Eng. &Tech.Journal, Vol.33,Part (A), No.

4, 2015
 
 

A Cumulative Damage Model for Fatigue Life Prediction

Based on Dynamic and Static Deflections

Dr. AlalKawi H.J.M

Electromechanical Engineering Department, University of Technology/Baghdad
Email:[email protected]
Dr. Nazhat S. Abdul_Razaq
Electromechanical Engineering Department, University of Technology/Baghdad
Email:[email protected]
Hasan H. Juhi
Electromechanical Engineering Department, University of Technology/Baghdad
Email:[email protected]

Received on: 4/9/2014 & Accepted on:2/4/2015

ABSTRACT
The main goal of this study is to report experimental evidence about the
accumulative fatigue damage behavior of CK35 steel alloy at room temperature and
zero main stress (R = -1) . A non-linear accumulative damage model based on static
and dynamic deflection is proposed, considering the loading sequence into account
and the constant S-N curve. Satisfactory predictions of cumulative fatigue life have
been observed when applying the proposed model to the two block loading sequence.
Comparison between LDR and proposed model has been made. It was found that
LDR can underestimate the fatigue damage and the proposed model showed good
agreement with the experimental results. A non-linear damage model seems to be a
proper choice for predicting the cumulative fatigue two block loading histories.

Keywords: fatigue damage accumulative, non-liner model, static and dynamic

deflection, CK35 steel alloy.

‫انموذج للضرر التراكمي لتخمين اعمار الكالل معتمدا على االنحرافات المستقرة‬
‫والداينميكية‬
:‫الخالصة‬
‫ تحت ضرر الكالل عند درجة حرارة‬CK35 ‫الھدف االساسي لھذه الدراسة ھو لتوثيق تصرف سبيكة الفوالذ‬
‫ تم اقتراح انموذج ال خطي للضرر التراكمي يعتمد على االنحرافات المستقرة‬.(R=-1) ‫الغرفة و نسبة اجھاد‬
‫ تم الحصول على تخمينات مقنعة عند استخدام‬. S-N ‫ اخذ بنظر االعتبار تتابع الحمل و منحنيات‬.‫والداينميكية‬
‫ مع االنموذج‬LDR ‫ تمت المقارنة بين‬. ‫االنموذج المقترح على تتابع االحمال ذات الحزمتين للفحوصات العملية‬
‫ لذلك يبدو ان االنموذج الالخطي للضرر ھو‬. ‫المقترح وقد اعطى االنموذج المقترح توافق جيد مع النتائج العملية‬
. ‫االختيار االفضل لتخمين الكالل التراكمي من نوع الحزمتين‬

  762
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

Notation:
Symbol Title
σ bending stress in (MPa)
P Force (N)
The arm of load equal to 135.7 in (mm)
dmin , d The minimum diameter of the specimen
a,b The fitting coefficient
I The second moment of inertia
R2 correlation coefficient
M Bending moment
E Module of elasticity
R Radius of curvature
N The applied cycles at σi .
D Accumulative fatigue damage
static , s Static deflection
, The applied stress at low and high stress levels respectively.
dynamic, d Dynamic deflection
X Variable quantity related to the loading stress level
Function of the applied load
Β The inverse slope of The S-N curve
S Slope of static deflection No. of cycles to failure
S Slope of dynamic deflection No. of cycles to failure
Nf Number of cycles to failure
xi Function of loading sequence
L The distance from the applied load p to fixed support
D exp Experimental Damage
D miner Miner Prediction Damage
D model Proposed model Damage

