i i ISSN 0110-3326 COMPUTER PROGRAMS FOR CONCRETE COLUMN DESIGN by D.J. King supervised by M.J.N. Priestley and R. Park. This report is based on a Master of Engineering Report by the same author. The project was sponsored by the New Zealand Concrete Research Association. Research Report 86/12 May 1986 Department of Civil Engineering, university of Canterbury, Christchurch, New Zealand.4 ACKNOWLEDGEMENTS ‘The writing of the computer programs presented in this report was carried out in the Department of Civil Engineering, University of Canterbury, under the overall guidance of its head, Professor R. Park. ‘The project was supervised by Dr M.J. N. Priestley and Professor Park, whose advice and guidance is gratefully acknowledged. ‘The financial assistance provided by the New Zealand Concrete Research Association is also gratefully acknowledged. Thanks are also due to Julia Mackenzie for typing this manuscript and to my room-mates who have a high tolerance of computer printers and radios and their willingness to listen to my problems.ABSTRACT. A suite of computer prograns has been written in BASIC for IBM personal computers for the design and analysis of Fectangular and circular concrete colunns and analysis of erbitrary concrete columns under uniaxial or biaxial bending, with or without an applied axial load. Three programs have bi n written using the ACT equivalent rectangular stress block for the design and analysis of rectan- gular sections under uniaxial and biaxial bending and circular Sections. These comply with the conditions set out in NZS 3101: 1982. Three programs have been written using the Mander model for concrete stresses. two are for the design and analysis of Fectangular sections under uniaxial bending and circular sections. The other is for the enalysis of symmetrical and general arbitrary Sections under uniaxial or biaxial bending, All three are able to do @ moment-curvature analysis on the section. Some comparisons have been made between the design charts Presented in the New Zealand Reinforced Concrete Design Handbook and the prograns. Generally the ACI programs agree well with the charts and the Mander model prograns give @ more economical design because of the more realistic concrete stresses. Full instructions for the use of the prograns have also been presented.CHAPTER T CHAPTER IT CHAPTER IIT TABLE OF CONTENTS ACKNOWLEDGEMENTS . ABSTRACT TABLE OF CONTENTS . Paid 96660 INTRODUCTION ae. AqM Boo 1.2 BACKGROUND . Fey) fe] 1.4 SIGN CONVENTION MATERIAL PROPERTIES 2 2.2 2.3 THE ACT EQUIVALENT BLOCK ‘THE MANDER MODEL 2 2 2 STEEL REINFORCEMENT STRESS STRAIN CURVES 2, 2 2 22 22.2 2.3 Bel 23.2 23.3 RECTANGULAR Confined concrete . « Cover concrete... Maximum concrete strain For ACI method... For the Mander nodel . Steel properties . . . THE DISCRETE ELEMENT APPROACH 3.1 3.2 GENERAL RECTANGULAR SECTIONS... . 3 a 2a 2.2 Concrete forces . . . Reinforcing steel . . STRESS PAGE aL aL 12 12 43. 15 16 16 16 1sCHAPTER 121 CHAPTER rv (continued) 3:3, CIRCULAR sections . |, 3-3-1 Calculating the seguent ‘area 3-3-2 Concrete forces 2. 3:3.3° Reinforcing steer. | | 3-4 ARBITRARY SECTIONS =. | | | 3-4-1 Method for determining position ofelements 2... | 3:4.2 Concrete forces... | pete =. Hettcorciacistcg | es 3-5 CALCULATING THE NEUTRAL axIS DEPTH 3-5-1 Uniaxial bending... |. | 3:5.2 Biaxial bending of rectangular sections eooend 3-5-3 Biaxial bending of general arbitrary sections... 3-6 MOMENT caLCULATION. 2... | | 3:7 PLASTIC cenrRorD 6. . | | 38 MOMENT CURVATURE ANALYSIS 2. . |, 3-81 Stopping conditions .. 38.2 The ideal moment capacity. , INSTRUCTIONS FOR USING ALL RECTANGULAR AND CIRCULAR PROGRAMS 42 GENERAL. Se 4:20 MAIN MENU OPTIONS ©. 4.2.10 Input Editor 2... | ease Sactlonidesion ee 4.2.3. Section analysis...) | | 4-3, CODE REQUIREMENTS ©. 4:31 General. 2... 4.3.2 For menbers not designed for seismic loading...) |. 4.3.3 Members designed for seismic loading. PAGE 18 1a 19 2 22 22 24 28 25 25 27 30 33 35 36 36 36 29 40 40 44 47 48 48 49 50CHAPTER V CHAPTER VI REFERENCES APPENDICTES INSTRUCTIONS FOR USING THE SYMMETRICAL AND ARBITRARY SECTION PROGRAM Sol) GENERA ee 5.2 MAIN MENU OPTIONS - 2.2... 2. 5.2.1 Input editor... 2... 0 5.2.2 Assemble elements of the section 5.2.3. Calculate plastic centroid. . 5.2.4 Store data on disk... . 5.2.5 Read data from disk 1... 5.2.6 List and delete data files on disk 5.2.7 Analysis... ee... 5.2.8 Moment curvature analysis... TESTING AND VERIFICATION OF THE PROGRAMS 6.1 GENERAL ©... ee 6.2 DESCRIPTYON OF THE SECTICNS USED... . 6.3 RESULTS 6.3.1 Desi... ...... 2.0 e Sede?) Amalyetee) eee 6.4 DISCUSSION OF RESULTS Grae Design es 6.4.2 Analysis... . 2 (6i5 CONCLUSION) te ee (S PUCICy coggneocaucus B. EXAMPLES OF SCREEN DISPLAYS AND PRINTOUTS C. LIST OF VARIABLES «2... 2... D. —- PROGRAM LISTINGS»... . 1 ee PAGE 52 53 53 58 59 59 59 59 60 61 61 63 67 69 70 70NOTATION 2 depth of equivalent rectengular stress block @ distance along the x axis to the neutral axis area of concrete core gross area of the section A, area of longitudinal steel Ruy axea of hoops b width of section b distance along the y axis to the neutrel axis bl width of concrete core concrete compressive force steel compression force © neutral axis depth from extreme compression fibre @iameter of the main reinforcing bars 4, depth of concrete core diameter of spiral or hoop stirrups eps iteration accuracy initial Young's modulus of elasticity for concrete Young's modulus of elasticity for steel eu sec fee/*te £, concrete stress fh concrete cylinder strength le Raximun strength of confined concrete £4, insitu unconfined conerete strengthlateral confining stress steel stress maximun steel stress after strain hardening yield stress of steel yield stress of confining steel hoops confinement effectiveness coefficient coefficient for defining steel stress-strain curve ideal moment capacity of the section moment when the steel first yields or concrete reaches a strain of 0.002 number of segnents applied axial load constant for the Mander model (set to 5) E/E, - &, eo! "Fe Boece) pou at centre line spacing between confining hoops clear distance between confining hoops steel tension force average centre line spacing between longitudinal bars average clear distance between longitudinal bars reinforcing steel coordinatessh “spall max