FEA for Dynamic Multybody Physics Fatigue 5

[Timms]

He

I am trying to decide What FEA package I should buy. I need to model the rear supsension of a motorcycle. I hope to be able to animate it and determine the stresses the rear swing arm, axel and differential will be subject to. I would like to model it in Solidworks but am wondering if Cosmos would be sufficient or if I need to use a program like Ansys. I have a weee bit of experience with ANSYS but perhaps not enough to save time on my learning curve. If I get Cosmos I would have to learn everthing from scratch. Could anyone give me some direction?

Blessings

 

[fkmeyers]

The problem with Raleigh damping is you can not specify it for individual components which you can in NENastran. NENastran supports Raleigh damping and it simply takes a percentage of the entire stiffness and mass and uses this for damping. In Nastran you can specify damping for individual materials and even make it frequency dependent like DMACX said. I was not aware that Algor was used at all in the automotive industry. Like DMACX said Nastran (and ABAQUS and slowly ANSYS) are pretty much the standards in the aerospace and automotive industries.

I also agree with what DMACX said about the NENastran support and training team. They seem very determined to make sure any issue is resolved and that you are up and running with the product immediately. I also have had nothing but excellent support from them and always recommend them to other companies. When we picked NENastran we looked at several other packages including some of the ones mentioned here in this thread and it was the clear winner in our comparison based on accuracy, performance, support, price, and existing industry usage. Other packages may look easy but you will outgrow them fast and when you need a specific capability and they do not have(like material based structural damping), you are forced to accept the limitations of the application or go to another one. Right now we have been able to do everything we need with NENastran.

 

[GBor]

I was not aware that Algor was used at all in the automotive industry

fkmeyers, you just blew my image of you...I thought you knew everything!

I agree with Frank on the Raleigh Damping. Although it is available, it is not the way for you to go and it does apply to the entire model (which is generally appropriate for me in dealing with uniform material structures). I'm not sure what happened in your WebEx, but I'm also not familiar enough with your particular application (sorry, not a motorcycle enthusiast!) Based on what I know of suspension systems, viscous damping in the shock absorber would be where I would have sent you.

Mechanical Event Simulation does allow for material damping at least in trusses, bricks, and tets. Not sure about any other element types. This could easily be incorporated into the simulation below without tremendously increasing the run time (depending on your hardware).

Here's a link to the Algor website with a simulation of a suspension system.

http://www.algor.com/products/analysis_replays/suspension/default.asp?sQuery=spring

Unfortunately, this simulation was only concerned about the spring, so everything else is modeled in "Kinematic elements", which don't calculate stress (saves processing time).

I would make one comment...you started this thread talking about COSMOS as a possible option. I was hoping to see someone comment on COSMOS's ability to handle something like this because I haven't seen it in a year or so, but no one has come to its defense. Draw your own conclusion.

In a problem as complex as yours, if you get a chance, I would love to see a comparison of NENastran, Algor, LS-Dyna, any of these programs especially if there is some test data to back it up. I've tested Algor's MES to physical test data and been very happy with the comparison. I've compared Algor's non-linear processor to NENastran's non-linear processor and to COSMOS's non-linear processor with some satisfactory results...not sure what this means, but perhaps it will help you in your decision-making process. The key to any of these is learning how to use them.


Garland E. Borowski, PE
Borowski Engineering & Analytical Services, Inc.

http://www.borowskiengineering.com

 

[GregLocock]

I am astonished to read that people think that the internal damping of steel components is of any great importance in a suspension when considering fatigue and shock damage.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GBor]

Greg,

Since this forum is supposed to be about technical ideas and not software marketing (which I feel like I've been doing this thread), the thought of an event simulation allows the input of a road profile that you can "drive" your tire over. Since it is an event simulation, you take one step back in the physics and let the software generate the force from the F=ma calculation. You could input a variety of road profiles from a variety of sources (vehicle test facilities - particularly military, US Forestry Dept., Wild Guess). If you already own the software, the analysis is not too expensive in terms of time (assuming you know how to use the software).

Now, believing the results? I guess that depends on your confidence in your analytical skills, but is there some reason why this wouldn't give a pretty good idea of how the system would respond without an ADAMS simulation?

 

[GregLocock]

Already done. It is not totally reliable yet, but high frequency tire models such as ftire show the way forward for the rest of us.

Tire companies already have their own high frequency, high fidelity, tire models, mostly non-linear FEA, and they can afford Crays on which to run them.

However, for predicting loads in the vehicle you also need a good model of the rest of the suspension system, in particular the shocks and the elastomers. High fidelity high frequency models of these do exist, but so far as I know nobody is routinely running such a model, with high fidelity models for all sub components, to predict road load durability from scanned road surface data.

There is one huge note of caution here. Given a repeatable road surface, a known tire, and 6 dof wheel force transducer data, one would have thought it possible to back generate a road profile to drive a different vehicle (possibly on the same tyres). That does not work to anything like the desired accuracy.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Timms]

Thank you all very much!

I have finally made a decision! I talked to yet another engineer at Algor and discovered that it does have the full capabilities that I need. They explained to me that I can create my own load curve that is dependent on the velocity in a nonlinear manner for dashpod damping. I will be able to accurately characterize the shock. I will also be able to do material too. The nice thing about Algor is it can do everthing in one system. I don't have to do a rigid body first and then transpose it into an FEA system, Which would limit me to a single time frame. I realize there are other packages out there that can do the same but Algor also offered incredible service that other companies that I have approached can't seem to offer. Also they are very reasonable with their price. All around I am impressed with them so far. Hopefully their software and service will continue to have the integrity that I am sensing from them. If someone is interested in an engineering website that shows a little more of their capability let me know and I will post it.

Blessings