SWAYBAR DESIGN AND FABRICATION
Designing and building a anti-sway bar for "SONEAT"
This will be a rear sway bar and it will be adjustable.
|
SONEAT (Dad's 1968 SAAB Sonett) has proven to be a very competitive car in vintage racing. It is a little bit underpowered compared to the top cars in its class but it has great stability and excellent grip in the corners. Braking is quite good as well. However, I know that there is still more to be gained in the handling and braking departments even without installing a limited slip differential.
The car currently rides on stock springs with 2 coils cut out. This lowers the car and increases the spring rate slightly. It still leans quite a lot despite these modifications. The amount of body lean forces me to run as much negative camber as I can possibly dial in. Despite running all the negative camber possible and using the rally-style left-foot-braking cornering technique, I still experience terminal understeer.
(Front) Tire wear is suffering. I am wearing out the inside half of the tire prematurely. I expect this is occuring under braking, not in cornering, though this has yet to be confirmed.
My plan is to design and fabricate an adjustable anti-swaybar for the REAR of the Sonett. With this, I can reduce the understeer and the body lean. I should have a more balanced car and may be able to reduce the negative camber at the front wheels. Hopefully, this will result in using more of the front tire under braking while maintaining high cornering forces.
I am aware that this will reduce the stability of the Sonett. I will have to adjust my driving style and use a bit more finesse in transitional handling situations.
09-24-2007:
Initial measurements suggest that the bar will be 30" wide. The length of the arms could be adjustable anywhere from about 6" up to 14".
3/4" (0.75") Anti Swaybar rates
Mike is helping me design the bar and sent me the following analysis.
Stefan here is some quick data. First is just to show you how much moving the lever length effect the spring rate, shown are 0.065 wall and .083 wall which will show you how much stiffer a thicker wall is. Also shown is solid for reference.
Here you will see what the different size bars you are thinking will give you. I have a meeting to run to, I will be back.
- Length of bar: 30in
- Lever length: 6in
- Density: 0.283 lb/in^3
| OD | Wall | ID | Spring rate (lb/in) | Bar wt |
| .075 | 0.0375 | 0 | 327.0 | 5.25 |
| .075 | 0.065 | 0.62 | 174.3 | 4.11 |
| .075 | 0.083 | 0.584 | 206.8 | 4.24 |
| .075 | 0.12 | 0.51 | 257.0 | 4.48 |
| .075 | 0.188 | 0.374 | 306.7 | 4.84 |
- Length of bar: 30in
- Lever length: 14in
- Density: 0.283 lb/in^3
| OD | Wall | ID | Spring rate (lb/in) | Bar wt |
| .075 | 0.0375 | 0 | 60.1 | 7.25 |
| .075 | 0.065 | 0.62 | 32.0 | 5.67 |
| .075 | 0.083 | 0.584 | 38.0 | 5.85 |
| .075 | 0.12 | 0.51 | 47.2 | 6.18 |
| .075 | 0.188 | 0.374 | 56.3 | 6.68 |
5/8" (0.625") Anti Swaybar rates
- Length of bar: 30in
- Lever length: 6in
- Density: 0.283 lb/in^3
| OD | Wall | ID | Spring rate (lb/in) | Bar wt |
| .0625 | 0.313 | 0 | 157.7 | 3.65 |
As you may have noticed, the shorter the lever the stiffer it gets and it gets there as squared function. It would be wise to shoot for a 170 lb/in in the middle of adjustment so we can go from there either way. I agree on some way to disconnect it quicly incase of rain, snow or sleet :)
That being said, the 3/4 bar with a .188 wall would work pretty good, 170 lb/in puts us at an 8" lever, we can bring the lever down to 6" to almost double the spring rate (306 lbs/in) or lengthen it to 11" to go under 100 lbs/in.
09-26-2007:
Stefan to Mike:
Question...
I was thinking about replacing the aluminum (sliding) block that my friend's design uses with a spherical rod end. I figure that will simplify the design because I could replace the bolt, the slider block, a bunch of washers and the rubber bushing inside the slider block... All with a rod end.
My question is... If I order a rod end with a 3/4" ID on the bearing... Will it slide over the 3/4" tube? Basically, is there clearance built in during manufacture? Or will I need to turn down the tube a couple thousandths to get sliding clearance?
I am looking at part 6072K28 on www.mcmastercarr.com. Is that OK do you think?
Stefan to Mike:
The axle measured 1.885".
As far as I can tell, shaft collars are available in 1/16th increments. The closest one that isn't too big is 1 7/8" = 1.875". That's pretty close and probably not worth re machining to make .01 larger. It isn't one of the specifically "weldable" ones, but I see no reason why we couldn't weld to it.
Also, from what I can find... Maximum axial load on a spherical rod end is somewhere between 10% and 15% of the maximum radial load (this is info gleaned from the Aurora bearing website FAQ). The rod end I picked on McMasterCarr (part # 6072K28) has an ultimate radial load of 11,550 lbs. So maximum axial load would then be between 1732.5 lbs and 1155 lbs. That sounds like plenty for our application.
Ordered the bending die from JD Squared.
- Part#: 100182
- Description: M3B Round Tube Die Set - 3/4" (2.25-180)
- Degree of bend: 180
- Minimum wall thickness: .058"
- Price: $135
10-04-2007:
I took a selection of photos of the underside of the Sonett. These are intended to aid in the planning and design stage. 
With what I see here in mind, I think maybe the way to go will be to mount the mounting blocks on the side of the trunk floor, rather than underneath.
-STEFAN Vapaa
|