FACTA UNIVERSITATIS

Series: Mechanical Engineering Vol.1, No 7, 2000, pp. 799 - 807

CALCULATION OF LOAD CAPACITY OF GEARS

IN RANDOM VARYING EXPLOITATION CONDITIONS
UDC 621.83:621.86.016 62.233.1/.9:621.86.016

Vojislav Miltenović, Dragan Milčić

The Faculty of Mechanical Engineering, Niš, Beogradska 14, 18000 Niš

Abstract. A high degree of accuracy in dimensioning and controlling the safety from
destruction can be reached by measuring exploitation load of the vital elements of the
construction. The choice of the adequate load spectrum is performed by introducing a
number of representative spectrums for particular operating conditions, each of these
having its probability of occurrence. The easiest way to solve the problem is to
introduce the standardised load spectrums with the aim of forming the catalogues of
the standardised load spectrums for representative mechanical systems.

INTRODUCTION
The requirements of the market are constantly growing more complex with regard to
productivity, quality and speed of mastering new products. Rapid development is the
cause of the increase of designing and constructional tasks with even greater degree of
complexity. The conventional "traditional" designing, based on empirical results and
intuition, does not suffice for successful following of progress in other areas of human
activity. In the engineering practice today, automatised designing is an imperative, which
is caused by the use of computers in the process of developing a product.
New tools for overcoming the flaws of the classical approach can be classified under
the term CIM- Computer Integrated Manufacturing- which defines the integrated flow of
all information both within and outside the company. CIM consists of modules such as
CAD (Computer Aided Design), CAM (Computer Aided Manufacturing), CAPP
(Computer Aided Process Planning), CAQ (Computer Aided Quality) etc.
The utilities and tools of CAD technology can be most widely applied in
automatization of the designing process and they appear as the support mechanisms in the
object modelling processes, in engineering analysis and documentation. The applicability
of the CAE (Computer Aided Engineering) is limited to the process of engineering
analysis (FEM analysis and simulation) while CAPP and CAM can be used in modelling

Received August 20, 2000

Presented at 5th YUSNM Niš 2000, Nonlinear Sciences at the Threshold of the Third Millenium,
October 2-5, 2000, Faculty of Mechanical Engineering University of Niš
800 V. MILTENOVIĆ, D. MILČIĆ

and designing (elaboration) operating procedures. The architecture of the intelligent

programme for designing toothed gearing, being developed at the Faculty of Mechanical
Engineering in Nis, is presented in Figure 1.
The structure of the system for designing toothed gearing is very complex and
heterogeneous. The system has been developed from a modular principle enabling to
perform, with the aid of the computer, certain activities and tasks of the designers. The
basic task of the system is to enable an integrated application of different programme
modules and systems developed by different authors and companies, which should be
used in automatization of certain activities in designing toothed gearing. This is the reason
why the software basis of a developed system rests on the maximal application of all
available standards in the area of data flow, communication and computers.
The programme for designing toothed gearing consists of 6 basic programme modules
integrated in a single totality:
1. the programme module for calculating gearing pairs,
2. the programme module for calculating roller and antifriction bearings (plain
bearing, sleeve bearing, slide bearing),
3. the programme module for choosing and calculating the shaft-hub joint,
4. the programme module for calculating and shaping the shaft,
5. the programme module for calculating intermediate belt gearing,
6. the programme modules for modelling and generating constructional documents
for integral parts of gearing (Auto CAD, GENIUS, Pro/ENGINEER)

Fig. 1. The activities and computer methods

applied in the process of toothed gearing design

The programme module for calculating the gearing pairs consists of:
ZPS1 - the programme module for calculating and optimisation of gearing geometrical-
constructional characteristics,
ZPS2 - the programme module for final calculation of cylindrical and cone gearing,
ZPS3 - the programme module for initial designing of worm gearing,
ZPS4 - the programme module for calculating the geometry and strength of worm
gearing pairs.
 Calculation of Load Capacity of Gears in Random Varying Exploitation Conditions 801