INTRODUCTION

F atigue cracks occur when the material is subjected to dynamic loading (cycle or
fluctuating loading) although the stress produced is below the yield stress of
the material. A method of determining the cycle stress based on its
deflection is presented in this work. The stress gradients contribute to both axial load
and bending moment. The accurate measurement of deflection has become an
important consideration in the design of the fatigue components [1, 2]. This paper
examines the fatigue behavior of CK35 steel alloy by investigating fatigue life and
strength depending on the dynamic deflections.
Recently, the fatigue behavior of metallic materials in high cycle fatigue (HCF)
region has drawn great attention because the development of modern industry often
needs metal structure such as railway wheels, turbine disks in aerospace [3].
Aid et al [4] proposed a new damage model based on S-N curve and the loading
history. They concluded that the model results are in good agreement with
experimental data.
Zhang et al [5] proposed a nonlinear fatigue damage model was applied to the
vehicle bridges life cycle. This model was providing to be an effective method in
predicting the bridges progressive fatigue damage due to random variables.

763
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

Al – Garni et al [6] tested AISI304 stainless steel under variable loadings

experimentally and using finite element analysis. It was found that as the load step
increased the summation of damage become around to unity. The finite element (FE)
used in above work gave a good correlation with the experimental results.
Zengah S. et al [7] proposed a damage stress model to evaluate the cumulative fatigue
data and they concluded that the application of the proposed model showed a safe
prediction of fatigue compared to the experimental results.
Fatigue and fatigue damage accumulation of P335N.L1 steel were assessed using
miner rule and non-linear damage models. It was found that miner rule produced
inconsistent result while the non-linear theory gave reasonable result and final
conclusion was that the non-liner damage theory proposed by Marco and Starkey
seem to be a proper choice [8].
A comprehensive overview of research activities of accumulative fatigue damage
theories has been presented by Yang and Fatemi [9].
The main object of this work is to propose a non-linear fatigue damage accumulation
and to report experimental evidence about this model.

Experimental Work:
Material selection:
Medium carbon steel CK35 is selected because it has a widely used in many
applications where better properties than those for mild steel are required. The
material was received from a state company – Mechanical Industries AL-Ascandarya
– and tested to determine its chemical composition and mechanical properties. Table
(1) gives the chemical analysis which is done in engineering center for testing and
recondition. The relevant mechanical properties are listed in Table (2).

Table (1) chemical analysis of CK35 steel alloy

CK 35 C Mn Si S p
Standard
0.32-0.39 0.5-0.8 0.15-0.35 0.035 0.035
DIN50114
Measured 0.33 0.75 0.25 0.025 0.013

Table (2) mechanical properties of CK35 steel alloy

CK 35 σu (MPa) σy (MPa) E (GPa) G (GPa) Poisson's ratio
Standard
550-700 ˃ 392 210 81 0.3
DIN50114
Measured 660 400 205 80 0.28

The mechanical properties listed in Table (2) were conducted in the Production and
Metallurgy Eng. Dep. University of technology using the tensile test rig (WDW –
200E). The specifications of the tensile test and specimen configuration have been
restricted according to the American Society for testing and Materials (AST81-8).
The above results are the average of three readings.

Fatigue test:
The specimens of the fatigue test were prepared according to the Machines manual as
shown in Figure (1). 20 specimens were manufactured and tested to generate the S-N
curve by an alternating bending specification of fatigue test machine (rotating
bending fatigue testing machine) shown in Figure (2)

764
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

.
Figure (1) Rotating bending fatigue specimen (cantilever beam) according to the
American Society for testing and Materials (AST81-8).

Figure (2) rotating bending testing machine

The application stress is calculated from the applied moment according to the
simple theory of a cantilever beam as:[10]

∗ 32 ∗
σ … 1
∗

Where:
σ is the bending stress in (MPa). P is the force in (N.). is the arm of load
P 135.7 mm. d is the minimum diameter of the specimen(mm).
Figure (3) shows the application of load with its arm.

(Load) P

Figure (3) schematic diagram shows the application of load with its arm.[11 ]

765
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

The number of cycles was recorded by using a digital counter and the relation
between bending and the number of cycles can be expressed in power law regression
as the following equation [12]. The results of Ref [12] were observed that the values
of R2 located between 0.99 and 0.97.

