head piles given in Table 5-2. For a group of piles, Deke [Lian G14) renre/[alnre nar] where a, is a displacement factor for lateral motion de- fined in an analagous fashion to a4. Approximate values for az may be obtained from Figure 5-5. ca Finally, the stiffness and geometric damping char- acteristics of the cap are represented as follows Ke = G80 (6-16) = be VGrvle Sa 17) where Say and Sig are factors given in Table 5-1. Uncoupled Rocking Motion Novak (ref. 10) has presented expressions for the uncoupled stifness and geometric damping constants for single piles and for pile groups undergoing pure rocking. In summary, ky = (Elid) fas 6-18) eg! = Exllon) fis Where J is the moment of inertia of the pile cross Section about the axis of rotation, and fr,x and fax are factors given in Table 5-2 For a pile group, (6-20) DN hat + ba? hace! — Decks!) + by where sy and z, are defined in Figure 5-6, and k,* and ket are ‘stiffness constants for single piles defined in Equations (5-5) and (5-12), respectively. In addition, kay = Edllre) faa 21) where fs.: is obtained from Table 5-2; and kgf = GaaiSyr + Oh (8/3) + (ea/r)® = ie/t) 15a (5-22) where 8 = h/t» and Sy and Sy are defined in Table 5-1, Note that it is necessary to include hap a cross stiffness term, in the solution. Note also that interaction factors (a) are not included in the solution for group stiffness because group action in pure rocking is not as prevalent as in the translational modes and because ade- quate studies have not been conducted to established Appropriate interaction factors for the rocking mode. Finally, ef = DY led pede? poded = Reve" tev! vd aad 6-23) where ¢-! and cet are damping constants for individual piles given by Equations (5-6) and (5-13), respectively, and ery = (Bxlitov) faa 24) ‘The factor fy.» is evaluated in Table 5-2. Further, ars VG v7 (Sea + (08°13) + (eal = 8(ee/1)) Se} (6-28) where Sys and Fy are defined in Table 1 Table 5-2 Values of fu ifuai frst frat tat fa for I/r, > 25 (after ref. 10)% Foundations 87 Gonerete Piles (1a/y» = 0.7) 2 Sas ‘Lone im fra Soa rr ‘0.0086 ‘0.0088 ‘302 ‘0439 0.0280 ors, Oss 0.380 0288 0.0888 O.0597 O.pe82 0.850 als ‘1810 0.25 0.0082 0076 0.195 35 ~0.0382 o.o168 Sows, os? 0388 Sores O.0388 0.0880 085 0308 onset te ale fins Sng fon 04 oor 0.0082 DaT6 = 00486 O08 0.0825, 0.308 =0.1400 0.05 ooo 0.308 02330 025 001 o.oors 0.180 ~9.001 0.03 o.as85 bs 01801 05, O10 0370 02168 *Values are appropriate for ae = 0:3 (See Table 61), but are approximately valid (10%) for 0.1 5 OK. Actual soil pressure = 382,318/15.75(27.5) = 883 pst <0.5 Sarr = 1,250 psf ‘Thus trial area of footing is OK. All other guidelines for trial sizing are checked and found satisfactory. A dynamic analysis check is then performed on the trial foundation. The various steps in this procedure are listed in Table 6-2. (text continued on page 97)94 Design of Structures and Foundations for Vibrating Machines Design Examples: Block Foundations 95 RANKSHAFT_ REVOLUTION | Fay ‘TOMB. VERT. FORCE. F,2 x 2 ol sana coe wt 355 cobaat ° = a M colle, VERT MOMENT Sty fNaq—*000 — oto ar atest cdstfe J SPn ] VERT. FORCE, Fy (P)=1889 cos ot bac: VERT. MOMENT*, Tg (P}: z ky VERE MOMENTR.Tg® IY 41008 | VERT. FORCE, F, (8) =288 esau | | a Megs NOMEN "79 (8 ‘ | we (HORE. FORCE,,F, (P)=725 sm wt ~~ ™s ° y Hone MOMENT wocan 4 p< ae cneree man 1000 a 71} a a7] . | — {0 crrronmey (@) Vortial Excitation (©) Horizontal Translation oe nes Ty(t) Figure 6-3. Unbalanced vertical and horizontal forces and moments. y Axis. oF coMe:6.6. OF MACHINERY je om TORSIONAL OSCILLATION,@-——} apeaipa 4 f ie cls Wy? 57475 tbs ‘ ¥ $|Cy| Y 1784.94 b se “ol c.s.rorces = SS CouERENSon 60. FORCES MoTOR Ep - | ___#4(@) Rocking Excitation _ al 2 Ty) id es “un 3 LL Tyo i ROCKING ¥ > — — ee Kx rn | &x s Wr OF FOOTING. i pn Bi + * — Wp 326243 bs. ali equvatent “|S Re 4 i fps cy f % - EQUIVALENT Mes K —|—e —- — tee ¥ AXIS OF PITCHING Ce9 i Oscicuarion. re. TY ni m,10088.29 obe27." about # XIS OF ROCKING i (€) Coupled Horizontal Translation & Rocking OSCILLATION, ¥ fai Oscilation Vz Figure 6-2. Modes of oscilation. 78 obout gh Figure 6-4. Foundation layout for reciprocating machine example problem.96 Design of Structures and Foundations for Vibrating Machines Table 6-2 Dynamic Analysis of the Trial Foundation Ser Pacancte "eDiecion ‘(Diver Getic ‘Ctiecoe) 1 a Te. aan GE ie mous eo Econ "Doon Ee + Seas ten pone istagea pete neo S2BRh TER, ATHBE Bead ct vente vy tts 0 eget Foret syns 6 rane smo na BARE ee the een 180 eyo ton Marin ‘oka en omm anon (© i ie maeus nate (Oqonatéening ates Dg 000 (9 na ding tsa Dyas unos (9 Tas ope ore py aaa pens eyo ones gt fw fie = nVs fe Nona 55 N ics © pn mee Imp = 1903 fag = 190 on Pimp Opt =e 502 oe ABE o0 min = yoo (Vain in Ete te eke aa pron so ee a 0 Comoe wen tas HABER, MACARRON eH, Xo Meele ida ae ~ mesa, Saamicion sett ES SEER SIR tate, “'Iiloton (0 Mel erent Tihdawart adios Tania" Cdeyaeoe 60 aia oor 0) = 330 a B= He (1) Tani ce Tabi re Stites TTS eta Lake TREE a “talon moments al sod nmi ww opi 9 dega Ie ao masu rs Howie heading ral, E. Dynamic Analysis Forces generated by the reciprocating compressor (Figure 6-4) will result in vertical (z), lateral (x), rocking (¥), and pitching (6) oscillation of the footing, While the forces generated by the motor only will tend to oscillate the footing in the vertical (z), longitudinal (0), torsional (8), and pitching () modes. However, the magnitude of the forces generated by the motor are small and, consequently, the resulting vibration response is negligible. Therefore, a dynamic analysis is only per- formed for the compressor forces. If an analysis is required for the motor forces, the tables and figures provided are applicable. F. Check of Design Criteria—Static Conditions (Gee Chapter 3 for Design Checklist) 1, Static bearing capacity: proportion footing area for 50% of allowable soil pressure. From D above, 883 < 1,250 psf, OK. 2. Static settlement must be uniform; CG. of footing ‘and machine loads should be within 5% of each linear dimension. The center of gravity of machine loads and footing coincide. 3. Bearing capacity: static plus dynamic loads. The magnification factor (Table 1-4) should preferably be less than 1.5. In the example problem, this factor is less than 1.5 for the primary operating frequency but is slightly higher for the secondary operating speed. ‘The sum of static and modified dynamic loads should not ereate bearing pressures greater than 75% of the allowable soil pressure given for the static load condi- tion Transmitted Dynamic Vert. Force, Ps = 2,127.1 Ibs. ‘Transmitted Moments Pry = #22,035.9 + 2,127.1 (2.75) + *870.7 (8.0) = 34,851.0 Ibs. fe Prp = 61,1076 Ibs-ft * Out of Phase by 90°, but this combination is con- servative, 2,197, Petre = 75.75) (275) _61,107.6 (6) 27.5 (15.75) Total static plus dynamic bearing presure = 883 + 162 = 950.2 pst < 0.15 (2,500) is OX. and 883 ~ 66.4 = 8166 pat (no uplift) is ako OK. 4 Settlement: static plus repeated dynamic loads. The combined C.G. of the dynamic loads and the static loads should be within six in. of the footing center of gravity. For rocking and pitching motion, the axes of rocking and pitching should coincide with the principal axes of 34,851.0 (6) 15.75 (27.5) a = — 664, 76.2 pst Design Examples: Block Foundations 97 the footing. The magnitude of the resulting settlement should be less than the permissible deflecting capability of the connected piping system. In this example, dynamic forces are small compared to static loads; therefore, settlement caused by dynamic loads will be negligible Limiting Dynamic Conditions 1. Vibration amplitude at operating frequency. The ‘maximum amplitude of motion for the foundation should lie in Zones A or B of Figure 3-3 for the given acting frequency. The maximum vibration amplitudes are 0,000717 in, and 0.000443 in. in the vertical and hori- zontal directions, respectively. These amplitudes fall in Zone A of Figure 3-3 at the operating frequency of 585 rpm and are, therefore, acceptable 2. Velocity is equal to 2x/ (cps) % displacement amplitude as determined in (1) above, This velocity should be compared to the limiting values of Table 3-2 and Figure 3-3 for at least the “good” condition, Velocity 2r (585/60) (0.000717) = 0.0439 in./see. This veloc- ity falls in the “good” range of Figure 3-3 and is, there- fore, acceptable. 3. Acceleration is equal to 42*/* x displacement amplitude as determined in (1) above. This check is only necessary if conditions (1) or (2) are not satisfied. ‘The acceleration should fall in Zones A or B of Figure 3.3. 4. Magnification factors should be less than 1.5. In this example-problem the magnification factors are less than 1.5 for the primary operating frequency and slightly larger for the secondazy frequency. 5. Resonance. The acting frequencies of the machine should have at least a difference of +20 with the reso- nance frequencies; that is, f < OB fm or > 1.2 fa. In this example, resonance cannot occur in the vertical and horizontal modes (due to the large amount of damping in those directions). In the rocking mode, 08 fu = 08 (1,934.3) = 1,547.4 cpm and 1.2 f, = 2,821.2 rpm Since the primary and secondary machine frequencies are 585 rpm and 1,170 pm, no resonance will occur in the rocking mode; therefore the design is judged accept- able. In the pitching mode, 0.8 f. = 0.8 (1,717.6) = 1,874.1 rpm and 12 f, = 12 (1,717.6) = 2,061.1 rpm. ‘The primary and secondary machine frequencies also fall outside of these ranges and, therefore, no resonance ‘conditions are possible. 6. Transmissibility factor (usually considered for high- frequency machines mounted on springs). The trans- missiblity factors should normally be less than 1. In the example, the T,, values are greater than 1 indicating that the dynamic forces are amplified. 7. Possible vibration modes: Vertical and horizontal ‘oscillation are possible modes since the force may act in98 Design of Structures and Foundations for Vibrating Machines those directions. Rocking (¥) oscillation is also possible since the horizontal forces act above the CG. of the foundation. Pitching ($) oscillation must also be con- sidered in this example since unbalanced moments are provided by the machine manufacturer. However, twist- ing or yawing oscillation is not considered since the horizontal forees do not form a couple in the horizontal plane. The horizontal translation and the rocking modes need not be coupled if: Vine Perel (foe X foo) <2/(3F) Using values from Table 6-2 for far = 1009.1, fop ,677.6 and (primary frequency) = 585 rpm, and substituting in the above expression, V( 008-1)" + (1,677) (1,009.1 x 1,677.6) = 1.156 x 10-* and 2/(3 X 585) = 1.140 x 10°, which appears to be a border line case (within 19). Hence, uncoupled mode analysis is OX. Environmental Demands 1. Physiological effects on persons. If the machine is located in a building, Figures 3-4, 3-5 and 3-7 are used to test the adequacy of the installation. In the example, Figure 3-4 indicates vibrations to be “barely noticeable to persons” at the operating frequency of 585 epm for a ‘maximum vibration amplitude of 0.000717 in 2. Psychological effects on persons. Use the same pro- cedures as in (1). When the machine is located close to people not connected with machine operations, an acous- tical barrier may be necessary. 3. Damage to structure. Use the limits given in Figure 3-4. The example check shows no danger. 4, Resonance of structural components (superstruc- ture above the footing). Avoid resonance with the lowest natural frequency by keeping the ratio of operating fre- quency to natural frequency less than 05 or greater than 1.5. No other structural components ace involved in the example. Conclusion, ‘The foundation is predicted to perform in an acceptable manner. The static and the dynamic analysis confirm the adequacy of the proposed founda- tion configuration and, therefore, the design as proposed is acceptable. ‘Nomenclature—Example 1 4; = Width of section i, ft ‘B= Length of rectangular foundation block, ft Bz, Bs, By, Ba, By = Mass (or inertia) ratio; vertical horizontal, rocking torsional and pitching vibration mode b, = Depth of section i, ft D= Damping ratio= C/Ce Dep Des, Dy, Dog = Geometric damping ratios; vertical, horizontal, rocking and’ pitching modes (D, = Internal damping ratio P(t) = Excitation force, ibs F, = Amplitude of excitation force, Ibs. F,=Maximum horizontal dynamic force, Ibs F,= Maximum vertical dynamic force, Is. Operating speed of machine, rpm je= Critical speed of the machine, rpm f= Natural frequency, rpm fn = Resonant frequency, 1pm fag = Resonant frequency in the rocking mode, rpm Ing = Resonant frequency in the pitching mode, rpm for = Natural frequency in the x-direc- tion, rpm foe = Natural frequency in the -direc- tion, rpm, fap = Natural frequency in the rocking direction, spn fay = Natural frequency in the pitching direction, tpn Shear modulus of sil, psi Acceleration of gravity = 322 f/ Dynamic horizontal force, Is Effective foundation embedment depth, fe 1y,1o,1p = Mass moment of inertia in the y (rocking), 0. (twisting) and (Pitching) ‘directions, "Ibs. sec*-fe Segment (i = 1, 2,3...) Spring constant, Kips/ie = Distance from center of mass to base of footing for segment i, fe bes kes bys ke, ky = Equivalent spring constants: verti- cal, horizontal, focking, torsional snd pitching modes ‘Width of base of machine founda- tion block, fe Dynamic magnification factor Mass, Ib. seet/ft Mass of sogment i, Ibs, sect/ft Rocking and pitching mass ratio factors for geometric dampin = Force transmitted in the horzontal direction, Ibs. Force transmitted in the vertical direction, Ibs ‘Transmnited rocking moment, f-lbs. Transmitted pitching moment, {lbs Bearing pressure due to transmitted dynamic force, pal Ratio of operating frequency to natural frequency = f/f, ro= Equivalent radius for rectangular footing, fe Rie = Horizontal distance from center to edge of footing in the x-direction, fe Horizontal distance from center to due of footing in the yaetion, ft Ry = Vertical distance from base to hori- zontal machine load, ft Allowable soil pressure, ks Unbalanced rocking moment, ft-lbs. Unbalanced pitching moment, ft. Ibs ‘Transmissbilty factor Transmissbility factor for primary operating frequency T, (S) = Transmisibility factor for second- ary operating frequency W = Total weight of machine plus foun- dation, Tbs. Weight of foundation, h isplacement response in the horic zontal direction in X;= Total displacement response in the pa hesizontal seston, tn. Z = Displacement response in the verti- cal =diretion, in, Total displacement response in the vertical e-irecton, in. x, a2,09, ag = Damping ratio embedment factor; vertical, horizontal, rocking and pitching modes B= Phase angle, rad Bu» Be Bo» Bo = Spring coefficients; vertical, hori- zontal rocking and pitching modes = Soil density, pef ‘tn ne Mo 96 = Spring constant embedment factors; vertical, horizontal, rocking and pitching modes = Poisson's ratio of soil (= Mass density of soil =y/¢ Ibs so'/ ‘sy We= r= Z o = Frequency of excitation force, rad/ en = Natural circular frequeney, rad, Example 2: Design of a Foundation Block for a Centrifugal Machine (ref. 2) ‘A. Machine Parameters The machine parameters necessary for the design of the foundation are defined in Chapter 3. The following data are required (all terms are defined where they first occur) Compressor: Weight (We) 35,270 Ibs. Rotor Weight (Wx) 2,100 Ibs. Operating Speed (/) 6,949 rpm (w) 721.7 rad/seo Design Examples: Block Foundations 98 Gritical Speed (f.) = Ist~3,400 rpm 2nd ~ 9,000 rpm Eccentricity of Unbalanced Mass é = 0.0015 in. (provided by manufacturer for the static condition) ‘The dynamic eccentricity at operating speed may then be calculated from IU (fffe)*} = 0.000472 in. Often, the manufacturer may claim a 2ero eccentricity for the rotor components. A design value can never- theless be selected from Table 3-1 as in Example 3 ‘which follows. Centrifugal Force Fy = (We/g)ea® = 1,359 Ibs. Turbine Weight (Ws) = 16,000 ibs. Rotor Weight (We) ‘545 Ibs. Operating Speed (/) = 6949 xpm 127.7 adfsec al Sped (f) ist~ 2,000 rpm Ind ~ 9,020 rpm Eccentricity of Unbalanced Mass, 1.0015 in, (which is again given by the manufacturer) Dynamic eccentricity at operating speed, ‘= 0,0015/(1 ~ (6,949/2,000)*] = 0.000135+ in. Centrifugal force Fy = 101 Ibs. ‘Total centrifugal force Fy = 1,359-+ 101 = 1,460 Ibs. Base plate: