I've been putting off installing the drivers side longitudinal and floor for over a year now. (see page 1 for longitudinal repairs and test fitting). The reason for the procrastination is that I've been afraid that the chassis may not be true. Roland's restoration book is adamant about NOT doing structural work on a rotisserie, so I decided to proceed with caution and check the chassis straightness before its too late!
I've been suspicious that something was amiss because when the casters are removed from my octagon rotisserie and the car is sitting flat it has a slight rock to it. At first I thought it must be the concrete floor, but it seemed to rock no matter where I moved it in my garage. Then I thought maybe the uprights on my rotisserie aren't equal length, but they are. This car didn't show any major wrinkles in the inner chassis to lead me to believe the car might be twisted. The car has been hit in the passenger nose and a previous owner had half of the nose replaced. Subsequently, I fixed some of the damage to the inner nose, battery box and diagonal brace. I would describe it as basic hammer and dolly repair (No hydraulic pulling or major heat needed.)
So, I started by measuring my garage floor flatness in 6 critical locations; Two front locations under the octagon uprights (sway bar attachments), two locations under the rear torsion bar tubes, and two locations under the rear octagon uprights. Tubular steel with both ends capped off where used to make a custom spirit level. I partially filled the tubes with automatic trans fluid and supported the tubes on machinist parallels at the measurement points of interest. Once the fluid stopped sloshing around, several minutes to be safe, I stuck my precision scale in the fluid and noted the level very carefully. Then I measured the far end. If the two measurements weren't the same then I shimmed it to make it level. Two tubes of custom length were used to get all six points reasonably close. I estimate that all six points are within a 1/16". This is the same method the Egyptians and Romans used to measure inclines or flatness. What does that have to do with 356’s???? I dunno, but I like the simplicity and accuracy. It beats a bubble level for sure. My garage floor needed a maximum of .100" shim. A far cry from the proper Cellete bench, but what is an amateur to do? Press on....
- Home made spirit level. Just a tube filled with ATF with windows at each end to measure fluid level with a metal scale.
- Shims required to level floor plus 1/8" spacers
A sharpened stud was threaded into the the left and right torsion cover bosses and tightened with a double nuts.
- Sharp measurement point
Once the car was supported on the floor and shims, I could measure how far off the suspension attachment points were. I measured just a tick under 3/16" or 5/32" vertical discrepancy.
I also compared the distances from front to rear suspension points using a trusty carpenter's square sliding on a square steel tube.
- Poor boy trammel bar
Porsche's workshop manual gives specific measurements and tolerances. Rather than making a pointy plug to fit tightly into the front torsion tube, I just touched the outer portion of the bearing carrier with the square. The factory tolerance is 2 mm or 5/64". My car had a variation of a tick less than 1/8". Not too bad considering my Rube Goldberg methods. The odd thing is that I would expect the passenger side to be shorter from the accident, but it was actually longer than the drivers side. Go figure.
It was fairly easy to twist the car back into shape. I clamped a long wooden beam to the rear octagon and rigged a floor jack to the front octagon. Then I rigged a vertical beam to the garage rafters to prevent the front octagon from rotating. I iteratively raised the jack on the front octagon so the rear beam was about 1" off the ground intitially. Then I stepped on it. Sure enough she groaned, squeeked and got better. Next time more height on the front jack lifted the rear beam 1.5” off and I repeated until all the twist was gone. Then I intentionally upset each front corner and measured the deflection. Once the deflections were equal, then that is as good as it gets.
- 4 x 6 beam attached to rear octagon. Jack at front octagon and vertical beam attached to roof!
Then I welded the longitudinal on while checking all dimensions periodically. I did rotate the car 45 deg to better access some of the bottom welds. Ron Roland recommends to begin welding at the longitudinal center and then work towards the front and rear closing panels alternating.
- Fully welded tude
- Tried to duplicate factory gas welds with a TIG where the edge of tude meets the heavy guage metal near torsion housing. Rosette welds will be ground down tomorrow.
There, now it’s done and I’m happy again.
I re-checked the front to rear measurements again and lo and behold, now it’s within factory tolerance. Hard to believe twisting would change that. Or maybe the heat induced by welding the longitudinal shrunk it into place. Also hard to believe. It could be my measurement error. The second time around I measured each side three times and kind of averaged my measurements. All were within a pencil width.
Now, lets back up a little a bit and make this long story even longer......I think the chassis may have warped when I welded the passenger side longitudinal several years ago. I did this on jack stands with the car fully loaded, ie engine trans, glass, interior, etc. Its very possible that I stressed the chassis unevenly when trying to access the bottom for welding. So, many thanks to Ron Roland for writing this great book with tons of sound advice. Indeed bad stuff can happen when doing structural metal work. He goes a step further and recommends using the doors to monitor chassis distortion. Good idea in my opinion, but I didn’t follow that advice. Mainly because my front fender edges are trashed and the doors need to be totally refitted anyway. I didn’t notice any difference in the way the doors fit after I did the passenger ‘tude on jack stands.
Also, I’m glad once again that I used the Emory octagon rack because it helped discover the problem even by accident. Anyhow, that’s enough blabbing for now. Thanks for reading.