… (2)

Where
a, b are the fitting coefficient [13]. The parameter (a) is related to static bending
strength while the parameter (b) is related to fatigue degradation and describes the
fatigue sensitivity. A correlation coefficient (R2) was used to verify whether the
experimental data are explained by power formula. The correlation coefficient can be
calculated by equations mentioned in ref [12]. The closer is R2 to unite the stronger is
the relationship between stress or deflection and number of cycles to failure.

Measurement of Deflection:
The experimental evaluation of deflection measurement was performed using the
electronic sensor (digital dial gauge) held with the fatigue test. The sensor was fixed
at the mid-span of the specimen. The properties of the electronic sensor (digital dial
gauge) are:
- time response 0.5 m/s
- resolution digital indictor 0.01mm
- measuring range 0-12.7 mm
- power , one silver oxide battery 1.5v
- operation temperature 0C ̊-40C ̊

Digital dial gauge

  sensor

Figure (4) the sensor holding on the test rig

Theoretical consideration:
Consider a shaft (fatigue specimen) of minimum cross section area ( min),
where dmin is minimum diameter of specimen. It is subjected to a vertical load (point
load) as shown in Figure (3).
The simple theory of elastic bending states that [14] .

… 3

Where
M is bending moment.

766
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

I is the second moment of inertia.

is the bending stress .
E is the module of elasticity and R is radius of curvature.
Using the Bernoulli – Euler's elastic equation which states that:

… 4

The bending moment equation is at section X (minimum diameter of specimen) for

the action of P to the minimum diameter which is equal to 135.7 mm.
Equation (4) becomes:

– … 5
6 2 3

Where
L is the distance from the applied load p to fixed support.
On integrating and solving the above equation with required boundary conditions, we
get the down word deflection of the beam as:

∗ 4.83 ∗ 10 … 6

.
L is equal to (156.7) mm based on the design of the fatigue testing. E = 200 GPa for
CK 35 steel alloy, I = 6.2 = 72.5 mm4

Results and Discussion:

The results obtained from the experimental work are listed and discussed in details
through this section. The experimental results include results of fatigue under
constant amplitude loading, fatigue under variable amplitude loading and deflection
measurement in static and dynamic loading test.

Constant fatigue test results:

The constant fatigue test was carried out at room temperature and stress ratio R = -
1. The results are graphically display in the form of S-N curve. These curves are
obtained by curve fitting the experimental data of fatigue test. The fatigue results with
the power law equation constants which express the fatigue behavior of CK35 steel
alloy and its correlation coefficient (R2) are given in Table (3). It is noted that these
equations have relatively high correlation coefficient which indicates that they are
well described by power law formula. This finding is in good agreement with ref.
[12]. Table (4) gives the analytic static deflection and the measured static deflection
using the electronic sensor (digital dial gauge) corresponding to the load in (N) and
bending stress in MPa .

767
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

Table (3) constant fatigue results with the regression parameters

Specimen no. σf Nf cycles Nf cycles a b R2
(MPa) average
1,2,3,4,5 175 164000,198000,210000 190666 1169 -0.156 0.9992
230000, 180000
6,7,8,9,10 200 95000,86000,90000 90333
92000, 89000
11,12,13,14,15 300 4000,5000,9000 6000
4500.8000
16,17,18,19,20 325 3000,5000,3500 3833
3500,5500

Table (4) analytic and measured static deflection.

P(N) load σb N/mm2 static mm (analytic) static mm
(measured)
86.15 500 0.41 0.46
69 400 0.33 0.37
56 325 0.27 0.24
51.7 300 0.25 0.20
34.5 200 0.17 0.19
30.15 175 0.15 0.165
17.23 100 0.08 0.13

Table (5) the static and dynamic deflection corresponding to the values of bending
stress listed in Table (3).

