Yung-Li Lee

Metal Fatigue Analysis Handbook

The local strain–life method was developed in the late 1950s and was shown to be more effective for low- and midcycle fatigue life prediction than the stress-life approach for the fatigue analysis of components. This approach is also preferred when the load history is irregular, where the mean stress and the load sequence effects are thought to be of importance. This method also provides a rational approach to differentiate the high cycle fatigue and the low cycle fatigue regimes as well as to include the local notch plasticity and mean stress effect on fatigue life.

Stress-Based Uniaxial Fatigue Analysis Using Methods Described in FKM-Guideline

The process of prevention of failure from structural fatigue is a process that should take place during the early development and design phases of a structure. In the ground vehicle industry, for example, the durability specifications of a new product are directly interweaved with the desired performance characteristics, materials selection, manufacturing methods, and safety characteristics of the vehicle. In the field of fatigue and durability analysis of materials, three main techniques have emerged: nominal stress-based analysis, local strain-based analysis, and fracture mechanics analysis. Each of these methods has their own strengths and domain of applicability—for example, if an initial crack or flaw size is known to exist in a structure, a fracture mechanics approach can give a meaningful estimate of the number of cycles it takes to propagate the initial flaw to failure. The development of the local strain-based fatigue analysis approach has been used to great success in the automotive industry, particularly for the analysis of measured strain time histories gathered during proving ground testing or customer usage. However, the strain life approach is dependent on specific material properties data and the ability to measure (or calculate) a local strain history. Historically, the stress-based fatigue analysis approach was developed first—and is sometimes considered an “old” approach—but the stress-based fatigue analysis methods have been continued to be developed. The major strengths of this approach include the ability to give both quantitative and qualitative estimates of fatigue life with minimal estimates on stress levels and material properties, thus making the stress-based approach very relevant in the early design phase of structures where uncertainties regarding material selection, manufacturing processes, and final design specifications may cause numerous design iterations. This article explains the FKM-Guideline approach to stress-based uniaxial fatigue analysis. The Forschungskuratorium Maschinenbau (FKM) was developed in 1994 in Germany and has since continued to be updated. The guideline was developed for the use of the mechanical engineering community involved in the design of machine components, welded joints, and related areas. It is our desire to make the failure prevention and design community aware of these guidelines through a thorough explanation of the method and the application of the method to detailed examples.

Assessment of the Mean-Stress Sensitivity Factor Method in Stress-Life Fatigue Predictions

The mean-stress sensitivity factor model developed by Schutz in 1967 is considered the most comprehensive method to account for the mean-stress effect on high-cycle fatigue life and strength, which is applicable to various materials (such as steels, steel castings, ductile irons, malleable cast irons, grey cast iron, wrought aluminum alloys, and cast aluminum alloys) and to a broad mean-stress or stress-ratio range. Even though the mean-stress sensitivity factor method has been frequently used in European-based fatigue commercial software and engineering design codes, such as the Rechnerischer Festigkeitsnachweis fur Maschinenbauteile (FKM) Guideline, it has not received a great deal of attention worldwide. Although the FKM Guideline provides extensive experimental test validation, very little evidence from independent sources have illustrated the validity of this method. Thus, it is the objective of this paper to examine the Schutz mean-stress sensitivity factor model by comparing the Walker mean-stress correction equation to the best fitting parameters published for steels, aluminum alloys, and titanium alloy. It was concluded that the Schutz mean-stress sensitivity factor model exhibits a reasonable degree of agreement with the Walker equation over a wide variety of materials.

Multiple Sinusoidal Vibration Test Development for Engine Mounted Components

Durability testing is required before vehicle launching to prevent failure before designed life. Limited publications were found on specifying vibration testing to validate durability and reliability of engine mounted components. These published test methods did not consider engine firing order effects and oversimplified the vibration profiles. In this paper, a new ordered multiple sinusoidal vibration test method is proposed to improve the existing procedures. The test method is designed to fulfill an infinite life durability requirement for engine mounted components subjected to a four-stroke internal combustion engine vibration. An innovative test development procedure, based on engine vibration field data, is illustrated in details in this paper. The ordered multiple sinusoidal vibration test method covers the choice of sweep type, sweep time, frequency range, vibration magnitude and profile, and test duration. Instead of obtaining the vibration magnitude directly from order analysis results, vibration magnitudes are determined by fully analyzing the vibration data in both time domain and frequency domain. Test profiles designed by enveloping method are proposed for a better represent of the engine excitations.

