Chassis analysis on ansys

[pentek123456]

Hi,

I was doing stress analysis on my chassis and the way we do our stress analysis is weird. I was wondering how other teams do the stress analysis of their chassis.

[pentek123456]

Hi,

I was doing stress analysis on my chassis and the way we do our stress analysis is weird. I was wondering how other teams do the stress analysis of their chassis.

[The AFX Master]

What about sharing how "weird" you're doing it?..

This forum is not for handouts.

[pentek123456]

what we do is that to find the back torsional rigidity, we apply 3000 N force in the Z-direction at one of the back suspension point and -3000 N force in z-direction at the back suspension. also we clamp the front suspension points. and we see the deflection that occurs in the back suspension points, and then we use the equation to find the torsional rigidity.

[thewoundedsoldier]

I've kinda wondered why teams are so interested in the torsional rigidity around the longitudinal axis. It makes sense that the greatest frame deflection will take place during roll (as opposed to pitch or yaw). However, doesn't the direction in which the roll load is applied matter as well? What if your springs don't point perfectly in the z-axis direction? Then won't your frame bend completely differently?

I would think it valuable to have a TR value for all three axes (or, more importantly, for all three on-track events: braking, turning, accelerating) and apply the forces appropriately based on the design of your suspension.

[Old Greg]

Soldier,

It's not a matter of the force in your springs, it's about the normal force acting on each tire. For a formula car, those forces are almost exclusively in the z direction (barring bumps and changes in road camber).

Of course, if your shock and/or rocker mounts are poorly braced then the orientation of your springs will definitely change the way your frame bends.

[thewoundedsoldier]

That doesn't make sense to me. Think about the sequence of events. The wheel hits a bump and experiences a sudden normal force in the z-direction. The a-arms allow the wheel to travel in some manner without taking much of the force. The vast majority of the force moves through the pullrod/pushrod and is pivoted at the bellcrank into the spring. The only force that the frame sees is from the spring and the bellcrank.

I'd argue that any good TR test would not just consider the spring direction but also the force on the bellcrank pivot (with respect to the frame).

I guess my argument is that the frame is just moving along in space--a sudden force on the wheel in the z-direction does not mean a sudden force on the frame in the z-direction. The frame only sees the load at the spring pickup, the bellcrank pivot, and to a very small extent the inner a-arm pivots.

Please do correct me if I am seeing it wrong.

[Kirk Feldkamp]

Well, based on the way the question was posed, and where I'm guessing pentek123456 is from, you may be going off on a tangent that's falling on deaf ears. There is a thread on here back in like 2005 or so in which a number of people described their setups for torsional testing. I believe Denny Trimble was a key player in that discussion. I think the Cornell SAE paper on frame design and testing also outlines proper assumptions and test rig layout. Basically, IIRC, you replace the spring/shock with a non-deformable dummy rod, and fix the rear hubs to the ground (through stands). Then, you put a pivot point under the car in the front, and apply a torsional load to the front suspension. Then you measure the frame deflection at certain points, and calculate what the twist angle is.

-Kirk

[pentek123456]

i would really appreciate it if you can find me the link to the discussion.

[Wesley]

And we would appreciate it if you use the search box first. Please.

[thewoundedsoldier]

I'm familiar with the 2004 thread titled "Chassis FEA results". I think our frame guy even used it as a basis for his TR test this past year.

I believe that all the methods brought up in that thread support my point that frame analysis must consider the load paths as seen from the frame. Z-direction forces aren't going to cut it.

I also had a discussion with our frame guy this past year about including the static loads in the test (driver, engine, etc). These probably have little effect, but might give a more accurate value.

[Old Greg]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">The frame only sees the load at the spring pickup, the bellcrank pivot, and to a very small extent the inner a-arm pivots.

Please do correct me if I am seeing it wrong. </div></BLOCKQUOTE>

You are correct, but you're only thinking in terms of the frame; Torsional rigidity depends on the entire system. Your load path starts at the tire.

PS: Don't underestimate the forces at the a-arm mounts.

[thewoundedsoldier]

Again, I only mean to challenge the way teams run TR analysis on the frame alone. If a team has a full CAD suspension setup with uprights, hubs, wheel centers and rims then that's awesome--definitely the right way to do it.

But as far as frame design goes, I think a lot of schools are shooting themselves in the foot by designing frames just so that the TR value looks good with longitudinal torsion.

Is it common practice for any of you to CAD the suspension components and set their stiffness to infinity and get a frame-only TR value that way? Apologies if that question is obvious.

[John_Burford]

Soldier

You question goes to "Why is torsional stiffness important" which a question for the vehicle dynamics people on your team. If the first words out of their mouth is something other than cross weight transfer then talk to someone else.

The other bending modes you mentioned are only important for ride comfort and NVH which is of no importants for FSAE cars.

John Burford