Fatigue Limit

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4
At a glance
Powered by AI
The paper discusses using finite element analysis to estimate fatigue life (S-N) curves of stainless steel AISI 304 in both air and corrosive environments.

The material being analyzed is stainless steel AISI 304.

Finite element analysis using ANSYS Release 13 to model rotating bending tests and estimate the S-N curves.

Proceeding Seminar Nasional Tahunan Teknik Mesin XII (SNTTM XII)

Bandar Lampung, 23-24 Oktober 2013

S-N Curve Estimation of AISI 304 in Air and Corrosive Environment


Using Finite Element Method

S. Huzni, J. Rahmaddireja, S. Fonna and M. Ridha

Department of Mechanical Engineering, Syiah Kuala University


Jl. Tgk. Syech Abdurrauf No. 7 Darussalam – Banda Aceh 23111, INDONESIA
Phone/Fax: +62-651-7428069, E-mail: syifaul@gmail.com

Abstract

Endurance limit that can be predicted from the S-N curve has become an interesting issue for reliability and quality
assessment of stainless steel especially in corrosion environment. However, the study of corrosion mechanism
related to S-N curve is not well established. This paper discusses application of finite element method for S-N curve
estimation of AISI 304 in air and corrosive environment. Finite element simulation was carried out using ANSYS
Release 13. An experimental set up using fatigue rotary bending machine based on ASTM E466 standard is used to
validate the simulation result. In the air environment, endurance limit obtained from finite element analysis is similar
to references. In the corrosive environment, even though the endurance limit obtained by finite element analysis is
lower than that obtained by experimental set up, the trend line of S-N curves for both works and analyses are the
same. Therefore, finite element simulation result can be used to estimate S-N curve as good as the experimental set
up.

Keywords: finite element method, fatigue life, AISI 304, and corrosive environment.

Introduction evaluation of fatigue parameters suitable for


estimating fatigue lives under multi-axial loadings.
Fatigue is a form of failure that occurs in structures The local cyclic elastic–plastic stress–strain responses
subjected to dynamic and fluctuating stresses. Fatigue were analyzed using the incremental plasticity
is one of the most common failure mechanism in metal, procedures of ABAQUS finite element code for both
estimated to be approximately 90% of the source of all smooth and notched specimens made of three
metallic failures (Callister, 2007). Furthermore, materials: a medium carbon steel in the normalized
fatigue can occur suddenly without preliminary signs condition, an alloy steel quenched and tempered and a
and result in catastrophic failures. Various methods stainless steel, respectively. For experimental
have been developed to address this issue. The most verifications, a series of tests of biaxial low cycle
common method is to use stress versus cycles plot, or fatigue composed of tension/compression with static
S-N curve. This curve plot fatigue strength with and cyclic torsion were carried out on a biaxial
respect to time-to-failure. In general, the purpose of S- servohydraulic testing machine. Comparisons between
N curve estimation is to avoid failure problems in numerical simulations and experimental observations
relation to safety, economy, durability and liability. show that the FEM simulations allow better
Austenitic stainless steels are the most commonly used understanding on the evolutions of the local cyclic
metallic materials in application requiring corrosion stress–strain. But, this paper not considering the effect
resistance because of their high strength and ductility of corrosion fatigue.
(Mc.Guire, 2008). However, austenitic stainless steels In this paper, application of finite element method for
has some of relative weakness, such as austenitic S-N curve estimation of AISI 304 in different media,
stainless steels are less resistant to cyclic oxidation i.e air and corrosive environment, is discussed.
than are ferritic grades and they can experience stress Specimen models are based on ASTM E-466. This
corrosion cracking (SCC) if used in an environment to study is expected to give information on the fatigue life
which they have insufficient corrosion resistance. The of AISI 304 in evaluating the application of the air and
risks of these limitations can be avoidable by taking corrosive environment. So, that further studies on the
proper precautions such as to know the endurance limit optimization of the design components can be done for
of the material before. this type of material.
Li et al (2006) has done simulation for cyclic
stress/strain evolutions and redistributions, and The Material and Experimental Method

