FEMFAT Agenda​​​​​​​​​​​​​​​​​​​​​

Carbon neutrality has become a significant target. One essential parameter regarding energy consumption and emissions is the mass of vehicles. Lightweight design of automobile parts improves the result of vehicle life cycle assessment (LCA) that considers the emissions generated from manufacturing to disposal, increases efficiency, and can be a step towards sustainability and CO2 neutrality. Weight reduction through structural optimization is a challenging task. Typical design development procedures have to be overcome. Instead of just a facelift or the creation of a derivative of the predecessor design, completely alternative design creation methods have to be applied. Automated structural optimization is one tool for exploring completely new design approaches. Different methods are available and weight reduction is the focus of topology optimization.  

This paper describes a fatigue life homogenization method that enables the weight reduction of vehicle parts. The applied CAE process combines fatigue life prediction and topology optimization. Since fatigue life prediction and failure locations are important indicators for optimization, its accuracy was validated by constructing FE model consisting of measurement system and assigning the local S/N curves that changes due to loading and manufacturing factors such as stress amplitude, mean stress, stress gradient, material composition and cast defects as material strength for each region via FEMFAT. 

An adapted design for a differential case was found, which does not sacrifice strength or stiffness properties of the component. Despite the very limited freedom for design modification, an interesting solution that saves nearly 20% of mass could be obtained, which demonstrates the potential of this approach to help achieve carbon neutrality through material saving. Other properties such as bolt loosening, stiffness, and castability were also considered. Verification with finite element analysis, fatigue assessment, and testing of the original and optimized components were performed. The lifetime results for virtual and real testing match quite closely and prove the effectiveness of this fatigue life homogenization method. Fine tuning of the simulation was also performed. Local material characteristics were considered based on filling and solidification simulations for the cast process of the component. 

Suspension components, such as cradles, are primarily engineered for durability, with additional considerations for stiffness, strength and crashworthiness. These components are pushed to their limits by stringent requirements for weight, cost, package and manufacturing constraints, resulting in minimal reserves for fatigue life. Furthermore, some OEMs are increasingly aiming to eliminate prototype testing, which necessitates a precise fatigue life evaluation. This methodology must also account for production variances to ensure reliability. At Magna Cosma, we have developed a proprietary evaluation method that addresses these challenges. Our approach includes advanced modeling techniques and customized S-N curves, specifically adapted to our manufacturing process. This leads to accurate prediction and enhance the fatigue life of suspension components leading to robust designs. Our approach is to test a series of weld specimens under cyclic loading to evaluate their fatigue performance. We use a finite element (FE) model that represents welds with 45-90° elements and uses a cubic stress approach to accurately capture the stress distribution within the welds. The L-N curves obtained from these tests are converted into S-N curves. These S-N curves are then stored in the FEMFAT weld database, which serves as a comprehensive data source for our fatigue analysis. These validated S-N curves and the corresponding stress evaluation methods form the basis for our fatigue analysis, which enables us to predict the fatigue life of welded components under different loading conditions

tbd

In electrical products, conductors such as bus bars and cables are used to transmit electrical signals and power. There is a risk of fatigue failure in these conductors due to vibration or repetitive loading on the product. However, a long-term reliability testing on prototypes is required to confirm that fatigue failure does not occur within the service life of the product. Therefore, a design method that can ensure fatigue strength is needed at the design stage. Toshiba TEC and Toshiba have developed technologies to predict the life span until fatigue failure of a conductor. We have utilized these technologies and have improved the fatigue life of various products. 

Wheel hub drives of heavy off-road vehicles are exposed to a diverse range of loads coming from the wheels. These load histories are highly transient, meaning that they do not lend themselves to simplified representations such as some kind of equivalent constant-amplitude or block loading. On the other hand, a capable FE model of the housing assembly will consist of a large number of 2nd order solid elements and non-linear contacts. A direct transient nonlinear simulation of the up to multi-hour load histories is impractical due to the infeasible solution times this would require, while a linear transient simulation approach, although much more performant, would be invalid on account of not being able to capture the changing contact behavior. In the presented case study, these difficulties are solved using the elastoloads approach, where the transient fatigue analysis of the housing assembly is conducted in FEMFAT ChannelMAX, using a basis of simple static FE loadcases, all the while preserving the non-linear behavior of the contacts and correct load paths throughout the course of transient event.