The programme module ZPS2 is intended for the final calculation of gearing and
offers the following possibilities:
− the calculation of precise geometry of cylindrical involute gear with outer and inner
serration for the defined parameters of the tool and the calculation of cone gearing,
− automatic choice of the centre distance for the previously defined standard sequence,
− automatic determination of the addendum air gap and shortening of tooth head for an
integral number of addendum circle radius,
− the calculation of factors of working conditions by means of load spectrum,
− the calculation of load capacity with respect to DIN-3990, AGMA- standard and
ISO-recommendations,
− the calculation of gearing elastic deformation,
− the calculation of load capacity with respect to seizing and pitting, violent and wear
fracture,
− the calculation of load capacity by simulating the exploitation conditions,
− the calculation of load capacity of cylindrical and conical involent gears with
straight and helical gear.
The paper deals with calculating the gear load capacity in random varying exploitation
conditions by simulating the exploitation conditions.

1. THE FUNDAMENTALS OF CALCULATION

The quality of design results is considerably higher with regard to the fact that the
application of information technology eliminates in a great degree the subjective sources
of error.
The most important constituent with regard to safety and reliability of MS is the load
capacity of its constructional elements. The load capacity can be as well be used to
measure the quality of the constituents. The calculation of load capacity can be performed
by comparing the operating and critical stress; therefore, the evaluation of quality can be
considered in terms of evaluation of exactness of the size of working and critical stress.
The accuracy of the operating stress calculation depends on the appropriateness of the
methods applied and the manner of determining the load. Whether a method is
appropriate is determined by its possibilities for identifying the state, position and degree
of stress. The quality of the methods applied is tested experimentally. The degree of the
operating stress depends on the load; it is therefore necessary to know the degree, course
of change and frequency, as well as the probability of occurrence of the highest stress
appearing during life for the observed part or element of the construction.
The current testing of the safety of mechanical parts against destruction is performed
by determining the nominal load, i.e. the nominal operating load, which corresponds to
the most frequent stress in the course of operation. The exploitation conditions, i.e. the
occurrence of load greater than the nominal is taken into consideration by way of the
factor of operating conditions chosen from the corresponding tables. A conclusion can be
drawn that this way of determining the operating load offers approximate values, which
leads to inadequate dimensions of the construction elements.
A high degree of accuracy in dimensioning and testing the safety from destruction can
be accomplished by measuring the exploitation load and identifying the load spectrum for
802 V. MILTENOVIĆ, D. MILČIĆ

the vital elements of the structure. The load spectrums can be obtained on the basis of
experimental MS in the process of operating for certain conditions; therefore each load
spectrum has its probability of occurrence. The problem of choice of the appropriate
spectrum is resolved by introducing a greater number of representative spectrums for
certain operating conditions, which allows for an adequate evaluation of all the inter
conditions. The load spectrums can be expressed in terms of a relative frequency of
certain amplitudes f (x) where
σ ai size of any stress amplitude
f ( x) = = (1)
σ a max max. stress amplitude

The load spectrum represents a basis for determining the critical stress as well. With
current methods of calculating the critical stress of dynamically strained pieces,
dynamical endurance, obtained in the continual relation of the lowest and highest stress
for the probability of destruction of about 0,5, is used. For a more precise calculation of
the safety of pieces against destruction it is necessary to introduce endurance
corresponding to the manner and number of changes in operating stress in the operating
life of the piece observed - operating endurance. Data about operating endurance are
obtained in experimental research of the test pieces or elements for particular spectrums
of operating stress in laboratories. The basic problem is a wide range and duration of the
research conditioned by the need to conduct the research on a greater number of load
spectrums and with a greater number of load spectrums in order to obtain a more realistic
picture about the size and laws of scattering of the research results. Therefore, this
problem is solved by finding the relation between the operating and basic endurance i.e.
by applying the hypothesis on the accumulation of damage of material. The basis of this
hypothesis is the supposition that every stress σi in a spectrum adds to damage
proportional to the number of circles to destruction Ni, where ni<Ni. Destruction occurs
when the sum of the damages reaches the value
ni
D=∑ = ∑ Di (2)
Ni
According to some hypothesis D = 1.
The experimental research of this hypothesis for different stress states has revealed
significant deviations. Therefore, the hypothesis has been altered in different ways
primarily with regard to the sum of damages etc.
The size of the load of the MS elements can be obtained by a computer simulation of
the exploitation conditions. The course of the life calculation is presented in Figure 2.
The starting point is the time function of the load change, discretization of which to
one or two parameters provides the load spectrum. By introducing the correctional
dependence between endurance and load spectrum (2) graphic dependence of the life for
a standardised load spectrum is obtained, i.e. the obtained output represents the available
piece resource in the form of interdependence of reliability and life. The basic problem is
how to obtain the time function of the load change. The best way is to measure load in
exploitation conditions. Taking into consideration that this is a long-lasting and expensive
procedure, such a function is usually obtained in the accelerated laboratory or polygon
research; it serves as a basis to obtain the corresponding load spectrums. For load
 Calculation of Load Capacity of Gears in Random Varying Exploitation Conditions 803

