Stresses by High G turn test on chassis in Solidworks ??

[PlatinumZealot]

Ok guys I have a problem. I modeled a chassis in SolidWorks and I want to stress test it for going through a 2G corner. The issue is that Solid works only has Static tests for my setup. (Drop test and Fatigue can not help here). And static test require that a restraint be placed on the model. Now in real life there is nothing restraining the car to the ground equally countering all the forces (because the car is accelerating)

In other words for external forces, You have weight and the force from the tyres.. but no restraints (no points fixing the model)

So how in the world do I model a Dynamic motions such as a cornering acceleration into a static Test that requires a restraint?!

Has anybody tested a chassis in solid works before?

[tomislavp4]

For tests like these you should use the Motion add-in, is that what you´re using?

[Tim.Wright]

You want to do the stress on ALL of the suspension components in one study? That will be a pretty massive study. Sounds like 'death by contact sets*' to me. I would suggest solving for the reaction forces for each member in isolation and apply those forces to an isolated FEA study.

I'm pretty sure these days solidworks can do a kinetmatic/kinetic analysis on a mechanism and use those forces to drive a subsequent FEA study.

However you do it, you can solve for these forces for a steady state corner by setting up an MBD model as follows;

Forces go in at the CG;
Cornering force (=ma)
Weight force (1G down)

Restraints could be
Vertial constraints at the contact patches (constrain to the road)
Longitudinal consraints at the contact patches (purely to 'hold the model still')
Lateral constraints on the two outside tyres
Leave the inside tyres free laterally

What you have there is a model which feeds forces into the CG and reacts them at the CP of the outside tyres only. This will give you half decent results and a place to start. Like any model, this is a simplification. You have to understand the inputs before you can make use of the results.

Tim

*Death by contact sets: A type of slow morbid death which occurs to an FEA operator due to the massively repetitive task of defining part contact conditions for every part-part contact in an assembly. EG a 200 tooth sprocket and a chain touching it at precisely 124 places.

[PlatinumZealot]

Thanks, I forgot to mention on thing..lol I just have the bare frame right now. It is just one single part. I haven't put on (made) the suspension parts, so there are no tyres.

You know like how you can transfer forces to other locations?
What I originally had in mind for the forces, was to have the forces from the suspension acting at the mounting points on the frame, which makes sense.. but then I remembered the implications of a static test.

OK.. You know what i will do.. I will try something in motion and return.

[Jersey Tom]

Not sure you're going to be able to do this one in Solidworks. Your constraints, by the way, are pretty easy. It's the applied load that's hard.

Constrain the frame either at the hubs if you have a suspension modeled (sounds like you don't), or at the suspension pickup points.

The loading is difficult because all the stresses here are from inertial forces. It's a distributed load, not a point load. If you're looking for stress levels, I don't think a point load case (at the CG) will tell you much.

Also you have the big issue of material properties. Calling the frame homogeneous 4130 or what have you, is not remotely accurate given the alloying and microstructure will change dramatically as you get from bare steel, into the HAZ, and then the weld pool.

This a FSAE car?

To be honest, it's hard to fail a chassis by just driving (not crashing). With even moderate trussing, tube thicknesses specified by the series for safety will make the thing well overdesigned.

Deflection is what you need to be looking out for, and that's pretty easy to pick out just by looking at each section of the chassis from hub to hub and looking for lack of triangulation.

[PlatinumZealot]

I am just doing this as an at home project.

Somehow I figured out how to do the crash test with decent results a few days ago. I designed a crash structure for the frame. After some tests I could only muster up a deceleration of about 30 g's in the first 16cm of deformation.I think Formula 1 is 10g's in the first 16cm.

I used a simple method (coincidentally similar to what you suggested =D> ) to estimate the inertial loads on the frame in the deceleration of the crash test, and since the vehicle is hitting a wall, i just put the restraints at the front. Assumed that all the different masses in the car where attached to the frame at certain points. Then applied pre-calculated inertial forces at those points. Crude but the results look Ok.

This is my crash test:

I tested the crashbox first (that 1200kg block in the back represents the car).The test resulted 30 g's deceleration in the first 16cm of deformation. THe crash box is Aluminum skin and PE Foam on the inside.



Frame:

You can see my crude estimates of the inertial loads. The restraint is at the front.



The deflections and stresses seemed reasonable here so I left the crash test at that. You see those two red members at the front of the tunnel, they were triangulated afterward.


The Turning scenario now, yup It's hard to think out. I thought of using the same method too.... but It didn't strike me to use the hubs/pickup points.


I will try it and return.

[bajanf1]

As Tomislav said this can be done in Solidworks CosmosMotion if you have your suspension points modelled in already. The lead suspension guy when I did Formula student created a simplified chassis model using the correct mounting points. Used simple A-arms, uprights and tyres, calculated weight transfer and steer angle (for particular radius corner) and we also had some tyre data so we got contact patch forces from it. Entering the loads for the simulation produced some interesting results that we could then import in the FEA. We were able to get forces and directions in all ball joints, so I assume you should be able to select some chassis members to see forces as well.

Saw what you are doing on 'wheels' Smikle, very good work man! i'm doing something similar as well to better understand vehicle dynamics and design.

[PlatinumZealot]

Thanks.

I am moving on to the suspension now. I will have to find some way to add weight to the frame (so that i acts like a full car) So lets see how it goes..

[im.freezin]

i am stuck in a similar condition....i hav made a design in solidworks but not able to test it...it has problems with creating a mesh....it is created using structural members....please help about the changes so that a mesh can be created...

[Jersey Tom]

If that's for FSAE, your chassis doesn't pass the safety rules, and I can tell you off hand it will have crap all rigidity.

Triangulation, my man. It's your friend.

Use a beam element mesh, not a solid or shell mesh. Should work.

[Ciro Pabón]

Great thread.

Just in case, after reading the previous posts, I spent half an hour googling about CAD software for engineering.

I ended downloading FreeCad and eDrawings Viewer.

[im.freezin]

@jersey tom...
yea its for fsae....removed the triangulations hoping the mesh mite get created....dont worry....we wont make a crappy one... ....got it done....thanx a ton

[im.freezin]

well this is what i got after i tested the chassis with the triangulations....tested it at 6g's...now can u please tell me wat does this test signify...m quite of a newbie n dont know much about this...help!

[Jersey Tom]

So you just did some random FEA without even knowing why you did it?

[Ciro Pabón]

What's FEA? And why there are so many squares in im.freezin second design after "triangulation"?

In the end, the answer to this thread original question is to fix the car at the outer wheels, isn't it? So, the car doesn't rotate, does it?

What's HAZ and the weld pool?