viii distance to the centre line of segment i from the extreme compression fibre distance to the end of segnent i from the extreme compression fibre depth of equivalent rectangular stress block as a proportion of the neutral axis depth concrete strain concrete strain at £1 concrete strain at £1, - usually 0.002 ultimate concrete compression strain maximum conerete strain minimum practical concrete strain steel strain steel strain at the onset of strain hardening spalling strain of cover concrete steel strain at maximum stress steel yield strain volumetric ratio of longitudinal steel to conerete core maximum allowable percentage of steel according to NZS 3101 minimum allowable percentage of steel according to NZS 3101 volumetric ratio of confining steel to concrete core curvature at M, : rf curvature at “first yield"Curvature when the extreme steel strain reached yield or when the concrete strain reaches 0.002 increment for the finite difference method of solving partial differential equations.CHAPTER 1 INTRODUCTION The aim of this report is to produce a suite of computes Programs for concrete column analysis and design, for use on @ microcomputer in a design office situation, the programs are te cover various section shapes including rectangular, circular, symmetrical arbitrary sections and general arbitrary sections subject to both uniaxial and biaxial bending. Different models for calculating concrete stresses, including the Mander model and ACI equivalent rectangular stress block, are to be used. 2.2 BACKGROUND The usual method for designing colunns is by using design charts or a hand analysis. Design charts are available fron the American Concrete Institute (ACI) and Concrete Publications Limited, New Zealand. The New Zealand Reinforced Concrete Design Handbook has used the equivalent rectangular stress block to Produce a series of charts for designing rectangular ana circular sections with uniform reinforcing steel. The New Zealand Concrete Research Association has produced a suite of reinforced concrete design prograns for circular sections with uniform steel and for rectangular sections with symmetrical steel. These prograns will design the area of reinforcing steel required for a given axial load and moment using the equivalent rectangular stress block to NZS 3101. Recent research by Brgndum-Nielsen and Kawakami et al has produced papers outlining programs for biaxial bending ofarbitrary section shapes. The progrins suggested by Bréndun-Nielsen are suitable for hand-helé calculators and are limited to an equivalent rectangular stress block for the concrete stresses. Kavakami et al use a nore realistic stress-strain curve for Concrete stresses. These programs are not readily avatiaple in New Zealand. Most columns designed for New tealand conditions require sone level of confinenent to be able to withstand plastic hinging ender earthquake loading. Confining the concrete core in colunne can enhance the concrete strength considerably. Several nodels for confined concrete have been developed over the years but Fecently Mander et al have produced a model which is suitable for ali section shapes. It is reasonably simple to use and produces realistic stresses for both confined and unconfined concrete. The scope of existing charts and prograns covers only regular section shapes under uniaxial bending with symmetrical Feinforcenent using a sinple nodel for concrete stresses. conputer Programs are required which can cope with non-uniform reinforcing arrangements, irregular section shapes under both uniaxial and biaxial bending. ‘the use of a more realistic model for concrete stresses such as the Mander model, which takes confinement into account, may also result in a more economical design. 1.3. scope The writing of seven programs was undertaken to cover the Gifferent conditions required. These are illustrated in Fig. 1.1 and listed as follows: (1) Uniaxial bending of rectangular sections using the Aci stress block. (2) Biaxial bending of rectangular sections using the ACT stress block. (3) Circular sections using the ACI stress block.,_E eee Se ee (4) Uniaxial bending of rectangular sections using the Mander model. (5) Circular sections using the Mander model. (©) Symmetrical arbitrary sections with bending in a direction Parallel to the axis of symmetry using the Mander model. (7) General arbitrary sections with uniaxial or biaxial bending using the Mander model. The first five prograns listed are written in a sinilar style and are almost identical to run. Each has the facility to @esign the required area of steel for a given axial load and monent and to analyse the section for a given axial load and area of steel. the programs using the Mander model also have a moment= curvature analysis routine. ‘The arbitrary section programs only have the facility for @ section analysis and a monent-curvature analysis. 1.4 SIGN convENTrON Compression forces are positive. Tension forces are negative.Uniaxial bending of rectangular sections Biaxial bending of rectangular sections Circular sections Symmetrical arbitrary sections c4 YO [Z| General arbitrary sections Fig. 1.1 The Different Conditions Considered.CHAPTER Ir MATERIAL PROPERTIES 2-1 THE ACI EQUIVALENT RECTANGUTAR STREEE DLOCK ‘The New Zealand Standard N2S3101 (2) suggests using fhe rectangular stress block to calculate the concrete compression force C, at an ultinate strain of 0.003. this nethod would be the most commonly used in design practice. The parameters for rectangular compressed ar, are shown in Fig. 2.1a and the equations are as follows: Cl = 0.8setab Qa ¢ where a = Be 8) = 0.85 for £1 $30 wpa 8) = 0.85 - o.008(£: ~ 30) tor £2> 30 mPa but not less than 0.65 ne depth to the line of force from the extrene compression fibre For non-rectangular compressed areas (Fig. 2.1b) the sane Pafeneters are used. Although they are not strictly applicable, ‘he nethod is probably sufficiently accurate for design Purposes,cs000 ® este al Lemon | fi ele shale it ‘Section ‘Strain profile ‘Stresses, o (a) Rectangular compression area. * ° * Loa Fauivalont stresses Achual concrete stresses Strain Section (b) Non-rectangular compression area. ‘The ACI Rectangular Stress Block. Fig. 2.12.2 THE MANDER MoDEi, 2.2,1 Confined concrete When a colunn contains stirrups or ties vhich are closely