Oh boy. I’m sure I’ll get flak from someone for tackling this one, but here goes anyway.
Let’s talk about tolerances for welding. For this post, we’ll focus only on the weld joint and follow the common theme of customer-vendor interaction.
Customer: “What kind of gap range can you tolerate in a weld?”
Vendor: “What kind of weld? Process? How deep? How wide? Material? Filler?”
Customer: “That doesn’t matter yet, I just need to know how big of a gap you can weld and how much it can vary.”
Like the last blog post, this is a trap! For the welding folks reading this, I KNOW you’re cringing with understanding. It’s like nails scraping on the chalkboard (if those are still used). For those of you looking for help interacting with your weld shop, please allow me to explain:
We work with welds that range from invisible to the unaided eye, to heavy section pressure vessel weldments that you can see from across the room.
That pressure vessel can, and does, tolerate a ¼” gap with a +/- tolerance of 1/16”. That’s a gap difference you can literally see from an arm’s length away, and it’s perfectly fine. That project uses high power, huge filler, and is relatively insensitive to sloppy tolerances.
Let’s look at the other extreme – that “invisible weld.” We have worked on a project forming a cylinder from shim material. This part is made with no filler, and at this scale we discuss gaps in the butt weld as a percentage of weld joint thickness. This material was not particularly sensitive to stress or heat- related cracking, so we were in a fairly good position to offer up to 25% of the weld joint thickness for localized gaps, with a length of no more than 20x the joint thickness. Sounds quite generous until I tell you that cylinder was .001” thick. Cue the stuttering, frustration-laden, concerned call with the designer.
Customer: “Those tolerances can barely be measured!”
Vendor: “The weld can barely be seen…..”
Ultimately, customer and vendor worked through the tolerance issue and developed a cut process that kept the edge tolerance well within the prescribed band, and all was right with the world.
I’ll cut this short at this point, but there will be a part 2 (and maybe 3) to discuss weld joint thickness, mismatch, and tight penetration requirements.