Cornell Chassis Stiffness Paper

[bmking425]

So I just completed a torsional stiffness test of our 2013 chassis as a datum for the 2014 chassis. However, I am getting numbers that are half the size I am expeciting (950 ft-lbs/deg). I have looked over the Cornell paper for help and I am confused with one of their equations.

On page 13 (http://www.theoryinpracticeengineering.com/resources/fsae/cornell%20design%20anal%20testing%20fsae.pdf), with the second method of testing where load is only applied to one side, they say that, "The torque acting on the car and resisted at the clamped rearbay is simply the force, P, times the lever arm, L2."
Then when they state the stiffness equation they show that K= (P(L_1+ 2L_2))/(tan^(-1) ((?A+ ?B)/L_1 )). Why does the torque become P(L_1+2L_2)? Can somebody help clear this up for me?

Thanks,
Brandon

[Michael Royce]

You need Bill Riley to answer the question. He wrote the paper!! He does come on the Forum occasionally.

[RobbyObby]

Originally posted by bmking425:
So I just completed a torsional stiffness test of our 2013 chassis as a datum for the 2014 chassis. However, I am getting numbers that are half the size I am expeciting (950 ft-lbs/deg). I have looked over the Cornell paper for help and I am confused with one of their equations.

On page 13 (http://www.theoryinpracticeengineering.com/resources/fsae/cornell%20design%20anal%20testing%20fsae.pdf), with the second method of testing where load is only applied to one side, they say that, "The torque acting on the car and resisted at the clamped rearbay is simply the force, P, times the lever arm, L2."
Then when they state the stiffness equation they show that K= (P(L_1+ 2L_2))/(tan^(-1) ((?A+ ?B)/L_1 )). Why does the torque become P(L_1+2L_2)? Can somebody help clear this up for me?

Thanks,
Brandon

Trust me, you'll be better off micro-analyzing your test setup then you will the maths. If you're using dial indicators to measure displacement, are they calibrated? What is the measurement error? How are you measuring the forse applied? Is your scale correct? How solidly is you chassis constrained? The quoted equation assumes the only deflections occur where the load is applied. This is not always the case. How did you come up with you're "expected value"? FEA? Are you're loads and constraints all correct? Did you use a simplified beam model or a mesh model? I apologize of you actually have gone through and analyzed all this already. But most likely you haven't. In the tests I've run, I've seen variations up to 300% from expected, without account for other variables. I've also seen values almost 25% off after performing te exact same test, twice in a row.

TL;DR: Analyze your test method and setup, before analyzing the maths.