weight (Ws) 9,000 Ibs. Total machine weight (Wx) = Wet Wr+ Wa 56,270 Ibs. B. Soil and Foundation Parameters ‘The soil parameters are obtained from the soil report prepared for the plant facility. The factors which are considered in the preparations of the soil report are listed in Chapter 3. In this example the following infor- mation is obtained from the soil report Plant Grade El, 1007-0" ‘Top of Foundation El. 1017.0" Recommended Foundation base, El. 98/4 Soil is Medium Firm Clay Soil Density (y) = 125 pel Shear Modulus (@) = 6,500 pai Poisson's Ratio (v) = 045100 Design of Structures and Foundations for Vibrating Machines Soil Internal Damping Ratio (Dy:) 0.05 Static Allowable Bearing Capacity (Sa) = 15 kal Also, the settlement at the allowable bearing pressure is negligible. ©. Selection of a Foundation Configuration A trial configuration is selected follawing the guide- lines listed in Chapter 3 under Trial Sizing of a Block Foundation. A shallow and wide footing is desired such that the combined center of gravity of the machines and of the foundation coincides closely with the centroid of the area of the foundation. A foundation block configura tion which satisfies this requirement is shown in Figure 6-5. Concrete Footing Trial Outline: ‘Weight of the footing (Wr) = 100,500 Ibs. Total static load (W) = machine weight + ‘weight of footing = 156,770 Ibs. Actual soil pressure = 156,770/12.5 (20) = 627 psf <05 (Sen) = 750 pst ‘Thus, area of footing is O.K. Weight of footing/weight of machine = 100,500/ 56,270 = 1.78 which is close to the suggested minimum of 20 for well-balanced centrifugal machines. All other ‘guidelines for trial sizing are satisfied and the dynamic analysis is then performed. D. Dynamic Analysis ‘The axis of rotation of the shaft is located 6 ft above the bottom of the foundation. The dynamic force acting at the axis of shaft is of the form F = meea* sin at (see Figure 6-5) which will excite the structure in three different modes, viz, vertical, horizontal, and rocking, Since the machine will operate at a constant speed in the steady-state condition, the amplitude meco® is con- stant, Thus, formulas associated with a sinusoidal force of constant amplitude Fy are used in the dynamic anal- ysis (F= Fy sin ot). A complete dynamic check is per- formed in Table 6:3. E. Check of Design Criteria ‘The foundation block is checked for the design criteria as described in Chapter 3. Static Conditions 1. Static bearing capacity. Proportion of footing area for 50% of allowable soil pressure. From C above, 627 < 750 psf is OK. 2, Static settlement must be uniform; C.G. of footing and machine loads should be within 5% of each linear dimension. The center of gravity of machine loads and foundation coincide and is O.K. 3. Bearing capacity: static plas dynamic loads. The magnification factor (Table 1-4) should preferably be Jess than 1.5, ‘The sum of static and modified dynamic loads should be within 6 in, of the footing CG. For 175% of the allowable soil pressure given for the static load condition = 627 + $74/(125 x 20) + 1420(6) / 20 (12.5)! = 682 oF 626 paf < 0.75 (1,500) paf is OK 4, Settlement: static plus repeated” dynamic loads. ‘The combined C.G. of the dynamic loads and the static loads should be within 6 in. of the footing C.G, For rocking motion the axs of rocking should coincide with the principal axis ofthe footing. The magnitude of the resulting settlement should be less than the permisible deflecting capacity of the connected piping system, In this example, dynamic forces are small compared to static loads; therefore, settlement caused by dynamic loads will be negligible. Limiting Dynamic Conditions 1. Vibration amplitude at operating frequency. The maximum amplitude of motion for the foundation sya- tem should lie in Zone Aor B of Figure 3-3 for the given acting frequency, Vibration amplitude (vertical) 2, = 01000019 in. at 6,949 pm. From Figure 33 this amplitude is within the safe allowable limits. Vibration amplitude (horizontal) at center of bearing area X= 0.000018 in. at 6,949 rpm. The amplitude falls in Zone A in Figure 3-3 and is, therefore, acceptable. 2 Velocity equals 2zf (cps) diaplacement ampl- tude az caleulated in (1) above. Compare with the limiting values in Table 5-2 and Figure 33 at leat for the “good” condition. Velocity equals 727.7 (0.000019) = 0.0138 in./sec. From Table 3-2 this velocity falls in the “smooth operation” range and is, therefore, ac- ceptable. 3. Acceleration equals 4'f¢ x (displacement ampli tude, as calculated in (1) above). Should be tested for Zone B in Figure 3-3. Note: ‘This check is not necessary if conditions (1) and (2) are satisfied, which they were in this example 4. Magnification factor (applicable to machines gen- crating unbalanced forces), The calculated values of M and M, (Table 1-4) should be less than 15 at resonance frequency. In the example, Min all modes is Jess than 1.5. 5. Resonance. The acting frequencies of the machine should have at least a difference of £209 with the resonance frequency of Equations of Table 1-4, (08 for = { 12 far). In this example, there is no res0- nance frequency in the vertical mode. In the horizontal mode, 1.2 % 1,221.4 < 6,949. In the rocking mode, 1.2 X 1,686.1 < 6,949. Therefore, a resonance condition does not occur C.G. OF MACH. LBS. AND FTG. Wm(MACH. WT.) 56270; Design Examples: Block Foundations 101 AXIS OF ROTATION be = 1’ bs = 2'-0" Figure 6-5. Foundation layout for centrifugal machine example problem. Table 6-3 Dynamic Analysis (Three Modes of Oscillation are Possible) eke toma i Teenie Ra an 1 ies en aaa =a oo _ = some in ga eB [tassios ne] y= Efecto saad en een ‘ata Te ete yn Sse emt tea | 7 ° tron met fe Monee ame bye 1 oer “sg fu=t0im Ig Rte fap sinh 8 na rawr ur= im fgg = ts atin ron ane wan © Ste Costa Ce eer 116) Tt mtn ron reemxwen coin xaond 10) gpa sgt — shame amplitude Orne = oipanits . 2 Rent et rene hoa a Straw “ Fis site, ag = Lab ate102 Design of Structures and Foundations for Vibrating Machines 6. Transmissbility factor (usually applied only to high-frequency spring-mounted machines). The value of transmissibility is calculated by equations of Table 14 and should normally be less than 1 for spring- ‘mounted machines having an inertia block. In the ex- ample, T, is less than 1 indicating that dynamic forces are not amplified, Possible Vibration Modes 1 and 2. Vertical oscillation or horizontal translation are possible modes as the force acts in either direction. 3. Rocking oscillation is possible since the point of horizontal force application is above the foundation mass CG, 4. Torsional oscillation. Since horizontal forces do not form a couple in the horizontal plane, this mode is not possible. 5. Coupled modes. ‘The horizontal translation and rocking oscillation are usually coupled. The coupled ‘modes may be considered as in example 3 which follows, Fatigue Failures 1. Machine components. Follow limits in Figure 3-4 and/or Table 3-2. 2. Connections. Same as (1) but check stresses using AISC code (ref. 13 of Chapter 3) when connectors are bolts or weld 3. Supporting structures. Use (2) for structural steel, For concrete footing, if reversal of stresses takes place and the amplitude is very high (such that the peak stress reversal is over 50% of the allowable stress), the main and the shear reinforcement (if any) should be designed for the stress reversal condition, In this example the amplitude of the dynamic forces js not large enough to produce any significant stress increase over the stresses caused only by the static loads. Environmental Demands 1, Physiological effects on persons. if the machine is located in a building, use the procedure given in condi tion Enviromental Demands under ‘Limiting Dynamic Conditions”, and use the limits from Figure 3-4. In the example, Figure 3-4 indicates no discomfort to people. 2, Psychological effects on persons. Use same pro- cedures as (1). If the facility is located close to people not connected with machine operations, use acoustic barriers. In the example, the machine is located away from habitable areas, 3. Damage to structures. Use limits in Figure 3-4 or 3.5. Example check shows no danger. 4, Resonance of structural components (superstruc- tures above the footing). Avoid resonance with lowest natural structural frequency by keeping the frequency ratio either less than 0.5 or greater than 1.5. In example, no structural components are involved, Thus, the trial design is acceptable and may be used to support the machine. ‘Nomenclature—Example 2- B= Length of rectangular foundation block, fe By, Bey By, Be = Mass (or inertia) ratio; vertical, hori- zontal, rocking and torsional vibration modes epth of section i, ft amping ratio Damping ratios; vertical, horizontal, rocking and torsional modes Internal damping ratio Eccentricity of unbalanced mass to axis of rotation at operating speed, in. Eccentricity of the machine's unbal- anced mass, in. F = Excitation force pm Gritical speed of the machine, mpm ‘Natural frequency, rpm Jy = Resonant frequency for constant force- amplitude excitation, rpm Jr = Resonant frequency for rotating mass- type excitation, rpm Resonant frequency in the horizontal direction, pri fay = Resonant frequency in the rocking di- rection, rpm G= Shear modulus, psi g = Acceleration of gravity, ft/sec? Dynamic horizontal force, Ibs. Mass moment of inertia, Ibs.sec*-ft Segment (1, 2,...) = Spring constant Distance from center of rotor axis to footing, ft ki = Distance fiom center of mass to base of footing for segment i, f¢ Equivalent spring constants; vertical, horizontal, rocking and torsional modes = Width at base of machine foundation block, fe M, = Magnification factor My max = Maximum magnification factor ‘M = Dynamic magnification factor ‘m= Total mass, Ib-sec#/ft h a hae Bes hs balanced mass Mas of segment i lumber of segments y= Force transmitted through spring 1= Ratio of operating frequency to natural frequency, f/f quivalent radius for rectangular foot ing, f ‘Ry = Horizontal distance from center to edge of footing Ry = Vertical distance from base to center of rotor axis, ft Allowable soil bearing capacity, kt Unbalanced torque, fb. 1, = Transmissibility factor Time, see Dynamic vertical force, Ibs Total weight of machine plus footing tbs, Wy = Base plate weight, Ibs, Compressor weight, Ibs. Wp = Weight of footing, bs. Wy =Total machine weight = We+ We+ Wa, Ibs. Wa = Rotor weight, Ibs. Turbine weight, Ibs. X,= Total displacement response in the horizontal x-direction, in. 2,= Total displacement respons tical =direction, in i Be, By = Spring confcients; vertical, horizontal rocking modes = Soil density, pof Poisson's ratio Mass density = 7/g,Ibs-sect/ftt Prequency of excitation fore, rad/see [Natural circular frequency, rad/sec the ver- Example 3: Foundation Design for Centrifugal Machines with Different Operating Frequencies ‘and Supported on an Inertia Block Tn some plant facilites, due «0 environmental con- siderations or poor sil conditions, it becomes necesary to limit the propagation and amplitude of the machine vibrations transmitted to the foundation (ref. 3), Tn those circumstances, the use of an inertia block sup- Ported on springs is recommended as a vibration isolator; see Figure 6-6. This type of supporting system requires that the piping which is connected to the machines be jointed with flexible couplings in order to absorb Without distress the resulting large movements of the Inertia block, This movement may be caused cither Design Examples: Block Foundations 108 due to a sudden surge condition during the operation of the centrifugal machine or when a resonance condi- tion occurs temporarily at start-up or shutdown of the ‘machine. The latter condition generally is more severe since an inertia block spring system generally has negligible damping resistance. ‘The inertia block spring suspended foundation is not recommended for heavy machines with large unbalanced forces. However, this type of system may be used when the machines are located on an elevated stecl-framed structure, In this example problem, a foundation system for a gas turbine/generator set is investigated, i.e, the ma- chine consists of an electric generator powered by a gas turbine. Both machines run at different operating fre- quencies and the step-down from the higher to the lower frequency is accomplished through a gear box. Foundations for this type of machine have been dis- cussed in Chapter 2, and the various steps required to complete the dynamic analysis are given below ‘A. Machine Parameter 1. Generator: Weight (Wx) = 28,150 Ibs. Rotor Weight (Wx) = 9,460 Ibs. Operating Speed (f) 1,800 rpm ° 188.5 rad/see Critical Speed fe 2,200 rpm Eccentricity of Unbalanced Mass, e ~ 001 in. (Table 3-1) Centrifugal Force Fy = (Wag) ¢u*= 871 Ibs. 2. Turbine: Weight (Wr) 6,305 Ibs. Rotor Weight (Wr) ‘567 Ibs. Operating Speed (f) = 8,990 rpm ° = 941.48 rad/see Critical Speed (fe) Ist ~ 2,885 rpm. 2nd ~ 11,670 spm Eccentricity of Unbalanced Mass at Operating Speed, e = 0.5 V'12,000/8990 mit (see Table 3-1) (00057 in, Centrifugal force Fy 567/386) x .00057 X (941.43)* = 742 Ibs. 3. Gear Box: Weight (We) = 13,045 ths, Unbalanced forces generated by gear box and cou- plings are assumed negligible. Total Machine weight (W) = Wat We + Wo = 57,500 Ibs. A layout of the equipment is shown in Figure 66.104 Design of Structures and Foundations for Vibrating Machines CG OF TOTAL EQUIPMENT GEAR BOX 13,045 Ibs GENERATOR 28,150 bs. INERTIA BLOCK (A) PLAN — EQUIPMENT LAYOUT COMBINED CG. OF EQUIPMENT AND INERTIA BLOCK CG. EQUIPMENT —C.G. INERTIA BLOCK (B) FOUNDATION AND EQUIPMENT LOADS Design Examples: Block Foundations 105 FOUNDATION C.G. EQUIPMENT, INERTIA BLOCK AND FOOTING COINCIDES "ed 2g & TURBINE CONNECTION To HAVE NEOPRENE BEARING PAD AND BOLT HOLES SLOTTED IN VERTICAL DIRECTION ~~ LATERAL RESTRAINT W6x20 (TWO EACH SIDE) (C) DETAIL LATERAL RESTRAINT Figure 6-6. Foundation configuration for centrifugal machine with an inertia block. 4. Center of Gravity of Unbalanced Forces: 2000 The unbalanced forces generated by the equipment are assumed to be acting at the center of gravity of the 1200 machine Joads and perpendicular to their shaft axis. 2 The shaft axes are shown in Figure 6-6B. The forces 4 of the two machines, when combined, are given by: E100 F(t) = 871 sin 188.5¢-4 742sin 941.4¢ ca ‘The plot of individual force functions, as well as the a combination of the individual force function, is given Bg, in Figure 6-7 B. Soil and Foundation Parameter on Soil is Soft Silty Clay Soil Density (y) = 110 pef Shear Modulus (G) = 3,500 psi Poisson’s ratio (v) Soil Internal Damping Ratio, (Dys) = 0.05 Static allowable bearing capacity (San) = 1.0 kst Figure 6-7. Plot of unbalanced centrifugal force.108 Design of Structures and Foundations for Vibrating Machines Settlement of soil for a 14 ft by 28 ft footing at 1 ksf bearing presture = 0.125 in, Water table Suctuates and is 3 ft below grade at cer- tain times of the year. ©. Selection of a Foundation Configuration ‘The guidelines listed under Trial Sizing of a Block Foundation in Chapter 3 are followed in selecting an initial configuration, Because the water table is 3 ft below grade, it is recommended that the footing be located at a shallow depth in order to avoid construction ‘complications and a relatively large contact surface area be used. In order to achieve uniform settlement, it is necessary that the center of gravity of the equipment plus the inertia block coincide with the center of gravity of the footing. A trial proportioning of the inertia block and the footing is shown in Figure 6-6. I. Inertia Block Trial Outline (Figure 6-6A) : Center of gravity of total equipment (y-direction) 16,305 (14.0833) + 13,045 (7.2917) = 6,505 F 13,045 + 28,150 5,608 fe =5 ft — 74% in. from centerline generator, or 11 ft — 2% in, from the left edge of the inertia block. Center of gravity of total equipment (x- direction) 1 (2.0) + 13,085 (1.0) 505 + 13,085 + 28,150 7940 fe 1% in, from centerline generator. ‘The inertia block has a uniform thickness of 8 in. (100 psf) and an additional thickness of 8 in, (100 pef) of dimension 10 ft x 10 ft in the region under the turbine, ‘The center of gravity of inertia block (y-direction) _ 10 x 10 x 100 (200) +25 x 10 x 100 (12.5) a 25 x 10 x II 0X 10 100 4.6129 1 4 ft— 7% in, (from left edge of inertia block). The center of gravity of inertia block (z-direction) =5 ft—0 in, (from bottom edge of inertia block) From Figure 6-6B, C, Combined C.G. of equipment and inertia block, 57,500 x (11.2292) + 38,000 (14.6429) (retin) ee et a = 12 {t—6 in, from left edge of the inertia block, Combined C.G. of equipment and inertia block, (x-direction) =5 ft—0 in, from bottom edge of inertia block 2. Footing Trial Outline: The footing plan dimensions should be larger than the inertia block in order to accommodate the support: ing springs of the inertia block and its lateral supports. Also, the resultant bearing pressure on the soil should be less than 50% of the allowable soil bearing pressure in order to minimize possible foundation settlements. A trial concrete slab size 14 ft wide by 28 ft long and 2 ft thick is then analyzed. The footing center of gravity Js made to coincide with the combined center of gravity ‘of the equipment and the inertia block, Weight of the footing (1) = 117,600 Ibs. ‘Total static load (1) = equipment weight + inertia block + footing weight = 210,100 Ibs. Actual soil. pressure = 210,100/14(28) ~ 0.5 Sarr 534 psf Thus, area of footing is OK. D. Dynamic Analysis ‘A mathematical model of this foundation was pre- viously discussed as Model 3 of Chapter 2. We have the following parameter calculations: Selection of Springs for Inertia Block 1. Vertical Direction. Try for transmissibility factor (7,) = 02. From the transmissibility equation of Table 1-4, and assuming damping to be negligible in the springs, D = 0, then the resulting equation is 0.02 = 1/|1- | or r= fifa = 70 For f= 1800 rpm, fy = 257.14 spm For f = 8990 rpm, fu = 1284-29 rpm ‘A natural frequency (fx) = 257.14 will be used, since a higher natural frequency will require a large number of springs Mass (m.) of inertia block + equipment (57,500 + 35,000) 7 38 = 289,64 Ibs.