Lessons learnt, you can't shrink with a MAP torch (need more focused heat of oxy/acetylene which I don't have), and you can't force one piece of metal to behave by doing the opposite to another bit of metal. Seem to recall I learnt that on the sub frame - and then forgot.
Days later, with all metal and my self in a relaxed state, time to finally tack and weld.
All looks good, run the ruler over it from top to bot...... muther fu@ker
So looks like the stretch went further than I thought and was compounded by the welding which has also stretched things a bit, including in the middle of the new repair section which is visibly raised in the middle.
Really pissed off now - walked away, vowing to fixed something else and come back to it in the future. But next morning went out and this is the first thing I see, mocking me. So with a calm head, and a bit of patience, I started to fix this by removing the stretched metal. Figured the best way was to start shrinking from where I started welding in the top LHC and move around, alternating top and bottom weld paths.
This was all achieved with the shrinking disk, slapper and dolly. Didn't take many pics as I was in a bit of a shrinking zen space. Did get this one which shows the one on the left starting to close up a bit.
Four hours of shrinking, weld in the inner repair, grind, repair a few pin holes and this is what I ended up with.
All waves gone, no oil canning, and any ripples are visually under 1mm. Doesn't look great compared to attempt 1 but it's straight now.
Lessons learnt part 2, must try to read the panel better. There was a slight stretch in original that could have been shrunk out before attempt one was welded in. Shrinking the edge on the repair to straighten it should have also been done at that point, but didn't yet have the shrinker, and doing it manually must have been put in the too hard basket. Then repair bottom edge should have been shrunk to match original. Sounds easy in hindsight.
(Not so rusty) Ol' Rusty - FB/EK Sedan