FWIW… have you tried triangular fillet pieces across all corners ?…

George.

When I made my welding trolley, opposite sides were clamped to the substantial steel plate, that would end up on top when finished. Sash clamps were then used to hold the two sides whilst the inter-connecting pieces were installed. This seemed to minimise the amount of distortion. At least all the wheels touch the floor, (if I find a level bit )

Pete

When you weld box section on one side only, it will pull out of shape. If you need to weld on one side only, it is best to clamp the box section to a steel beam with a packing piece under the middle of the box section so when its welded it will pull straight.

When I welded up the stand for my Tormach CNC mill I used 50x50x2. It was welded using oxy-acetylene. I don't recall any particular problems with distortion on the frame itself. Welding the sheet metal tanks was a different story. The fact that I've got bracing diagonals may have helped reduce distortion.

Irrepective of possible distortion adding some diagonals to your frame will effectively make the sides a truss and will considerable stiffen the frame. That means you can probably miss out some of the welds, further helping with distortion.

Andrew

Provide levelling feet. You can say they are there to account for an irregular floor. You won't have to admit to a warped weldment.

Found the pics I wanted,

Back in 2012 I decided to scrap a 2 meters between centers TOS lathe for a few reasons the main ones being it was taking a lot of space up, it had some wear and it was attracting the wrong sort of work [ 1 tonne armatures etc ] when it was working which wasn't often.

Once I had started on the project it soon became obvious what the "it will come in handy one day" was.

Virtually all this had been there that long I had no idea it was there so it got skipped.

Machine dragged out and loaded up for a trip to its new home.

The scrapyard.

Which left this.

So once the cupboards had been scrapped, racks moved,  we got to the relevant part for this post.

3 metre long bench out of 40 x 40 x 3 stick welded with a crappy buzzbox as I couldn't get any of the hard wired Migs into this shop.

And I finished up with this,

Laser cutter on one end, bolt rack on the other and space for two working small CNC's with the keyboards on slide out trays on top of the toolboxes. Various metal short end drawers at one end, wall painted and all new electrics plumbed in. Unfortunately it has got to come out in the grand down size and it won't go in the new shed in one piece so I reckon a quick chop with a 1mm slitting wheel where it changes from cupboards to open space where the tool boxes fit, move in two pieces and stitch back together again and a swift repaint to credit a brand new workshop

Edited By John Stevenson on 24/07/2017 21:13:14

So can I, I reckon Grizzly bear has done himself a favour if he can't

Just five minutes after J.S. had posted the pictures, only the top half inch was visible, the rest was green & yellow horizontal stripes, then a few minutes later just "'X's".

I've seen them today, nice one John.

Regards, Bear..

"… but 50*50*2 is roughly twice as flexible as 40*40*3"

Is this observation relating to flexibility in bending? Or some other loading type (e.g. torsion)? Or localised loading on the thinner 2mm plate?

I ask since resistance to bending is a function of the flexural rigidity EI. For the same material, the comparison is formed by the ratio of the second moments of area of the section – (bo^4 – bi^4)/12 for a square tube. For the ratio of SHS second moments of area for the dimensions quoted, this becomes (50^4 – 46^4)/(40^4 – 34^4) = 1.45. It appears that the 50x50x2 will be 45% stiffer in bending than the 40x40x3 section.