1 FATIGUE LIFE PREDICTION OF THERMO-MECHANICALLY LOADED ENGINE COMPONENTS Csaba Halászi , Christian Gaier, Helmut Dannbauer MAGNA Powertrain, Engineering Center Steyr GmbH & Co. KG Steyrer str. 32, A-4300 St.Valentin, Austria Keywords: thermo-mechanical fatigue, engine, Sehitoglu INTRODUCTION Nowadays engine components are subjected to higher loads at elevated temperatures than before, due to the increasing requirements regarding weight, performance and exhaust gas emission. Thus, fatigue due to simultaneous thermal and mechanical loading became determinant among the damage forms. At the same time, there is the need to reduce development times and costs to handle the growing number of model variants. Therefore, the development of suitable simulation tools, which reduce the number of necessary component tests, seems to be very rewarding. The problem of thermo-mechanical fatigue (TMF) life prediction has received considerable attention in recent years, with efforts principally concentrated on the prediction of TMF under uniformly repeated loading conditions. Several researchers have developed models to treat this problem, generally based on isothermal (IT) considerations. However, isothermal tests do not capture all damage mechanisms that operate under variable strain-temperature conditions. As Sehitoglu emphasizes, a deeper understanding of the different micro mechanisms affecting the behavior of materials under isothermal and thermo-mechanical loading conditions is needed. THERMO-MECHANICAL FATIGUE Thermo-mechanical fatigue (TMF) is the case of fatigue failure due to simultaneous thermal and mechanical loading. The life prediction of TMF loading cases has received considerable attention in recent years mainly in engine and gas turbine development. The fluctuation of complex thermal and mechanical strains is usually determinant for fatigue life of machine parts. The mechanical strain arises either from external constraints or externally applied loadings. Thermo-mechanical and low cycle fatigue (LCF) can show a lot of similarity, mainly because of the presence of cyclic plastic strain. The cyclic thermal load occurs by nature in a small number of cycles, but the stresses generated by the restrained thermal expansion may be far beyond the elastic limit. In engine parts the superposition of a LCF/TMF effect due to start-stop cycles and a HCF effect due to the combustion cycle is to be observed. Throughout a thermo-mechanical cycle, one crosses different temperatures. The temperature dependent processes that occur during a common TMF cycle are plastic deformation, cyclic aging,
14
Embed
fatigue life prediction of thermo-mechanically loaded engine ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
FATIGUE LIFE PREDICTION OF THERMO-MECHANICALLY LOADED
ENGINE COMPONENTS
Csaba Halászi, Christian Gaier, Helmut Dannbauer
MAGNA Powertrain, Engineering Center Steyr GmbH & Co. KG
TTTrrraaannnsssiiieeennnttt ssstttrrreeessssss---ssstttrrraaaiiinnn aaannnaaalllyyysssiiisss using visco-plastic material model incl. temperature and strain rate dependent cyclic material behaviour (e.g. Chaboche or Sehitoglu)
TTMMFF ddaammaaggee aannaallyyssiiss Using Sehitoglu‘s TMF life prediction model involving about 20 material parameters
11
Figure 6 – RPM and temperature history
The calculated maximum temperature was about 340°C at the bowl rim at the end of the
maximum loading step. The next figure shows the temperature distribution at the end of the
maximum loading step.
Figure 7 – Temperature distribution and temperature history of node 7377
Stress-strain analysis
The transient stress-strain analysis was performed by ABAQUS using the temperature results of the
transient heat transfer analysis. The stress-strain behavior of the material was simulated using a
simple elastic-plastic material model considering the temperature dependent heat transfer
coefficient and stress-strain relation.
Mechanical constraints were applied in the middle-lower section of the pin. Contact between the
pin and the piston hub was considered as well.
n
T
nidle
nmax.load
Tmax.load
Tidle
Warm up to idle Maximum load Cool down
60s 60s 60s
t
t
12
Pressure distribution was applied above the 1st oil ring for each time increment to simulate the
operational combustion pressure. Neither side nor mass force was considered.
