Welding an HSS tube inside of an HSS tube

[UnneutralAxis]

Strange situation here. I have a contractor that is supplying some structural items for an industrial application. They ran this scenario by me.

Don't bother with the whys. I got caught up in the whys for about an hour and I'm not sure why it HAS to be this way yet either.

A 18' long HSS10x8x0.5 has 2" diameter holes @ 2' o.c. on the 10" deep faces. They want to insert another smaller HSS tube into the 10x8 and weld them together at the 2" hole locations. They originally sent me a sketch showing an HSS7x4 inside the larger tube (no theoretical clearance at all). I told them they would be extrememly lucky if the small tube would fit inside. Now they want to get a smaller interior tube and weld plates to it to give approx. 1/4" of play between the two faces. This way they can insert the smaller tube into the HSS10x8 and not have too large a gap for welding.

I am thinking if you add the tolerances of each tube together, as well as accounting for straightness, there would need to be a pretty good gap to ensure the fit of an 18' long tube inside another. Possibly too much of a gap to weld acceptably for the entire length? What are your thoughts?

 

[connectegr]

The maximum gap allowed by AWS would be 3/16" if the intent is to fillet weld or plug weld the 2" holes.

Why not fabricate a heavier tube from plate, if section modulus of available tubs are insufficient for a limited depth? Sorry that is another "why" question. But, plate is cheaper than HSS, grade 50 would be stronger, and the corners could be PJP welded.

http://www.FerrellEngineering.com

 

[UnneutralAxis]

connectegr,

You are right with me. I suggested several things like this, your suggestion being the first.

This is going on a piece of equipment in a very large factory. The original one was similar to this I'm told, so they want to recreate it and management will have it no other way. I have not observed the equipment, so I have no idea why it is like this to begin with or any idea of loading, etc.

Here is what I just came up with if it has to be this way. Use a smaller tube (6x4x3/8 or what ever size it has to be) and possibly have several thicknesses of bar stock on hand. Scoot the smaller HSS all the way flush to one side of the outer HSS and then slide long pieces of bar stock in the gap on the other side until you find the greatest thickness you can push in there. Then take the whole thing out and weld that bar stock thickness to the HSS. Then reinsert and weld at the 2” holes.

Of course, this all assumes that the numbers check out on this.

 

[connectegr]

In addition to mill tolerances, an 18' section will probably not be straight. There is also a raised welded seam on at least one interior face. The faces may also be concave slightly. As you know we don't work with Swiss watch components. You may be lucky and be delivered two great pieces and find a perfect fit. But most likely the shim thickness will vary over the entire length and may not be required at all.

How about a HSS 10 x 8 x 0.75" with HSS 6 x 4. This would be a total gap of 1/2". If centered the gap should +/- 1/4". If this is a not structural application, the additional 1/16" may not be an AWS issue.

http://www.FerrellEngineering.com

 

[dhengr]

That sounds like some hunk of junk a plant or industrial engineering dept. draftsman came up with which utilized some material they had laying around. It met their needs and has worked over the years, so there is little desire or incentive on the part of management to change anything. They need a replacement, what do they know about good engineering and good design. Maybe there is a real reason for that configuration, in which case I’m all wet, but I find it hard to imagine. Many people believe slopping rolled sections together is cheaper and better than designing and fabing a built-up section.

As engineers, it is not our job to go out of our way to antagonize clients or their management, but if we are trying to do a good job, it might be our duty to nicely explain the errors of their ways. By your and connecteng’s suggested detail (built-up section) you can build it straighter, stronger for the actual loading conditions, and for less cost than all the farting around their ill conceived design costs in materials and fab. time and effort. You can put the material on the faces where it’s needed, even using a thinner, lesser grade stl., on the other two faces. And, as likely as not the corners can be PJP with only one edge (pc. corner) being preped. or simpler yet, maybe just fillet welds. The edge of the thinner pc. might butt to the inside of the thicker pc., or the thinner pc. might partially lap the edge of the thicker pc. for the fillet welds. At one of the thicker flanges you can fillet weld on both sides of the webs to the flange if that would take concentrated loads into the web better. You are potentially adding value (improving it, or saving money and time) to their equipment, but you have to sell it. You can maybe make believers (in you) out of them. You’ll become something other than just another fabricator to them.

 

[JStephen]

If I remember right, there is a considerably variation in the corner radius on rectangular tubing. If one piece has the minimum radius and the other the maximum, they may not fit together like you assume they would.

 

[slickdeals]

It may not be related to the original question, but.....

I have seen this done on a recent curtain wall project for a stadium. There were areas where wind loads were higher and the vertical mullion size had to remain the same for architectural reasons. The curtain wall contractor (Crown Corr) has supposedly done this on other projects as well.