spaced, the strength of the concrete core is enhanced by the confining pressure provided by the ties. Mander et al (4) presented an analytical model to calculate the longitudinal stress in the concrete, £, (see Fig. 2.2). An advantage of this model is that one equation defines the total curve. finer 2 Goo (2.2) alts” where fh. = co, ce rs 24) mos Met, ‘ « a ee (5) z te gs (2.6) ® . (2.7) B= 5000 AEE (2.8)\e— 1. cover CONCRETE 1 ! | | I Ein E, Eto Espot Ec bod Fig. 2.2 Concrete Compressive stress Strain Model as Proposed by Mander.foe Se etae 29) €'G5 7 0-002 (2.10) R = 5 asa recommended average. has been used Here the cylinder strength of concrete, £1 instead of the ineitu unconfined concrete stienyth, £1). ‘the £1 is the value used in a design office and there appears to be only small differences in the values of each. For circular sections: (see Fig. 2.3) ea (1 ~ 0.5874.) * 7 for circular hoops (2a) (2. ~ 0.5874) x aces for spirals (2.22) : (2.13) ee 7 Qa4y For rectangular sections: e Fig. 2.4) (= 0.58" ~ 0.58174) Q= eo) (2.15) where Nbars = number of longitudinal bars W' = average clear distance between bars.- “Das (2.16) bas a (2.17) Gecle. brae For the program for rectangular sections, an average value for p, has been used rather than the method proposed by Mander where the octahedral shear stresses are used to calculate £1. Also, average values for the bar spacing have been used instead of the actual. This requires slightly less input and was considered sufficiently accurate for the purpose. 2.2.2 Cover concrete The cover concrete stresses follow the curve for unconfined concrete where the confining stress £1 = 0. lence flo et : en: he 2 (ns) fhe Ely (as) fe (2.20) sec aos . co @ piece (2.21) 7 Feo The curve 4s Linear between 2¢2, and co, where eo, = 0.0064. This value was taken as an average value from reference 4. for S 268) - (2.22) : i |eax (2.23) for 22, < €6 § Cspant fect] |? ee ee ceerr: cued a) (2.24)2.2.3 Maximum concrete strain ‘the maximum concrete strain is calculated from the simplified equation rather than using the energy balance method. Snax 7 Fgpann + 904082) (2.28) where coe fee z fgpatr 0-906 2.3 STEEL REINFORCEMENT STRESS STRAIN CURVES 2.3.1 For ACT method ‘A simple bi-linear relationship is assumed for the prograns using the ACI rectangular stress block as shown in Fig. 2.5a, where f, 7 Bt, fore, sey (2.26) and fet, fore, >e (2.27) Ei a & The yield strain ey = 3 where f, = yield stress.. = eee 13 2.3.2 For the Mander Model In the programs using the Mander model, strain hardening of the steel can be taken into account as a user defined option. ‘The shape of the curve is shown in Fig. 2.5b and has been taken from reference 1. = Ee, for es (2,28) fo f, for «<< c (2.29) . y y eh Me, - eg.) +2 lel - e,, (160 - m a” fen s~ Sah £, 7 fy [eote,- e,,08 2° -atzors ay ae for fon Se % Sgn (8, £1308 + a? eee au/*y' usr? where om andor tee.14 stress g strain (a) As used in the ACI programs. #* Steel stress Steel strain & fen fu (b) As used in the Mander model prograns. Fig. 2.5 Reinforcing Steel Stress strain Curve.2.3.3. Steel properties 1s Table 2.1 Steel Propert: vrup sree | nxcn sanenora vaeuo stness LL naxzwm staess sygitrad | 25 6 Ls voune's wovunus (mPa) | de input | as input YIELD STRAIN 4/8 Lv ie 5 STRAIN HARDENING . we 3.24 Cy *h ¥ Ge waxamoM STRAZN fy [Onde | one The normal design values for £, and 380 MPa tor nigh strength steel. usually taken at 200,000 MPa. re 275 MPa for mild steel Young's Modulus isCHAPTER III ‘THE DISCRETE ELEMENT APPROACH 3.1 GENERAL Because the equation for concrete stress cannot be readily integrated, the column sections have to be divided into a number of elements. The concrete stress is then calculated using the s assumed strain at the centroid of each element and the stress to be constant over each elenent. Since the stress-strain curves are different for confined and unconfined concrete, the areas of core and cover concrete have to be defined. The confined core of the section is taken to the centre-line of the confining ties or stirrups. 3.2 RECTANGULAR SECTIONS For uniaxial bending, the neutral axis is always perpendicular to the direction of bending, so the elements or segments become thin strips. ‘The cover concrete at the ends of the section is divided into three segments. The fourth segment has its centre- Line coinciding with that of the main reinforcing steel and a Jength equai to the main reintorcing bar Giameter 4, plus the tie Gianeter. The boundary between the 3rd and 4th segments is on the centre-line of the ties, This is illustrated in Fig. 3.1. Segments 4 to N-3 contain both core and cover concrete. Note that the area of reinforcing steel within a segment is subtracted from the core area of that segment. 3.2.1 Conerete force To calculate the total concrete force, the strain at the centreline of each segment is determined using equation 3.1.Fig. 3.1 Division of Rectangular Sections into Segnents. j——5 bors ‘ 6 bors us +5 bars ce Actual Modeled in program ~~ 3 Fig. 3.2 Location of Reinforcing steel in the Segnents.Concrete strain ee Se . 5 iy 7 ve GB. ‘he stresses for both confined and unconfined concrete are calculated and the total concrete force C_ for segment i equals ci” Meorert * Fetcoreri * Meoverni * feteoverdi {°?? If the neutral axis lies within a segment then the forces are calculated using the part of the segment between the neutral axis and the previous segment. 3.2.2 Reinforcing steel Using the segnents set up, the reinforcing steel is, in effect, smeared around the perimeter of the section on the centre-Line of the main reinforcing bars. The total steel fat each end is lumped into segnents 4 and N-3. The side steel remaining is distributed evenly amongst segments 5 to N-4. The area of steel in each segment acts at its centre-line. Fig. 3.2 illustrates this. 3.3. CIRCULAR SECTIONS For circular columns the section is divided into segments in a similar manner to rectangular sections (see Fig. 3.3). ‘The cover is divided into three elements and the other elements contain areas of both cover and core concrete. 