Can anyone give me a hand on my ARB calculations? I'm using the equations given in the OptimumG tech tips found here http://optimumg.com/documents/techtips/Springs%20&%20Dampers_Tech_Tip_2.pdf

I've gone over my numbers about a million times and tried to check for inconsistent units, typos, etc. and its no use, I get a negative value for the required ARB stiffnesses. I refuse to believe that this is possible with 100-135 lbs springs, shooting for 1 deg/g roll gradient.

I don't know how to attach my spreadsheet, but if anyone is interested in taking a look, I can email it to you. I would appreciate it a lot if someone could tell me what I'm doing wrong.

Thank you

Edit: This may be a terrificly stupid question, but are you supposed to get a negative number for the required ARB stiffness?? They don't give an example on OptimumG so I have nothing to go by. I'm pretty sure Milliken gets positive numbers, that's why it's probably a dumb question haha.

I posted the spreadsheet for Ross on our site:

http://www.seas.upenn.edu/~fsae/docs/ARBcalcs.xls

-Matt

I have seen a spread sheet which calculates required ARB stiffness' based on a target roll gradient and spring stiffness'. A negative number meant that you are already getting the stiffness you require from the springs alone.

If you don't reckon this is possible with your springs, maybe theres an error, like your motion ratios being upside down or something.

Tim Wright
Suspension & Vehicle Dynamics
Curtin University

This message has been edited. Last edited by: Timo,

well, I did find a missing factor of g, so we're back in the positive range for now.. although a little high.
Now thats back into the realm of the other guys (like Ross who originally posted).

On that note, anyone care to share their target roll gradients? From what I've seen, 1-1.5 deg/g is fairly common.

1.5 is fairly soft compared to most teams. For a first year car I'd say do the standard .8-1.0 and see how it works.

Check here- I found a mistake after the 2nd tech tip was released and corrected it on the 4th one.

I checked your spreadsheet, the "missing g" is because you entered "vehicle weight" as a mass instead of force, and hence had to add in a "g" to convert it to a force (the tech tip says to put W, vehicle weight, in Newtons).

If you don't check your units, you can't have any pudding

The negative number begins at the roll rate due to the ride springs. Here's the confusion (it's my fault): the equation has a negative sign (-H * W), which holds true if you enter the "H" as a vector (so you would enter a negative number if your cg is above your roll center, because SAE +Z is down). I didn't make that clear at all when I defined "H". So get rid of the negative sign in "-H * W", and it will be correct up to that point (I didn't check the equations after that one).

Thanks Matt. We caught the "missing g" later last night. As far as the confusion with the negative sign, I don't think it actually affects the calculated ARB stiffnesses, it just tells me that my roll gradient from the ride springs is negative instead of positive.

I just put the spreadsheet up that we revised last night, after finding units mistake.

and its our first year with ARB - 3rd year overall

ARB? Anti roll bars maybe?

Hi,
I have read the forums and the Optimum G docs but we are still confused. We are trying to do a simple statics calculation to see how much stress is taken by the ARB but when we do this we get that the bar will fail miserably, even though we are starting with a 4130 5/8" .065 bar. Which from what I am told should be too much. Any ideas? We have the ARB roll rate at 4580.8 in-lb/deg. This came from calculations in the RCVD book. (Ch16)
Used Cosmos too, but it says we are getting a deflection of 18" at the ends. Probably not constrained right :/
Not asking for the answer just a push in the right direction.
Thanks

Nube,

Determine the rotation that the swaybar will see during steady state cornering. Apply that rotation to the swaybar to determine stress (no FEA needed, just a hand calculation).

That'll get you started ...

Thanks for the response. We figured out that there would be about .5 " of deflection at the ends per degree of body roll. That seems reasonable, but when we calculate the dimensions from that and the stress it just doesn't seem right. The length of the bending arm is too long and the stress is way too high. We checked units they are fine.

Nube,

Are you converting the end link displacement (linear) to degrees of twist (angular) for your torsion calcs correctly? Assume rigid end links for rough calcs.

Depending on the formula you are using make sure that you are using the proper units for twist, confusing radians and degrees would give a result like you describe.

Best,
Drew

Ok, I appreciate all the help. I just was wondering if off the top of your head, if I want a ARB roll rate of approximately 1000 in-lb/deg from the front/rear separately, does a 1" .134 tube seem too big? Because it seems too big to us, but this is the first time we have run a sway bar so we don't have much to compare to...shooting for .8 deg/g roll gradient and 2.5 Hz natural frequency in the rear.
Thanks

The dimensions of the bar depends on your motion ratio, the active bar length, lever arm length and desired rate.
This may be a stupid question, but have you ensured that you have converted the tensile yield stress of the 4130 tube to a shear yield stress?

We know the displacement we would see with our motion ratio, we know the bar length, the lever arm is our variable, and desired rate we know. But what do you mean by converting the tensile yield stress to shear yield stress? Because we calculated the stress in the moment arms due to bending and calculated the displacement. Then we figured out how much displacement would be going into the torsion section of the bar and from there we figured out the shear stress. And we know the yield properties of steel in tension. Am I stupid?
Thanks for the help.

Shear yield stress = 0.577 x Tensile yield stress

You can't just get your caluclated shear stress from the calculation of angle of twist and compare it to the published yield strength of the 4130 tube you have.

If you've done this and I've completely missed your point, ignore my last 2 posts