Stressed engine in ANSYS?

[SeanTLindsay]

I'm working on our 2005 chassis right now, and I was wondering how other teams modeled the effects of a stressed engine.

I was thinking of just modeling some plain 1" tubing connecting the front and rear engine mounts so I could be assured that the actual torsional stiffness would be higher than the number I get from Ansys.

[Denny Trimble]

We've modeled our engine with rigid bar elements connecting every mount with every other mount, and we've had very close correlation with our FEA and testing. But, we don't fully stress the engine, so the forces aren't very high. If you were to rely completely on the engine for the stiffness of the rear of the car, more accurate modeling of the case stiffness might be needed.

One of our sponsors has a CMM machine, I'm tempted to take some case halves in and see what kind of a model I can get...

[Brent Howard]

Sean and Denny,

One really quick, but probably not extremly accurate, way to get the effect of the engine might be to use a solid model of the engine and shell it. I have a model of the f4i in solidworks that is just a solid block, but it would be easy to shell to the average thickness of the case.

Brent

[SeanTLindsay]

Denny: What tube dimensions did you use?

Brent: If you know how to import and shell a model in Ansys, you're way ahead of me http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Denny Trimble]

I modeled the engine members as "rigid links", an element type in ANSYS for linear elements. The rest of the tubes were beam elements, specified by a line in 3D space, and section properties (OD, wall thickness, material).

[Brent Howard]

Sean,

I'm not too sure about Ansys as I have never done any FEA, but if you call me at work tommorow I can explain how to do the engine model. I think it is on the school drive, and shelling is easy. Ask Steve for my work/cell number.

Brent

[Kevin Hall]

Importing into ANSYS is a nightmare. IGES files are terrible to work with. What I ended up doing is running beam4 elements from each engine mount to the other, so that it was triangulated in every possible direction, with each member approximately 10X stiffer than 1" .065" wall tubing. The results worked out extremely accurately in terms of deflections.

[NJM.]

Another quick and easy way to model the engine is to create an area element. I created nodes where the pick ups on the engine are and created a 4 sided area between these elements. I did this on all sides until I had a 6 sided box representing the engine. I then gave it big stiffness properties because hopefully you will not be deflecting your engine very much. The results were spot on. Here is a picture of what it looks like. This is our rear chassis in isolation.

http://www4.ncsu.edu/~njmckay/rear_chass_2.jpg

hope it helps

[SeanTLindsay]

How did you model a stressed floorpan? I'm not sure how to merge the nodes of a meshed area with those of the space frame. Would a cross brace of equivalent strength to the floor pan in shear be sufficient?