Table (5) the static and dynamic deflection corresponding to bending stress
σb (MPa) static (mm) dynamic (mm) Nf cycles Av.
175 0.13 0.18 190666
200 0.19 0.22 90333
300 0.20 0.27 6000
325 0.24 0.33 3833
Note that the above data of deflections are the average of five readings.

350
300 S‐N curve under static 
250 stress
200 σf = 1168.9Nf‐0.156
150
stress 

100
50 σf = 672.69Nf‐0.157
0 S‐N curve under dynamic 
0 50000stress 100000 150000 200000 250000
number of cycles to failure

Figure (5) the typical S-N curves

768
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

The proposed model:

Marco and Starkey [15] proposed the first nonlinear load depending damage
theory, written by a power law:

… 7

Where
xi is a variable quantity related to the loading stress level.
Pereira et al [16] proposed a nonlinear model for low-high and high-low loading
sequence.

1 For (low - high) loading stress level … (8)

α

1 For (high – low) loading stress level … (9)

Where
is a function of the applied load, to be obtained from the experimental data.
Alalkawi et al [17] presented a nonlinear damage model taking into account the effect
of loading sequence and shot peening treatment using medium carbon steel.

∑
… (10)

Where
x represented the effect of loading sequences and shot peening treatment (β). x
can be defined as :

x= β For low to high loading sequence … (11)

And x = β For high to low loading sequence …(12)

Where
β is the inverse slope of the S-N curve.
Following the above researchers, damage can be defined for two loading sequence
as:

For low-high loading

D ∗ … (13)

Where
δ , S are static deflection and slope of static deflection and No. of cycles to
failure curve respectively.

769
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

Figure. (6) Shows the δ -Nf curve , the best fit equation which describers the
experimental data may be taken the form :
.
δ 0.904 ... (14)

δ , S are dynamic deflection and slope of dynamic deflection No. of cycles to

failure respectively. Figure (6) gives the relation between the δ and Nf , the
experimental data may be described the following equation :
.
δ 0.443 ... (15)

0.35
0.3
deflection (mm)

0.25
0.2
0.15 static
0.1
dynamic
0.05
0
0 50000 100000 150000 200000 250000

number of cycles to failure 

Figure (6) deflection-number of cycles to failure curve (static &dynamic)

Equation (13) can be equated to the ratio of number of cycles considering the
effect of loading sequence as:

∑ …(16)

Where
xi is a function of loading sequence.
n is the applied cycles at σi .
And Nf is the cycles to failure at σi .

Combine equation (13) and (16) to get:

δ S
… 17
δ S

Where
R is the number of programs and , are the applied stress at low and high
stress levels respectively.

770
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

For high – low loading:

Here, damage may be defined as the inverse of equation (13)

D ∗ … (18)

Equation (17) can be written in form:

δ S
… 19
δS

The application of proposed model to experimental results:

Table (6) gives the fatigue life prediction using the proposed model compared to the
experimental data.

Table (6) fatigue life prediction using the proposed model*

Spec. No. Loading Nf D R Nf D R
sequence(MPa) Ava. exp. exp. model model model
1,2,3,4,5 200-250 39000 1.245 3.9 35836 1.05 3.5831
6,7,8,9,10 250-200 34667 1.107 3.46 32757 0.9598 3.2757
11,12,13,14,15 175-300 15334 1.293 1.53 12795 1.0573 1.2795
16,17,18,19,20 300-175 14334 1.208 1.43 11684 0.9655 1.1684

*For more details, see appendix (A) and calculation.

The most common methods to investigate the accumulative fatigue damage is the
block loading experiments. The accumulative damage under two blocking usually
evaluated by miner rule or linear damage rule (LDR). The experimental damage (D
exp.) in Table (6) is greater than unity while Miner damage is equal to unity. But the
nonlinear damage model proposed give D model greater than unity for low-high
loading sequences and slightly less than unity for high-low loading sequences. This
means that, the LDR produces clearly inconsistent prediction. For the present model ,
the loading histories beginning with a low (L) block loading are more damaging than
sequences starting by high (H) blocks [18] while miner prediction states than D=1 for
all cases of loading histories .