3.3.1 Calculating the seqment area To calculate the area A of the shaded portion of the circle shown in Fig. 3.4 the following formula is use aia)? cos (1 - 2) - Hee yWaia.y-y? (3.3) oe aia 2 It y is the distance to the end of a segment then the total area ofSegment i is: Msegmentyi "Ay for dia 19 section diameter To calculate the core area of segnent i; Meorei ~ Ay area minus the core area, 3.3.2 using equations 3.1 and 3.2 as where dia = ALTA, Concrete forces Concrete forces For the program using the ACI stress fo the centre of compression from the extrem Aianeter to centre-line of ties for segments 4 to ne \i-1 f0F Segments 4 to n-3 e- 2 fe sindy - (2 - yy? , Vary ~ y?} ya ere n radius of section By. area calculated by Eqn. 3.3 reo (a - 2) © compression fibre i, (3.5) 4) described for rectangular sections. block the distanceDividing Circular Sections into Segments. re fe Fig. 3.4 Area of a Sector of a Circle. Fig. 3.3 Fig. 3.5 Definition of 6.nee = ee ee ee ee, ee ee =—_— = ‘The total area of reinforcing steel ig the section at its centre-; “smeared* around tine. A proportion of the total steel area is Put into each segment depending on the arc length, at ‘he steel centre-line, within the segment. (See Fig. 3.5.) 6 2 cos"h(a - ¥ ~ cover = 4/2), ae) ¥ radius to the steel centre-line, ‘the aren of steel in segnent 1 is therefore: Asisegnent)i ~ *s(tota2)3.4 ARBITRARY SECTIONS For both the synmetrical and general arbitrary colunns, the section is divided into a nunber of rectangular elements which form a grid between the extrene boundaries of the section (see Fig. 3.6). The section will have a region of core concrete and may contain voids with cover concrete around the void itself. All these are defined by co-ordinates. once the grid has been set up, each element has to be checked to see where it lies in relation to the section. For example an element may lie outside the section, within a region of cover concrete, within the core or within a void. If the centroid of the element lies with the region of cover concrete, for example, then the whole element is said to lie within that region. 3.4.1 Method for determining position of elements Fig. 3.7 shows the triangle ijk. The line ij represents two consective points, 1 and 2, on the bourdary of the section shown in Pig, 3.7b and k repre: nts the element A which lies outside the section. Using the notation shown, the area of the triangle ijk ist Area = blab, ~ asd) 3.7) If this area is positive then the point k is said to be on the left hand side of the line ij and if the area is negative then the point k is on the right hand side of the line ij. Pdnt k must be on the right hand side of every consective point on the section boundary to lie within the section. This is the case for element B. Note that element A is on the right hand side of the boundary Lines 2-3, 3-4 and 4-1 but is on the left hand side of line 1-2, so it must lie outside the section.Fig. 3.6 Pig. 3.7 Forming an Arbitrary Section into a Grid. (a) tefangie 3x. 4 (b) Section 1-2-3-4, Determining the menses +This method is used to determine the position of each element centroid in relation to the section. If the section shape has a re-entrant such as a T or L section then the re-entrant must be input as a void. (See Fig. 3.8) If this is not done and the section is defined by the actual section co-ordinates, then only those elements within the shaded area will be recognized as being within the section. The void may have cover concrete around it. Fig. 3.8 A Re-entrant section 3.4.2 Concrete Forces concrete stresses are taken to be uniform over each elenent for the strain at the centroid of the elenent. Each element i has a corresponding area of cover concrete and core concrete. For elements within a region of cover concrete: Aicoverya ~ AF@a of elenent Mcorels ° For elenents within the core: a 27 (cover )i aT Area of elene: For elenents outside the section or within a voi ®cover)i *(core)é25 The total concrete force: c ©” FW corersetcore)t * * cover) fetcover) 3.4.3, Reinforcing steel with an area of steel at each co-ordinate, Either individual bars or groups of bars may be modelled. 3:5 GALCULATING THE NEUTRAL AXIS DEPTH SAUCULATING THE NEUTRAL AXIS DEPTH of the neutral axis depth. 4 general the neutral axis has to be in such a position that (i) internal e section is satisfied, quilibrium of the Sections and the symmetrical arbitrary sections, the secona criteria above can be ignored, ‘The neutral axis is assumed fo be Perpendicular to the direction of applied nonent and the symmetrical nature of the section ensures tha t this criteria is satisfied. ‘when equilibrium has been reached then the neutral axis depth is said to be correct. Equilibrium is satistiea by the following equation: C+e en = p cee, where C, = concrete compressive forces C, = steel compressive forces tT. steel tensile forces P= applied axial loadFor these programs, the neutral axis is calculated by an iterative method using linear interpolation. The latest value for ¢ is based on the previous two values and the axial load calculated from these. as) where ¢), c,, P, and P, ari vious two values for the ere cy, Cyr Py 2 are the pre two values for th neutral axis depth and the axial loads calculated from then. For the rectangular and circular section programs the initial values for c, and c, arer 2 a) 7 section depth 1 4 (P+ ageyay 0.8s7¢ section width y The resulting axial loads P, and P, estimate c, is calculated using Equation 3,9, The axial load is then calculated. If c, is not the P, is then calculated. If cy te are established and a new rue value then c, = Cyr 17% P, = Pye, = 6, and PL = P, and c, calculated again. ‘The iterations stop when: lap] < eps [rt (3.108) [acl << eps (3.108) where eps is the iteration accuracy. When the applied load P is negative, i.e. tensile, the iterations may not converge. To avoid this problem the value of c, is halved or doubled accordingly until ? lies between P, and P,. For further calculations of c,, c, and c, are selected so that P 1 always Lies between P) end P,. For symmetrical arbitrary sections, the initial value for sag 1 z 0.85°£, .section width27 ana 2 or 7 so that P lies between P, and P,. Again, for further calculations of c,, c, and c, are selected so that P lies eulations of ¢,, ¢, s jetween P, and P, so that convergence is assured. between P, and P so thi 9 3.5.2, Biaxial Bending of Rectangular Sections This procedure for the calculation of the neutral axie Position and the moment capacity has been taken di: rectly out of reference 5. ‘The notation used here is slightly different. % two dimensional Newton-Raphson iteration 1s used to calculate a and b (see Fig. 3.9) Quan (3.22) where £, = (EA, fy, + £, Jy aa ~ pyisine CAS EL X, +f, Ix ah - Pecos8 = 0 (13) £7 TALE tea ne = 0 aa be Gas+ Mc 8 stant He al peter & cpreve SotoneSSow ncn Pig. 3.9 Biaxial Bending of Rectangular Sections Using the ACI Stress Block.29 The forward difference method is used to calculate the partial differential equations, for example: at l Fert Rey lavay = eta 16) the area of compressed concrete 7 a7 and the integrals Jyan = 38? av? - Ps - 2007 (3.18) Jxgan = 202 aoa = 9 - 703 ~ 210) (3.19) where ye i- @ $0 (3.204) 88 ee a- gy fo (3. 