spectrums defined in this manner it is possible to obtain the corresponding load spectrums
for similar systems.

Fig. 2. The relative life for the standardized load spectrum

2. THE CALCULATION OF THE GEAR LOAD CAPACITY FOR THE CRITERIA

OF ENDURANCE OF THE TOOTH FLANK AND ROOT

The calculation rests on the fact that each gear revolution has a corresponding
damaging effect. The effect of damage depends on the size of stress; it can be disregarded
with small stress. The life obtained is a measure denoting the available resource of the
material. For such a calculation to be possible, it is necessary to have a load spectrum,
exact characteristics of durability of material and applicability of appropriate hypothesis
on the accumulation of damage of material. Our calculation is based on Palmgren-Miner
hypothesis where D ≈ 1.
The load spectrum is aproximated with corresponding blocks whereas each torque has
a corresponding number of cycles of load change ni. In Figure. 3, besides the load
spectrum, the Veler curve of the tooth load is represented.
804 V. MILTENOVIĆ, D. MILČIĆ

Fig. 3. Load spectrum

For each of the load spectrum levels Ti the corresponding contact stress is determined
according to:
2000 ⋅ Ti u + 1
σ Hi = Z H ⋅ Z E ⋅ Z ε ⋅ Z β ⋅ Z B , D ⋅ K vi ⋅ K Hβi ⋅ K Hαi (3)
d1 2 ⋅ b u
where:
ZH – zone factor, Zβ - helix angle factor,
ZE – elasticity factor, Zε - contact ratio factor,
ZB,D – single pair tooth contact factor for the pinion, for the wheel,
Kvi – dynamic factor, KHβi – face load factor, KHαi – transverse load factor.
Aplication factor KA has not been used in the expression (3), since it is taken into
consideration via the load spectrum.
The characteristic of the endurance of material is determined by the Veler curve i.e.
the damage line.
For example, for C5426 endurance characteristics have been presented in figures 4.a
and 4.b.
The inclination of the Veler curve (Fig. 4a) is:
N BD
log 5 ⋅ 107
log
N BS 105 = 10,19
p= = (4)
[σ H ]s 2384
log log
[σ H ]D 1295.8

The number of load change cycles to the fracture for any of the levels Ni equals:
log N i = p ⋅ {log[σ H ]s − log( S H ⋅ σ Hi )} + log N BS (5)

The tooth root stress for each of the load spectrum levels Ti is determined by
2000 ⋅ Ti
σ Fi = YFa ⋅ Ysa ⋅ Yβ ⋅ Yε ⋅ K vi ⋅ K Fβi ⋅ K Fαi (6)
d1 ⋅ b ⋅ mn
 Calculation of Load Capacity of Gears in Random Varying Exploitation Conditions 805

2400 2384

2200

2000
σH lim N /m m 2

1800

1600

1400
1295,8

1200 4 5 6 7 8 9
10 10 10 10 10 10
NL

a)
2400
2255,8
2200

2000
σFlim N /m m 2

1800

1600

1400

1200

1000
902,3

800 2
10 103 104 105 106 107
NL

b)
Fig. 4. Damage lines for damage statistical probability of 10%

The characteristics of the endurance of material are determined by damage curve

presented in Fig.4, whence
N
log BD 3 ⋅ 10 6
log
N BS 103 = 8,7378
p= = (7)
[σ F ] s 2255.8
log log
[σ F ] D 902.3