1
Proceeding Seminar Nasional Tahunan Teknik Mesin XII (SNTTM XII)
Bandar Lampung, 23-24 Oktober 2013

ANSYS Release 13 was employed in this work


Material which is running under windows 7 ultimate. In
The chemical composition of the AISI 304 is given in ANSYS Fatigue module, the available type of fatigue
Table 1. Figure 1 show geometry of specimen is used analysis is Stress Life and Strain Life. Stress Life is
in this experimental set up. Before fatigue test, the based on empirical S-N curves and take into account a
cross-section was mechanically polished using variety of factors. Strain Life is based upon the Strain
progressively finer grades of abrasive paper followed Life Relation Equation where the Strain Life
by buff-finishing to make sure no scratch on surface of Parameters are values for a particular material that best
specimen. fit the equation to measured results. The Strain Life
Relation requires a total of six parameters to define the
Table 1 Chemical composition (wt.%) strain-life material properties; four strain-life
C Cr Fe Mn S Ni P Si N parameter properties and the two cyclic stress-strain
0.08 20.00 74.0 2.0 0.03 12.00 0.045 0.75 0.10 parameters. And the six parameters required for a
Strain Life analysis are Fatigue Strength Coefficient,
Fatigue Strength Exponent, Fatigue Ductility
Coefficient, Fatigue Ductility Exponent, Cyclic
Strength Coefficient, and Cyclic Strain Hardening
Exponent. The Strain Life Relation equation is
shown as:
∆𝜀 𝜎𝑓 𝑏 𝑐
Figure 1. Specimen configuration. 2
= 𝐸
(2𝑁𝑓 ) + 𝜀𝑓 (2𝑁𝑓 ) …………………….. (1)

Where:
Experimental Methods Δε = Total Strain Amplitude
Δσ = Stress Amplitude
E = Modulus of Elasticity
Fatigue tests were performed using four points rotary Nf = Number of Cycles to Failure
bending fatigue test machines operating at frequency
of 50Hz shown in Figure 2. Fatigue tests were The two cyclic stress-strain parameters are part of the
performed with five different loading conditions;12 kg, equation shown as:
10 Kg, 8 Kg, 7 Kg, and 6 Kg. The corrosive 1⁄
∆𝜎 ∆𝜎 𝑛
environment is simulated using 3.5% NaCl solution. ∆𝜀 = 𝐸
+ 2(𝐾 ) ………………………….… (2)

Where:
Δε = Strain Amplitude
Δσ = 2 x the Stress Amplitude
E = Modulus of Elasticity
K = Cyclic Strength Coefficient
n = Cyclic Strain Hardening Exponent

Solid Modeling

In Simulation monotonic and cyclic properties for


AISI 304 are given in Table 2 that obtained from Colin
(2010). Load from experiment is converted to moment
Figure 2. Rotating Bending Machine.
in simulation.
Linear elastic model with the static structural analysis
Simulation Method
is chosen for the analysis. Two moments are applied
on two sides of the specimen with different direction,
A standard specimen has been analyzed by finite
and cylindrical support, which is chosen to hold a
element method, subjected to a condition similar to
specimen in tangential direction, is set to free as shown
experimental test. Since the test conditions were the
in Figs. 3.
same, it is possible to compare the results and evaluate
whether the finite element method could be suitable to
Table 2. Monotonic and cyclic properties for AISI
predict fatigue life of AISI 304.
304
monotonic properties SS304

2
Proceeding Seminar Nasional Tahunan Teknik Mesin XII (SNTTM XII)
Bandar Lampung, 23-24 Oktober 2013

Modulus of elasticity 196 10 3243.40


Yield strength (0.2% offset), Sy (MPa) 208 20 2574.29
Ultimate tensile strength, Sa (MPa) 585
Percent reduction in area, %RA 84 50 1896.75
Strength coefficient, K (MPa) 680 100 1505.45
Strain hardening strength, n 0.214 200 1194.88
True fracture strength, 𝜎𝑓 (MPa) 2051 2000 554.61
True fracture ductility, 𝜀𝑓 (MPa) 186 10000 324.34
Cyclic properties SS304 20000 257.43
Cyclic modulus of elasticity, E’ (GPa) 196
Fatigue strength coefficient, bilinear fit, 100000 150.55
′ ′ 330/1890 200000 119.49
𝜎𝑓1 /𝜎𝑓2 (MPa)
Fatigue strength exponent, bilinear fit, b1/b2 -0.0373/-0.0204 1000000 69.88
Fatigue ductility coefficient, 𝜀𝑓′ 0.1325
Fatigue ductility exponent, c -0.3738
Cyclic strength coefficient, bilinear fit, Result and Discussion
434/4742
𝐾1′ /𝐾2′ (MPa)
Cyclic strain hardening exponent, bbilinear
0.1106/0.5121
fit, 𝑛1′ /𝑛2′ Fatigue life analyses of five different loading
Cyclic yield strength, 𝑆𝑦′ (MPa) 220 conditions was performed by using ANSYS Release
13. The fatigue life contours of specimen shown in
Figs. 4.

