This falls into the category of "do it for fun, not to save money".

You will spend a great deal of time and the results will almost surely be less than the commercial parts. Your out-of-pocket expenses will likely exceed the cost of a set of used bars.

If you do this, be sure to have the bars properly heat treated. Failing this, the bars will either break or bend permanently.

Proper operation of a sway bars is widely misunderstood. The "how stuff works" article linked above is packed with misinformation.

jluet is on the money. Sway bars do not cause weight transfer on a given axle, they transfer weight between axles. This is how they are used to balance handling.

They can also be used to trade wheel independance for roll stiffness.

I agree with all of the above with one minor exception that doesn't apply to independent suspension transaxle cars like Porsches. On a live rear axle front engine car a rear sway bar is often used to reduce suspension to body "twist" under straight line acceleration and therefore is actually transfering force within one axle. In this case the bar is used to transfer more of the engines torque to the pavement by trying to spread it more evenly across the rear tires. That is the only case where the bars effect only a single axle to my knowledge.

Heat treating is definately important or the bars will break or permanently bend as Chuck suggested. Also you looked up the material in Mcmaster-Carr, while a good source to find stuff they are expensive in many cases, find a metal supplier nearby for a better price and stay away from annealed metal, it bends easier but cannot be heat treated. Making your own bar would be a fun endeavor but you have to realize that doing it yourself will likely not produce a bar that is as good as a used one. Sure you can calculate rigidity etc. but a some point you're actually have to measure it to confirm those calculations.

i'm not sure about this. i've read the book. :D

anti roll bars (correct term) are just that and their primary purpose (at least originally) was to reduce body roll.

it must have been it says so in the book. ;)

you can transfer weight front to rear by adjusting the rates of the coils or t/bars. you want oversteer you go for a light front bar.

yes you can also do this with the ARB's (and i guess alot of guys prefer that route).

now, wheres that damn book.........

Woo Hoo !! I just scored a turbo rear sway bar for just $56 (including shipping). Now if I can get a matching front bar I'll be in business.

Of course, this just means I can safely go crazy trying to fabricate bigger DIY sway bars because I will have a good inexpensive used set as a "fall back position" ;).

So seriously (now strictly for informational purposes), if one gets some 4140 heat treated hardened steel rod or tubes, how difficult would it be to bend it to the proper shape? And how might one get all of the bends on the same geometric plane? I know Eastwood sells a tool for bending pipe. Would it be strong enough or might the hardened steel break? And getting the thing to lie flat after the bends are done might be a headache. Hmmm...

"Heat treating is definately important or the bars will break or permanently bend as Chuck suggested."
True, a spring temper is probably what's needed.

"...stay away from annealed metal, it bends easier but cannot be heat treated."
True it bends easier in it's annealed state, but untrue it can't be heat treated. Knifemakers do this all the time. Annealed high carbon steel is steel without the heat treat or hardening process applied, or that has had the hardening "removed". There is also a process called normalizing which is somewhat similar to annealing that allows the molecular structure of worked steel to settle into a more stable structure. Heat treating is typically a two part process. The steel is brought to full hard (martensitic crystalline structure) where it becomes temporarily non-magnetic, and is immediately quenched in oil or air. After it cools it is reheated to a lower temperature and allowed to soak for an hour or two. A peacock blue color on the steel for a spring temper would be approximately correct. You could do this with a torch if you could keep the entire bar consistently hot until you quenched it (not easy with such a large piece of steel). Cryogenic freezing of the steel is also sometimes done to maximize the conversion of the original austenitic structure to martensite.
I'd "$pring" for a ready made bar if I were you.

IROC, can you give more detail on modulus of rigidity please.

I'll try. The equation for determining how much rotation you get in a round bar from a given amount of input torque:

Theta = TL/GJ

Where:

Theta = the angle of twist
T = the input torque
L = the length of the bar
G = the modulus of rigidity
J = the polar area moment of inertia

J = pi(D^4 - d^4), D is OD, d is ID

Per Mark's Handbook, G (the modulus of rigidity or shearing modulus) is given as:

Cast steel: 11.3 x 10^6 psi (not good for anti-roll bars, I'm sure)
Cold rolled steel: 11.5 x 10^6 psi
Stainless steel: 10.6 x 10^6 psi
"All other steels": 11.0-11.9 x 10^6 psi

So, as I mentioned previously, for all intents and purposes, the G term in this equation is the same for "all steels". I'd pick 11.5 and run with it.

Using the above equation, you can - in theory - design your own anti-roll bars if you put alot of thought into it. Moment arms and the attachment point on the suspension, etc. is going to play a factor....

Mike

no substitute,

Thanks for the correction, I always thought annealed metal was metal that had been "overheated" to the point that it remained soft and therefore could not be rehardened. I really wasn't certain of it however and you've given me another ounce of knowledge for my collection. ;)

OK, now who wants to stamp out their own new body shell?

BTW, I think Ti is more springy than steel. Think about how much wt. you could save....

This crossed my mind too, a titanium swaybar. 6/4 (aluminum/vanadium content as part of the alloy) titanium has a natural spring to it, and I'm guessing it's about 3/4 as heavy as steel. Didn't they use Ti of some type for the rubber axles on the 935? In fact, aren't there already titanium swaybars?

You didn't mean a titanium body shell did you?