Search Results

Spot Welds are a category of welds used extensively in automotive structures, normally for metals. The fatigue analysis of such spot welds can be evaluated using (a) the Point 2 Point (P2P) method where a beam or bar is used to connect the 2 surfaces being joined, (b) a more modern approach where the 1D element is replaced with an “equivalent” brick element, or (c) a third approach that falls somewhere between where a “spider” and circular ring of elements, is used to represent the spot weld. In all 3 cases there is an assumption that the cross section is circular. For some specialist cases such as plastic connectors, the cross section is not circular so a new user defined weld is proposed. This paper will describe the approach that is based on the concept that a user generated tensor line can be used (equivalent to the theoretical Force/Moment to stress algorithms built into the P2P approach) along with special S-N curves create for different joint shapes.

The fatigue analysis of a full car body requires the sheet metal (sheet fatigue), spot welds (spot weld fatigue) and seam welds (seam weld fatigue) to be thoroughly evaluated for durability. Traditionally this has always been done in the time domain, but recently new frequency domain techniques are able to perform these tasks with numerous advantages. This paper will summarize the frequency domain process and then compare the results and performance against the more usual time domain process.

In most aspects of mechanical design related to a motor vehicle there are two ways to treat dynamic fatigue problems. These are the time domain and the frequency domain approaches. Time domain approaches are the most common and most widely used especially in the automotive industries and accordingly it is the method of choice for the fatigue calculation of welded structures. In previous papers the frequency approach has been successful applied showing a good correlation with the life and damage estimated using a time based approach; in this paper the same comparative process has been applied but now extended specifically to welded structures. Both the frequency domain approach and time domain approach are used for numerically predicting the fatigue life of the seam welds of a thin sheet powertrain installation bracketry of a commercial truck submitted to variable amplitude loading. Predicted results are then compared with bench tests results, and their accuracy are rated.

For many automotive systems it is required to calculate both the durability performance of the part and to rule out the possibility of collision of individual components during severe base shake vibration conditions. Advanced frequency domain methods now exist to enable the durability assessment to be undertaken fully in the frequency domain and utilizing the most advanced and efficient analysis tools (refs 1, 2, 3, 4, 5). In recent years new capabilities have been developed which allow hyper-sized models with multiple correlated loadcases to be processed. The most advanced stress processing (eg, complex von-Mises) and fatigue algorithms (eg, Strain-Life) are now included. Furthermore, the previously required assumptions that the loading be stationary, Gaussian and random have been somewhat relaxed. For example, mixed loading like sine on random can now be applied.

A previous SAE paper (ref. 1) did a comparative study of automotive system fatigue models processed in the time and frequency domain. A subsequent paper (ref. 2) looked at relative random analysis under base shake loading conditions. This paper proposes to merge these two analysis procedures to implement a new “Loads Cascading” procedure. The objective of this paper will be to show how loads (accelerations, displacements, forces) can be cascaded (transferred) from input load position such as road load data (RLD) body loads to some internal location, for example a battery pack location. Also note that the response from one “module” could form the input to another, therefore, once the loadings are in the frequency domain, the possibility exists to “cascade” the loads through a system. For example, from the chassis, to the subframe to attached components.

Today in the automotive industry, there is a continual reduction in available development time. There is also an urgent need to reduce cost and weight, to adapt to customer and legislation which drives to an increase in design complexity. These challenges are sometimes made harder by the late availability of hardware and this creates the need to extend and continually improve the established CAE methods which are used to develop automotive parts. This holds especially true in the field of exhaust systems and their components, which experience loads from various sources like temperature, engine or road. In the field of road excitation the use of dynamic transient simulation and subsequent damage calculation is state of the art in terms of simulations methodology.

It has been recognised since the 1960's that the frequency domain method for structural analysis offers superior qualitative information about structural response (refs 1, 2, 3, 4); But computational and technological issues have held back the implementation for fatigue calculation until now. Recent technological developments (see refs 5, 6, 7, 8, 9) have now enabled the practical implementation of the frequency domain approach and this paper will focus on the accuracy of the approach when compared with the traditional time based (transient dynamic) approach.