The Mises stress distribution at maximum loading is shown in the next figure.
Figure 8 – Mises stress distribution and Mises stress history of node 7377
Damage analysis
The damage analysis was performed with FEMFAT HEAT using transient temperatures, stresses
and strains and time information from the ABAQUS analysis. The concept of the TMF analysis
with FEMFAT is displayed on the next figure.
Figure 9 – TMF damage analysis
Results
• Damage distribution
• Analysis protocol
FE analysis Input Data
� Transient stresses
� Transient strains
� Transient temperatures
� Time information
Material data
FEMFAT HEAT
Temperatures
stresses strains
13
No creep damage was considered because of the lack of material parameters for creep damage.
On the other hand creep damage is not relevant for this component, because it is governed by OP
loading. The next Figure displays the damage distribution of the mechanical and environmental
damage for one TMF load cycle.
Figure 10 – Distribution af the mechanical (b) and environmental damage (a)
It’s clearly seen, that oxidation damage is dominated for this load case.
FATIGUE LIFE PREDICTION OF AN ALUMINUM CYLINDER HEAD
Another example is the life prediction of an aluminum cylinder head. This analysis is briefly
reported here.
The transient heat transfer analysis and the forthcoming sequentially coupled transient stress-
strain analysis is performed by ABAQUS using a temperature dependent elastic-plastic material
model for the simulation of the behavior of the AlSi7 aluminum cast alloy.
The damage analysis was performed by FEMFAT HEAT considering fatigue damage,
environmental and creep damage. The results of the damage analysis are shown on the next Figures
indicating that fatigue and oxidation damage is dominated. The creep damage is not significant.
Figure 11 – Distribution of the total (a) and mechanical damage (b)
a) b)
a) b)
14
Figure 12 –Picture Distribution of the oxidation damage (c) and creep damage (d), respectively
CONCLUSION
In this paper, it is shown how to predict the fatigue life of thermo-mechanically loaded components
using Sehitoglu’s method. The method covers most of the effects and influences that operate under
variable thermal and mechanical loading in engine components. The damage calculation method
includes the calculation of the classical fatigue (mechanical) damage, the environmental and creep
damage. The method is implemented into the fatigue software FEMFAT allowing a fast, efficient
and high quality assessment of thermo-mechanically loaded components. The software is tested on
engine components like cylinder heads and pistons. In most cases the results showed good
correlation between analysis and measurement in terms of critical location and fatigue life. The
method is under evaluation for other applications like exhaust gas manifold and brake discs.
Furthermore, a method is presented for the determination of the required material parameters
using parameter identification based on isotherm LCF and TMF test data. Until now the material
parameters were determined for about 6-7 materials using this method including aluminum alloys
and cast irons.
Finally, the application of the life prediction method was demonstrated on two examples. In
these examples a simple elastic-plastic material model was used, however in many cases a visco-
plastic material model is needed for more accurate FE-results, which is essential for the fatigue
analysis.
REFERENCES Bäck, T. – Schwefel, H.-P. (1996): “Evolutionary Computation: An Overview”, in Proceedings of 1996 IEEE
International Conference on Evolutionary Computation, May 20-22, 1996, Nagoya University, Japan. IEEE, pp.
20-29.
Nelhiebel, A. (2005): Determination of the material parameters of a TMF damage model from experiments, Diplom
thesis
Price, K. – Storn, R. (1995): Differential Evolution – a simple and efficient adaptive scheme for global optimization
over continuous spaces. Technical Report TR-95-012, ICSI, March 1995
Sehitoglu, H., et al. (2002): “Thermomechanical Fatigue Analysis of Cast Aluminum Engine Components”,
Thermomechanical Fatigue Behaviour of Materials Vol. 4, ASTM STP 1428
Sehitoglu, H. – NEU, R. (1989): “Thermomechanical Fatigue, Oxidation and Creep: Part I and II. Experiments”,