-sec*/in. Total spring constant (fs:) (257.14 x zal) ere 60 173,762 Ibs. fin. ‘Try 20 spring with a spring constant for each 173,762 spring = —5— ‘There is a commercially available spring of 8,800 Ibs. force for a I-in, deflection. The dimensions of the springs are: height = 9.0 in., width = 5.25 in,, length 3.0 in,, and maximum solid load is 10,912 Ibs. for a L.2tin. deflection. Use 20 springs for total ky; = 176,000 Ibs. fin. Then, fa = (60/2e) 176 000/239.64 = 258.8 rpm. 2. Horizontal Direction (Figure 6-6C). Lateral re straint is provided by W6 X 20 vertical posts, two on 8,688.1 Ibs.fin. each side. The posts are fully fixed at their bottom and are connected to the inertia block by the provision of slotted holes in the vertical direction so that the oscil- lation of the vertical springs is not effected. A neoprene bearing pad layer is inserted in the connection in order to absorb high-frequency vibrations. The spring constant in the lateral direction is given by: E ay = 1.6132 x 10° bs in Mass (m,) ofthe inertia block and equipment 239.64 Ibssec*/in. © fa= (60/2x) V7,613,200/230.68 3. Rocking Oscillation about Point 0 ( ‘Two rows of springs, each containing ten springs, are located at a distance of 60 in. on either side of an’ axis passing through point 0. Due to this arrangement, the inertia block is capable of rocking about that axis. ‘The spring constant for the rocking oscillation y is, thus, a function of the vertical spring constant and is given by: iy = Dye? where ky is spring constant of each row (10 x 8800 = 88,000 Ibs/in.), ¢ is the distance from the axis to the row (=60 in.). sy = 2 X 88,000 x 60 x 60 683.6 X 10° bs-in./rad In order to calculate the natural frequency, the mass moment of inertia for the inertia block and equipment ust be calculated and is given by t= Erm Machines = (57,500/386) (60)* = 536,269.0 Inertia block = (25,000/386) (12)? + (10,000/386) (20)* = 19,689.0 y= 536,269 + 19,689 = 555,958 Ibsin.-sec* (60/25) (635.6 X 105 /555,958 322-4 rpm “fa Therefore, the natural frequencies of the inertia block- equipment-spring system are 258.8, 783.5, and 322.4 rpm in the vertical, lateral, and rocking modes. Table 6-4 lists all computations for the single-degree-of-freedom system, E. Dynamic Analysis as a Mul lass System ‘The calculation of the natural frequencies for a two- ‘mass model (see Figure 1-20) is given by: Design Examples: Block Foundations 107 nsf e bth oh SEY +] The terms with subscript 1 stand for inertia block plus equipment, and 2 for the footing, The calculations for the footing only (Id’ x 28° block) is performed in Table 6.5 1. Vertical Oxcilation 4 = 176,000 Ibs.fin, my = 23964 Ibsen Kea = 3.07 X 10° bs fin, mz = 904.35 Ibs-sec*/in where (Table 6-5) and m,are the spring constant and mass of the foundation block, respectively 691.85 oF fey = 251.2 rpm tx? = 10,707.93 0 f= 988.2 rpm 2 Horizontal Oxcillation: Key = 1.6132 10°lbs in, m= 299.64 Ibs-sect/in, kee = 2.3588 10* bin, m= 304.35 Ibs-sec*/in, fou = 5845 rpm foe = 1282.1 rpm F. Discussion of Dynamic Analysis 1, Natural frequencies. The values calculated for a single-mass model and a two-mass model reveal that for the vertical mode there is no difference in the calculated frequency when either model is used. Tis is because the natural frequency of the modgl, including the inertia block, has 2 natural frequency of less than half the natural frequency of the footing in the vertical mode (258.8 vs. 717.5 rpm, Table 6-5, respectively). Therefore, the fundamental frequency ofthe coupled model has small difference withthe lowest frequency calculated as individual uncoupled models (251.2 rpm vs. 258.8 rpm, respectively) This fact can be demonstrated by using Southwell-Dunker- Iey’s formula, 1 et substituting, fs = 258.8, fa = 717.5, then, fe= 2435 xpm, which differs less than 3.29% from the calculated coupled model frequency and 6.3% from the calculated single mass frequency. Similarly, for the rocking mode, the fundamental frequency calculated by using f, = 3224 and fs = 883.2 into the Southwell-Dunkerley’s formula will be 30285 rpm, which differs by 6.5% from the calculated frequency of the single mass model. Ifthe coupled equations of horizontal and rocking modes are solved, then the lowest frequency is found to be close to 250 rpm. Therefore, due to the significant difference in[ i t i 108 Design of Structures and Foundations for Vibrating Machines Table 6-4 Dynamic Analysis of Single-Degree-of-Freedom System (Machine plus Inertia Block Only) Step Vertical Horizontal Rocking No. Parameter Source Excitation Excitation Oscillation, oo fe Natural Frequency, 4/— 258.8 cpm 783.5 rpm 822.4 rpm Magnification factor Mf Table 1-4 0.284 = 0.008 Dynamic force F() Figure 67 27.0 20 (tbs) 4. Vibration Table 4 800) SX = 1.205 X 10 v= 2B x 1* amplitude = 1089 10+ ()Z(y = 8,900) v= 70% x 10+ 421 x lo radians) Ga) ¥ x Ru, 5 Components of rocking ( f ~ 1,800 (c) 2.792 X 10'* x 60 - ‘oscillation A x 10 fm 8900 (a) 7.027 x 10° x 60 216 x 10% - (ay Z 6 Resulting vibration +e) 2167 X 10-* = ‘amplitude HOF) Gn) : 7. Trangmissbilty factor 7, Table 14 Ty (J = 1,800) = 0.021 T, = 0.284, Py = 208.80 T, = 0.088, Py = 1,725.0 ‘and force Pee 1830 Mansmitied Po TAU 5 8000) = 0.001 7, = 0.008, Px = 501 T= 0.001, Pum _ 445 = OH | “8 —— ‘Total 209.74 ‘Total 1,769.5 otal 19.08 (lbs.) (bin) tbs.) the natural periods of the two mass elements, it is per- iissible to assume that the individual elements act inde- pendently of each other, i, in the vertical and rocking ‘modes the inertia block and the bottom footing masses can be analyzed on the basis of an equivalent one-degree- of-freedom or uncoupled system. However, in the horizontal mode, both mass ele- ments have nearly equal natural frequency; the inertia block has a natural frequency of 783.5 rpm, and the footing has a natural frequency of 628.9 pm. The fre- quencies of the two-mass coupled mode are 584.5 and 1,282.0 xpm for the inertia block and the footing, respec- tively. From Dunkerley’s formula, the lowest frequency is 490.5 rpm, which is quite low compared to 783.5 rpm obtained by considering the inertia block plus the ma- chine as an individual element. ‘Therefore, a coupled ‘model investigation is justified in the lateral direction. 2. Response calculations: Because the équations of motion of Model 3 for these foundations are linear, the dynamic response generated by each of the two com- ponents 871 sin 198.5¢ and 742 sin 944.4¢ ofthe excitation force can be combined using the principle of superposi- tion. This procedure has been used in Steps 4 to 6 of Table 6-4. However, in that table, the inertia block clement was considered to be acting independent of the footing. This uncoupling was found to be justified for the vertical and rocking modes, but for the horizontal mode, an analysis based on coupling of m; and ms: is required. ‘The following equations give the response values and consider the effects described above: nts) m F sin (ot = #9) m (a = el) (oF — a al) Design Examples: Block Foundations 109 Table 6-5 Dynamic Analysis of Footing. Only (14° x 28° X 2-0") Step Vert Horizontal Nor___ Parameter Source Exeltation Excitation 1. Equivalent Table #2 nat mW W = 210,100 Ibs Ty (Machine) | = 107,481.0 2 Massand 4g Cneria block) = "21/0288 mass moment 2 m= 210,100/82.2 m= 6.5248 Ty (Footing) 645018 of inertia a irae ‘ 9.5248, (Summation) 21, 198,088.4 Ms-ooe'/te Issectte + mbt} Mass ratio Table 43 B= 0277 By = 0.183 Geometric) Table 43 D, = 0k De = 0214 damping ‘ato. 0.208 Fternal soil dats Neatgible Neslisible Day = 0.950 cainping Sorin Figure 41 240 = 105 By = 040 Pent 6. Equivalent Table -1 2830 x 10° by, = 1702.12 x 10 spring iste Ibsefi/rad constant 1% Natural (60/2) VE 628.90 rpm $85.20 rpm Freeney fy 8 Resonance Table 14 = 902.40 rpm fay = 954.00 rpm frequency fu 4 Fsin (wt — Comparing the amplitudes of xy (t) with the values of an = —HEain (ot sparing the amplitud (8) with the nama (oF = wi) (a) In these equations, » (operating speeds) = 188.5 and 9414 rad /sec ‘of (square of the mass m; circular frequency) = 313331 o:* (square of the mass ms circular frequency) = 16,647.57 1m, (total mass of inertia block and equipment) = 289.64 Ibs.-sc*/in, 1m, (mass of the footing) = 304.45 Ibs-sec*/in. 4k (horizontal spring constant between m, and m) ey = 1.6132 X 10 Ibs 4 (horizontal spring constant of soil) = key 2.3575 X10" Ibs./in. F (amplitude of the dynamic forces) = 871 and 742 Ibs 4 and gy are phase diflerence and Substituting the above parameters, (0) = 1.336 x 10-* sin 188.5¢ +3.521 X 10“ sin 941.4¢ in, (1) = 2.682 X 10-*sin 188.51 0.0022 10% sin 941.4¢ in, x in Step 4 of Table 6-4, it’ may be observed that the response values do not change significantly, using either of the two assumptions. Therefore, the assumption of independent behavior of the inertia block in all modes of oscillation is a valid step. Furthermore, in Figure 6-7, which shows the plot of the total centrifugal force, it is found that the total forcing function follows approxi ‘mately the path of the curve: sin 188.5, and has an amplitude of 1613 Ibs. Using this function in the response equations, 247A x 10-+sin 188.5¢ in, 5831 x 10-*sin 188.5 in. The above values are much higher than the response values calculated by the summation of the individual forcing functions. Therefore, the analysis of Table 6-4 gives more accurate results. ‘Another point worth discusting isthe exclusion of the damping term when solving the coupled equations of motion of Model 3. Since the natural frequencies of the modes, fx = $345 and 1,282.1 rpm are quite different from the operating frequencies, f= 1,800 and 8,990 rpm, the influence of damping on the response values is neg- ligible for all practical purposes.110 3. Design of Structures and Foundations for Vibrating Machines ‘Transmissibility factor: From the foregoing dis- cussion, it can be stated that the factors calculated in ‘Table 64, considering the footing as a stiff support for the inertia block, is a valid assumption. Transmissbility factors were not calculated for the footing because the forces transmitted from the inertia block were of very small magnitude. However, for design purposes the values obtained for the inertia block are assumed to be trans mitted to the toil without any amplification or reduction. G. Check of Design Criteria (as listed in Chapter 3) 2 Static Conditions: (a). Static bearing capacity. Proportion footing area for 50% of allowable soil pressure. From C above, 534 pel = 500 pal (allowable). (b). Static settlement must be uniform; C.G. of footing, inertia block and machine loads coin- cides, and thus settlement will be uniform. (©). Bearing capacity: static plus dynamic loads 84 + (19/14 (28) } [(1770/12) 6/28 (14)*] 535 paf < 0.75 (1000) pst is O.K. (4). Settlement: static plus repeated dynamic loads. The increase in pressure due to dynamic loads is less than 1 paf and thus would not create uneven settlement. Limiting Dynamie Conditions (refer to Table 6-4) (a). Vibration amplitude at operating frequency. Inertia block: Z, (vertical vibration ampli- tube) = 00028 in. at f = 1800 rmp. From Figure 3, ths falls within the safe allowable limits. xy (horizontal vibration at centerline of bearings) = 00030 in. at f = 1800 rpm. From Figure 3.3, the amplitude falls in zone A and, acceptable. Footing: The dynamic forces transmitted through the inertia block are very small and thus vibration amplitude is also negligible. (b). Velocity equals 2xf (cps) displacement amplitude as calculated in (a) above. Veloc- ity = 2x (1800) (1/60) 0.0003 = 0.0565 in./ sec. From Table 3-2 this velocity falls in the “good operation” range and is, therefore, acceptable. Velocity check by RMS (root mean square) method, when response involves more than one frequency: Using response values of m: of Model 3, therefore, WIBES x 1.336 x 10}? + (9414 X 3.52 X 107)* = 0.0254 < 010565 in./sec. calculated above, thus is O.K, ‘Machines: (16,305/386) [(85.5)* (c). Acceleration: 4x* (30)* (0.0003) = 10.66 in,/seet (A). Magnification factor: From Table 6-4 this value is less than 1.5 for all modes of oscil- Tation and, thus, is acceptable. (e). Resonance condition: The natural fi cies (Table 6-4) in all modes of oscillations for the inertia block and footing are less than 0,8(1800). Therefore, no resonance condition ‘occurs at the lower operating speed. This ratio is also true for the critical speed of the machine rotor. Thus, the foundation is class- ‘fed as low-tuned or under-tuned. (£). Transmissbility factor: This factor is less than 596 in the vertical and rocking modes of the inertia block and, thus, meets the normal imitation, However, in the horizontal mode when acted on by the lower frequency (f= 1800), T, was found to be 1.234 > 0.05 normally used. This happened due to the use of a structural member (W6 X 20) asa vibra- tion isolator. Use of structural member as a lateral restraint is a required feature in this type of system in order to maintain the sta- bility of the inertia block in ease of failure of the springs. In any case, the lateral force transmitted to the footing is small and can easily be absorbed by the lateral soil in contact with the footing. 3. Possible Vibration Modes: (a) and (b). Vertical oscillation or horizontal translation is a possible mode as the force acts in either direction. (c). Rocking oscillation is possible since the point ‘of horizontal force application is above the foundation mass C.G. (d). Torsional oscillation is possible as the forces ‘generated by the two machines are of different frequencies. However, it is estimated that the natural frequency of this mode would be too low compared to the acting frequency such that the response values would not be of much significance. In case an analysis is required, then the following steps are given: (1). Mass moment of inertia about the ver- tical axis through center of gravity. + (14.5)4) + (28,150/386) [(83.5)° + (9.5)'] + (13,045 /386) [(2.5)* + (4)*] = 833,475.0 lbs.-in. sec Inertia block: 25,000 /120 X 120 , 300 x 300) co 2) 10,000 coy + 'G&R? (90)* = 773,816.0 Ibsxin.sec* 138,475.0 + 773,316.0 1,606,791.0 Ibs. in. sec* (2). Spring constant of vertical posts (W6 X 20) using the weak axis. (3) (4) (5) (6) 4 poss) = SBR 3X30X10°X2X13.3 ey = 258,503 Ibs. /in. hy = Det = 2X 258,508 (60)" 1.8612 X 10" Ib (= in) eae = 825.0 rpm Forcing function: The centrifugal force of the generator which is in phase with the peaks of the turbine’s centrifugal force at f= 1,800 rpm will not form any significant torque couple, The other peaks of the turbine force will form a torque couple; ie., four out of six peaks (Figure 6-7). Conservatively, it may be assumed that the turbine centrifugal force will form a torsional moment and may be given as T. = 742 X 60 sin 941.44 = 44,520 sin 941.4¢ Ibs.