Fatigue Predictions of Various Joints of Magnesium Alloys

In this project, a front shock tower of a passenger vehicle is developed with various magnesium alloys and joining methods. To predict the fatigue life of the joints in the structure, fatigue tests of various joint specimens including friction stir linear welding, self-piecing rivet joint with and without adhesive, and friction stir spot welding were conducted. The magnesium alloys used for the specimens are AM60 (cast), AM30 (extrusion), and AZ31 (sheet). Various finite element modeling techniques were attempted for simulating the various joints. Fatigue life prediction method for the joints was performed using the stress-life curve approach. The finite element modeling technique and the fatigue prediction method will be verified with fatigue tests of the actual front shock tower structure subjected to variable amplitude loadings in near future.

Assessment of the Poisson Ratio Effect on Low Cycle Fatigue (LCF) Behavior of Shear-Cracking Mode Materials

This paper evaluates the effect of different Poissons ratios and an equivalent Poissons ratio formula on the crack initiation angle and life estimation by using the three shear-strain based fatigue damage models (Brown–Miller, Fatemi–Socie, and Lius Virtual Energy) on strain-controlled biaxial cyclic loading tests of specific tubular material specimens. The four shear failure mode materials studied were Inconel 718, 304L stainless steel, normalized 1050, and quenched/tempered 1050 steels. This study was motivated by the fact that for all the strain-controlled fatigue testing, the induced transverse strain was not measured, but assumed to be a negative product of the axial strain and the Poisson ratio, where the axial strain was known and given, and the Poisson ratio was either assumed to be a constant or a variable represented by a function of elastic and plastic equivalent strains. Therefore, the assumption of a Poissons ratio in the transverse strain calculation was assessed in this study. It is concluded that the use of different Poissons ratios in the three fatigue damage models would have some effects on the crack angle prediction, depending on the material type, and little effect on the fatigue life estimation for all materials investigated in this study.

Fatigue Behavior of AM60B Subjected to Variable Amplitude Loading

Magnesium alloys are considered as an alternative material to reduce vehicle weight due to their weight which are 33% lighter than aluminum alloys. There has been a significant expansion in the applications of magnesium alloys in automotives components in an effort to improve fuel efficiency through vehicle mass reduction. In this project, a simple front shock tower of passenger vehicle is constructed with various magnesium alloys. To predict the fatigue behavior of the structure, fatigue properties of the magnesium alloy (AM60B) were determined from strain controlled fatigue tests. Notched specimens were also tested with three different variable amplitude loading profiles obtained from the shock tower of the similar size of vehicle. The test results were compared with various fatigue prediction results. The effect of mean stress and fatigue prediction method were discussed.

Vibration Fatigue Testing and Analysis

There are two commonly used vibration test methods in the engineering industry to validate a product to meet its durability and design life requirements, sinusoidal vibration and random vibration. The choice of the test method depends on the source of the dominant forcing input. For example, sinusoidal vibration is often used for testing engine accessories, manifolds, and components mounted on manifolds that are primarily experienced engine firing loads. Random vibration is usually applied to vehicle body, frame, and chassis subjected to random road load inputs.

Rainflow Cycle Counting Techniques

The goal of fatigue analysis is to assess the damage or life of a structure that is subjected to a complicated, variable amplitude loading history. The basic processes include: (1) selecting a fatigue damage model, (2) determining the required fatigue properties, (3) decomposing the variable amplitude loading history into a spectrum of simple constant amplitude loading cycles, (4) calculating the fatigue damage contributed by each loading cycle, and finally (5) assessing the damage or the life of the structure by the Palmgren-Miner linear damage rule ( Palmgren, 1924 ; Miner, 1945 ).

Strain-Based Multiaxial Fatigue Analysis

Strain-based approaches for multiaxial fatigue parallel the approaches for stress-based multiaxial fatigue. Whereas a stress-based approach is most suitable for high-cycle fatigue, the strain-based approaches are most commonly used in mid- to low-cycle fatigue regimes. For certain loading conditions like proportional loading, some equivalent stress–strain approaches can be used to determine fatigue life. For general loading conditions, the critical plane approach can be used.