Techniques for calculating fatigue life from random structural responses were first proposed in the 60's but these early methods were limited to narrow band responses (ref 1). When used for wide band responses these same techniques could become very conservative. In order to reduce this conservatism much effort was devoted from the 1980's onwards to develop methods that worked more accurately for the wide band situation. Several methods now exist for the wide band case and these typically exist alongside Finite Element (FE) based random analysis tools like Nastran, Ansys or Abaqus to take the PSD's of stress response and return the Rainflow cycle count and fatigue damage (ref 2). Several problems still exist with todays design methods. Firstly, for large models, these stress transfer functions have to be generated and stored for subsequent use in the fatigue life calculation and these files can be very large.

This paper presents a fundamental conceptual change to the traditional CAE based fatigue analysis process. Traditional approaches take the responses from a stress solver and these are then transferred into a secondary fatigue analysis step. In this way fatigue is, and always has been, treated as a post processing step. The new conceptual change described in this paper involves combining the two separate tasks into one (stress and fatigue together). This results in a simple, elegant and more powerful Durability Management concept. This new process requires no large data files to be transferred, no complicated file management and it is likely that whole fatigue calculation process can be done in memory. This makes it possible to perform optimization with fatigue life as the constraint. It also facilitates full body fatigue life calculations, including dynamic behavior, for much larger models than was previously possible.

Conventional approaches for the fatigue life evaluation of automotive parts like headlamps involves the evaluation of random stress conditions in either the time or frequency domain. If one is working in the frequency domain the fatigue life can be evaluated using one of the available methods like the Rayleigh (Narrow Band) approach or the more recent Dirlik method. Historically, the random stresses needed as input to these methods have been evaluated by the FEA solver (eg Abaqus, or Nastran) and these “in built” stress evaluations have limitations which relate to the fact that the stress conditions are complex and so the common “equivalents” for stress like von-Mises or Principal have not been available. There have also been limitations in the location and method of averaging for such stresses. In addition, the fatigue calculation approach for doing the evaluation has been constrained to the linear stress based (S-N) method.

The objective of this paper is to demonstrate that frequency domain methods for calculating structural response and fatigue damage can be more widely applicable than previously thought. This will be demonstrated by comparing results of time domain vs. frequency domain approaches for a series of fatigue/durability problems with increasing complexity. These problems involve both static and dynamic behavior. Also, both single input and multiple correlated inputs are considered. And most important of all, a variety of non-stationary loading types have been used. All of the example problems investigated are typically found in the automotive industry, with measured loads from the field or from the proving ground.

The time domain is currently the most widely chosen option in fatigue testing to fully represent random events occurring in multiple simultaneous input channels. In vehicles for example, time domain tests can represent the same conditions of the road, by applying the same loads at the hard points of the vehicle along a time history. The main drawback of this methodology is the extensive testing duration and hardware cost. Time domain based fatigue tests are composed of a complex hardware, which requires servo motors to work, in order to induce the specific amount of load at a specific time window. These tests are time consuming, since they require the same length duration of the event they are reproducing, times the required repetitions. The frequency domain method for fatigue testing, on the other hand, requires simpler hardware, since there are no need for servomotors and the test length is reduced, since there is no need to run the full event times the required repetitions.

Frequency domain methods of analysis are now being used for the evaluation of fatigue for large vehicle systems and these methods offer advantages over equivalent time domain approaches in a number of ways, including analysis efficiency and the usefulness of derived results. One big potential advantage is to be able to do localized sub-component analysis using “cascaded” loads. Such sub-components can be analysed with refined parameters such as more sophisticated damping or a different frequency range. Local parts can also be re-analysed at a different phase in the design program. This paper will demonstrate the approach and show examples of the method.