Results 1 to 13 of 13
Thread: stitch welding - when and why?
11-08-2009, 12:35 PM #1
stitch welding - when and why?
I did a bunch of welding at work this week for the first time in awhile. I'm the welder in our shop, not because I'm really any good but simply by default. The project was to plan and build some simple rolling carts, just angle iron frames with flatbar stiffeners and some casters, to set some material racks into so they could be rolled around the shop. Job came off pretty good, but as I was welding on them it brought to mind a question.
One section included a long butt joint, which out of habit I welded with a series of stringer beads about 2" long spaced about 2" apart. I developed the habit of doing this because Dear Old Dad, lo those many years ago, taught me that a long butt weld, when subjected to yield loads, would be weaker than several welds spaced apart along a joint of the same length.
When I went to work as a welder's helper a few years later, I would occasionally hear welders talk about "stitch welding", usually in connection with structural work, and assumed this must have been what they were talking about. All we did in that shop was pressure vessel welding, so I never got to actually see stitch welding being applied in practice. Every butt joint we welded was on pipe or rolled plate, and was naturally a continuous 3-pass full penetration seam. Which leads me to wonder, when is stitch welding best applied, and why is it that a continuous weld is weaker? Or was I even taught right?
11-08-2009, 01:00 PM #2
I have done this also, though with stick welding (and some TIG), but not so much with MIG. My thinking is that there is less heat input in an individual area, which in some applications could reduce warpage and heat affected areas. I'm not sure that the process itself is inherently capable of creating a stronger joint, but it could speed the process up of welding a long section without badly warping and overheating the workpiece. It's possible (I'm not sure though) that by welding shorter beads that are not touching other beads that you could limit crack propagation, but I don't know that for sure.
11-08-2009, 01:01 PM #3
I think it is primarily to control distortion and residual stresses.
11-08-2009, 01:33 PM #4
"back stitching" is often used to weld thinner materials to prevent burn through. This method is preferred over a single long bead which can get too hot for the parent metal thickness.
11-09-2009, 04:08 AM #5
I use back stitching for the same reasons noted by USMCPOP and JL Sargent. However, what you are referring to is called skip welding and can be specified by weld symbol. It is also applied when less distortion is needed.
Many, many times I have had to convince an engineer to use skip welding when adding gussets to a weldment. When skipping we always wrap the corners. We also weld in the same direction on both sides and weld so the welds are opposing each other.
11-09-2009, 05:07 AM #6
11-09-2009, 05:12 AM #7
I appreciate the responses, guys. I'm basically a hack welder, but like to learn as much as I can.
Walter, out of curiosity, what does the symbol for a skip welding callout look like?
Edit- freakish coincidence, USMCPop, the guy who provided the picture in your link has my exact name, John Allen. Crazy, huh?
11-09-2009, 05:55 AM #8
The skip information is added to the reference line either above or below depending on the weld location (arrow side or opposite side). The weld length is the first number and the center to center distance is the second number.
So, this symbol tells me to do a 1/4" fillet weld, 2" long with a 4" center to center distance all on the opposite side of the arrow.
11-09-2009, 06:49 AM #9
The link provided by USMCPOP shws pretty clearly the differences between intermittent welds (sometimes called either "stitch" or "skip" welds) and "back-step" welding, which is an entirely different technique.
"Back-step" welding results in a continuous bead, but it is produced in short segments that are run opposite to the direction of the weld. It is effective in controlling distortion when a continuous bead is needed for appearance's sake.
I only used back-step welding once, when a customer needed a 16" X 3/8" strip six feet long in a hurry. I back-step welded two 8" X 3/8" strips together using this technique, and did a lot of grinder work on the one face that needed to be pretty. It came out perfectly -- it was impossible to tell it wasn't originally a solid strip. At least from the face, which was all that was needed. Glad they didn't need both sides to look pretty.
I think a continuous bead this length would have warped like crazy no matter how it was clamped, so it's a nice trick to know.
I learned about it in "Welding Fabrication and Repair" by Frank Marlow -- useful book.
11-09-2009, 07:09 AM #10
11-09-2009, 09:18 AM #11
But it's nice to have the technique in the old bag of tricks
11-09-2009, 07:46 PM #12
The TWI article is a good one and the TWI is an excellent resource.
My first job as a welding engineer was as a joint detailer literally going through hundreds to thousands of welding blue prints and specifying and writing all the weld symbols. A lot of the time I was given the joint strength requirements and had to figure out the welding detail from that.
now as to why stitch welds are used. Distortion control is generally the first reason. In general lower heat= less distortion. And distributing that heat evenly, by skipping around, back stepping etc can help control distortion. You can also control distortion by preheating a part (another discussion)
In general the other reason joints are only intermittently welded is economics and strength.
For example a simple stiffener like a gusset on the leg of stand almost never needs a full weld. Assuming the material is the same thickness. The main member will fail by buckling or torsional stress way before the stiffener would fail. So placing full length welds gives you no strength advantages and lots of headaches and wasted cost. I remember specifically taking out pass after pass
I would say about 90% of intermittent welds (Besides distortion) are specified for this reason. The service condition stress is way lower than what a full welded joint would provide. So why bother?
finally occasionally in the post welded condition or by using a different filler material a weld joint will have a higher strength than the base metal In that case fully welding will not increase strength and actually may decrease it. In those cases intermittent welds are called out, but that requires a decent amount of pencil and paper engineering.
Now just to throw you through a loop. There remains some very special cases when a LOWER strength filler material will always fail AFTER the higher strength base metal... but it wouldn't be interesting unless I waited awhile until I posted the reason. (most likely some of you already know why)
11-09-2009, 08:34 PM #13
Your last line immediately made me think of brazing, as used in bicycles. Yes, this isn't welding, since it doesn't bring the parent metal to it's melting point.
But, fatigue is what you are referring to. A lower strength metal is usually more ductile and doesn't work harden as quickly.
But, the HAZ is the bigger issue. Unless a post weld heat is done, the HAZ will artificially increase the hardness and the weld will fail at the seam between the weld and the parent metal.