Where to measure chassis stiffness?

[rwolcott23]

We are currently testing our chassis torsional rigidity. Where is the best (or standard) place we should be taking deflection measurements during torquing?

Thanks,
Bob

[rwolcott23]

We are currently testing our chassis torsional rigidity. Where is the best (or standard) place we should be taking deflection measurements during torquing?

Thanks,
Bob

[Lyn Labahn UW-Madison]

All along the length of the chassis! This can help you determine any weak areas that need improvement in future designs. As long as you are setting up the chassis to test it, why not take multiple measurements.

[BeaverGuy]

What my teammate was reffering to was the reported chassis stiffness measurement. We have a profile of the displacement along the chassis. When you give a reported chassis stiffness at what points are you reporting it. For example the stiffness when measured at the chassis will be higher than the stiffness measured at the hubs because the suspension deflects also.

[GTmule]

It's my feeling that the final number should be from the hub, as suspension stiffness is an integral part of chassis stiffness. The profile is great to have along the chassis, but as a point of comparison, the measure at the hub is what really matters.

[Lyn Labahn UW-Madison]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by GTmule:
It's my feeling that the final number should be from the hub, as suspension stiffness is an integral part of chassis stiffness. The profile is great to have along the chassis, but as a point of comparison, the measure at the hub is what really matters. </div></BLOCKQUOTE>

Yes, I agree the total "installation stiffness" including the deflections of the suspension is what the cars "sees" and affects the performance. However, if a design judge where to ask you of the stiffness of just the frame itself, which is important to judge the design of the frame, then you better have this number as well.

[markocosic]

I'd load the chassis/take the deflection measurements from the suspension mounting points, taking angle of twist between the front and rear suspension mounts for a given loading as the frame/chassis stiffness. Seems logical given that that's where the major loadings will be?

[Jeff The Pyro]

the *most* important reason to have a torsionally stiff chassis is to help the suspension work correctly. ideally you would want to replace the shocks/springs with solid struts and measure the torsional rigidity at the hubs. the problem is you will see a small amount of initial deflection just from the tiny amount of slop in the hardware. The best way to get around this is to put a small torsional preload on the chassis to take up all this slop and then take the measurement from there. if you wanted to get really crazy, you could take measurements at small increments of torque, graph that, and then look at the graph and figure out where the function becomes approximately linear and take the stiffness from there. It's also important that the suspension points dont move a lot in relation to each other during any loading... but if you're confident that your chassis is well designed this shouldnt be much of an issue, however i'm guessing that if you went ahead and measured the stiffness at the inboard suspension points as well and compared the two and came up with a logical conclusion based on these two results, the judges would probably be more than satisfied.

From the judges' standpoint, the important thing is can you come with a rational explanation for why you chose the route that you chose.

[Denny Trimble]

It's just as easy to build a torsion test fixture that loads through hub plates, if not easier, than building one that attaches to suspension pickups. And, the loading will be more accurate.

If you're bolting to suspension mounts, how do you control the exact force vectors on each point? I don't think you can, accurately, without the control arms, pushrods, and dummy shocks on the car.

Here's our fixture:
http://students.washington.edu/dennyt/fsae/torsion_screenshot_DT-5-10-04.jpg

And, you can get chassis stiffness (frame twist) as well as installed stiffness (hub twist), which is very useful.

[Lash]

Denny, I built a similar structure that was clamped to the table (i'm assuming you clamped yours too). The big issue what I had was the deflection of the bar. At one point while I was loading it, the deflection (on the side that caused the chassis to be pushed up) was less then it was with the previous weight. It seemed impossible but if you account for the deflection of the bar it will pull the front of the chassis downwards, and at the same time twist it. Our moment arm was huge...(like 7 feet or so from the pivit) Worked out to be putting 679 ft lbs on the hubs!

[karsten]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">All along the length of the chassis! </div></BLOCKQUOTE>

iam not convinced that this makes sense
if there is not straight structural member
going from front to rear.
deflection is caused by angle x lever(not sure
if its the right engl. term)?? the lever is
the distance from a axis (if you think
about a torsion beam) to the measure point.
in our tublar space frame you cannot measure
the deflection with the same lever in the
front/middle/rear (its not lika tube with constant cross-section). you can take that into
account but we got a result you cannot compare
to other cars. its just good to check fea-modell.