To my other TR6 pages

September 20, 2016

Hanging Panels/Door Gaps

[Click the pics for a larger view]

So after finishing the rust repair on the front and rear wings, I set out to get the door gaps right before doing final body work.

After blasting all the remaining paint off the wings and doors, I primed them with an epoxy primer.  This not only protected them from any further rusting, but it made them a lot more pleasant to handle.  Unlike the wings, the doors had virtually no rust.

I hung the doors onthe rebuilt hinges, and installed the bazillion or so "spire nuts" for the wings.




With all the new and fabricated pieces and parts, some of the panel mounting holes didn't align well at first.




The door gaps on the first fit-up were not horrible, but needed some work.




The rear body mounts were hanging about 5/8" above the frame, and increasing this gap would raise the rear of the tub, and close the rear door gap. The 1-1/2" spacer was too much, but 1-1.25" seemed about right.  




Unfortunately, this is a familiar situation for TR6 rebuilders.  Now, I had checked the so-called "kick-up" of the rear frame rails before I began work on the frame, and it was a smidge over 4 inches.  I believe the correct value is about 4.5 inches, but in addition to whatever tolerence the factory allowed itself, stresses on the frame over the years tend to cause the rail ends to sag, especially if it is weakened by rust, so many if not most frames will have less kick-up than factory original.

Another thing that can cause the frame rail ends to sag is welding on the bottom surface of the frame.  Welding tends to shrink metal, and shrinking the bottom face of the frame, where most of the welding was, will tend to bow it to some extent.  I tried to minimize this by controlling the heat as much as possible, but I assume that the extra sag was from my welding operations.

Spacers work perfectly well to compensate for sag and lift the tub back up to the level it needs to be to get good door gaps, but there are a couple of other problems with having rear frame rails hanging too low.  First, there is a bumper bracket that has to mate with both the frame and a boss on the rear valance panel of the tub.  If the frame rails are too low, the bracket can't be installed properly.  Second, rails hanging too low will be obvious when looking at the back of the car.  I believe that on a new car, the frame rails hung just slightly below the bottom edge of the rear valance, but hanging an extra inch lower, they would be obviously out of place.  Further, since the muffler and tail pipes are hung from the frame, drooping frame rails will cause the tail pipes to hang lower under the valance and bumper than normal.

One recommended fix for this condition is to reverse the frame bow by cutting the bottom and side surfaces of the frame rails in the cruciate plate area, raising rear of the frame to the correct position, and rewelding.  Though I acknowledge that this may be an absolutely correct way to mitigate the problem of drooping rear frame rails, I didn't really savor the idea.

I expect that the solution I finally applied might be controversial.  That's fine with me, because nothing I report here is a recommendation for action.  Here is my logic:

The other main component attached to the rear frame rails that might be affected by sagging is the rear suspension.  If we assume that all of the droop of the rails was due to welding at the narrowest part of the frame, we can think of that as a pivot point for the sag.  A little trig suggests that the sag at the rear suspension attachment would be about one fourth of the value at the rail ends, since the rail ends are about four times further away from the pivot.  Since my sag appeared to be about an inch, this puts the sag at the rear suspension at about 1/4".  This is probably absolute worst case in my situation, since I also had welded repairs behind the rear suspension which would have contrubuted to the sag.

Another data point is the body mounting points at the top of the rear suspension bridge.  With the nominal compliment of mounting pads and shims in place, there was only about a 1/8" gap remaining.  This suggests that the actual sag at the rear suspension is likely closer to 1/8".  This, I believe, is getting close to factory tolerence.

So maybe it's rationalization and maybe it isn't, but I decided not to worry about how the sag was affecting the rear suspension.  But I did have to fix the bumper mount situation.

First, removed the silencer to get some room to work.




Then measured the height of the rail tails to get the baseline.  Contrary to what the pic shows, I made the measurement with the rear of the car on jack stands.




I then cut the bottom and sides of the frame rails just where they straighten out to be parallel with the centerline of the car.




Then jacked up the extreme ends of the rails by 1-1/8 inches.




Inserted a filler piece of 14 gauge steel, welded 'er up, and patched the paint.




Back on its wheels, the frame rails are barely visible, and the exhaust pipes hug the valance.




Put in the nominal spacers on the rear mounts, and the gaps came out darn close to parallel on both sides.  




Though the gaps were pretty even from top to bottom, they were too wide.  It looked worse in person than in the pic above.  There is only so much you can do by shifting fenders, since they have to line up with their respective valances.  Even moving the fenders in towards the doors as much as I dared, the gaps remained too wide.

After trying everything I could think of with no relief, and coming up empty on my Google search for door stretchers, I finally fell back to an old body trick to reduce gaps--welding a 1/8" rod to the edge of one of the panels.  I chose to do this to the doors.

After grinding away the primer from both sides, I shaped a piece of rod to the right curve for the top of the door.




Tacked the rod in, then added rod to the rest of the door edge.




Then a test fit to see if it was going to work.




It helped a lot, but there was still a wide gap at the very top of the rear of the door.  I welded a shorter piece of rod there.  The lower part of the rod would later be tapered down to nothing.




It looked like this would do the trick, so I welded up both edges of the door.




But just tacked the back side to try to limit the amount of heat.




Then ground the outside smooth.




The other side went through its own similar process.  The gaps can be fine tuned later.  They might be a hair tight in places, but there is now plenty of meat on the edge to work with without risking breaking through the skin.




One other thing to address is the "dive" often found on panel edges.  Dive is a deformation at the edge of a panel where it is crimped back on itself or onto some kind of support.  The dive was present on the factory panels, but the welding certainly didn't help it any.  




The dive is a subtle thing, but it can interrupt an otherwise razor straight reflection along the side of a car.  A little hammer and dolly work around all  the edges of the doors and fenders  brought the lips back into line.




Looking pretty decent from this angle.




But there is still one fly in the ointment.  Though the gaps look promising, the passenger side door's lower rear corner was sticking out by about 1/4".




I messed with this a lot--shifting hinges, trying to twist the door--all to no avail.  One thing I did notice is that if I leaned on the top of the B post, pushing it in slightly, the gap at the lower corner got smaller.  So it appeared that the B post was leaning a little outward.  I pushed as hard as I could, but the B post would just spring back to its wayward position.  I needed some mechanical muscle.

I fashioned this plate out of scrap.  The holes were drilled to match the soft top frame mounting holes at the top of the B post.  The bent bracket was bolted to the seat belt anchor point at the base of the opposite B post.




I attached the plate and bracket and hooked up a come-a-long between them.




Before doing any coaxing, I noted the distance between the inside surfaces of the B posts as a baseline.




This is all the movement it took--less than 1/4".  Of course, I had to pull the post nearly double that to have it spring back to this point.




I re-primed the panels to hide any evidence that any work had been done, and hung them back on the car.  They seem to go back on a little different every time, but I'm confident that in the end, I'll be able to get them to my liking.




Now, on to the actual body work on the panels--fixing dings and dents, and the distortion from all the patch welding.

Comments to Ed at elhollin1@yahoo.com

To my other TR6 pages