-in. Magnification factor M = = 0.0013, Transmissibility factor T= 0.0013 Response value of inertia block (longi- tudinal direction) : 1/1) M (Ts/kx) € = 0.0013 X [44,520/ (1.8612 X 10*)] x 60 = 1.87 X 10 in. (negligible) Design Examples: Block Foundations 114 (7). Force transmitted to the foregoing (Te X T)/2e = (44,520 X 0.0013)/ 2 (60) = 0.482 Ibs. (negligible) ‘Therefore, torsional mode oscillations are not significant. (e}. Coupled modes: The degrees of freedom for each of the masses were found to be acting independent of each other. Because of the linearity in the equations of motion, the prin- ciple of superposition is used to find the total response, Possible fatigue failure checks and environmental demands are also found to be satisfactory and the foundation is judged to be adequate. ‘Nomenclature—Example 3: A= Dynamic amplitude a, = Width of section i, ft Length of rectangular foundation block, ft Br, Br, By, By = Mass (or inertia) ratio: vertical mode, horizontal, rocking, and torsional vibra. tion modes b:= Depth of section i, ft D= Damping ratio Dz, De.Dy,Do= Damping ratios: vertical, horizontal, rocking, and torsional modes D,= Internal damping ratio ¢= Eccentricity of unbalanced mass to axis of rotation at operating speed, in. or half the distance between the vertical springs for calculating the equivalent value of ky and Modulus of elastic Excitation force, Ibs a= Amplitude of excitation force, Ibs, Operating speed of the machine, 1pm Je= Critical speed of the machine, rpm = Equivalent fundamental frequency, rpm Ja = Resonant frequency for constant force- amplitude excitation, rpm esp = Resonant frequency in horizontal (x) vertical (2), and rocking (y) modes a= Natural frequency, rpm fou fon= Natural frequencies of masses m;, me in coupled model, rpm fuufe=Natural frequencies of masses m,, ms in uncoupled model, pm psi112 Design of Structures and Foundations for Vibrating Machines G= Shear modulus, psi Acceleration of gravity, ft/sect Dynamic horizontal force, lbs Moment of Inertia of vertical post of strong axis (x) and weak axis (9), in IyyTo= Mass moment of inertia of inertia block in rocking (¥) and torsional (@) modes, Tbsinesec? Segment (1, 2, ...) Spring constant Distance from center of rotor axis to base of footing, f Distance from center of mass to bate of footing for segment i, ft aw Kye kor = Equivalent spring constants: vertical, horizontal, rocking, and torsional modet L= Width at base of machine foundation block, ft 1 = Height of vertical pot, in, -M, = Magnification factor wc = Maximum magnification factor "M = Dynamic magnification factor ‘m= Total mass mq = Unbalanced mass sm, = Mass of segment i = Number of segments PyyPr= Force transmitted through springs in horizontal (x), vertical (2) directions, i Ibs. a= Moment transmitted to springs in rock- i ing oscillation (y),Ibe-in. i P,=Force transmitted through spring mounts, Ibs. i [Ry = Horizontal distance from center to edge i of footing, fe i y= Vertical distance from base to center of f rotor axis, ft i i 1 = Ratio of operating frequency to natural frequency, {/fs A r= Equivalent radius for rectangular foot- i ing, ft ‘ $o31 = Allowable soil bearing capacity, ksf i Ty = Unbalanced torque, ft-lbs. I TR, T= Transmissibility factor #= Time, see Ve ww Dynamic vertical force, Ibs. fetal weight of machine plus footing, Ibs. We = Weight of gear box, Ibs. Weight of generator, Ibs. Weight of turbine, Ibs. ‘Total weight of machines, Ibs. Wr = Weight of rotor, Ibs. X,Z,¥= Displacement amplitude in horizontal (x), vertical (2), and rocking (y) modes X,.Z,= Total vibration amplitude in horizontal (x) direction at machine axis level and in vertical (2) direction at footing level, Y,, ¥2= Displacement response of masses ms, ms in coupled model, in. Be, Bes By = Spring coefficients: vertical horizontal and rocking modes Soil Density, pef Ratio of unbalanced mass to total mass m./m Poisson's ratio p= Mass density = y/g, Ibs.-sec?/ft* Frequency of excitation force, rad/sec Natural frequency, rad/sec References 1. Arya, Suresh ©., Drewyer, Roland P., and Pincus, G., “Foundation Design for Reciprocating Compressors,” Hydrocarbon Processing, Vol. 56, No. 5, May 1977. 2.Arya, Suresh C., Drewyer, Roland P., Pincus, G., “Foundation Design for Vibrating Machines,” Hydro- carbon Processing, Vol. 54, No. 11, November 1975. 3, Winterkorn, Hans F., Fang, Hsai-Yang, Foundation Engineering Handbook, New York: Van Nostrand Reinhold, 1975. 4. Richart, F. E., Jr, Hall, J. R,, Jr, and Woods, R. D., Vibrations of Soils and Foundations, Englewood Cif, NJ: Prentice-Hall, 1970. 5. Newcomb, W. K., “Principals of Foundation Design for Engines and Compressors,” Transactions, ASM.E., Vol. 73, pp. 307-318, 1951 7|Computer Analysis and Applications: Elevated Foundation ‘The availability of electronic digital computers having ‘great calculating speed and analytical power has re- sulted in substantial advancement in the engineering art of analysis and design of structures supporting dynamic machines. Increasing machine weight and speed coupled with lange costs have made the rule-of- thumb approach and hand computation either unsafe for too conservative for many structures. Modern com- puter programs yield, among other factors, the natural frequencies, the deformations, and the forces in the Inc., Channelview, Texas. Photo by 113114 Design of Structures and Foundations for Vibrating Machines structure (ref. 1). These quantities were either ignored, conservatively assumed, or calculated in a simplified approximate way in precomputer times. For example, an equivalent static strength analysis of a structure sup- porting a centrifugal machine is usually made for the following loading conditions: Total vertical load plus 0.5 of the full load acting in the vertical direction, 2. Total vertical Joad plus 0.3 of the full load acting in the transverse direction. 3. Total vertical load plus 0.1 of the full load acting in the longitudinal direction. These approximate machine load factors (0.5, 03, 0.1) are fairly accurate for an equivalent static analysis ‘when the ratio of machine acting frequency to natural frequency r in the specified direction is less than 1, greater than 1, and much greater than 1, respectively, as shown in Figure 1-36. However, the factors 0.5, 03, and 0.1 are derived for a machine with an acting fre- quency of 1,800 rpm and considered very stiff in the longitudinal direction. The selected ratios of 0.5, 03, and 0.1 are generally consistent with a highest rigidity in the vertical direction, a notso-high rigidity in the transverse direction, and no dynamic load component in the longitudial direction, Even though the ratios, 0.5, 0.3 and 0.1 are approximate, they are useful in that being conservative, a safe structure will result. Since the strength is usually not a controlling design factor, many designers have traditionally used equivalent static loads in their strength check. A dynamic analysis, coupled with the help of the computer, will give the true dynamic forces that act on the structure in addition to the natural frequencies and displacements In many cases, the structure or soil parameters are known only within certain limits. For example, the shear modulus of the soil may vary by 25% or more at points below the foundation as described in Chapter 4. The effect of these variations may be studied by ‘making additional computer runs and varying the parameter in question. Thus, the behavior of the struc ture may be predicted for probable ranges of parameter values. This feature of computer use’ is important since the possible variation of some parameters may be rather wide and strongly affect the results. Computer coding and sofware applications for the solution of structures supporting dynamic loads are con sidered in this chapter. An analysis of a dynamically loaded structure is performed to obtain the following information: 1. Forces and deflections in members and joints for all static loading conditions. This will determine if the structure is statically safe or if deformations exceed tolerable limits. For structures supporting dynamic machines, the members are usually very large and massive and the stresses and deflections will be well within tolerable limits. This situation is a direct result of initial trial sizing of the struc- ture where the mass of the supporting structure is made several times the mass of the machine as described in the section on Trial Sizing of Elevated Foundations (Table Tops) of Chapter 3. 2. A dynamic analysis is also performed to determine the natural frequencies or eigenvalues of the struc- ture, the mode shapes or eigenvectors, and the dis- placements and member forces at a number of time intervals, ‘The dynamic analysis technique used in most com- puter programs js called a normal mode (or modal) technique which results in the calculation of the fre- quencies and mode shapes which in turn are used for the response calculations. The method is termed normal because the equations of motion (one per dynamic de- gree of freedom) are transformed to a new coordinate system called normal coordinates, resulting in uncoupled linear equations leading to a relatively efficient solution process. The primary purpose of the dynamic analysis is to ascertain possible resonance conditions, that is, to deter- if any of the structure natural frequencies coincide with the machine acting frequency or any of its critical speeds. A true dynamic analysis is sometimes replaced by a static analysis by using the Rayleigh method to calculate the lowest natural frequencies. The calcula- tion of the Raleigh frequencies is a very simple and in- expensive feature when used with a computer static analysis, and some designers will only perform a static computer analysis with Rayleigh natural frequencies calculation. However, only a complete dynamic analysis ‘will provide the necessary information for predicting the behavior of « structure supporting time-dependent loads. Example Problem An example of computer coding for the solution of an elevated foundation is given in the following pages. ‘This example has been selected to illustrate the use of the popular computer software package, STRUDL (Structural Design Language, part of the MIT-devel- ‘oped Integrated Civil Engineering System, ICES) ap- plied to the analysis of an elevated foundation (also called a table top) ‘The structure shown in Figure 7-1 is analyzed using the software package mentioned above. The structure trial dimensions are selected to meet certain preliminary criteria as described in Chapter 3 under Trial Sizing of Elevated Foundations (Table Tops), Computer Analysis and Applications: Example 115 Figure 7-1. Typical elevated pedestal foundation (table top). A. Machine Parameters “Total machine weight ~ 150,000 Ibs. ‘Turbine speed = 6,949 rpm or » = 727.7 rad/see Compressor speed = 6,949 rpm or » = 727.7 rad/sec Turbine rotor weight = 159 Ibs. Amplitude of turbine force = (W/g)eot 159/(82.2 X 12)] X (1.012,000/6,949/1,000) (727.7)8 86 lbs where ¢ is obtained from Table 3-1. Compressor rotor weight = 4,328 Ibs. Amplitude of compressor force = (W/) ea? abheny 22x20 ViR00076, 515/000) x 217)" 794 Ibs. B. Soil Parameters Shear modulus at the expected bearing pressure, see Chapter 4, Shear Modulus, G = 6,500 psi Poisson’s ratio, » = 0.45 Coefficient of subgrade reaction, ky = 120 Ibs./in.* Soil density y = 115 pef Allowable bearing capacity = 2,000 paf Predicted static settlement = 0.2 in. at 2,000 psf 6. Selection of Foundation Configuration Selection of a tial configuration‘ accomplished by following the guidelines deseribed under ial Sizing of Elevated Foundation (Table Tops) of Chapter 5 1. Machine and piping reqiirements dictate the plan arrengecent of the top of the Foundation, 2s show in Figure 7-1, 2. A iat foundation is recommended by the soi consaltant. The column spans are 12 ft and 8 fej thus the mat thicknes fat Feast, ¢=007 (10)"= 151% Try a B4ft mat, The thickness of the mat will also be atleast one tenth ofits largest dimension to asure rigid behavior. ‘The relative ste dimension, Equation (51), i i Ee b= laa ype ‘The modulus of elasticity of concrete is 3,122,000 psi, and its Poisson's ratio is 0.17; therefore, 3,122,000 x (36)*]>* T ) 120 = 101.0 in. = 8.42116 A flexible mat is one whose outside dimensions exceed 31, or 25.3 ft in each direction and is loaded over a small area, as described in Chapter 5. Therefore, a 24 ft by 30 ft by 3 ft mat may Design of Structures and Foundations for Vibrating Machines 12, Column resonance check shows no column reso- nance with the acting machine frequency (6,949 1pm). For example, for all the columns, P= 44.34 psi Computer Analysis and Applications: Example 117 Computer command, Explanation [cam + pe ot computer Program j be considered as rigid since the load is spread = 168:n, (clear height of columns) Tyre rae re : . out over most of the mat. A, = 4400 (3,000)*29/ 4434 TSB Se arse mom efinition of Structure 3.5. The crosesectional dimensions of columns and = 3,842 rpm ‘Thus, the trial design is judged satisfactory and the ‘dynamic analysis for the proposed configuration is ‘beams are selected according to these guidelines. 6. The ratio of mass of structure to mass of ma- chine is then performed. 428,000/150,000 = 2.82 ~3 OK. ‘The idealized computer model is shown in Figure 7-2 t 7. The mass of the top half of the structure is 0.8 where numbered joints have been located at member times the mas of the machine (the ratio should intersections and at other points of interest such a8 preferably exceed 1.0). loading points. The structure is idealized as a Model 8. The maximum static pressure is 6 type D described in Chapter 2. ‘The global coordinate (423,000 + 150,000) /(24 x 30) = 796 pst faxes are selected according to the right-hand rule with i $0.5 X 2,000 psf OX. axis Y being vertical; each member is also numbered f 9. The center of resistance of the soil is found to (numbers within cirele in Figure 7-2), and springs are coincide with the centroid of all superimposed placed at joints in contact with the scil in the vertical : loads (structure plus machine). and horizontal directions. ‘These springs represent the 10. The center of column resistance found as shown resistance that the supporting soil offers to displace- igure 3-2 coincides with the center of gravity ment, and the equivalent spring stiffness is calculated os y : of the equipment plus the top half of the struc- using soil properties as described in 4 below. inte “Stoe Static Loading Conditions ture. ‘A flow chart of the steps that occur during the com- [isto aarone | 11, Column and beam deflections are checked in puter analysis is given in Figure 7-3 and represents the computer analysis that follows typical analysis regardless of the software package being area Pianos bee Description of Structure Geometry Description of Structure Stiffness seson Elastic Constants Non Description of Dynamic Masses oe, of the Model Lumped at Joints Description of Masses Participating in Motion and Types of Motion Print Dynamic Input Data alii mad due to Dynamic Forces Calculate and List Mode Shapes LIST CNAME GSOEBPVEDTORO. Normalized to Unit Maximum Amplitude Termination ofthe Output Figure 7-3. Computer program flow chart. Static Analysis, Print Static Analysis Results Type of Dynamic Forces Acting Calculate and List Natural Frequencies of All Mode Shapes Figure 7-2. Computer model of elevated pedestal foundation.118 Design of Structures and Foundations for Vibrating Machines used. The chart lists the specific stages in an ICES- STRUDL analysis, and these are described in detail. Other software packages would include similar stages, but the exact commands and their order would be different. Example—STRUDL Coding STRUDL is a command structured language where the user can deseribe the structure in simple, almost conversational statements. The commands are given in logical order, that is, the geometry of the structure, the topology or connectivity, the member properties, the ‘material constants, and the loads are described prior to the analysis. A fixed format is not required in either the exact order of the commands or in a precise align- ment on a computer card. A summary of STRUDL. commands is given in Ap- pendix B, and the reader is referred to the ICES- STRUDL user's manuals for additional information (ref, 2 and 3). The steps in coding a problem for STRUDL solution are given below (lines in capital letters are actual commands-one per computer card) ‘These steps are noted in the printout on pages 121-157.