200) A The symbol { means that if the expression is neg is set equal to zero. ve then it ‘This procedure applies directly to the ACI equivalent rectangular stress block method for calculating the concrete stresses. The function £, is the equilibriun equation which satisfies the first criteria and the function £, satisfies the second criteria for the position of the neutral axis depth. Initial values for a and b are chosen so that the compressed area of concrete equals the applied axial load plus half the total steel force and in proportion to the applied moment. (PRETT ETT TR bo — |F| Jost, LiConnents Initially small values of 4, in Equation 2.8 were used to solve the partial differential equations. The errors involved with this method are in the order of age so it was felt that a value of 0.01 mn would be suitable for 4. Problems with convergence were encountered though and with some experimenting, it was found that the most appropriate values for 4 are between 10 and 20 nm for typical sections sizes. It appears that the local slope of the curve is not @ good estimate for the general slope. Problens were also encountered as a or b approached infinity, ise. uniaxial bending. To avoid division by zero errors and for fa reasonably quick convergence, the ratio of the two moments has been selected to be not less than 1 to eps. This will give a finite value to both a and b and hence a moment in both directions but the moment in the direction perpendicular to the principal direction of bending is usually small and the results are accurate enough for the purpose. Generally, convergence on the neutral axis position is quickest as the ratio of the monents H, and M, approach 1 to 2 and for average axial loads. The time for convergence is increased as uniaxial bending is approached, for very small or very Jarge axial loads and with a greater nunber of reinforcing bars. 3.5.3 Biaxial Bending of General Arbitrary Sections For general arbitrary sections a similar procedure outlined in section 3.5.2 can be applied using Equations 3.11 through to 3.15. the forward difference method is used to solve the partial differ- ential equations. For example: 1 5 Uey(a + 0) = flan G2) For each estimate of the neutral axis depth, the position of the most extreme compression fibre has to be established.‘The distance z (shown in Fig. 3.10) is the minimum distance from the origin, perpendicular to the neutral axis, to any point Xyry. within the section boundary. Bos min (y,cos8 + x sing) (3.22) nce this has been calculated the strain at the origin ¢ is: or calculated the st: 93M 19 9) aesing. *(0,0) ~ Seu (a.sinB-z) eed and the strain at point x,y is: Fony) ” £¢0,0) "2 ~ ee Using this strain profile the concrete and steel stresses can be calculated for each elenent and functions f, and £ calculated using equations 3.13 and 3.14. For general sections the initial estimate for the neutral axis position has to be reasonably close. the progran attempts to provide a good estimate using the following: andConments If there are problens with the neutral axis iterations then it may be necessary to provide an option to input initial estimates to ensure convergence. There also nay be problems if either o or b are negative, To avoid these the area of compression should be nearest the origin. 3.6 MOMENT CALCULATION (a) Uniaxial benaing ‘The moment capacity of the section is celcvlated about the extreme compression fibre. For rectangular sections using the ACI stress block (see Fig. 3.118): Mos 71(0.85£" ab/2 + TEA teres 7 P94/2) (3.250) and for rectangular sections using the Mander model (see Fig. 3.11b): _ ° Mos “MEE Nye, + TE ALy. - P.e/2) (3.25b) ci Nci¥ es sisi gs where d= overall depth of rectangular sections or overall diameter of circular sections. For circular sections the sane equations are used except that for the ACI program the distance to the line of concrete force is not 2 but is as defined in section 3.3.2. (>) Biaxial _bendin: For biaxial bending of rectangular and arbitrary sections the moment capacity of the section is calculated separately for the two directions. For rectangular sections using ACI stress block:4—_—_ segment i Fig. 3.12 (a) ACT Stress Block. "EXTREME COMPRESSION F BRE (b) Mander Mod Notation for Moment Calculation.38 Mo= -1(0.85£' fx dA + Ee a - a x c J c sisi a eee) = -1(0.85£! da + - Py (ouesty [yeas te ay, - ay (3.26) where fy.@& and Jx aA are calculated using Equations 3.18 ana 3.19, For arbitrary sections the two moments are: = lr a + AUX, > Pex x CEA MMe PL) (3,278) = ne ay, + -P 7 CEA, + Tay, = Peyiagg) (2.27) where x and are the co-ordinates of the mere Xs oe ON Yay age Are thy ‘dinates of th plastic centroid. 3.7 PLASTIC CENTROID For arbitrary sections, the axial load P is taken to act at the plastic centroid, the plastic centroid is defined as “the centroid of resistance of the section if all the concrete is compressed to a maximum stress and the steel is compressed to the yield stress with uniform strain over the section"(1). ‘The confined concrete within the core is assumed to have @ compressive stress of £1. and the stress of the cover concrete pressive st: f EL, and ths 8 us is calculated at a strain of 1. using the stress-strain curve for cover concrete. The steel stress is taken equal to the yield stress, EM, + *2Meore*: * "Peover*i) Xract (3.28) a x, + tf" + TEa, eae ot t peceteore : cover! end sinilarly for y.,54-3-8 MOMENT-CURVATURE ANALYSIS For the programs using the Mander model for calculating concrete stresses, a monent-curvature analysis can be done. This involves incrementing the concrete strain fron approx. c.,, and calculating Uw moment and curvature for each strain increment. The analysis stops when one of the stopping conditions has been reached. At the present, the analysis is available for cracked sections only, i.e. the concrete has zero tensile strength, but tensile capacity could be added at a later date. ‘The minimum practical concrete strain is calculated by: ete ol see (3.29) The programs start the analysis from a concrete strain of 0.0001. If ¢,,, is greater than this then 0.0001 is added to the starting strain until it is greater than ¢, This shortens the calculations slightly for the neutral axis iteration as the neutral axis depth is generally quite large for low strains and also keeps the strains in round figures. 