The Stovebolt Site Forums The Shop Area Paint and Body Shop Frame alignment

Just sandblasted and powdercoated a donor 3100 frame. It is from a straight truck and it looked ok, no cracks no visible bends.
Now checking measurements on the bare frame, it seems to have a 1" difference between the rear diagonal measurements specified in the book.
Can this much difference be fixed by rigging a come-along on the long diagonal, and a jack on the short diagonal?

Now you get the fun job of figuring out where it's out of square. You need to figure that out before you can figure out where to push/pull on the frame to straighten it out.

Thanks. There is a 1" measurement difference is in the diagonals between the rear spring mounts.
The diagonals of the whole frame are also off by 1"

Been here on my '46 project too. After talking to a bunch of experienced street rodders I decided my best option was to find a body shop with a frame machine. They used the info in the shop manual to pull it back to the specs. It's been some years ago, but as I recall the cost was less than $500.

Thanks - Since the frame is sitting bare in the garage, it seems worth a shot to try pulling across the long diagonal with a chain hoist/come-along. Or (simultaneously?) pushing across the short diagonal with a jack? Anybody have any luck with this? Suggested anchor points or hardware? Downside or potential damage to watch for?

Also looking into local body and frame shops as plan B.

Well if both frame rails are Exactly the same length, your idea may work. If they are not exactly the same length, you will have to factor that in. Only has to move 1/2", each rail....theoretically. I would put a rack load on it, and using wood for protection, hit it hard all around with a BFH to "vibrate" the rivets. Especially bang the ends of frame rails toward where you want them to "move". Keep very accurate measurements from start to while banging/tightening. Probably have some spring back so go past 1" a little. To DG's comment, keep checking level/straightness of members carefully all along the process. Maybe heating the frame will help, not too hot, just 200 deg MAX...or maybe just heat cross members? Heating too long or too much will affect temper of frame, not good. I don't know if it's "bent" somewhere, I am assuming some of it was as delivered and some movement over time. If it was hit or crashed, the process may be only possible by a pro. I assume you don't have any springs or axles on it. Did you measure before removing all that and after? Are you sure 1" is all that bad if body was mounted OK, etc.? Don't rush into this until you see other similar frames are much better and you trust your measuring.

Maybe run some string lines from each end about 1/2" off the corner to help find any twist and give you a known straight line for visual reference as you bump.

That's a good point. It would not be beyond prudent to string the rails and ends and use a carpenter's square to check every cross member for 90 deg, so that you may discover something about the geometry of the "rack" before a plan is established.

Thanks - helpful ideas. Taking time on this.
The rails measure and look symmetrical - a few small dents, but no visible curves.
Checking with a carpenter's square is great idea - it shows my rear cross-member is out of square with the rear ends of both rails; consistent with the short and long measurements of the "diamond".
The front cross-members arent supposed to be square, since the rails converge. Measuring those angles would take some head-scratching, so will stick with measuring diagonal distances.

Appreciate the caveat about checking for potato chip twist. Doesnt seem to be any yet.

Could the frame joints just be racked at the rivets? If so, it seems that pulling in the horizontal plane might pop them back in line without bending anything else?

I would be interested in knowing what the tolerances were back in the day. Maybe an inch is just about right? I haven't heard any info on this issue. Where to measure, what the actual tolerances are... etc.

Good work, good thinking. Lets hope that it will "swivel" back. measure everywhere, record and try it a little to see if it starts to move. It will be a semi lucky, miracle. But worth a calculated try for sure. I think you have the same basic thoughts we do. You are the guinea pig.

"the book" would be the assembly manual, which has frame checking details

Bill

The TF Assy Man I have has no tolerances for frame. The body gap tolerances I have seen are generous but those include moving parts and adjustable gaps like doors. I suspect frames were in a jig while riveted, which renders dead nuts dimensions and geometry.
The only reason I asked if 1" was acceptable is that if the body fits, and it does not dog track, maybe it's not actually visible? probably wrong on that thought.

Maybe this would help
At least it gives some point ideas to measure from.

I guess I would be more concerned about doing more harm than good on a riveted frame. The integrity of a rivet is that it completely fills the hole very tightly. Goofing around with it may loosen things up and cause other issues? Just a thought.

manual CASO linked says 3/16" for the items listed

Bill

That's +/- 3/32. Tight Tol if you ask me. But like I said, riveted on a jig, piece of cake.
I believe that W46 is going to take it slow and will know to stop before any damage is done. Probably pretty hard to damage with a come-a-long. Just make sure the frame "tails" don't start to bend inward at a point after the last rivet.