* 1, Computer program and structure definition: STRUDL ‘EXAMPLE’ ‘STATIC AND. DYNAMIC ANALYSIS OF At TABLE TOP} ‘TYPE SPACE FRAME 2. Geometry of the structure (Figures 7-1 and 7-2) Units are defined since the default internal units are inches and pounds. Each joint is described by its num- ber and its x y 2 coordinates. Numbering of joints should be selected so that the difference in joint number at each end of all members is a minimum. The term SUPPORT identifies the joint as a support. Note that support joints have been identified under each column and midway across the slab edge and transverse equiv alent “beams.” UNITS FEET IPS JOINT COORDINATES i 00 09 00 SUPPORT 2 00 0.0 10.0 SUPPORT. 3 80 0.0 0.0 SUPPORT 4 80 0.0 100 SUPPORT. 15 80 170 50 16 80 170. 100 2 200 cr) 5.0 SUPPORI 3, Structure topology which indicates the connectivity of the members in the structure. Each command is of the form IJK and means that member I goes from the start joint J to the end joint K, The positive sense of the forces acting on the member follows the right-hand rule when the first (axial) axis is oriented from start to end of member. MEMBER INCIDENCES 1 1 9 2 2 ou 3 3 IF 37 4, Restraint conditions at the joints. Since a sup- ported joint is assumed rigidly supported (fixed), it is necessary to release these restraints and describe the stiffness (force per unit displacement) of springs at- tached to the joints. ‘The spring stiffness is a function of the supporting soil properties and the bearing area around the joint that acts against the soil for this rigid mat as described above during trial sizing of the struc- ture. Chapter 5 gives procedures for calculating the spring constants of rigid or flexible mats. For ex- ample, at joints 1 and 2, the contact area is (2.5 + 7) ft along Z times (5 + 4) ft along X = 85.5 sq ft. The total foundation contact area is 24 ft times 30 ft = 720 sq ft. The total foundation spring stiffness in the vertical direction (soe Chapter 4) is ky = GBs VBL 4e/(1—v) 6,500 144 X22 x V2E KIO _ ips /fe smn 4 22x 100,462 kips/ ke where the terms G, v, and me (equal to unity in this ‘example) are defined in Chapter 4. Therefore, for nodes 1 and 2, kya = ky = 100,462 X 85.5/720= 11,980 kips/ ft. Note that the ¥ direction in the computer example of Figure 7-2 is the vertical soil direction previously de- noted as the Z-direction in the soilspring constant equations of Chapter 4, At the risk of some confusion, the Y-direction is selected vertical in the computer anal- ysis. due to certain global-local axes advantages. Alo, rotational restraints are assumed to be non-existent since they are generally negligible. Further discussion of the calculation of the spring constants in the case of larger flexible mats is presented in Chapter 5. * The symbol + denotes continuation of the previous line on & Single card, UNITS KIPS FEET JOINT RELEASES 1,2 MOMENT X YZ KFX 8649. KFY 11930, KFZ 8649.4 25 MOMENT X ¥ ZKFX 4046. KFY 5581 KFZ 4046.t The third command listed above means that joints 1 and 2 have X, Y, and Z rotational freedom and linear springs with 8,649 kips/it in the global X- and Z-direc- tions (horizontal) and 11,930 kips/ft in the Y-global direction (vertical). The horizontal springs are placed at exterior joints although equivalent horizontal springs could have also been distributed among all joints in contact with the soil. No significant difference would be detected in the results 5. Gross-sectional properties for the member, The cross-sectional area and the moments of inertia for each member about its own local coordinate axis are given. The local x-axis is always directed along the member while the local z-axis is parallel to the global Z-axis for columns and is horizontal for beams, provided that the global Y-axis is set vertical. For further details on Jocal-global axes relationships, the reader is referred (o the MIT STRUDL User's Manual, Volume I (ref. 2). Modifications are necessary if the local Z- and ¥- axes are not parallel to the global axes. The areas AY and AZ are disregarded and the analysis will, therefore, assume the members to have no shear stiffness, For frames consisting of relatively long and shallow mem- bers, no significant difference can be detected by neglect ing the shear stifiness of the members. MEMBER PROPERTIES 1,2 PRIS AX 60 IX 4.7 IY 45 1Z 2.0 35, 36, 37 PRIS AX 125 IX 17.9 IY 26. 1Z 65 ‘The second command means that members 1 and 2 fre prismatic, have crosssectional areas of 6 sq ft, torsional moments of inertia of 4.7 ft’, moments of inertia around the local y-axis of 4.5 ft, and moments of inertia of 2 ft" around the local z-axis. Note that far these 3 ft x 2 ft columns the local z-axis is parallel to the global Z-axis and, therefore, x B/12=20 i, 6. Material Properties. The material properties in- clude the modulus of elasticity 2, the shear modulus of elasticity G, Poisson's ratio, and the material density, ‘These constants are 3,122 ksi, 1,384 ksi, 0.17, and 0.0868, "osJeu in., respectively. Density is specified in IbsJew in. ‘Computer Analysis and Applications: Example 149. for convenience; however, some STRUDL packages teat density as mass density, and units and magnitude cof mass density must be entered, UNITS KIPS INCHES CONSTANTS E 3122, ALL c 1334, ALL POISSON 017 ALL UNITS POUNDS INCHES CONSTANTS DENSITY 0.0868 ALL 7. Static Loadings, A number of basic loading condi- tions and combinations of the basic loadings are con- sidered. UNITS IPS FEET LOADING 1 ‘FULL LOAD ACTING IN THE VERTICAL DIRECTION} JOINT LOADS 42 FORCE Y — 384 25 FORCE ¥ -150 (other loadings and combinations follow) RAYLEIGH LOADING 3 ‘FULL LOAD ACTING. IN THE TRANSVERSE DIRECTION JOINT LOADS . 1,2 FORCE 384 25 FORCE Z 150 A loading for which the RAYLEIGH frequency is to be calculated is denoted as shown above for loading 3. This loading includes the weights of the structure adjacent to each joint and the weight of the machine applied in the direction of the desired vibration mode, in this case, the transverse of Z-global direction. Thus, the analysis will include a Rayleigh-Ritz calculation of the natural frequency for the structure in the transverse direction. Loading combinations as follows would be ‘added by some designers in a static check as discussed previously. These loadings are quasi-static, that is, con- servative equivalent static loadings which may be used for design check of deflections and member forces ‘Total vertical load + 03 X total load acting in the transverse direction (Z) ‘Total vertical load + 0.1 X total load acting in the longitudinal direction (X) ‘Total vertical load + 0.5 X total load acting in the vertical direction (Y)‘These are denoted Loading Combinations 4 through 6, respectively, in the computer printout. LOADING COMBINATION 4 ‘FULL VERTICAL LOAD PLUS 0.3t FULL TRANSVERSE LOAD’ —+ COMBINE 1 1.0 3.03 LOADING COMBINATION 6 ‘FULL VERTICAL LOAD PLUS 05+ FULL VERTICAL LOAD’ —+ COMBINE 1 1.0 105 ‘The first command specifies that loading combination 4 consists of loading | times 1 plus loading 3 times 0.3, 8. Listing of all data. A printout of all internal data is requested with the command PRINT DATA ALL. 9. Geometry plotting. Structure geometry plots are requested as a further check of the input data. 10, Static Analysis. The following command is used to build and invert the structural stiffness matrix and to solve the problem for all loadings. STIFFNESS ANALYSIS (REDUCE BAND ROOT) ‘The command within parentheses is optional and gen- erally results in a more efficent algorithm for large problems. 11. Output of results, Results are printed by using the LIST commands, such as UNITS KIPS INCHES CYCLES SECONDS LIST RAYLEIGH OUTPUT BY MEMBER, LOADING LIST 4, 5, 6 LIST FORGES DISPLACEMENTS REACTIONS ALLt Results for loadings 4 through 6 are requested. This step completes the static analysis. The dynamic analysis includes the following additional steps: 12, Mass acting at each joint. The structure's mass and the machine mass are taken to act at the structural joints in the three linear directions only. Inclusion of rotational inertia has a negligible effect on the results, ‘The structure mass may be computed internally and automatically lumped at cach joint and the machine ‘mass is then added at the machine support joints. UNITS POUNDS _ INCHES INERTIA OF JOINTS LUMPED INERTIA OF JOINT ADD 9, 11 LINEAR ALL 69.88t INERTIA OF JOINT ADD 15 LINEAR ALL 117.754 n of Structures and Foundations for Vibrating Machines INERTIA OF JOINT ADD 17, 18 LINEAR ALL, 28.47t INERTIA OF JOINT ADD 20 LINEAR ALL 73.76 Note that the added machine mass at joints 9 and 11 is 27,000/ (32.2 x 12) = 69.88 Ibs-sec/in., and simi- latly for joints 15, 17, 18, and 20, see Figures 7-1 and 7-2. The structural mass is included through the INERTIA OF JOINTS LUMPED command. 13. Dynamic degrees of freedom-—only translation modes are considered, The commands that accomplish this are DYNAMIC DEGREES OF FREEDOM JOINT 1 TO 25 DISPLACEMENT X YZ 14. Damping ratio, The damping ratio for each de- _gree of freedom is given by DAMPING RATIO 0.10. 75 Where total average damping of 0.10 has been speci- fied for the X, ¥, Z translatory degrees of freedom at each of the 25 joints, or 3.x 25 = 75 times. Note that, realistically, those degrees of freedom associated with foundation movement have a damping ratio in the 0.15-0.20 range, whereas damping for the remaining degrees of freeciom may only be in the 0.05-0.10 range. ‘Therefore, 0.10 is used throughout as an average value. Chapter § gives a further discussion on the choice of soil damping 1 15, Dynamic forcing function. Forcing functions in the vertical (Y-direction) and in the transverse (Z- direction) ate applied at the joints where they occur. "These forcing functions include a force amplitude equal to the unbalanced machine force, a frequency given by the acting machine frequency (in radians) and a phase angle of 1.5707 radians (90°) for the transverse func- tions, ie., the transverse function is 90° out of phase to the vertical function in this centrifugal machine. The dynamic forcing functions are applied at the centerline of the centrifugal machine shaft, joints 10, 15, and 20. ‘One half of the turbine force acts at joint 10 or 0.5 X 286 sin 727.7t, one half of the turbine and compressor forces act at joint 15 or 0.5 X 286 sin 727.78 +05 X 7,794 sin 727.7t, and one half of the compressor force acts at joint 20 or 0.5 X 7,794 sin 727.78 UNITS RADIANS SECONDS POUNDS. INCHESt DYNAMIC LOADING 7 ‘CENTRIFUGAL FORCES't JOINT 10 LOAD FORCE Y FUN SIN AMPL 143. FREQ 727.7t f JOINT 10 LOAD FORCE Z FUN SIN AMPL 143. FREQ 727.7 PHASE 1.57071 JOINT 15 LOAD FORCE Y FUN SIN AMPL 4040. FREQ. 727.7+ JOINT 15 LOAD FORCE Z FUN SIN AMPL. 4010. FREQ 727.7 PHASE 1.5707+ JOINT 20 LOAD FORCE Y FUN SIN AMPL 3897. FREQ 727.7t JOINT 20 LOAD FORCE Z FUN SIN AMPL 3897. FREQ 727.7 PHASE 1.5707+ 16. Time Periods. The time span and time incre- ments for the dynamic analysis must be specified. The integration time periods should include, as a minimum, 12 steps per single complete operating frequency cycle, in order to achieve a +59 accuracy in the results (ref. 4), that is, for a frequency of 727.7 radians/sec, the integration time periods should not be greater than T (one cycle) = 2x/727.7 = 0.00863 sec; then, ‘At = 1/12 = 0.00863/12 = 0.000794 see ‘The following command includes 10 complete cycles of ‘machine operation with 12 steps in each cycle, However, 3 complete cycles of operation may be sufficient to study. the response of the structure, INTEGRATE FROM 00 to 0.0863 AT 0.000719 Computer Printout in ICES-STRUDL Computer Analysis and Applications: Example 121 17. Listing of dynamic data, A printout of all dy- namic data is obtained with the command PRINT DYNAMIC DATA ALL 18. Dynamic analysis. The actual dynamic analysis is obtained with the following command with the part within parentheses being optional. Only the first 20 ‘modes are included in the analysis to reduce computing time with negligible loss of accuracy. DYNAMIC ANALYSIS MODAL (REDUCE BAND ROOT) 20t 19. Output of dynamic analysis. The natural. fre- quencies, the modes (normalized), the displacements, and forces for each time increment are requested with this command. The first 20 modes are requested to con- serve paper but more may be printed, UNITS KIPS INCHES CYCLES SECONDS LIST DYNAMIC EIGENVALUES 20 NORMALIZE EIGENVECTORS LIST DYNAMIC EIGENVECTORS 20 LIST DYNAMIC DISPLACEMENTS ALL LIST DYNAMIC FORCES ALL 20. End of analysis. The last command in the job is FINISH (text continued on page 157) ‘STRUDL FECAMPLEY «STATIC AND OMANI ANALYSIS OF A TABLE=TOP* @ Sassachosents NE emorbots matsuoune rss - : Wenn Somes tereaeegaaas Beabbeoes srogestee Design of Structures and Foundations for Vibrating Machines 029) celahterer S909 220038391 sroge8 erenstsne ‘Computer Analysis and Applications: Example 157 ovoa1e7ae neno7s9—n77enss9 Srogeeats Ttonzase Nena? Interpretation Of Results Computer output results include the following: Member Forces—Static Loads. Axial forces, shear forces, torques, and moments at each member end are tabulated for the design loading conditions, loading 4 (full vertical load plus 0.3 full transverse load), loading 5 (full vertical load plus 0.1 full longitudinal load), and loading 6 (full vertical load plus 0.5 full vertical load). ‘These approximate equivalent static loading con- itions yield conservative estimates of the maximum dynamic plus gravity loads, forces, and deflections as escribed previously. ‘The members and deflections are then checked for the maximum applied loads. In the usual case, this step includes selection of the longitudinal reinforcement, which is often the minimum code value and which is otherwise determined by the largest axial force and moment. Transverse reinforcement is also selected during this stage using the tabulated maximum shear and torque values. Displacements—Static Loads. Displacements of the joints, both support joints as well a8 free joints, are tabulated. Certain tolerable Timits on deflections may be established based on attached piping or other e ‘ment, and the deflection values are checked against the tolerable limits. Note that the tabulated values include the effect of gravity loads. The incremental deflections ue to the 0.3, 0.5, and 0.1 equivalent static loads over and above the deflections due to gravity only are very { small and, therefore, negligible in this example problem, Dynamic Eigenvalues—Natural Frequencies, The inst 20 natural frequencies (out of a total of 75 arising from 25 joints with 3 translatory dynamic degrees of freedom cach) are printed. Each frequency corresponds to a mode shape discussed below. The first frequency (transverse mode) is given as 3.682 cycles/sec, section 19 of the computer printout, which compares to a Rayleigh-calculated frequency of 3.985 eycles/sec, sec- tion 11 of the computer printout. In general, Rayleigh- obtained natural frequencies are close to but higher than the corresponding true natural frequency as ob- tained from a dynamic analysis. Since the acting ma- chine frequency is 727.7 radians/sec (see 15 above), which equals 6,949 rpm or 115.8 cycles/sec, the first 20 natural frequencies for the structure are well below the acting frequency, and the structure is said to be low tuned (undertuned). The undesirable range of natural frequencies is between 0.8-1.2 of the acting machine frequency as given in item 2(f) of the checklist table in Chapter 3. This check assures that no resonance con- ition will be encountered during machine operation, Additional checks are performed, not necessarily during the computer analysis phase, but during the preliminary design to assure that no resonance condition exists be- tween the natural frequencies of vibration of individual columns and beams with the acting machine frequency, as in item 12 of the trial sizing procedure above, Dynamic Eigenvectors—Mode Shapes. The first 20 eigenvectors or mode shapes which are normalized to a maximum unit value are listed. These serve to identify the physical direction for each mode. For example, for Mode 1 joints 9 through 21 show a near-unity Z displacement. Therefore, the first mode occurs in the transverse Zdirection. This