After reading a thread a few weeks ago on building a DIY sway bar I figured it would be a good engineering / fabrication challenge. Pics found HERE
I had previously thought that sway bars were some magical form of controlling body roll and were made out of pure on unobtanium. After a few minutes searching around on google I found that most sway bars were made out of mild or chromoly steel. I pulled out my old Strength of Materials book and notebook and did some basic calculations - angle of twist, torsional shear stress, and moment of inertia. It appears that the modulus of rigidity and yield shear stress of most steels are fairly similar. I went out to the car (the POS cavalier) and began some basic measurements. The next step was to make a couple Solidworks models of my designs. The moment of inertia calculations showed that the bar stock ends would be more resistant to bending than the bar with rod ends. The Solidworks model showed that the bar end model also had less mass. The final straw for the bar end design was after I calculated the bending force of 0.875in Cold Rolled Steel rod. I tested bending a piece of 9/16 rod I had laying around with a 10ft pipe, and determined that I was using ~100lbs of force. The 7/8in bar would require 53% more force to bend, so that was given up. I went down to the local steel supply and after looking through their piles of “scrap” I had to place an order for the 7/8in cold rolled steel rod. My only option was to purchase 12 feet of rod, whether I wanted to take all 12 feet or not… $60 for some steel rod, atleast I found some 1 x 0.5 in bar in the scrap. For $74 I had enough material for one or two more sway bars.
Costs: 12 ft – 0.875in cold rolled steel rod - $63 10 ft – 1 x 0.5in steel bar stock - $11 Energy Suspension frame bushings # 9.5158 - $22 2 – grade 12.9 bolts, nuts, washers - $05
After looking at some pictures of the various rear sway bar options for the cavalier, all the designs were flawed. They either mounted the bar to the bottom of the torsion beam, which reduced ground clearance by 2+ inches, or they mounted the bar on the back of the torsion beam which caused it to rub on the e-brake cable and ABS wheel sensor wires. I decided to fix this by welding on two 0.25” steel plates to the back of the beam. They overhung the beam by 5/8”. I would be able to reduce the number of fasteners required by bending the lower bushing ear around the plate, and the exhaust was still the lowest part of the car.
The next step was to cut the center rod to an approximate length and install the frame bushings. The plates were marked and drilled, so I could not install the center section to mock up the ends.
The bar stock ends were cut excessively long at 16 inches, even though the Solidworks model said 12in was correct. I then marked the directions and locations on the bar where it was to be bent. I chucked on end into the bench vise at the first mark for the bend. I then clamped the remainder of the 10ft steel bar stock 90* to the piece being bent at the second mark for the bend (usually 1 to 1.25in). All 155lbs of me found it quite easy to bend the bar stock, and the ends took less time to make than any other part of the bar. Once I had one end fitting nicely I then marked the bend locations on the other bar and went to town.
Once both bar ends were done, the center rod was trimmed and the ends were chamfered with a flap disk to aid welding penetration. The center was bolted up again and the ends were tack welded; then the bar was taken off and fully welded.
All I had to do next was to cut the two plates for the ends to attach to the lower shock bolts and I could take it out for a spin! While all the calculations said that theoretically it would work, we all know that theory and practice don’t always mirror each other. After a short prayer that all my hours of fabrication wouldn’t end in a spectacular failure, I started up the turd and took it out for a spin. We figured that the bar would work, assuming my welds between the bar ends and the center rod held. I wanted to make certain that they didn’t fail so I ran another pass.
SUCCESS! I haven’t done any skidpad tests, but the bar seems to be holding up fine. While I won’t get the cavy confused with my RX-7, the body roll is significantly reduced.