3.8.1 Stopping conditions There are five conditions which could stop the analysis. ‘These are as follows: (1) if the moment capacity falls below 80t of the maximum moment capacity of the section; (2) 4f the concrete strain exceeds ¢. (Equation 2.25); (3) 4f the moment capacity is negative; (4) if the steel strain exceeds the upper limit of ©. (5) if the number of cycles exceeds that specified. 3.8.2 ‘The Ideal Moment Capacity At the end of the analysis the ideal moment, M,, is calculated. zabn (7) defined M, “to be the maximum moment reached hefora theaf curvature exceeds five tines $". ‘This definition is adoptea for these programs. Fig. 3.12 illustrates the two cases which could occur. 4) 9nd MY are the curvature and moment calculated when the extreme tension steel reaches the yield strain or when the extreme concrete compression strain reaches 0.002. this Second condition applies for columns with high axial loads where the steel has not reached yield before the concrete has reached @ strain of 0.002. For case two, M, is established by an iterative method. Once MY and $" have been calculated, the first estimate: ¥ ¥ = 1.25 x 59° 6. ta oy 3.30) Linear interpolation between calculated points is used to determine H,, for $,1. Then yield curvature ¢, {8 calculated by (3.31) is not equal to 54 then ¢., = 5: M,, interpolated If ¢, is qual to 56, ip = 58, Pe vig ial and the cycle continues until ¢, = 5¢, ( @ tolerance). For case one, M, is equal to the maximum monent as it is greater than the moment at a curvature of 54,ia steal yields oe ‘concrete sirin reaches 0.002 + hy CURVATURE Fig. 3.12 The Ideal Moment Capacity.39 CHAPTER IV INSTRUCTIONS FOR USING ALL RECTANGULAR AND CIRCULAR PROGRAMS 4.1 GENERAL ‘The five programs covered under this chapter are: (2) Uniaxial bending of rectangular sections - ACI stress block. (2) Biaxial bending of rectangular sections - ACI stress block. (3) Circular sections - ACI stress block. (4) Uniaxial bending of rectangular sections - Mander model. (5) Circular sections - Mander model. These programs will calculate the required area of reinforcing steel for a given axial load and moment or calculate the monent capacity of the section for a given area of steel. ‘The two prograns using the Mander model also do a moment curvature analysis. ‘There are also facilities to store and retrieve the input data on disk. Some code checks are available in the prograns using the ACI stress block. The units used are as follows: Section Dimensions - mm Material Properties - Pa Loads > kW and xm Steel Areas ~ om’ The function key £10 is set up as an escape key. If the wrong menu option is selected, or to stop calculations, then press the £10 key and RETURN at a prompt or just £10 ie the program is running to go back to the Main Menu.4.2 MAIN MENU OPTIONS ‘The options available on the Main Menu are: (2) put zaitor (2) Print Input Data on screen (3) Print Input Data on Printer (4) Section Design (5) Section analysis (6) Moment Curvature Analysis (Nander model only) (7) exit 4.2.1 Input Editor ‘The following options are available on the Input Editor: (2) Input section dimensions, (2) Input material properties, (3) Input ioaas, (4) Input reinforcement details, (5) Set control parameters, (6) Store data on disk, (7) Read data from disk, (8) List ana delete data files on disk, (9) Exit to the Main Menu. For entering data, the program will prompt the user for the required input. Within each prompt, the current value for the item is shown in square brackets. This is the default value and can be entered by pressing RETURN only, without typing in data. If the value is to be changed then type in the new data and press RETURN. If a mistake is made then it can be corrected by deleting the required characters before RETURN or by going back to that optio: from the INPUT EDITOR. 4.2.1.1 Input section dimensions Input the section dimensions and cover to the main reinforcing steel as shown in Fig. 4.1.Main bar diameter Tie diameter Tensien steel area (5 bars) Side steel area (6 bars} DEPTH Compression steel area (5 bars) oe ee | steel ratios - tension: compression: sides 5 Se ele tare 1 ie (a) Rectangular sections. cover Main bar diameter Tie diameter Total steel area @ bars OVERALL DIAMETER (b) Circular sections. Fig: 4.1 Input Parameters for the Prograns;4.2.1.2 Input material properties ‘The following values are initialized when the programs are run. £ = 275 MPa ¥ EB, = 200,000 MPa = 30 MPa Identifying mild steel or high strength steel in the Mander model programs enables the relevant stress-strain curve to be set up to include strain hardening of the steel. The default option is mild stel 4.2.1.3 Input loads Input the ideal axial load and noment applied to the i section, i.e, the axial load and moment have been divided by the relevant strength reduction factor(s). | ‘he applied moment is required for section design only. For the biaxial bending program, the design moments in both the x and y directions are required. 