For those looking to add to their welding knowledge base - and who aren't following my RacerBenz build on Grassroots Motorsports - here's a video I did today. I'm lifting it from my post on Grassroots, where welding of the roll cage started today.
One point here is that it's possible to use welding distortion to your advantage! Another point is that, holy jeebus, welding REALLY moves things! Plan carefully!
And I threw in a weld porn shot of one of the cage welds. Martin is something else...
Enjoy:
Ahead of welding, angles were checked on the main hoop and A-pillar legs. Why? You've all heard of the distortion welding causes so it's a good idea to plan your welding order. This allows you to better control what moves more easily and what doesn't. Turns out the left A-pillar was tipped in a bit more than planned (like 1/2 a degree) and the right side wasn't tipped in enough, by about the same amount. It's not something anyone would likely have ever noticed but why not do all that can be done to make things as close to perfect as possible? (BTW, yes, the bases of all the legs are tacked to the rocker boxes.)
So... we braced the right side A-pillar against the chassis and wedged the not-yet-final-length dash bar in between the A-pillars, "encouraging" the left side out at the top. That side, upper area was welded. Then we removed the dash bar and went after the top area of the right side pillar. With no bracing, when the weld cools, the welding distortion (shrinkage) "should" pull the pillar in, improving the angle.
Did it work? Take a look at this video and watch it happen as the weld cools! I started the video just after the welding stopped. Look real quick when the video first starts and you'll see the angle value starts at 89.4 degrees.
We checked again at the end of the day and it settled at 89.8 degrees. Nearly 1/2 a degree. The good news is it's now the same angle as the other side. The plan worked!
https://www.youtube.com/watch?v=bpy3HT-9hnw