SS316 W10x33 Built up member

[jagui007]

Hello All!

We are making W10x33 using 316SS plate. Our concern is the flanges stressing with the heat at the welding stage and warping out of shape. So our suggested solution is to bend or break a relelief on to the top and bottom flanges so that when they heat up they will bend back to flatness. Does anyone know if there is an easy way to look up what the bend dimension should be?


Thanks in advance!

 

[dhengr]

Jagui007:
This type of built-up member is made and welded every day without any great problems. Any problems are pretty much std. fabricating issues, which any good shop should know how to handle. I haven’t worked much with SS, but I don’t think it should act much differently than our regular structural steels, on this account. Some thoughts on the matter....
1.) Your built-up member is about 10" deep, has a .25 - .375" thick web, and has 8" x .5" t&b flgs. How long is it? w.r.t. your question, it matters how you sequence your welding, and an experienced shop should know this, and be willing to experiment a bit with it.
2.) In your welding, you should not try to replicate the full penetration and large radius at the web to flg. juncture, you see on a rolled shape. This large amount of welding is generally not required for structural design reasons or good welding design practice, and is usually the cause of excessive distortion problems, both across the flgs. and longitudinally on the member. I would be surprised if you needed anything larger than .25" fillets on your beam.
3.) Build you members on a strong-back which is several times stiffer than your beams are. You do not normally pre-bend the flg. pls. to counter this problem, you gag them down to the strong-back to account for this distortion if it is serious enough. You also gag the partially fit-up member down longitudinally to account for the longitudinal camber which welding causes, but you start with a straight web pl. We actually cut camber curvature into the web pls. of a deep pl. girder and fit the flg. pls. to this curved web, but you should not do this with this light beam.
4.) I’ve done these types of calcs. over the years, but at best they are good approximations based on a lot of engineering and welding experience and judgement; and also based on a bunch of variables which are difficult to pin down exactly. The welding industry has some literature and methods for tackling this problem, if it really becomes a problem.
5.) To a good extent the solution to these types of problems was a matter of good fab. shop knowledge, experience and practices on std. weldments. Then, when I thought there might be problem, some of my calcs. and guesses; then a meeting of the minds (engineering & shop) as to a starting point on the first weldment, and a little adjustment on the second, if needed. We didn’t often do large runs of one thing, so we had to get pretty good at our initial guess. If you have the time and quantity to do a little experimenting and refining, you can finally jig that stuff to come out straight as a string, in all directions.
6.) Weld sequencing is another matter, for another post.