4.2.1.4 Input reinforcement details t Details of the reinforcement must be entered before running an analysis on the section. They-are not required for the design of the steel. (a) Uniaxial bending of rectangular sections ~ ACI. Enter the main bar diameter, tie diameter and the steel areas as illustrated in Fig. 4.12. (>) Biaxial bending of rectangular sections ~ ACI.I 4] I en “ ' co 4.2.1.5 43 ‘The total steel area is calculated using the number and diameter of the bars. Note that the number of bars must be divisible by 4 as the Program assumes an equal number of bars in each fa Circular sections - act. Enter the main bar diameter, tie diameter and total longitudinal steel area as in Fig. 4.1b. Uniaxial bending of rectangular sections - Mander, Enter values for the following: (4) Munber of bars, (44) average bar dianeter, (ma) (444) average centreline distance between bars, (mn) (iv) Steel areas, as per Fig. 4.2, (mm) (v) average nunber of tie legs per side, (vi) Tie diameter, (mm) (vid) Tie spacing, (mm) (viii) Tie strength, (MPa). This data is required to calculate the confining pressure f+ and k, for the Mander model. average values, rather than actual are adequate for the purpose, Circular sections ~ Mander. Enter values for the bar diameter, total steel area, tie diameter, tie spacing and strength, Hoop and spiral ties also need to be distinguished between for the calculation of k, Set control parameters ‘The program initializes the following value:number of segments - 30 iteration accuracy - 0.001 (uniaxial bending) (eps) 0.01 (biaxial bending) These values can be changed if necessary. 4.2.1.6 Store data on disk To store the input data, type in the file name (maximum of eight characters) and the disk drive required at the relevant Prompts. If no drive is specified then the conputer will default to the current diive. the program also provides a .DAT extension to the file name to identify it as a data file. The data files are stored in a suk-directory under each program. This is to avoid trying to load the wrong data file into a program. If the sub-directory does not exist then the program will create one and then store the data. If no characters are entered for the file name then the Program will go back to the INPUT EDITOR. 4.2.1.7 Read data from disk To load the input data, type in the file name and disk drive as in the procedure for storing data. 4.2.1.8 List and delete data files on disk ‘This will list all the .DAT files vithin the sub-directory of the current program. A prompt will then appear for the name of the file to be deleted. If no files are to be deleted just Press RETURN to go back to the INPUT EDITOR. 4.2.2 Section Design ‘This option will calculate the required area of reinforcing steel for a given‘axial load and moment. Before the reinforcement can be designed, the section dimensions, material properties and loads must be entered.ae Se 45 Initially the minimum steel content of the section is 0.88 of the gross area and the moment capacity calculated. If this moment is greater than the input value then the results are printed on the screen and printer. Note that this coresponds to the minimum column reinforcement allowed by N28 3101. If minimum steel is not sufficient the steel area is incremented using linear interpolation until the moment capacity equals the input moment (plus or minus eps times the input moment). The printed output includes information about the moment capacity and neutral axis depth at minimum steel, the required steel area, moment capacity, neutral axis depth and percentage of steel. Also the number of bars for a range of sizes is printed. For the ACI programs, the volumetric ratio for confining steel and information on the code checks is printed (see section 4.3). (a) Uniaxial bending of rectangular sections - act If the section is to be designed for seismic loads then type "Y" at the prompt. This is to determine whether to use section 6.4 or 6.5 in NZS 3101 for the code checks. Estimates for the main bar diameter and tie diameter are required so the section can be divided into segments. These values are not too critical and could be several sizes out without making much difference to the final answer. The ratios of the tension steel area to compression steel area to side steel area have to be entered. Press RETURN after entering each number. For uniform steel around the and 1:1:2 will be suitable but depend on the total nunber of bars. A ratio of 1:1:1.2 section ratios of between will generally cover most situations, (see Fig. 4.1a). The design is not particularly sensitive to thé side steel ratio. ‘The steel areas can be increased in one or more of the specified regions:>) te) (ay 1, Tension steel 2. Compression steel 3. Side steel To increase the steel in the required region, type in the number . For example, the combination 123 will increase the steel uniformly over the whole section but 12 will inerease the steel in the tension and compression regions only. The side steel will be left as it was for minimum steel. The numbers can be typed in any order. ‘The program will not allow steel to be increased in the compression region only. Biaxial bending of rectangular sections - ACI Type "Y" if the section is to be designed for seismic loading. Enter estimates for the number of bars (must be divisible by 4), bar diameter and tie diameter. As the steel is assumed to be distributed evenly around the section, the total steel area is increased if the minimum moment is not sufficient until the moment capacity in both the x and y directions equals the applied moments (+ eps x moment), cular sections ~ ACI Type "Y" if the section is to be designed for seismic loading. Enter estimates for the bar diameter and tie diameter. The total steel area is increased if the minimum moment is not sufficient. Uniaxial bending of rectangular sections and circular sections - Mander Enter estimates for the bar diameter, tie diameter, tie strength, f 10s) and k,- trength, £,, 9, (RHOs) and k, —wet should 47 The default value for £,, is the same as the input value for f. y The default value for.p, is calculated using Equations 6.22 and 6.23 of NZS 3101 for circular nections, or 6.24 and 6.25 of NS 3101 for rectangular sections, with £,, equal to the input value. The default value for k, of 0.7 for rectangular sections is fairly typical but values for thi could range fron 9.5 to 0.8 depending on the shape uf the section, the number of reinforcing bars and the tie spacing. As the section shape becomes long and thin k, will decrease. Also the fewer bars there are, the lower the value for k,. For circular sections, the default for k, is 0.9. typically, values for k, could range between 0.85 to 0.95, mainly depending on the spacing of the stirrups. The results are not particularly sensitive to k,. Enter the steel ratios for rectangular sections as outlined in section 4.2.2(a). Options are available to include the full strain-hardened stress-strain curve for steel and to choose the ultimate concrete compressive strain. Section Analysis Before running a section analysis, the reinforcement details be entered using the INPUT EDITOR. ‘The moment capacity of the section is calculated for the input reinforcement. An ultimate concrete strain of 0.003 is used for the ACI programs and the options of steel strain hardening and choosing the ultimate concrete strain are available on the two Programs using the Mander model. The actual values for p, and k, are calculated for the inputreinforcement and these are printed along with the moment capacity, neutral axis depth, percentage of steel, extreme steel strain and section curvature. 