CONCLUSIONS:
1- The fatigue damage accumulation for two block loading sequences of CK35
steel alloy was investigated at room temperature and stress ratio R = -1 .
2- The LDR rule trend underestimated the accumulative fatigue life.
3- A non-linear proposed model based on static and dynamic deflections was
suggested and it gives satisfactory fatigue life predictions compared to the
experimental results.

771
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

Appendix (A):
For low-high loading sequences

… 1

where calculated from the deflection equations which are obtained
experimentally (see Figure. (A1))

and …(2A)

Figure (1A) Relation between stress and deflection

Table (A1) gives the values of corresponding to applied stress used in
cumulative fatigue damage testes

Table (A1) against applied stress

200 0.226 0.3384
250 0.2748 0.415
175 0.2017 0.3001
300 0.3235 0.4915

While Ss and Sd where obtained from deflection-No. of cycles to failure . see Figure.
(2A)

0.4
deflection (mm)

0.3

0.2
static
0.1 dynamic
0
0 50000 100000 150000 200000 250000
number of cycles to failure 

Figure (2A) the deflection – NO. of cycle to failure curve

772
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

The value of D corresponding to applied stress are given in Table (A2)

Table (A2) damage D corresponding to applied stress level

200 1.05 0.951

250 1.041 0.9598
175 1.057 0.9452
300 1.035 0.9655

Table (A3) shows a comparison between LDR and proposed model based on the
experimental results
Loading Nf exp. Nf miner Nf Proposed D D model
sequence(MPa) cycles (cycles) model (cycles) miner

200-250 39000 31318 35836 1 1.05

250-200 34667 31318 32757 1 0.9598
175-300 15334 11858 12795 1 1.0573
300-175 14334 11858 11684 1 0.9655

It has been showed that the damage expressed by the right hand side of equation
(11) (proposed model) is greater than unite , see Table (A3) , because the
accumulative of damage at low stress level is much greater than at high levels
[19][20] while the damage is less than unity if the test stars at high loading sequence
equation (13) . For this reason, LDR is sometimes dangerously optimistic [5]

Calculation:
200-250 (MPa)

. ∗ .
1.05
. ∗ .

Appling equation (11)

D 1.05

0.06531 . 0.254 . 1.05

0.293R = 1.05 R = 3.5831
Nf = R [ 10000] = 35831 cycles

250-200(MPa)
D -1 = 0.9598 Appling equation (13) gives 0.293 R = 0.9598
R = 3.2757
Nf = R [ 10000] = 32757 cycles

175-300 (MPa)

773
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

. ∗ .
1.0573
. ∗ .
Appling equation (11)

1.0573
0.0258 . 0.8175 . 1.0573
0.8263R = 1.0573 R = 1.2795
Nf = R [ 10000] = 12795 cycles

300-175(MPa)
D -1 = 0.9655 Appling equation (13) gives 0.8263R = 0.9655
R = 1.1684 Nf = R [ 10000] = 11684 cycles