is as expected since the structural stiffness is the lowest in the transverse direc- tion. In general, the modes are ordered according to the stiffness of the structure in each direction. Other listed values give the deformation of the structure and a plot of the structure vibrating at its first frequency158 Figure 7-4. Structure vibrating at the first fre- ‘quency mode (transverse). mode may be obtained either manually or using a builtin STRUDL plot package; see Figure 7-4. The second mode shows joints 9 through 21 with near unity longitudinal displacements; thus, the second mode occurs in the longitudinal X-direction. In particular, all modes ‘occurring in the direction of the applied dynamic forces (transverse and vertical in this example) are important for response studies. Some of the higher modes are coupled since they result from combination of basic lower modes. Dynamic Displacements. Displacement of all joints in the structure at each time period are listed. The upper joints show the largest displacements. For example, joint 20 has a maximum displacement of 0.0002851 in, (0.007 mm) in the Y-direction occurring at time equal to 0100576 sec. This maximum amplitude of displace- ‘ment is then located on Figure 33 at the given machine- acting frequency. ‘The point falls in zone B (minor faults, correction wastes dollars) which indicates satisfactory performance. Figures 3-6 and 3-7 also indicate that for a maximum displacement of 0.007 mm and a machine speed of 6,949 rpm (115.8 cps), the predicted structure behavior is satisfactory. The amplitude of vibration at all other joints, being smaller, is also satisfactory. Note that many structures supporting centrifugal machines show a largest amplitude of vibration in the transverse Design of Structures and Foundations for Vibrating Machines direction, However, in this example, the representation ‘of the soil support as springs results in the amplitude of vibration being the largest in the vertical ¥-direction. Maximum Velocity. The maximum velocity is (6,949 x 2» X 0.002851) /60 = 0.207 in./see ‘This maximum velocity falls in the “slightly rough” range of Table 3-2. However, the combination of velocity and machine speed fall within the acceptable zone B of Figure 3-3, The designer may consider increas- ing the base dimensions and ze-analyzing the revised structure so that all design criteria are met. Dynamic Forces. The dynamic forces acting at the ends of each member at each time increment are listed. The forces are very small, for example, the dynamic axial force is 1,132 Ibs. in column member 5 at 0.0237 see. The moments are likewise very small, and the struc- ture is considered adequate for supporting these small ‘dynamic loads ‘Since the structure was dimensioned initially to meet the design requirements listed in Chapter 3, including providing a sufficiently large soil bearing area, the pro- posed design is satisfactory and the structure meets its intended purpose. Note that all items listed in the design checklist of Chapter 3 are explicitly considered during the initial tial sizing phase or implicitly considered dur- ing the computer analysis. The design checklist serves fas a reminder of important factors to be considered during a step-by-step hand calculation, References 1. Arya, S. C., Drewyer, R. P., and Pincus, G., “Mathe- ‘matical Modeling and Computer Simulation of Ele- vated Foundations Supporting Vibrating Machinery,” Mathematics and Computers in Simulation, Vol. XIX, No. 4, Amsterdam: North Holland Publishing Co., December 1977, pp. 262-271. 2. ICES STRUDL I Engineering User's Manual 6-91, Volume 1: Frame Analysis, School of Civil Engineering, Massachusetts Institute of Technology, November 1969. 3. ICES STRUDL Il Engineering User's Manval 70-77, Volume 2: Additional Design and Analysis Facilities, School of Civil Engineering, Massachusetts Institute of Technology, June 1971 4. Mes, M. J., “Maximum Time Step for Dynamic “Analysis,” Ocean Engineering, pp. 40-58, November 1976. Appendix A Solution of Multi-Degree-of- Freedom System Introduction ‘A mullidegree-of-freedom system may be defined as a system in which more than one independent motion is posible. These independent motions may either be associated with a single mass, or a single independent motion may be asociated with each of the several masses in a system. In the former type of system, the motion of the mass may either be coupled or uncoupled. ‘The coupled motions of a single mass are described by the of motion where determination of the response (vibration displacements) of the system involves the solution of a set of simultaneous equations. An example of this type of system is shown in Figures 1-49 and 1-46. ‘The response values x and @ (associated with a single mass) appear in both equations, and the pair is said to be coupled. The characteristic of uncoupled motions is described by the model of Figure 1-47. The response Yalus snd @ appear individually in the equations of ‘motion, and each of those equations can be independent oto ea be independently ‘An example of a multidegree-of-feedom sjsterm having several masses associated with one type of independent ‘motion is shown in Figures 1-50 and 1-51. The inde- Pendent motions x, and se in the vertical direction are associated with masses m and mg, respectively. ‘These ‘ypes of systems always undergo coupled motions, and that effect is due to the coupling of masses m, and ma through the spring f,. Thus, the two equations of motion for this system have to be solved simultaneously since al sy appetr in both of them. Note that it not ‘necessary for each mass to have only one type of motion. 1 the masses ave capable of osillatng in the horizontal rection, then each of the masees m, and mg will have {wo types of motions, ie, vertical and horizontal mo- tion. In fact a two-dimensional planar system has up to 159 three degrees of freedom (two linear and one rotational direction) per mass. For a spatial system, each mass can oscillate in six directions (three rectilinear and three rotational directions) and has six degrees of freedom, Therefore, the number of degrees of freedom is not necessarily equal to the number of lumped masses. Dynamic Analysis A complete dynamic analysis of a system is normally performed in two stages. The initial stage of investiga- tion involves the determination of the natural frequen- cies and the mode shapes of the system, The natural frequencies and mode shapes provide information about the dynamic characteristic of the system. For instance, the lowest value of the natural frequencies (generally called fundamental frequency) indicates the relative degree of stiffness built into the system. In addition, it is also possible to compare the lowest natural frequency with the frequency of the acting dynamic force so that 2 possible resonance condition may be prevented. This requirement can be achieved when the ratio of the operating frequency to the lowest natural frequency does not fall within a given range (normally the undesir- able range is at least 0.8-1.2). The determination of the mode shapes in a multidegree system has further signifi- cance. The mode shapes (see definition of first (funda- mental) mode of the Terminology section in chapter 1) gives the deflection pattern that the system assumes when it is left to vibrate after termination of the disturbing force. Generally, it is the first mode which dominates the vibrating shape, and the higher mode overrides (when superimposed) that shape. The first mode will also indicate the particular mass or masses which have the maximum amplitude of oscillation in a given160 Design of Structures and Foundations for Vibrating Machines irection from their state of rest (static equilibrium). ‘This serves to indicate the relative degree of structural stiffness among various points of the system. This examination of the mode shapes in the vibrating system is considered a valuable step in adjusting the vibration amplitudes at critical points by varying the stiffness, mass, and damping resistance of the system. A practical example can be observed in the operation of a washing machine during the rinse cycle. If the load is acciden- tally lumped to one side of the dram during centrifugal motion, severe vibration beyond a predetermined magni tude can occur. Special sensor cutoff switches are then activated in order to prevent damage to the machine. ‘Activation of the contact switches is actually set accord ing to the fundamental mode of the spindle-drum assembly. The next stage of analysis is a response calculation of the system caused by the dynamic force. This solution is quite tedious and time-consuming for multidegree-of- freedom systems; however, for a system with three degrees of freedom, the response calculation can be accomplished by using hand calculators. Use of com- puter programs is recommended for systems with more than three degrees of freedom. This part of the analysis gives the displacement, velocity, and acceleration of the ‘masses and also the internal forces in all members of the system. Determination of Natural Frequencies ‘and Mode Shapes. Natural frequencies and mode shapes are obtained by employing one of following methods. A three-degrees-of- freedom system shown as Model 6-C (part a) of Chap- ter 2 (Figure 2-9) is used as an example. The parame- ters for that model are 9.662 kips-sect/ft 160.2 kips/fe 7.80 kips-sect/ft 2,843.3 kips-sec?-ft 72,835.0 kips/ft 4,703,707 kipsft/rad D.=050 Dy=0.20 H= 170K Determinant Equation Method All natural frequencies and mode shapes are obtained at the same time in this method. The general form of the equations of motion for a three-degree-of-freedom system is given by (ref. 3, Chapter 1) anata + masks + mais + cake + cris st ewis + haunt + hate + hate = Fi mais + mais + mais + cain + cnte tenia + daass + hata + ks = Fe () al) masa + maa + mais + casks + ein peawis + havts + heats + hats = Fa ‘The natural frequencies depend on the mass (mej) and stiffness (kis) terms; therefore, the damping (cs)) and applied forces (F,) terms are omitted, Equations (Al-1) are then reduced to mati masts + maa +b hats + hte + hints = 0 matt + masks + mania sb hss + haste + bra (a1-2) + mais + mais ot havts + heats + hats = 0 {tis assumed that the free-vibration motion of masses is simple harmonic (see definition of modes in the Terminology section of Chapter 1), which is expressed for a multidegrees-of-freedom system as xc = Agsin (at + 9) (als) 123 where i Substituting Equation (A1-3) into Equations (Al-2) land omitting the sine term, the following set of equa tions is obtained: (bu — mus") Ar + (haa = mu!) de + (ka — mut) As (has — mas) As + (oa — mane!) An + (be — mast) dy = 0 oe (hor — mas) Ar + (be maia") As + (ka — mas?) Ar = 0 Equations (Al-#) are a set of algebraic equations and have a nontrivial solution for Ay only if the detet~ minant [hag mye | = 0 5 tous, (ka—mue*) (ba—mnw") (ke —mya?) A) =| amas") (ba—mas") (ka—myge") | = 0 (x-mas) (ba mys") (nme) (ats) ‘The expansion of determinant (A1-5) gives the char- acteristic equation of the system. This equation will be of third degree in the frequency parameter («*) and has three roots representing the three basic frequencies of the system, Having determined the three natural fre- ‘quencies, the mode shapes are obtained by making use of Equations (Ai-4). For each of the three values of the ratios of (4s/A:), (4s/4:), and (4y/A;) are evaluated and yield the three mode shapes forthe sytem. Example. For part (a) of Model 6-C, the equations of motion without damping and applied forces terms (@) mils + (be sb tas — hts + lp = ) (8) mia — bss + hse — bbl = 0 (ais) © mab + 16 + ky = 0 J Substituting the assumed solution of the form (0) = Asin (ot +4) () a = asin (wt +4) (a7) Ov = Ay sin (wt +8) into Equations (A1-6), (0) (he + hy~ mu") Ay — bad + byl Ay = 0 (0) ~ bya + (by — mus) — hy Ay = 0) (A1-8) (© ~ mately + (ba ~ Lud )Ay The frequencies of the system are given by the condi- tion 4 = 0, where A is the determinant of the square matrix in Equation (A1-8). The expansion for a deter- ‘inant Solution of Multi-Degree-of-Freedom System 161 a nae Fagl=) an ae an (aL-9) ban ae as = cu (ain0s ~ arate) ++ am (onan — onan) + au (ana — arson) In order to reduce the amount of computation, the values of aj; are calculated by substituting the actual values of masses and stiffness. an = (ket hy — myst) = (78,995.2 — 17.800") an = hy = —6,160.2 ay = byl = 6,160.2 X 17.0 = 104,723.4 an = ky = —6,160.2 a = (by ma*) = (6,160.2 — 9,662.4) ag = hy = ~6,160.2 X 17.0 = —104,723.4 an =0 am = —mHa = — 164.254" ax = (ky = Lya") = (14,703,707 — 2,843.3 0") Substituting these values in the expansion of Equa- tion (AI-9) and setting « (78,995.2 — 17.802){9.0577775 x 10” = 17515296 x 10% — 142.06721 x 10% + 27,471 9642 — 17,201237 x 10%] 6,160.2§9.0577775 x 10° — 17.515296 10%] + 104,723.4{1.011837 x 10° = 0 (At-10) Rearranging Equation (A1-10) results in a cubie equa tion in 2, 2 — 10,872.9922" + 31,653102 x 10% — 1,3491454 x 10" = (arin Solution of the cubic equation yields three roots: (a) = 5120 (b) = 44810 (atta) (0) .= 5.8800 ‘The square root of the 2's is then the natural frequen cies (in rad/sec) of the system,162 (2) oy = 22.63, (Ho = 66.98 (al-i3) © & = 76.68 ‘The mode shapes are obtained by substituting the above frequency values in any (AI-8) equation. For Equation A1-8(a) and (b) using the given values of masses and spring stiffness, (@) (78,995.2 — 17.8004) Arg — 6,160.2405 + 104,723-445 = 0 (alg) (©) ~6,160.24, + (6,160.2 — 9.66205) Ay ~ 104,723.445_ = 0 In the equations above, the subscript m takes the values of 1, 2, and 3 and gives the amplitudes 4’s for each of the three modes. Since the right sides of Equa tions (Al-14) are zero, unique values of the 4’s are not obtained. However, it is possible to obtain the relative values of all amplitudes, in other words, the ratio of any two amplitudes. When one amplitude is assigned an arbitrary value, then all others are fixed in magnitude. ‘A set of such amplitudes defines the mode shape, and therefore, the modes shapes are not dependent upon the absolute true values of amplitude. Using = 1 and sub- stituting «,? = 512.0 in Equations (A1-14), (© 69,881.64n — 6,160.24n + 104,728.4451 = 0 { (ats) (8) ~ 6,160.24u + 1,213.34 — 104,728.444. = 0 Assuming an arbitrary value of Ay, = +1 and solving Equations (A1-15) simultaneously for Az. and Ayy, the amplitudes of the first mode are obtained An = +1, An = +12.8809, Ay = +0.0904 The notation adopted is that the first subscript of the A identifies the mass, or point on the structure at which the amplitude occurs, and the second subscript desig- nates the mode. Using the value of n = 2, and substitut- ing o,? = 4,481.0 in Equations (Al-14) yields, Design of Structures and Foundations for Vibrating Machines (0) 766.6041 ~ 6,160.20 + 104,723.44). = 0 (ALi6) (®) ~ 6,160.24y — 37,135.240 — 104,723.44), = 0 Again setting Ai = +1.0 and solving Equations (a) and (b) of (A1-16) simultaneously for Azz and Ay, the amplitudes of the second mode are obtained: An = $1.0, dn = 0.16, Aye = —0.002091 Similarly, the amplitude for the third mode can be ‘obtained by following the procedures given above for the first two modes, Those amplitudes using w:* = 5,880.0 Ay = +10, dy = 0.560248, Aya = +0.212150 The computation of natural frequencies and mode shapes were performed using small electronic hand cal- culators with eight significant digits. However, in order to solve the cubic equation (Bquation (Al-11)), 15 sig- nificant digits are required to obtain more “exact” resuls ‘This type of accuracy is normally not available in small hand calculators and if sufficient. number of significant digits are not retained in the calculation of frequency, then large errors are present in the mode shape ampli- tude results. Therefore, itis important that the values fof mode shape amplitudes be checked using the octho- gonality conditions. The orthogonality condition of nor- ‘mal modes (see definition in the Terminology section of Chapter 1) is expressed by the following equation (ref. 1, Chapter 2) EM, An Ary = 0 (a7) where J and n identify any two normal modes of the 5) tem, and the subscript r refers to the rth mass out of total of m masses Expanding the series of Equation (A1-17) for the first and second mode, ie,,= Land n = 2; then Miduds + MaAndn + MrAndsr (al-18) Substituting an appropriate value for each term, 17.80, My = ma = 9.662, 