Many structural components of road or railway vehicles are exposed to random vibrations during operation. To ensure structural integrity throughout their service life, it is crucial to accurately assess fatigue damage from expected loads. Two main approaches are commonly used for this purpose: time-domain-based and PSD-domainbased methods. While the time-domain model uses numerical integration of diKerential equations, the PSD-domain model applies a statistical frequency-domain description using 2. order moments. The former can capture the eKects of non-stationarity, but the numerical solution can become computationally very expensive for long observation times and large structures. The latter is computationally very eKicient but only valid for stationary Gaussian loading. Given that many vibrations aKecting vehicle components are non-stationary, these analysis methods often prove inadequate. To address these limitations, the Decomposition into Gaussian Portions (DGP) method is proposed. This approach is based on the concept that a non-stationary process can be approximated by a decomposition into multiple Gaussian segments, which collectively retain the same fatigue damage potential as the original signal. The DGP method is based on the framework of higher order spectra. It models the statistics of non-Gaussian signals by including higher order moments and their frequency characteristics. A key advantage of this method is that it enables damage estimation for each Gaussian segment using well-established, PSD-domain techniques. It employs a minimization algorithm to identify these Gaussian segments. 

Examples are presented to illustrate the relevant phenomena based on the application of FEMFAT. Firstly, an academic example explains the impact of multiple eigenfrequency resonances induced by non-stationary excitation on the resulting fatigue damage. A synthetic signal, comprising multiple non-Gaussian frequency bands aligned with the structure's eigenfrequencies, is defined as the excitation. Fatigue damage estimation is subsequently performed through a PSD approach directly based on the structure's response, as well as by initially applying the DGP method. For validation purposes, damage is additionally assessed utilizing the time-domain method. The results indicate that traditional PSD-domain methods underestimate fatigue damage, whereas the DGP method provides a more accurate approximation.

Secondly, the practical application of the DGP method is showcased through on a real structural component subjected to a measured non-stationary excitation. It becomes evident that the exclusive application of PSD-domain-based methods leads to an underestimation of fatigue damage, whereas the DGP method yields suKiciently accurate results. Moreover, it is significantly more resource-eKicient compared to the time-domain-method.

In conclusion, it can be stated that the DGP method provides a robust approach for conducting accurate and time-eKicient fatigue damage estimations of non-stationary structural responses

A live session with FEMFAT including tips and tricks for the usage and optimising the workflow

Location will be announced soon

Calculation of Solder Joint Fatigue during vibrational loading with FEMFAT MELCOM 

• Generation of solder geometry 
• Automated FE-Modelling of PCBA with ODB++ and Pick&Place file 
• Calculation of solder joint fatigue with FEMFAT 

Discover a completely new equivalent stress method with the Complex Invariant Hypothesis (COIN LiWI). This workshop introduces this new approach that leverages the principles of Linear Wave Interference (LiWI) and complex invariants (COIN). Traditional invariant concepts (e.g. the signed von-Mises approach), while simple, often fall short in accuracy and can result in unphysical discontinuities, especially for multi-axial applications. Join us to explore how COIN LiWI can overcome the limitations of conventional invariant methods. 

Unlock the full potential of modal fatigue analysis with the latest advancements in the HARMONIC tool. This workshop will guide you through the functionalities of HARMONIC, a powerful signal generator for modal lifetime analyses in ChannelMAX. Learn how HARMONIC generates modal participation factors for each mode using transfer functions from harmonic response analysis. You will also discover how to specify various vibration processes, from constant sine waves to sinusoidal sweeps and measured/simulated load-time signals. The session will highlight the newest features, including digital sensor options and multi-sine definitions. Join us for an in-depth look at how HARMONIC can streamline and improve your vibration analysis workflows.

“The workshop "Innovations and Enhancements in FEMFAT inside ANSYS"  provides an introduction to the most important developments and improvements in the FEMFAT extension within ANSYS Workbench. The workshop covers the following topics: 

1. Introduction and overview: Short introduction and overview about FEMFAT inside ANSYS 
2. New Features and Tools: Introduction of  the most important new features and tools in FEMFAT inside ANSYS. 
3. Application Examples: Practical examples showcasing the application of new features.”