Figure 3. Geometry and boundary condition.

For corrosive environment, boundary condition is Figure 4. The fatigue life of AISI 304 in cycles for
approach by using analytic formula to get the stress life 12Kg load.
parameter that used for engineering data in ANSYS.
The result shown in Table 3. The S-N curve from fatigue analysis in air condition
for AISI 304 is shown in Figure 5. It can be seen from
log 𝑁 = log 𝑎̅ − 𝑚 log ∆𝜎……………. (3) the curve that the fatigue limit for AISI 304 was at
323.34 MPa. This result is close to result from INCO
N = predicted number of cycles to failure for data published by NiDI (Keytometal). This data shows
stress range ∆𝜎 endurance limits from reverse bending fatigue tests as
log 𝑎̅ = intercept of log N-axis by S-N curve can be seen in Table 4.
M = negative inverse slope of S-N curve S-N curve in Figs. 6 shows comparison of results
∆𝜎 = Stress Range obtained from simulation and experiment in corrosive
environment. Even though the curve obtained by finite
element analysis is lower than that obtained by
experimental set up, the trend line of S-N curves for
both experiment and finite element analysis are similar.

Table 4. Endurance limit data for common AISI


Table 3. Alternating Stress Mean Stress stainless steels
Cycles Alternating Stress (MPa) AISI Type Endurance Limit, MPa

3
Proceeding Seminar Nasional Tahunan Teknik Mesin XII (SNTTM XII)
Bandar Lampung, 23-24 Oktober 2013

301 240 and Engineering. United State of America.


303 240 [2] McGuire, Michael F. 2008, Stainless Steels For
304 240 Design Engineering, ASM International, United
310 215 State of America.
[2] Li, B., L. Reis, M. de Freitas. 2005. “Simulation of
600 cyclic stress/strain evolutions for multiaxial
550 fatigue life prediction”, Department of
500 Mechanical Engineering, Instituto Superior
Maximum Stress, MPa

450 Te´cnico, Av. Rovisco Pais, 1049-001 Lisboa,


400 Portugal.
350 [3] Surpui, Dumenico. 2011. Stainless Steels, Gruppo
300 Lucefin, Italy.
250
[4] Nakamura, Yuki. Isono, Kenta. 2011. “Fatigue
200
Properties of Metastable Type 304 Stainless Steel
150
Depending on Test Temperature and Frequency”,
Department of mechanical Engineering, Gifu
100
1,E+03 1,E+04 1,E+05 1,E+06 1,E+07 1,E+08 University & Toyota National College of
Cycles Technology, Japan.
Figure 5. S-N curve simulation in air environment. [5] Colangelo, VJ. Analisis of Metalurgi Failures,
John Willey and Sons, New York, 1973.
[6] Bannantine, J., Comer, J., Handrock, J. 1990.
“Fundamentals of Metal Fatigue Analysis”, New
Jersey, Prentice Hall.
[7] Stephens, Ralph I., Fatemi, Ali, Stephens, Robert
R., Fuchs, Henry O. 2001. “Metal Fatigue in
Engineering”, New York, John Wiley and Sons,
Inc.
[8] Lampman, S.R. 1996. “ASM Handbook: Volume
19, Fatigue and Fracture”, ASM International
[9] www.keytometal.com, “Fatigue of Metal: Part
Three”.
[10]Colin, Julie, Ali Fatemi. 2010. “Fatigue
Behaviour of Stainless Steel 304L Including Strain
Figure 6. The comparison of S-N Curve simulation Hardening, Prestraining, and Mean Stres Effect”.
and experiment in corrosive environment. France.

Conclusion

From the result of this study, it can be seen that finite


element simulation produce a good agreement
compare to experimental result. Finite element method
simulation is able to provide insight and prediction of
fatigue life comparable to experimental works. In the
corrosive environment, even though the endurance
limit obtained by finite element analysis is lower than
that obtained by experimental work, probably by the
presence of micro crack, the trend line of S-N curves
for both works and analyses are similar. Therefore,
finite element simulation result can be used to estimate
S-N curve as good as the experimental work.

Bibliography

[1] Callister Jr, William D. 2007. Material Science

You might also like