REFERENCE:
[1] Yin Zhang , Ya-Pu zhao ((An effective method of determining the residual stress
gradients in a micro – cantilever)) Microsyst Technol 12,357-364 (2006)
[2] Cargi Majumder , Kaushik Kumar ((deflection and stress analysis of a cantilever
and its validation using ANSYS)) International Journal of mechanical Engineering
and research , Vol.3 , No.6 , PP.655-662 (2013).
[3] Sun, C. Xle A. , Zhad Z. Lei , Hong Y. ((accumulative damage model for fatigue
life estimation of high strength steel in high – cycle and very high – cycle fatigue
regimes)) FFEMS ( Fatigue and fracture Engineering and structures ) 35-638 (2012)
[4] Aid A. , Amrouche A. , Bouiadjra B.B. , Benguediab M. , Mesmacque G.
((fatigue life prediction under variable loading based on a new damage model))
Materials and Design 32, 183-191 (2011)
[5]Zhang W, Cai C.S. , Pan F. , (( nonlinear fatigue damage assessment of existing
bridges considering progressively deteriorated road condition )) Engineering
structures 56,1922-1932 (2013).
[6] Al-Garni A.M. , Abdulellatife A.K. , Hamed M.A. (( effect of load step and cyclic
ratio on accumulative fatigue damage for stainless steel 304)) , to be published
(2015).
[7] Zengah S. , Aid A. , Benguediab M. , ((comparative study of fatigue damage
models using different number of classes combined with the rain flow method))
Engineering technology and applied science research , Vol.3 , No. 3 , 446-451 (2013).
[8] Pereira H.F.G. ,De Jesus A.M.P. , Ribeiro A.S. Frenandes A.A. ((fatigue damage
behavior of structural components under variable amplitude loading )) Mecanica
Experimental Vol.17 , PP 75-85 (2009).
[9] yang L. , Fatemi A. ((cumulative fatigue damage mechanisms and quantifying
parameters: alternative review )) Journal of testing and
evaluation,Vol.26,issue2,PP.89-100(1998).
[10] Al-Alkawi H.J.M, Shakir Sakran Hassan, Salah F. Abd-El-Jabbar.((
Experimental Study of the Effect of Shot Peening on Elevated Temperature Fatigue
Behavior of 7075-T651 Al. alloy)) Eng. & Tech. Journal, Vol.31, No.3, 2013.
[11]Jack A. Collins ((failure of materials in mechanical design )) by John Wiley and
sons, second edition 1993.
[12] Al-Khafaji M.B.H. ((Experimental theoretical study of composite material under
static and dynamic loadings with different temperature conditions)) Phd thesis ,
university of technology (2014).

774
 
Eng. &Tech.Journal, Vol.33,Part (A), No.4, 2015 A Cumulative Damage Model for Fatigue Life
Prediction Based on Dynamic and
Static Deflections

[13] Al-Alkawi H.J.M., Amer Hameed Majeed ,Saisaban Ali Fahad ((Applying Non-
linear Damage Model for Predicating Corrosion Effect on Fatigue Life of (Carbon +
Glass) Fibers / PMMA Composite)) Eng. & Tech. Journal, Vol.30 , No.16 , 2012
[14] Hearn E.J. ((mechanics of materials)) Pergamon press Ltd (1970).
[15]Marco S.M. , Starkey W.L. ((A concept of fatigue damage )) Translations of the
ASME , 76 (4) PP. 627-632. (1954).
[16] Pereira H.F.G. , Jesus A.M.P. , Ribeiro A.S. Frenandes A.A. ((fatigue damage
behavior of structural components under variable amplitude loading)) Mecanica
Experimental , Vol.17, PP. 75-85 (2009).
[17] Alalkawi H.J.M. , Alsafar S.A.K. , Abdul-Jabar H.A ((A new cumulative
damage model for fatigue life prediction under shot peening treatment )) Al-
Khwarizmi Engineering Journal , Vol.10, No.2 ,P.P. 57-64 (2014).
[18] Pereira H.F.G. ,De Jesus A.M.P. , Ribeiro A.S. Frenandes A.A.((cyclic and
fatigue behavior of P355NL steel under block loading )) ASME (2007).
[19] Miller K.J. , Mohammed H.J. , De Los Rios ((fatigue damage accumulation
above and below the fatigue limit )) European group of fracture publication No.1 ,
EGFI , Instn , Mech.Eng. London (1986).
[20] Fatemi A. and yang L. ((cumulative fatigue damage and life prediction theories ;
a survey of the state of the art for homogenous material )) Int. J. fatigue , Vol.20,
No.1 PP.9-34 Elsevier science Ltd. (1998).

775