[ROVERNET - UK] Alum/bonnet straightening
gofanu at usachoice.net
Sat Sep 2 17:19:34 BST 2006
Got a ver kind response from Mark Coorparoo off list re this, including
a question on "aluminum brazing", copy follows, including my response to
Paul's comments about "shrinking hammers" that got lost earlier. Might
be of interest.
Thanks for the kind words. I was just re-reading it and realized that I
have in essence managed to get everything you need to know for repairing
crash damage in there. All else is just a matter of analysis of damage
sequence, and application of those principles. The real trick is
learning to "read" stress in a panel. All of this applies equally to
steel panels too; the only difference is that it is easier to use heat
on steel, as it is a bit more forgiving of excess. I have straightened a
number of severely damaged vehicles with NO parts, and an absolute
minimum of cut/weld. On the Tinman site, you can see a truly stunning
repair of a badly crashed Aston Martin by Kent.
"Aluminum brazing" is not very useful, since it's not as strong as
welding, gets hot enough to anneal the metal, and is much harder or
impossible to coldwork after, which is necessary to restore the
required characteristics of the alloy. You can do structural welding by
MIG or TIG, but for panel work you need to gas weld. This requires the
correct flux, which is extremely difficult to obtain due to govt
regulation of "hazardous material", and correct goggles. Kent White, aka
Tinman, has made the flux available again after many years, and has
developed a truly superior goggle glass, now used by neerly all aluminum
panel workers. If you are serious, study the Tinman site; Ron Fournier's
books are superb as well.
I wrote some more on the evils of "shrinking hammers" in response to
Paul's comments, which got lost on its way to Rovernet, here it is:
Agree with comments about going gently and working it out as you said.
By shrinking hammer I meant one with a grid pattern on it, possibly
"teeth" as you define. For thin Al yes it would be too severe. The
shrink is only the Last step, to help fix the ding.
I have seen the Al shrink done so I know it works.
Yes, that's the booger. It's nasty for metal, induces vicious stress
points, and makes it impossible to metal finish. Yes, it works, but it
is totally unnecessary, and is only a fix for the incompetent and
impatient; the only useful function is to convey cash to the purveyor.
People have been "raising" gold, silver, copper, steel, or aluminum for
thousands of years without such, and you can be sure that nobody was
filing away vast chunks of precious metal - it's all precious when you
hammer sheet from ingots by hand - because they were too lazy to learn
the technique of shrinking, or "gathering" as it is usually called.
Filing is the only way to metal finish an unpainted object with hard
edged damage, and teeth marks are not acceptable. With no such marks,
planishing with a polished hammer and dolly will give any level of
smoothness you like, while removing NO metal. While shrinking can be
done with a steel hammer on a steel dolly, it's easier and quicker to
have one or the other be wood, preferably end-grain hardwood. The end
grain grabs the metal just like the teeth on the useless tool, with no
damage. The trick is to hammer bumps down in such a way that a portion
of the movement is in the (local) plane of the sheet, which compresses
and thickens it, reducing the area.
If the ding we speak of is in the leading edge, as seems the case, then
it is likely that not only is no shrinking required, but some may have
already occurred in the formation of the ding, If the ding is the result
of a pointy thing hitting it, then there may be some stretching, but
that can be removed and more "shrink" can easily be introduced in the
preliminary straightening. That leaves it a bit shrunk for final
finishing, which is the ideal case. A key rule is that when using a
steel dolly and hammer, a "thunk" is shrinking metal, and a "tink" means
the hammer has hit the dolly with the metal squashed between, which
More information about the rovernet