The LMP360 needs swaybars. The calculated roll stiffness is currently a paltry 2.5 degress / g which seems low, except at 2G's that consumes all of the available shock travel. It needs roll help.
I've done preliminary maths and determined I need the rollbars to contribute ~300-400 lb/in of spring coefficent to the wheels to get roll stiffness to around 1 deg / g which should keep the car fairly flat. The car is getting a 3rd element installed at the same time (I will post about all that math separately), but I am leaning towards a T-bar swaybar with the 3rd element attached. This would drive the swaybar MAX travel to match the 3rd element shock travel of 2.5". However Loading to this amount of twist would basically be impossible.
It looks like a 3/4" Bar 12" long, with a 10" T-arm would get me in the ballpark for torsional stiffness required. The stress during up to 3G cornering would remain bellow 50Ksi, so Fatigue life will be a thing but Yield should not occur if 4130 is used for the bar. 1020 would probably do the job, but If 3G cornering is a legitimate thing 1" tube may be required with longer arms.
Bar stiffness math came from here: https://calcdevice.com/bar-spring-id247.html
Youngs Modulus for 4130 is 29700000, 1020 29000000.
Poissons Ratio for 4130 is 0.29, 1020 I used 0.27.
Ok so that's the background of What I'm up to.
Has anyone fabricated a sway bar? I'm thinking 4130 Tube with plates Tig welded to the end.. But? Is this a recipie for disaster? I'm okay if I need to replace the Torsion bar every few years to prevent fatigue failure.
I also may look to use the OEM subaru 360 Torsion bars in a "traditional" U bar arangment. They have Hex keyed ends so it would be easy enough to fabricate around without welding.
I don't want to spend huge money, and my space constraints prevent me from using any OEM bars.
There are a couple of companies that will do this, usually to produce sway bars for production cars with no aftermarket, Addco is one?
Another option would be to look into sway bars for circle-track race cars which are a standardized shape and see if you can make those fit (Edit: These are sometimes called Ultra4 sway bars because they're also very popular in that class).
Coleman and Schroeder make splined bars just for this. Typically mini sprint / sprint car and NASCAR stuff.
I'd look into adapting a used set of splined bars before welding some ends on tubing, personally. We run splined sway bars with one off arms on almost everything we do. And they've helped us with goofy stuff for hidden torsion bar suspension prior as well.
I've seen a lot with arms that were just welded to the ends of the tube. Seem to work OK. For a different take, here's a guy whose Youtube channel I follow:
https://www.youtube.com/watch?v=nw-CsHWxHKI
He milled flats on the ends of a U-shaped bar, then milled a groove in a flat lever arm that then bolts to the bar on both ends. Might be more durable (?). Of course, I assume you want to stay away from a solid bar..
Trent
PowerDork
3/8/23 6:37 p.m.
I used to have a side gig reproducing the rear sway bars a company called The Rabbit Hutch used to make for A1 chassis VW's. It was a 38 inch long piece of 1.5" 0.09" wall DOM with brackets welded to the ends. They worked great and I wouldn't be surprised if there were still some out there in use 25 years later.
I reproduced a sway bar for a Lola this way about 5 years ago, come to think of it most of the SVRA vintage race cars I see in this area run tubing with arms welded on the ends. These guys aren't the types to consider heat treatment afterwards.
I'm aware of multiple Locosts (paging SkinnyG) using welded DOM (not even 4130) bars that have not failed with regular use. I'm convinced that, much like cutting springs, it can work surprisingly well if you actually understand what you're doing... In other words, keep the loads in check relative to the HAZ properties. Heck, even the Caterham CSR looks to use a welded front sway bar, although it may (or may not) be post-weld heat treated.
For a slightly different take, perhaps something like a (interference fit) sleeve for the arm, and a similar depth solid plug for the tube. Put a hole through the sleeve, tube, and plug that gets bolted through, then weld only the outboard end of the sleeve to the tube. Heck, maybe do two bolts and no welding.
An option that could have brackets clamped on rather than welded would be to make it from rec/sq bar stock. It'll be a bit heavier than equivalent tube, but could partially be made up for by using a proper 'spring' steel since you wouldn't be harming the heat treatment. Think more along the lines of air-cooled VW front torsion springs.
Finally, there is a perpetual supply of used Schroeder torsion bars with splined ends on ebay. The 7/8 spline rear bars are readily available in 22.5 and 26 inch lengths.
You are trying to build a T-bar sway bar? Something like this? This video unfortunately doesn't have animation.
Hey,
When I was designing/installing A-bars on my E-mod autocrosser, I used a company called (I think) Stock car products out of Richmond VA. I can't get their website to work at the moment.
