Designing sway bar - very upset!

I can't for the life of me find my C. Smith "to Win" books, four years and two moves after I last saw them! This is getting VERY serious...

Can someone out there help me out with the formula for tubular anti-roll bar rates, while I decide where else I might look - or if I need to order new copies.

Thanks

K

Kirk,

Neither Tune or Engineer to Win give a formula...perhaps Prepare (which I don't have)

For a solid bar:

Rate (Pi * G * D^4)/(16 * B^2 * L)

Where G= Modulus of elasticity of material

D= Diameter of Bar

B= Length of lever arm (measured 90 degrees to length of bar---not along the lever arm) Adjacent not Hypotenuse

L= Length of bar

4130 C34-C38 has a G of 1.2 x 10^7

I obtained this formula from a suspension engineering class I took years ago--pre FSAE type programs.

You will be guessing on the modulus of elasticity, so perhaps this will get you close enough to at least get the right OD bar...you can then tune with the lever arm. Geometry of the links and mounting bushings will also effect things, but at least it will be a good starting point.

Perhaps ADDCO has a site with a formula or a sprint car website/chat forum might be of some help.

on edit--for a hollow bar my notes say to substitute (O.D.^4 - I.D. ^4) for the "D^4" in the solid bar formula.

Using a 1" solid bar vs 1" hollow w/.125" wall gives a change of about 40%

Happy engineering!

Daryl DeArman


[This message has been edited by Quickshoe (edited May 18, 2004).]

I had trouble finding the books...A couple of book stores told me that they where out of print...I even gave them ISBN numbers. I finally got a good person at Borders and they where able to find me the book.

Steve Smith autosports has them..

------------------
Chuck Baader
E30 ITA under construction
Alabama Region Divisional Registrar

I am considering hollow sway bars for my car to lose weight. I have removed every legal item, and am still 108 pounds over. The problem is that I am getting conflicting information. My fabricator tells me that a hollow bar is torsionally stronger than an equal diameter solid bar. Roland de Marcellus, in his book “Handling What It Is And How To Get It” states that a hollow bar is approximately 65% less than the same size solid bar. Who’s right? The weight savings from two one inch solid bars would be substantial.

Chuck
S.E. Region
#34 ITA

Kirk,

I don't have my books nearby, but here's an idea: Assuming Daryl's formula is correct for a solid bar, just do 2 calculations.

That is, do one calculation for a solid bar with the same OD as the proposed hollow bar, and do a second calculation for an imaginary solid bar that would fit inside the hollow bar (OD of imaginary bar is ID of hollow bar). Then subtract the two numbers to get your resulting stiffness...

And, I might add, please make it MUCH stiffer before taking any more cornering pictures or I might get seasick

Chuck,

It all depends on WHAT they're made of. Assuming exactly the same materials (unlikely), a hollow bar of exactly the same OD will be a little softer than a solid bar. But, A hollow bar can be made with a slightly larger OD to have the same stiffness as the solid bar while saving weight.

You already know what the lightest,stiffest bar is,so what is the issue? Do you think you can make a lighter effective bar?

Dick Shine

Substitute (D^4 - d^4) for D^4 in your formula above and it will work.

D = outside Dia.
D = inside Dia.

I believe I got this from an old Steve Smith book. I just happen to keep that formula in a little notebook I always seem to have handy

As far as the index of elasticity, I think you will find that there is such a small difference in the indices that the difference between mild steel and chrome moly will be so small that they will be inconsequential. It is the method and degree of heat treat that will make the difference.

You are right on, Mark; The STIFFNESS of chromoly is virtually identical to mild steel. What makes it useful is that it can deflect a great deal more before it plastically deforms. Essentially - it makes for a better spring. Remember though - it loses it's properties when welded - in fact, it becomes brittle if you don't anneal it, at which point, it is virtually identical (in the heat effected region) as mild steel. (assumming you don't then reheat treat the bar). Bottom line - only bother using chromoly if you are bolting together a swaybar.

back up from the dead !!!


when calculating a sway bar like this(not my pic)


what do you use for the lever arm length?

is it from the center of tube to the mounting holes? what about it being mounted on both sides or is that negligible?

thanks

The lever arm length will be the distance from the center of the bar diameter TO the end of the arm that is providing the twisting force on it (trailing arm?).

...and depending on what car it's for, you'll come back and want to do the math again for the additional rear bar.

The answer to Shine's question above was, "The issue is it doesn't give us enough rear roll stiffness."

K

Man, Dick had to wait a looong time for THAT answer, LOL!

so would it basically be from the center of the bar to the center of the stub axle? or the actual end of the trailing arm?

pic of the type of beam it will be going on(not my pic)

Yes, in a straight shot from the center of the stub axle to the center of the bar, not following the curvature of the trailing arm.

thanks, now I am getting numbers that look reasonable

Yes, in a straight shot from the center of the stub axle to the center of the bar, not following the curvature of the trailing arm.[/b]

Click to expand...

Well, not exactly.

The problem with the rear suspension design of the V-dubs is that it *is* a rear swaybar all by itself. Adding Shine's bar to that design creates two "problems":

One, the change in torsion rate is not nearly as large as one might think. The "real" change in torsion is not a function of lever arm vs. torsion rate, as it is for most cars, it's a function of percentage change in torsion of the whole system, which is a lot less than one might imagine.

Two, that supplemental bar does not act on the same axis as the main beam (no co-axial). Therefore, the force is a function of the lateral relative deviation of the stubby "legs" it's mounted on, rather than in torsion. That bar is not being torqued about its axis, it's acting in tension, resisting the arms moving via stretching.

It's more of a complex function than you might think, but to break it down into its simplest form, you *might* be able to calculate the tension value of the torsion bar itself over a lever arm of the linear distance from the bar's axis to the beam's pivot axis (short). But it would only be an approximation.

Bottom line, it's not really doin' much at all, except maybe stiffening up the main beam a bit, and it's possible that it may very well be creating some minimal toe changes as the front of those "stubbies" are pulled closer together during deviation. You'd get about the same action as welding in a piece of tubing to those triangular stiffening tabs. To increase torsion you're better off "boxing" in the beam itself like we used to do on the old Rabbits...