2,848.3 +10, dy = +10, dn = +12.8809, =0.1600, du = Ay, = +0.0904, = —0,002091 on the left side of equation 17.80 X (1) (1) + 9.662(12.8809) (0.16) + 2,843.3(+-0.0904) (—-0.002091) = —2,650298, which is not zero and thus indicates errors in the results calculated above. Similarly, the orthogonality condition for the first and third modes (I= 1, n = 3) is checked by expanding Equation (A1-17) into a serie Mrdudn + MaAndn + Mya (Al-19) Substituting yy = +10, 4z, = ~0.560248, Ays = +0.212150, and the values of other terms which are given above into the left side of Equation (A1-19): 17.80 x (1) (1) + 9.662(12.8808) (—0.560248) “+ 2,848.3(-+0,0904) (+-0.212150) = +2.6040, Which is again not zero, thus confirming that errors are present in the frequency calculations. ‘Therefore, it is desirable that natural frequencies be calculated using digital computers for a system having more than two degrees of freedom, Stodola-Vianello Method Calculation ofthe natural frequencies and characteris- tic shapes become cumbersome even with hand caleu- Tators for cases where the degrees of freedom exceed more than two, as is evident from the above example. ‘The reason is that round-off errors are relatively impor. tant when the equations include terms of very large and very small numbers. Accuracy is lost in rounding off Scant figres during the process of solving fr the roots of the characteristic equation, It, therefore, has become common practice to resort to numerical, iterative (ie, tial and error) procedures, such as the Stodola- ~ Vianello method (ref. 1, Chapter 2). The various steps in the solution are as follows (2) Assume a characteristic shape, icy a set of 4y _ values (ee Equation (Al-4)); (2) using one of Equa: {fons (Alt), efor fy 9) wing the vomating (N-1) equations, obtain a new shape by solving for the (Nl) (sin terms of the Neh 4; and (4) use the new Computed shape as the revised assumed shape in the Solution of Multi-Degree-of-Freedom System 163. next cycle starting at (2) again, The procedure is to be repeated until the computed shape is the same or very close to the last assumed shape. In Step (1) it is usually convenient to asign a unit value to the first value of 4. ‘The procedure outlined above ean best be described by applying the method to the solution of Equa- tions (A1-8) ‘The equations of motion (A1-8) can be rewritten as (©) ohms + (be + i) Ay — fad + by, = 0 ©) ~chinuAs — bad “+ nds — lnlly = 0 ) © ~8'hyAy — sinlde they = Multiplying equation (A1-20b) by Hand adding to Equation (A1-20c) in order to eliminate the term aimyHAs from Equation (Al-20c), the resulting equa- tions are rearranged in a convenient form: he Oe dan = — Bde + By (alan) © wi Aye = Where the added subscript » indicates that the equations apply to any mode. Substituting the values of mi, ma, 1s key, Fy, and H in the Equations (A1-21), (0) oh Aig = 4,497.98 4ay — 346.08 406 + 5,888.34 Ay (8) oi Aug = ~637.57ag + 637.57 Arg — 10,838.69 4, (A1-22) © Age = 36.83Aig — 36.8845 + 5,797.49 Aon164 ‘The iteration procedure converges on that of the highest, or third mode. The following steps are used 1. Assume values of 4’s (amplitudes of the mode) such that the amplitude of mass 1 = +1.0, ie, Aww = +10; 2. Substitute the values of ’s in the right side of Equation (A1-22a) to compute the value of wy" 3. Substitute the value of oy? and trial values of A's in the right side of Equation (A1-22b) to compute the value of Aen; 4 Substitute the value of wx? and the trial values of A’sin the right side of Equation (A1-22c) to obtain the value of 4px; Use the new values of zm gy along with the value of Av = +1.0 as new trial values of the 4’s and follow Steps 2 to 4. This process is to be continued until con- vergence is achieved, i, the difference between previous and new trial values is negligible, For example, the amplitude values for mode shape three, ie., n = 3 Ay = +10, de = 15, Ay = +05, ‘Substituting these in Equation (Al-22a), BED = 449799410) ~ 546.08(—1.) + 5,883.34(+-0.5) 2, @ = 7,898.72 Substituting this value for ws? and the trial values of 4's in the right side of Equation (Al-22b), 7/898.72(dy) = ~687.57(++1.0) + 637.57(—1.5) — 10,838.69(-+0.5) ot day = —0.8879 Finally, from Equation (A1-22c), 7,898.72(44s) = 36.83(+1.0) — 36.83(—-1.5) = §,797.49(+0.5) or dee = +0.3786 ‘Therefore, the first estimate of as* is 7,898.72, and the next set of trial values of ’s are + 1.0, -0.8879, and +0.3786. This procedure is to be repeated from Step 2 to 4 until convergence is achieved, as shown in Table Al-. The value of 4’s computed are in fact the amplitude ratios with respect to Ays which is arbitrarily taken = +1.0, Proceeding further to the second mode, the ortho- sgonality conditions given by Equation (AI-17) are used Design of Structures and Foundations for Vibrating Machines in order to reduce the number of equations by one. Expanding the equation for the second and third mode, ie,1=2and n= 3 E MiAndmn MsAusda + Midaades + Madrsdss = 0 Substituting My =m; = 17.80, Mz = my = 9.662, 2,843.3, dy = +1.0, dn = —0.6011, +0.2412, in the series. 17.804) (41.0) + 9.662(4s2)(—0.6011) + 2,843.3(4n)(+0.2412) = 0 (or Axe = 0.025955 Ave + 0.0084694n, (Al-23) which is equal to Ags. Substituting the expression for Ags into Equations (a) and (b) of (A1-2) and using n = 2, yields (@) od Aus = 4,285.2384y — 296.2540, ¥en (0) 6 An = —356.254m + 545.7440 Equation (A1-24) is iterated in a similar fashion as was done for Equation (A1-22) in order to find o:%, Aug and Azz. Then, Equation (A1-23) is used to obtain Ags. Asa fist trial, assume Aye = +1.0 and Azy = 10.5. Using Equation (A1-24a), 3 (41) = 4285.28(+1.0) — 296.25(+0.5) or a = 4,494.36 Using Equation (A1-24b), 4434.36 (du) or dn = —356.25(+1.0) + 545.74(+0.5) 1419 Subsequent cycles of iterations are given in Table Al-1, where it may be seen that the values converge rapidly in four cycles. In the fourth cycle, Equation (A1-23) is used to obtain Ags ‘The determination of the frst mode is made directly {rom the orthogonality conditions by applying the condi tions to the first and the second modes, and then, to the first and the third modes. Expanding Equation (A1-17), Solution of Multi-Degree-of-Freedom System 165 Table A1-1 ‘Stodola-Vianello Procedure for Model Shown in Figure 2.9, ‘Third Mode Trial Valu Computed Values os Ay Teal Equation uation Now An Ate diez) Gr p +10 Fos 782 “F050 z +10 Fora 65 ogee i #0 yous ie “Fo. ‘ $0 foo oat +ozr00 5 #0 +o.zr00 aot e4 oper 8 ro Fozoer eae fost : io 03501 Sima foam & tio fo2s07 Sigrtt Fov3t0g ° to Fo2a00 S102 33 Fopes 0 $15 Fo2ue Sos Fope " $10 Fons sora Foe Second Mode “Trial Values 1 Computed Value oe An Au eat Equation tention Equation Now Au Gis ab) ar T +10 4.90 2 +10 Suara a +0 Sai4t t +10 iin 0.02070 Summary e As 4s Ay (eadtanafoce) __F(600) Tiled Mode Sora 10 ~oson Faas Tra5 240 Second Mode Soin ro = ome = Soza tes 1045 Fist Mode "15.88 tio wasn FRI Bn 3a 3 Assuming As; = +10, these two equations are solved (@) 2 MyAnda = Midudss + Mendes simultaneously to provide i + Midd (ares) (0) 2 Mada = Miduds + Mederde j Ax = 12.6098, dy = Ayr = 0.08151 “ + Mrdadss Substituting the values of Ms 9.662, My = I, = 2,843.3, du = +10, dn = 0.0946, dus = Age = —0002676, dy = +1.0, Aa 0.6011, dis = Aix = 0.2412, Equations (Al-25a) and (b) yield my = 17.80, My = my (@) 17.80(4u) (1.0) + 9.662(4n1) (0.0946) + 2843.3(4n) (—0.02676) (b) 17.8042) (41.0) + 9.662(4x) (—0.6011) + 2843.3(An) (40.2412) = 0 (A126) ‘The value of ax? may be computed by making use of any one of Equations (A1-22). Using Equation (Al-22a), of (41.0) = 4,437.93(+1.0) — 346.08(+12.6894) + 5,883.94(+0.08151) of = 525.93 (at27) If Equations (A1-22b) and (c) are used and the last computed mode is substituted, ? is found to be 517.70 and 515.68, which for normal purposes is considered to be an accurate solution. A summary of the complete solution is given in Table Al-t.166 Design of Structures and Foundations for Vibrating Machines Steady-State Response Analysis Calculation of the maximum amplitude of vibration for a steady-state condition is often the main item of interest in an engineering dynamics problem. The amount of computation work is quite extensive when a response analysis is required for a system with more than three degrees of freedom. In those eases, computer pro- grams such as ICES-STRUDL, NASTRAN, ANSYS, or NISA may be used. Hence, the investigation of the three-degrees-of-freedom system considered above is extended to the calculation of the mode shapes and fre- quencies. There is a variety of methods for finding the response in this type of problem; hawwever, a modal anal- ysis technique is used here. This technique has become the current state-of-the-art (ref. 1, Chapter 2). This method of analysis consists of calculating the response for each normal mode individually and then superimpos- ing the individual responses to yield the total solution. There are some limitations on the applicability of method: The system has to be linearly elastic and the dynamic forces acting on the masses must follow the same time variation, ie, if the applied forces are har- ‘monic, then all the forces must have the same acting frequency. However, these restrictions can be relaxed if numerical methods are used in the solution of the modal equations. For a lumped multimass system having j masses M damping constant C, associated with mass 1, 5 springs, and NV normal modes the modal equations of ‘motion for the nth mode is derived by the use of the Lagrange equation. At any instant in the system, the total Kinetic energy i a Kad 6M, Gs), (a128) the total strain energy in the springs is 0=3,%64(2 4). (ar29 the total energy dissipated by the dampers is Ge D=—-2 CE atm (AL-30) and the total work in terms of displacement is (ast) In equations A-1.28 to A-I.31, ayy and dry ate respec- tively the displacement and velocity component of mass 1 associated with the nth mode, dy it the distortion of spring g (i, the relative displacement of its ends) in the nth mode, and k, is the stfiness of hat spring “These equations are based on the fact that any displace- ‘ment or velocity is equal to the sum of the modal com- ponents The squared series in Equation A-1.28 is equivalent to the sum of the squares of all modal components of dry plus tice the sum of all cross products of these components. When summed over all masies, the total of these cross products must be zero, according to the orthogonality condition given by Equation A-1.32. 2 BM, apn try = 0 (A132) ‘This orthogonality condition is true for the displacement and for the velocity vectors. Thus Equation A-1.28 may be written as (41-93) Similarly, by the same reasoning the cross product terms of the series in Equation A-1.29 will also be zero and thus, is reduced to fi N UH=2 WEE dn (1-34) For each mode it is convenient to select a modal displacement Xq so that all individual mass displace- ments may be expressed in terms of this one variable. X, is usually taken as the displacement of one arbitrary selected mass. Thus (0) on = ($8) = Hd Wan = ke (2) = han © Aan = Xu C/K) a Al (A1-35) where App and Aagy are constants for a given mode. The resulting equations of K, U, D, and W, may therefore be written as iN ©D=-2G2 KX Ae iN Wem ERE Ke dey (A136) ‘The Lagrange equation for a conservative system is given as 4 (aK)_aK oY ab _ ate ( ) aq 8g 8g, Age arsn ai a5, where qi and 44 are the generalized coordinates of the system, which in Equation A-1.36 are X, and X,, Substitution of expressions of Equation A-1.36 into ‘A-1.37 leads to the equation of motion, gid j 2 Mr Ain Xa B Cy Aig t+ XyE by Ady J 2 Fn (A138) ‘These equations of motion are analogous to the single- egree-of-freedom system equation and are associated as follows i 2M, Aig = cquivalent mass 3 2 G, diy = equivalent damping constant 2 by Aly, = equivalent spring constant i 2 F, dry = equivalent force Where X= displacement of nth mode, ¥,= Falf(t)] = forcing function acting on Arm = constant of mode n at mass 7, ben = distortion of spring g of mode n. Nv Solution of Multi-Degree-of-Freedom System 167 Solution of Equation (A1-38) for a normal mode n is SO % Fad i ie (Al-39) oy EM, Aim Xe where M is the magnification factor given in Table 1-4. ‘The total displacement for the mais r is given by superimposing the NV modes: y 2 Fada XO 2, dn (We fi FO — (At4o) on M, Ame The parameters Fr, M,, and f(t) are specified for each particular problem, and if the modal frequencies and characteristic shapes are available, then the solution of Equation (A1-40) follows. Example. The response of a three-degree-of-freedom system shown in part (a) for Model 6-C in Chapter 2 (Figure 2-9) subjected to dynamic loads will be deter- mined. The modal frequencies and mode shapes were previously obtained and the results listed in Table Al-1 are used. The equations of motion containing damping terms and the forcing functions for the system are as follows: (0) mii + Ga + hess hy (se—21— YH) (b) mats + ky (—m1—-YH) © Fa () (AIAL) ONE + Cb + mill + boy = Fa OH ‘The following values of the forcing functions are «ven for this example as: Fa = Fu ff = 0 Fa) = Fa f() = 8080 sin 727.71 kips Ty () = Faf@ ~ 137.360 sin 727.71 kips-t, where the acting frequency is 0 = 727.7 rad/see The values for the masses are 7.80 kips-sed /ft | my = 9.662 kips-sec'/fe Ty = 2843.3 kips-sec!te168 Design of Structures and Foundations for Vibrating Machines Table AL-2 Modal Response Analysis, Mee, ng Thea Mew aie Mee gol ke ae tate We cy g ati: 0 moen "8 oe de tees © alta ata Es a a sees EEE 2 Hem tty sao nat asses a, re ee cae 1 bso 8 etm 8 24 Ta stam Cam seas ar ae SE 5 sae ae a 1 i 0am am gam 2 oe er eo rc Sanaa Sma? Sine 53 anne ae aus ass ++ Silat adh — tae Cares ‘The inclusion of damping in a multidegree-of-freedom system is accomplished by assuming a certain percentage of critical damping in each mode. This factor is nor- ‘mally accounted for while computing the value of the ‘magnification factor (M) for that mode, which are then superimposed. In multi-degrees-of-freedom problems requiring a steady-state response analysis, damping is generally assumed in terms of the damping ratios rather than as damping coefficients such as Cz, and Cy used in Equation (Al-41). These damping coefficients are in- cluded in the equations of motion, when it is required that 2 free-vibration (transient) analysis be obtained. ‘Therefore, in this example problem, damping ratio (D) for various modes are assumed as follows: Mode 1, = 00 (ao damping in the structure) Mode 2, D = 0.50 (damping in the lateral mode) Mode 3, D = 0.20 (damping in the rocking mode) (Al-42) For the sake of clarity, Equation (A1-40) is expanded Fuh Padrt Fade a niece Pot] xo Fy Fada Fads ++ asco fi tater tate | (A143) ‘Similar expansions can also be written for x; and x». Numerical solutions of these expansions are tabulated in Table Al-2 where x, corresponds to the deflection at mass my (at the base of the footing), x, corresponds to the mass my (at the top of the structure), and a, is the rotation of the mass Iy. The function f(t) in this example problem is equal to sin a, which when super imposed with the maximum displacements, gives the final solution: (a) %: = 00011 x 10-%sin 727.76 ft (b) s2= 1.5800 X 10° sin 727.7¢ ft (c) p= 0.09195 x 10° sin 727.7¢ rad (AL44) Appendix B Summary of ICES-STRUDL Commands ‘The following ICES-STRUDL commands are described in refs. 2 and 3 of Chapter 7. Some special symbols are used to denote options. Among these are: —Underlines: The portion of the word which is required is underlined and the rest of the word is optional, —Braces {}: A set of braces indicates that a choice exists. Any one, sometimes more than one choice, can be made, —Parentheses () + Any item in a parentheses may be omitted. The element inside a parentheses is optional, there- fore, the meaning of the command does not change if the item is omitted or included Asterisk *: An asterisk located outside and in front of a set of braces indicates that more than one choice may be made. Arrow —> : An arrow located in front of an element inside a set of braces indicates that if the user does not ‘make any choice, the element indicated by the arrow will be assumed (default value) For convenience, the format of the “list” element is given first. The symbols v, v: . values. Yq denote decimal alphalist where: integers . ny TO ny alphalist, = ‘a (‘ar’) integerlist, ia (i) 1. EJECT Command EJECT 2, FINISH Command FINISH 3. DEBUG Command OFF MAP