The number of load change cycles to fracture for any of the levels Ni is
log N i = p ⋅ {log[σ F ]s − log( S F ⋅ σ F i )} + log N BS (8)

The algorithm of the calculation of the blocks safety degree SH and root of teeth SF is
presented in Figure 5. The calculation is iterated until the damage is in the scope of
0,95<D<1,05.
806 V. MILTENOVIĆ, D. MILČIĆ

Example: Cylindrical gear pair z1 = 20; z2 = 73; b = 106 mm; a = 580 mm; mn = 12 mm;
β = 60; gear material C5426 cemented/C5431 induction hardened; life Lh = 50000 hours;
nominal power P = 900 kW; input number of revolutions n = 1080 min-1 ; load spectrum
defined in the table 1:
Table 1.
Load block Relative time participation Ti/Tn
1 0.15 0.9
2 0.4 1
3 0.23 1.2
4 0.12 1.6
5 0.1 1.8

Fig. 5. The algorithm of the calculation factor of safety

from tooth breakage and factor of safety from pitting

The results of the calculation have been presented in Table 2.

Table 2.
i ni Ti [Nm] σHi [N/mm2] σFi [N/mm2]
1 8100000 7161.75 735.61 190.53
2 21600000 7957.50 770.82 209.84
3 12420000 9549.00 836.80 247.90
4 6480000 12732.00 957.34 325.14
5 5400000 14323.50 1013.84 364.66
SH = 1,458
SF = 2,1899
 Calculation of Load Capacity of Gears in Random Varying Exploitation Conditions 807

3. CONCLUSION
The following conclusions can be drawn:
1. In order to obtain the exact results of the calculation of operating load it is necessary to
know the size, course of change, frequency as well as the probability of occurrence of the
greatest loads appearing during life for the part of the construction being observed.
2. The size of load of MS elements is obtained by experimental research or a computer
simulation of the exploitation conditions.
3. The calculation of gears is based on the fact that each gear revolution has the corresponding
damaging effect, i.e. the calculation has been made for Palmgren-Miner hypothesis.
4. An integrated, intelligent system for simultaneous designing of gearing has been
developed at the Faculty of Mechanical Engineering in Niš. This programme enables to
perform calculations for gears by simulating the exploitation conditions.

REFERENCES
1. Miltenović V., Ognjanović M.: Mašinski elementi-II - elementi za prenos snage, Mašinski fakultet Niš -
Beograd, 1995.
2. Niemann G., Winter H.: Maschinenelemente, Band II - Getriebe allgemein, Zahnradgetriebe- Grundlagen,
Stirnradgetriebe, Springer-Verlag, 1985.
3. DIN 3990 Teil 41
4. Miltenović,V., Milčić, D.: Proračun nosivosti zupčastih prenosnika simulacijom eksploatacionih uslova.
Naučno-stručni skup "Istraživanje i razvoj mašinskih elemenata i sistema" - IRMES ‘98, 10. - 11. septembar
1998., Beograd, s.235-240.

PRORAČUN NOSIVOSTI ZUPČANIKA U NESTACIONARNO

PROMENLJIVIM EKSPLOATACIONIM USLOVIMA
Vojislav Miltenović, Dragan Milčić
Visok stepen tačnosti pri dimenzionisanju i provera sigurnosti protiv razaranja može se postići
merenjem eksploatacionih opterećenja vitalnih elemenata konstrukcije. Izbor merodavnog spektra
opterećenja vrši se uvođenjem više spektara reprezenata za određene radne uslove pri čemu svaki od
njih ima svoju verovatnoću pojave. Problem se najlakše rešava uvođenjem normiranih spektara
opterećenja sa težnjom da se formiraju katalozi normiranih spektara opterećenja za reprezentativne
mašinske sisteme.