I just checked the Coleman racing products website, and they show pictures of A-bar parts that look identical to the ones I used several years ago. See attachment:
https://www.colemanracing.com/Chassis-Suspension/Sway-Bars/SWAY-BAR-1-14-x-48-SPLINE
You might want to call them and see what they offer in terms of Bar diameters and lengths.
BTW: To corner at 3 g's lateral you need tons of aero assist and horsepower to generate those forces.
Good luck! Someone here will probably have a better and more comprehensive response.
Rog
In reply to VolvoHeretic :
The goal is basically this. Except there will be much schenanigans involving a second set of bellcranks to get my currently perpendicular belcranks to be able to actuate the 3rd element this way without re-doing the entire belcranks.
Found this. Cody Loveland with the Enviate Hypercar used a simple welded tube Tbar very similar to what I envision. So this seems like a good route.
Because your bellcranks are vertical... look at something like a 1987-1990 Formula 2000 Reynard used. Real simple system. Easy to adjust, and you can adjust from the cockpit with a simple cable mechanism.
I can't find any pictures in my files... I'll keep looking.
Look at this photo for ideas. Bars are just different size tubing. I carried about three differnt bars. One arm is adjustable. It is a blade (right side) that can be rotated 90 degrees from the cockpit. When adjusted "flat" it is soft, when adjusted "vertical" it is full "hard". Lots of variations in between. I might start a race full hard, as tires go away, soften it. The red cable is attached to an arm on the blade that can rotate it.
This is not my car, and it looks like a pretty small bar, maybe 1/2". I ran bigger diameter bars on a winged car.
In this picture it is adjusted about 1/2 way between soft and hard.
nocones said:
The LMP360 needs swaybars. The calculated roll stiffness is currently a paltry 2.5 degress / g which seems low, except at 2G's that consumes all of the available shock travel. It needs roll help.
I've done preliminary maths and determined I need the rollbars to contribute ~300-400 lb/in of spring coefficent to the wheels to get roll stiffness to around 1 deg / g which should keep the car fairly flat. The car is getting a 3rd element installed at the same time (I will post about all that math separately), but I am leaning towards a T-bar swaybar with the 3rd element attached. This would drive the swaybar MAX travel to match the 3rd element shock travel of 2.5". However Loading to this amount of twist would basically be impossible.
It looks like a 3/4" Bar 12" long, with a 10" T-arm would get me in the ballpark for torsional stiffness required. The stress during up to 3G cornering would remain bellow 50Ksi, so Fatigue life will be a thing but Yield should not occur if 4130 is used for the bar. 1020 would probably do the job, but If 3G cornering is a legitimate thing 1" tube may be required with longer arms.
Bar stiffness math came from here: https://calcdevice.com/bar-spring-id247.html
Youngs Modulus for 4130 is 29700000, 1020 29000000.
Poissons Ratio for 4130 is 0.29, 1020 I used 0.27.
Ok so that's the background of What I'm up to.
Has anyone fabricated a sway bar? I'm thinking 4130 Tube with plates Tig welded to the end.. But? Is this a recipie for disaster? I'm okay if I need to replace the Torsion bar every few years to prevent fatigue failure.
I also may look to use the OEM subaru 360 Torsion bars in a "traditional" U bar arangment. They have Hex keyed ends so it would be easy enough to fabricate around without welding.
I don't want to spend huge money, and my space constraints prevent me from using any OEM bars.
I have, I took a tube of 4130 to the local spring shop ( Think big trucks) and had him heat treat it for me.
I was slightly too stiff so instead of .125 wall I went, I think it was .094 wall and most tracks I used about 6" of the 10" arms. I used a sliding clamp arrangement so if I felt a shade more or less I didn't have to select a given hole.
All of my bars so far have been merely cold rolled mild steel. As long as it doesn't exceed its range of elasticity, it will work.
I have a former student now friend and mechanical engineer who fabricated a sway bar using 1" SCH40 pipe with welded arms for his race car, and still runs it now years later.
This stuff does work, and I'm cheap.
Nocones:
Since you are headed down the T-Bar path, I would definitely consider splined torsion members. This will allow you to tune the roll stiffness in a more controlled manner and more easily replaced spares if you run itno issues at track. Also consider adjustable drop link connection/torque arm positions for tuning. If you really want to get into the weeds, look for adjustable blade T-bars. Basically a thin beam on the torque arm axis that can be rotated into/out of plane with the 3rd spring motion to control the roll stiffness. I have had hands on all of the items mentioned above back when I was doing CART/IRL cars.
