Currently i have road load data containing wheel transducer forces FX,Fy,FZ,MX,MY,MZ and strain gauge data at critical stress concentration zones.
i want to develop generic load cases at wheel centre which can be used for future assesment of same vehicle suspension.
This is a typical problem of "fatigue under variable amplitude loading".
The problem is rather involved and rarely tackled even at level of coupons. When components are under concern (namely, the suspension system) empirical approaches are used, mostly proprietary. Normally one requires a failure criterion for the material under study and a "rate equation" for the degradation of properties with the cyclic loading. Such rate equation are generally obtained under "constant amplitude loading" and then used for the real case with some gross approximations. One of them is the existance of uneffective level of loading. That is to say that below such level of stress no properties degradations are observed. This makes easy subtracting from the real amplitude spectrum those peaks and valley not contributing to the properties degradation. In this way the loading spectrum is enormously simplified. My first suggestion is to measure the fatigue behavior of the whole system under constant loading conditions in order to obtain a Wholer-type curve for the system of suspensions. Then try to adopt a "rate" equation for the properties degradation with cyclic loading in order to model the residual properties of the same system under variable amplitude loading. Under this perspective, sometimes it is assume the validity of the Palmgren-Miner rule, depending on the material under study. These are just few suggestions, or at least the first procedures I will adopt towards similar problems. However the task is rather complicated and I don't really know if rapid and reliable solutions can be found.
There are some software packages that help bridge the gap from test data to simulation. nCode DesignLife is an example, link below. These tools can be very helpful but confidence is always gained through verifying assumptions in the test laboratory and on the vehicle. Historically I have had better accuracy and confidence looking at direct strain measurements and equivalent damage correlation than attempting to just just loads and acceleration data.
http://www.ncode.com/fileadmin/mediapool/nCode/downloads/events/UGM_13/UGM_SignalProcessingForDesignlifeUsers_Munson.pdf
Any time “generic” loading is used overall accuracy for the system becomes more difficult as you include more parts of the system. Consider that a single loading condition for laboratory or simulation cases may be “accelerated fatigue” for that one part of the system could be completely unrealistic for adjacent components in the same system thus the need to verify assumptions through at least some physical testing.
- Badges
- Science topic