Torsion Bar Autopsy (Graphic)

dtw · 06-04-2005, 09:56 PM

As some of you already know, I picked up an '86 Carrera a few weeks back. The car was sold with both rear torsion bars broken. The car had been parked for several years in this condition, so the shocks had long since given up and the rear of the car was squatted down, basically resting on the tires.

According to the PO, the car was involved in a sideswipe with a guardrail around '99 or '00. Both wheels on the driver side apparently received some crash forces in the same plane as the axles/spindles. Here's the strange bit - the rear bars didn't break immediately. I'm not sure how long they survived - it may have been 2 or 3 years. The PO was fortunate in that the first bar broke while he was navigating a parking lot. The second one broke on the lift after the car was taken to the shop. PO said that the shop crew were in another room eating lunch when the bar broke with a loud BANG, scaring the crap out of all of them.

My two theories on why the bars broke:
1. Driver rear suspension pushed inward during collison, causing a bind in the torsion bar, creating a stress point and eventual failure. This doesn't explain why the passenger side also broke.

2. Suspension was compressed way beyond normal travel in the wreck. Torsion bars exceeded their elastic limits, causing overstressed areas in the bars and eventual failure. This would explain why both sides broke. Also makes sense considering the 2nd bar let go in the shop on the lift, with suspension extended beyond normal travel, accelerating the stress failure.

As you can see in the photo, the bars broke at the inner splined ends (thus spring plate bushing failure and subsequent torsion bar corrosion is ruled out - the bushings are in decent shape). You see two inner splined bar ends, but only one mating torsion bar. The other bar was thrown out before I got the car. I had to remove the second one and remove both inner splined ends, which were still anchored in the center of the torsion bar tube. This was NOT easy or pleasant!

Just posting this for the cool/rarity factor, as it is not every day that you see two broken rear torsion bars. Also wondering if anyone has seen anything similar, and would love to hear other theories on what caused the failures.

Imagine if these had let go at speed?!?!??! Hey Noah - I think it is time for some new Sander bars in your ride, man...

lateapex911 · 06-04-2005, 10:03 PM

Dave...funny you should post this...I am just swapping rear bars now, and upon reading the manual on how to R&R them I came across the line "If the bar is broken remove the other bar and drive the broken end out with a suitable tool", to which I thought...what happens if BOTH are broken???

So, what was your technique??

patkeefe · 06-05-2005, 07:26 AM

Check the torsion tube for distortion. If it got bent up in the crash, it would stress the bars in a non-axial manner, which could lead to failure you described. Also, check the wheel wells to see if there are any distinguising marks which would indicate excessive vertical compression of the suspension occurence during crash. I would think that the outer ends will take some compression due to the rubber bushings, but not the inner ends...no give in there except to pry the splined carrier out of alignment.

The fragments look like fractures from axial compression...kind of like some of my bones. Evulsion fractures.

Pretty nasty looking bars.

trak ratt · 06-05-2005, 07:33 AM

Not an engineer but the discolorations looks like an aged stress crack. This I think would support your suspension compression theory. I’ve seen fronts that have similar discoloration at failure point. These usually resulted from bush failure, chipped paint then tiny rust caused hole and finally failure. I read somewhere that once the bars surface is compromised they deteriorate rapidly. Also glade they didn’t fail at speed!

David · David's Garage

The pictures aren't clear enough to do a thorough examination, but it looks like a classic torsional failure mode. So the cut in the coating causing a corrision induced stress riser sounds plausable. I just installed new Elephant racing bars and even though they're zinc plated, I put a liberal amount of grease on them to prevent corrosion.

dtw · 11:39 AM

Jake,
I went to the local home improvement big box and got some long wooden stakes and a 4lb mallet. Pounded the stakes into the tunnel enough to get the stubs offset a little bit in the tunnel. Then I went in with the broken torsion bar and pounded the stubs out from one side. Sounds a lot easier than it actually was, it took an hour to get them out.

Discoloration: These bars broke a few years ago, and so I'm not sure if the discoloration happened before or after the failure. Wish I knew more!

I just got the car running again yesterday, and after gingerly enjoying the car on backroads for a few hours, I dropped it off at my mechanic for compression and leakdown on the engine, as well as a thorough inspection of the rear suspension members.