hammer type for flame straitening?

[coppercone2]

The process where you heat up a spot of metal red hot, then hit it with a hammer, is not documented well with regards to what the profile of the hammer should be.

In animations I see that they kinda show the hot spot to just be a elevated hill that is flattened with a flat hammer.

Is this correct? should the hammer maybe have a slight round to crater it a bit?

I am hoping someone just knows how to do this right.

[Stray Electron]

  That is (or used t be) a well known process in automobile body work and in some other forms of metal working. But striking metal (hot or cold) tends to expand it so it will increase the amount of curvature instead of straightening it.  To shrink the metal and to reduce the curve, you heat it to near red hot and then quench it with cold water.

  There is a very old shipyard worker in Japan that I have seen at work and he makes compound curves out of 6" thick steel that way for use in ships using nothing but an oxy-actelyne torch, cold water and a very big hammer and he is considered one of their living National Treasures.

  I have used that technique for small repairs in automobile sheet metal and I made my own hammer by taking a good quality 12 ounce (IIRC) ball peen hammer and literally tying a 36+ inch long string to the back of it and tying the other end to the tail stock on my lathe. Then I put a sanding disk in the lathe chuck and run it at low speed while swinging the hammer back and forth with the hammer face against the sanding disk. I used 80 grit and then 200 grit and then 600 grit sand paper on the sanding disk and gradually smoothed the face of the hammer to a very smooth, convex surface with a 36 inch radius. 

[coppercone2]

well the technique of hammer flattening is well documented. It works because I did it on a less warped panel.

https://www.boconline.co.uk/en/images/Fundamentals-of-Flame-Straightening_tcm410-113398.pdf

But I am not doing it optimally, because I don't know what hammer to use. Page 20


The water method sounds very interesting and less loud

The idea is it rises like a hill, then you flatten it, then it contracts and solidifies as a smaller hill (convex bump) that is on both sides of the metal but localized, because when you heat it, its a concave on the bottom.

[coppercone2]

around 8:45 on this video


I thought I saw a video of them doing a bridge girder with a hammer and torch but im not sure. Maybe you are right about the hammer being not applicable to thicker metal. Actually it looks like they don't hammer the bridge girder. But they do seem to hammer thin metal.

And I used this method without a brace/jig welded around it on the top part of the chassis, it works to flatten it out. That was 2 years ago and I used a copper rounded hammer, but it was slow/tedious. I just worked on it on top of a metal plate.

[coppercone2]

for your water method, how much heat do you need? with my transformer I can reliably make red hot dime sized sections. So just make a hot spot and pour water on it? Inwards spiral?

The thing warped like crazy when I welded angle irons on it to make a flange for the top part to screw into. Actually bent the damn angle irons I thought they would keep it strait. I bet its a lost cause but I thought that about alot of projects. Need to get rid of like 1/2 inch bend in a 2 foot long by 14 inch section