REGISTERS > COMMON POOL ALL DEBUG 169186 Design of Structures and Foundations for Vibrating Machines 58. DAMPING Command General Form: > RATIOS PERCENTS ve (in) DAMPING | \ vs (i) - 59, PARAMETER Command PARAMETER (‘parameter’) {parameter specs | ‘paneer fpanmeter ea | ‘panne? {parm es } Bement: ‘parameter’ = alphameric parameter name (up to 8 characters) vs (FOR) (MEMBERS) list parameter _ ) y, ay specs ae . v, ALL BUT vs (FOR) (MEMBERS) list 60, CHECK CODE Command CHECK (CODE (FOR) ) MEMBERS list Index A ‘Accelerating bodies, 12 ‘Acceleration, 12 ‘Agarwal, S.L., 79 ‘Amplification factor, 20 ‘Amplitude, 13, ‘Amplitude, free, 12 Analysis types of, 13—14 Analytical computer models, 113 ff development of, 32-45 Anderson, D.G., 67 ‘Angular frequency, 18 ‘periodic motion, 20 3 Balancing, 14 Boum ‘ena 50 fed med, 35-36 anne 30 Sites, 30 bet iequeney, pring, 14 eratigo, 88 Black foundation 18,83 axgn cheat 334 Sesgn example, 91=112 mode 34 alse, 9-20 loc-ype Foundation utr, See Block fundations Boundary conditions, 14 * oundas c Cansever (nde, 35 artesian ordinates, 15 Cental inches sign of design example, 99 eat frequen 618 cular natural fequeney, 6, 1 orc rater, 6,18 cciunaion in 6-68 ceficent damping, 18 Colum resonance, 31 Column ates, 50 187 Compreto, eiprocat design example, 92-99, Comper analysis 13,113 ff. coding, 118 example, 118. Tow ear, 117 interpretation, 157=158 reatont for, 113104 Cnsitest miss, 20 Constant damping, 15 Constant Sprig sifless, 27 Constraint conditions, 14 Continnovs mass, 20 Continuous system, 28 Cooedinates ‘ype a, 15-16 ‘coupled modes, 2122, 55 CCitzal damping, 6,15 ‘olution uations for, 7-8 cite speed, 26 rosie test, 63 D Damped harmonic frequency, 18 Damped natural frequency, 18 Damping, 32-33 ‘coefficient, 15, critical, 67, 15 ddashpot, 15, ‘geometric, 70~71 ‘material, 70~71 types of, 15 Damping constants, 15 for flexible mats, 79~80 Damping ratio, 6, 15, 70-71, 80, 83 computation of, 59 sgeomemetric ‘computation of, 82 in modeling, 32~33 obtaining, 78 for pile foundations, 81 Dashpot damping, 15, Deep foundations, 80-89 Design checklist, 51-54, 96, 110—111 Design conditions, 46~54 Design eriteria, 54, 97, 100, 110188 Design of Structures and Foundations for Vibrating Machines Determinant equation method of frequency and mode shape determination, 160-163 Differential equations types of, 16 Displacement, 12, See also Displacement amplitude Displacement amplitude, 13. See also Displacement Donovan, N.C., 89 Dynamic analysis, 14, 97, 100, 106 ‘multi-degree-of freedom-system, 159-168 Dynamie balancing, 14 Dynamic conditions, See also Dynamic design conditions limiting, 97-98, 100—102 soil ‘problems of, 3962 Dynamic design conditions, 5253, Dynamic desiga factors, 5 Dynamic equations of motion ‘ypes of, 1617 Dynamic equilibrium equation, 33 Dynamic force, 17 Dynamic load, 17 Dynamic response, 26 Dynamic system, 17, 28 E Eccentricties machine, igenvalues, 16 Eigenvectors, 16-17, 22~24, 117, 121 Elastic halfspace model, 58 Elastic spring stiffness, 28, 32 Elevated foundation. See also Elevated frame foundation struc- ture; Elevated pedestal foundation example, 114—158 trial sizing, 50 levated frame foundation structure, 18 Elevated pedestal foundation, 1. See also Elevated foundation typical (model), 36-38 Embedment. See Footing embedment Environmental demands, 49, $4, 98, 102. Se also Geotechnical ‘considerations Equations of motion, 20. See also Differential equations, Dynamie equations of motion development, 33~34 for forcing function, 11 in modeling, 34~38 Equivalent forcing function (F() ‘calculation for, 4 Equivalent lumped-mass model, 20 Equivalent mass (m,) ‘calculation of, 24 Equivalent spring constant (&,) calculation of, 4 Equivalent spring stiffness, 28 Equivalent system, 28 Excitation. See also Excitation frequency rotating mass-type solution for, 1112 sources of, 8-11 types of, 17-18 Excitation frequency, 19 47-48 F ‘£40, See Equivalent forcing function Factor damping, 15 Fatigue, 5354 failures, 102 Field shear modulus ‘determinations of, 62—64 Finite elements, 32, 40, 79 First mode of vibration, 22 Fixed beam (model), 35-36 Flexible mat foundations, 83 Flexible shaft, 27 Footing embedment effects of, 72, 92-99 Forced vibrations steady-state solution of, 8—11 Forcing frequency, 19. Forcing function (F(@)), 4 ‘equation of motion for, 11 and free vibration, 6 in modeling, 33 Foundation analysis theoretical approach to, 1~2 Foundation coafiguration selection of, 93, 100, 106 Foundation mass, 50 Foundation response modification of, 78~79 Foundation~soil interaction, 71-72 Foundation structure (for machine), 18 ‘pes of, 48-49 Foundations, 77-90 block, 1, 18, 83 design checklist, 52~S4 design examples, 91—112 model, 34 trial sizing, 49-50 categories of, 1, 78 deep, 80-89 sizing and construction of, 81 design of, 1 elevated, 1, 18, 3638, 50, 113 elevated pedestal, 36-38, 113, cembedment, 72, 92 Mlexible mat, 40, 83, 115 forms of, 1 machine mass ratio, 49 mat, 18, $0, 79-80, 83 rigid mat, 83 structural systems, 1 table top structure in, 1, 18, 113 ‘mathematical model, 4, 6-8 he Frequencies. See also Fundamental frequency; Rayleigh’s Trequency circular, 18 circular natural, 6 natural, 19 determination of, 160-168 resonance, 26 types of, 1820 of vibration, 16 Fundamental frequency, 19 Fundamental modes, 23, ¢ Generalized coordinates, 15 Geometi damping, 70-71 Gentes contain, 47-52, 5=76 otechnal requirements, Sev Gece! consierations Ghazzaly, O.1., 81 " a Start foundations in, 64¢5 Grigg, Rs 8188 H Hardin-Drnevieh equations, 66-67 Harmonic components, 9 Harmonie excitation, 17 Harmonic motion, 21 Horizontal motion in pile foundations, 86 Hudson, W.R., 79 Hwong,'S-T,, 81 1 ICES STRUDL commands summary of, 169-186 Ideatized system, 28 Idris, LM, 66~68, 70 Impulse excitation, 17 Inertia block (mode, 35, use of, 103 Inertial excitation, 17 Initial conditions, 15 J Jobsis, A.C, 89 K 4, See Equivalent spring constant L Laboratory shear modulus determination of, 63—64 Lagrange’s equation, 33~34, 166, 167 Linear differential equations, 16 Linear spring stiffness, 28, Linear system, 28 Load factor, dynamic, 17 ‘Loose granular soil (sand) stratum effect of, 7475, Lowest modes, 22 Lumped mass, 20 analysis of, 37 technique for obtaining, 24 Lumped-mass spring-dashpot system, 28 Lumping of mass in modeling, 32 ™M Machines Properties of, 4647 Fequirements for, 46—47 fete factor, ibration-Severity-Data, 54 Magnification factor, 101i, 13, 20 Mass, 20, calculations for, 2~4 consistent, 20 ‘continuous, 20 of foundation, 50 Index 189 lumping of, 20 In meeting, 32 technique fo obtaining, 24 ed (node), 36-7, 38 Snglesmped mode 36 twovtomped oe, 3738 Mat foundations 130, 83, model 333 ‘eral spring and damping constants for, 79— Material damping, 70-71 * Mathematical ede calculation of parameters for, 2-8 formuliton of ¢=11 Matrix method analysis, 14 MDOP. Se Mute depreot-feedom sytem Nodal nay Modal mult-degre tamped-ass analysis, 37, 1 Mode shapes, 157 : eee determination of, 160-168 Model clastic halfpace, 58 mathemati Calulton of parameters for, 2-8 formulation of ¢=11 Modeling alermative, 38 Modeling techniques, 32-33 Modeling pes, 33-38 Modes of vibration, 83, 92-93, 102. See alo Mode shapes topes of, 21-24 Motlon, 6, 20-21 tonontal in pie foundations, 86 verti in pile foundation, 8183 Matidepreoffeedom (MDOP) system, 28-3, 107,159 oltion of 139168 Maltuped mass with coupled soistructure iteration, 38 i F Natural fequeces of motion, 19 jdsininaton of 19016 tar! fequenes Of vibration, 16 Node, 4 Node points, 24 Node viraig ystems, 28 Nonlinear spring snes, 28 Nonlinear snc, 28 Normal coordinates, 16 ‘Nonmal mod 23~24 Nortey, De 6? Novak, My 86-21, 84, 86 ° O'Neil, M.W., 81 Operating regency, 19 Orthogonality: a coniton, 162,166 Osclion, 24 Se alo Oxilator tests Onsiltor est, 68 Oreamping olan fusions for, 8 Overturned oondation suture, 18 Pp Particular integral, 9-10 Peak-to-peak (double amplitude of vibration), 24 Pedestal foundation, elevated (model), 3638, 113, \190 Design of Structures and Foundations for Vibrating Machines Period, 24 Periodic excitation, 9, 17 Periodic mation, 20 Phase, 24-25 Phase angle, 6-12, 2425 Physiological effects. See Environmental demands Pier foundations. See Pile foundations Pile cap, 81, 84 Pile foundations, 49~$0, 80-89 Pile groups, 82-86 Poisson's ratio selection of, 71-72 and soil density, 71-72 typical values, 72 Principal coordinates, 16 Principal modes of vibration, 22-24 Procedures, design, $4 Prychological effects. See Environmental demands R Ratio damping, 6, 15 Rayleigh wave lengths, 63, Rayleigh’s frequency, 1920, 114, 119 model, 36-37 Reeiprocating compressor design example for, 92-99 Reciprocating machines, 92-93 design for, 49 Resistance calculation of, 4 center of columns, $1, 116 of soil, 49-50, 116 Resonance, 12, 25-26 column, 51 condition, 25—26 frequency, 26 Resonant column test, 64 Response, dynamic, 26~27 Toundation ‘modification of, 78-79 steady state, 10, 2627 transient, 6,27 Richart, FLE., Jr, 72 Rigid mat foundations, 83 Rigid staff, 27 Rocking equivalent spring, 38, 111 Rocking motion in pile foundations, 86-88 Rotating-mass-type excitation ‘dynamic aystem subjected to, 11—12, 17 s Sands, 74-75 foundations in, 64-65 Saturated clays foundations in, 65—68 Saul, W.E., 89 SDOF, See Single-degree-of-reedom system Seed, H.B., 66-68, 70 Shaft critical speed of, 27, ‘Shear modulus. See also Shear strain magnitude calculation of for structure-soil interac correlations, 64-68 field determi analysis, 68-69 tions of, 62-64 laboratory determination of, 63-64 and pile foundations, 81 sil, 2-69 field procedures for, 62-64 laboratory procedures for, 64 typical values, 69 ‘Shear strain, 6970 ‘Shear strain magnitude Selection of, 69-70 ‘Simple harmonic motion, 21 Simultaneous differential equations, 16 Singh, J.P., 89 Single-degree-of-freedom (SDOF) system, 2, 28 in layered soils, 73-74, model examined, 412 ‘Sinusoidal excitation, 17 Sinusoidal motion, 21 Skempton, A.W., 67 Soil loose granular, 74-75 stilt, 72=74 Soil density | and selection of Poisson's 1 Soil dynamics problems of, 59-62 Soil—foundation interaction, 71~72 Soil parameters, 47—49 evaluation of, 59-62 Soil shear modulus, 62-69 Soil spring stiffness, 28 Soil—structure interaction, 38, 71—72 Soil tests, 63 : Southwell-Dunkerley formulae, 107 Spring absorbers (model), 35 Spring constants, ‘equivalent calculation for, 4 evaluation of, 58-39 in modeling, 32 obtaining, 78 vertical for flexible mats, 79-80 Spring-dashpot system. See Lumped-mass spring-dashpot system Spring stiffness, 27-28 Static analysis, 14 Static balancing, 14 ‘Static con 100 Static design conditions, 50, 52, 114 Steady-state response, 26-27 ‘method of frequency and mode shape determination, 166168 Steady-state solution of forced vibrations solution equations for, $11 Stiff shat, 27 Stiff underlying stratum effect of, 72-74 ‘Stodola-Vianello method, 163165 Stokoe, KH, Il, 72 Strain magnitude selection of, 69-70 Stratum Tose granular soil effect of, 74-75 stiff underlying effect of, 72-74 io, 1-72 Structural system of foundations, 1 Structure~soil interaction, 71-72 analysis of, 68-69 STRUDL computer coding, 118~121 ‘commands, 169186 Subharmoni¢ motion, 21 ‘Superharmonie motion, 21 T Table top foundation structure, foundation ‘Terminology, 12—31 Testing methods, 63, 88~89 Theory of vibrations fundamentals of, 2 ‘Transient excitation, 18 Transient motion, 6 Transient response, 27 ‘Transient vibrations al model, 4, 6~8, 16 lity factor, 26, 31 18. See also Elevated Trial sizing of block foundation, 49-50 of elevated foundation, 0 Two-lumped mass, 16, 23, 37-38 u Uncoupled modes, 22 ‘Undamped system solution equations for, 6-7 Underdamped system solution equations for, 7 Index 191 Undertuned foundation structure, 18 v Velocity, 12 Vertical motion in pile foundations, 81-83 ibrating machine ‘supported by a cantilever (model), 35, supported by a fixed beam (model), 35-36 supported by block-type foundation (model), 34 supported by mat-type foundation (model), 3435 supported on inertia block and vibration isolated from foundation (model), 35 Vibration amplitude, 13 Vibration analysis, 14 Vibration limits, 32-54 Vibration modes, $3, 92-94, 102 types of, 21-24 Vibration response, Wibraten ts, 38 bration theory fundamentals, Vibrations ” “2 forced pease saan of, 811 Solution of, 4, 6-8 transient, 4 68 Viscous damping, 15 w Whitman, R.V., 63-64, 71 Woods, R.D., 67Design of Structures. 7° {2% 2208 together traational and new concepts and procedures for analyzing and designing dynamically loaded and Foundations structures. With Design of Structures and Foundations for x S Vibrating Machines, practicing engineers and students now have Vibrating Machines @ text which integrates theories of vibration, geotechnical engineering (including soil dynamics and half-space theory), computer coding and applications, and structural analysis and design. The many concepts and procedures used in the design of structures supporting dynamic machines and ultimately supported by the soil until now have been unavailable in a single source ‘The design process in this fleld has gradually evolved from an approximate rule-of-thumb procedure to scientifically sound procedure. In Design of Structures and Foundations for Vibrating Machines, stato- of-the-art techniques are employed in actual design problems by using simplified step-by-step routines. In addition, at every step of investigation a brief description explains the physical meaning of the parameters used and the role they play in the design process. ‘The introductory chapter reviews fundamentals. Chapter 2 describes alternatives of modeling dynamically loaded systems. Chapter 3 considers and lists the information necessary for design. Chapters 4 and 5 describe the geotechnical aspects of the problem, and Chapter 5 specifically considers flexible mats and foundations. Finally, Chapters 6 and 7 include examples of different types of structures ‘supporting dynamic machines. About the Authors ‘Suresh C. Arya is principal engineering specialist in the Technology Section, Civil-Structural Group, with the CE Lummus Company. He is responsible for all vibration and dynamics analysis of structures and foundations made in Lummus’ Houston office. He has a Ph.D. from the University of Texas with 18 years of experience in structural and dynamics analysis for the petrochemical industry, nuclear power plants, and marine terminals. He isa registered professional engineer in Ohio, Texas, and California, Michael W. O'Neill is associate professor of civil engineering at the University of Houston. He has a Ph.D. from the University of Texas, and has served as a geotechnical engineer for Southwestern Laboratories, Houston, ‘Texas, with responsibilities in consulting for clients involved in designing foundations and earth structures. He is ‘a registered professional engineer in Texas. George Pincus is professor of civil engineering at the University of Houston. He teaches structural engineering and directs graduate student research in structural mechanics and structural analysis design, and he supervises {full-scale laboratory testing. He has a Ph.D. from Cornell University, and is a registered professional engineer in ‘Texas, Kentucky, Florida, West Virginia, New Mexico, and California. ‘The Practial Vibration Primer Charles Jackson “Thisbook provides a working knowledge ofthe fundamentasrequiedto evaluate mallunctionscausedby excessive vibration in machinery. ISBN 0-87201-891-1 Introduction to Offshore Structures: Design, Fabrication, installation Witta J. Gra Ph. ‘This concise, practical guide tooshore platform engineering covers tubular steel and concrete gravity platforms, Itdescribes geotechnical design, construction and placement of platorms, design lads and force, corosion protection, fatigue analysis, catwalks, hepors, ané more. [SBN 087201-604'3, Equipment Design Handbook for Refineries and Chemical Pants, Frank L Evans. ‘Volume 1, 2nd Edition describes by example ll the design procedures needed for major rotating equipment items: crivers, compressors, ‘jectors, pumps and process rerigeration, ISBN0-87201-200-2 ‘Volume2, 2nd Edition covers maior nonvotaling equipment ems: trad heaters, nea exchangers, cooling towers, vessels, separators and ‘accumutalors, flare slacks, piping, sewers and valves. ISBN 0-87201-267-0 Cover Design by David Price Gulf Publishing Company Book Division Box 2608 Houston, Texas 77001 ISBN 0-87201-294-8