4.3 CODE REQUIREMENTS Chapter Six of N2S 3101: 1982 governs the design of menbers for flexure, with or without axial loads. 4.3.1 General ‘The following clauses of NZS 3101 regarding material strengths have been complied with: 63.1.3 Maximum usable strain at extreme concrete ‘compression fibre shall be assumed equal to 0.003, 6.3.14 Stress in reinforcement below specified yield strength fy for grade of reinforcement used shall be taken as E, times steel strain, For strains greater than that corres- ponding to fy, stress in reinforcement shall be considered independent of strain and equal to fy. 63:16 Relationship between concrete compresive stress distribution and concrete strain may be assumed to be rectangular, trapezoidal, parabolic, or any other shape that results in prediction of strength in substental agree- ‘ment with resulls of comprehensive tests, 63:17 Requirements of 63.1.6 may be considered satisfied by an equivalent rectangular concrete stress di- tribution defined by the following (@) Concrete stress of 0.85 fc shall be assumed uniformly distributed over an equivalent compression zone bounded by edges of the crosssection and a straight line located parallel to the neutral axis at a distance @ = fic from the fibre of maximum compressive strain (©) Distance ¢ from fibre of maximum strain to the neutral axis shall be measured in a direction perpen- dicular to that axis. (©) Factor B, shall be taken as 0.85 for concrete strengths f’, up to and including 30 MPs, For strengths above 30 MPa, By shall be reduced continu ously atthe rate of 0.04 for each § MPa of strength in excess of 30 MPa, but By shall be not taken a lest than 065.4.352 49 For Menbers Not Designed for Seismic Loading Checks are made against the following clauses of NzS 3101: 64.1.5 For columns and piers the maximum design axial load in compression at a given eccentricity shall not exceed O85 GP, for members with transverse reinforce- ment conforming to either 64.7.1 (a) or 64.7.2 (a) and composite members conforming to 6.4.12.7, nor 0.£0¢P. for members with transverse reinforcement conforming to either 64.7.1 (b) or 64.7.2 (b) and composite members conforming to 64.128, where Fo * O85 fe Ag—AudthAg = + + (E61) 64.6 Limits for longitudinal reinforcement in columns ‘and piers 64.6.1 Area of longitudinal reinforcement for non- composite columns or piers shall not be less than 0.008 nor more than 0.08 times gross area A, of section. 64.62 Minimum number of longitudinal reinforcing bars in columns and piers shall be six for bars in a circular arrangement and four for bars in a rectangular arrange- ment, 64.1 Limits for transverse reinforcement in columns and piers, 64.7.1 Spiral or circular hoop. reinforcement for colums and piers shall conform to 53.29 and 5.4.1 and Shall be placed a8 follows: (©) When a svength reduction factor @ of 09 i wed: (1) Volumetric ratio py shall be not less than the vue wen oy A, fe = 045(—8 Jf... (Eg. 63) a= oat Ve 63) were fal nl exceed 400 MP. 6.4.7.2 Hoop or tie reinforcement not of circular shape for columns and piers shall conform to 53.30 and 54.2 and shall be placed as follows: (When a strength reduction factor ¢ of 0.9 is used: (1) The total area in each of the principal ditec- tions of the cross-section within spacings, shall not be less than here fyy shal not exSed 400 MPa50 For clause 6.4.1.5 the maximum allowable axial load is taken as 0.8 P, for a strength reduction factor of 0.9. the maximum axial load of 0.85 P, for a strength reduction factor of between 0.7 and 0.9 has not been considered as the applied axial toad must be less than 0.1 £14, Note that the volumetric ratios calculated by Equations 6.3 and 6.4 of NaS 3101 are a maximum and may be reduced according to clauses 6.4.7.1(b) and 6.4.7.2(b). 4.3.3 Members Designed for Seismic Loading Checks are made against the following clauses of NZS 3101: 6.5.1.5 For columns and piers the maximum design sail load in compression at a pen ecentrcty shal not exceed 0.7 9 fog, unless Ht can be shown that isles than 0.7 $F, where Fo = O85 fe (Ag—Ay)* Syge «(Ba 613) 6.54.2 Longitudinal reinforcement in columns and piers shall be as follows: (2) Area of longitudinal reinforcement shall be not less than 0.008 Ap (b) Area of longitudinal reinforcement shall be not greater than 0.06 4g for Grade 275 steel nor greater than 0.045 for Grade 380 steel, except thal in the region of lap splices the total area shall not exceed 0.08 Ag for Grade 275 steel nor 0.06 Ag for Grade 380 steel6.5.4.3 Transverse reinforcement in columns and piers shall conform to 5.5.4 and 55.5 and shall be placed as follows: (@) In potential plastic hinge regions, as defined in 6.5.4.1, when spirals or circular hoops are used: (Q) Volumetric ratio A, shall not be less than the reater of: ‘Ag \ fe P, =0.45(“8— 1) 7¢ [0541.25 ec. noolt- )ilonagt) (Eq. 622) or fe os fe except that, where permitted by 6.5.4.3 (c), 6 may be reduced by onehalf (b) In potential plastic hinge regions, as defined in 65.4.1, when rectangular hoops with or without supplementary cross-ties are used: (1) Total ares of hoop bars and supplementary eross-ties in each of the principal directions of the eross-section within spacing sp shall not be Jess than the greater of (Eq. 624) ay =0.i2ynr foseras fe th Ag, (Eq. 625) ‘except that where permitted by 654.3 (c) Ash may be reduced by one-half ‘The confinement calculated by the programs is the full amount given by the appropriate equations in NZS 3101. Clause 6.5.4.3(€) allows a 50% reduction for colunns of frames protected against plastic hinging by capacity design procedures.