mark 4B-03bench · Steel noteschecked June 4, 2026
Shop weld or field weld: what the drawing actually says
Bench check on a weld callout: where the symbol puts the weld, what the field flag changes, and what the crew does when the drawing and habit disagree.
Noted by Ray Tobin, shop foremanchecked by Hollis Grant6 min read2 sources

The drawing came back to the bench with a pick already in the air. A beam for a school gym job, a shear tab on the web, a callout reading 5/16 fillet, 12 in. long at 12 in. on center. Nothing on that line is unusual, and the iron was fit and tacked and waiting. The argument was about the flag, or about the absence of one, and it had already gone from the deck to the office and back twice before somebody laid the sheet on my table. The question underneath it was simple enough: does this weld get made here, on the bench, or out there in the air, by somebody else.
The check, before anybody answers anything
Protocol first, because the answer is worth nothing without it. I read the sheet in this order: revision block, general notes, the symbol, then the member itself where it sat on the fit-up skids. The instrument list was short. A 25 ft tape for the weld lengths and the spacing between them, a caliper for the fillet sizes on the two tabs already welded that morning, the shop's WPS binder, and the mill reports for the two heats involved, an ASTM A992 W16x26 and an ASTM A36 plate tab. I read the callout against the general notes on sheet S-1 and not against the detail alone, and I read the tail before I ever looked at the flag.
What the tail of a weld symbol is telling you
The tail is the small fork at the far end of the reference line, and it is where a callout names its authority. On this sheet the tail carried the process and the electrode, E70XX, which is a 70 ksi minimum tensile electrode, and a WPS number that points into the binder. When the tail is empty, the process and the electrode come from the general notes instead, and the note on S-1 says all welds are to be made to a WPS qualified under AWS D1.1. That is a citation of a standard, not a decoration. It means the procedure was written and qualified before the first arc, and the tail either points at it or defers to it.
What the tail does not tell you is where the weld happens. That was never its job.
The flag is the whole argument
The flag sits at the junction of the leader and the reference line, and it means one thing: field weld. No flag means shop weld. That is the entire convention, and it is worth more money than most people on a job stop to realize. A 5/16 in. fillet welded at the bench costs a fraction of the same weld made 30 ft up, with a welder in a lift, a fire watch, and a hot work permit. A shop weld gets made in position, with the member rolled flat, by whoever is at the bench that day. A field weld gets made in whatever position the structure allows, in whatever weather New England is offering that week.
On the sheet on my table there was no flag, which put the weld in the shop.
When the drawing and the habit disagree, who wins?
Here the habit and the drawing agreed, which is not always the case. On a hospital wing addition two winters back, the flag was on a column splice and the shop had already fit and welded a stiffener that the detail left to the field. The fit was easier on the bench, so the bench took it. That is a conversation, not a call one crew makes alone. What settled it was an RFI carrying the sheet number, the detail number, and a photograph of the callout, not an opinion about which way is better. The answer came back in three days and the stiffener stayed in the shop, with the field weld moved to the remaining seam. Three days against a fab shop's schedule is real money. It is still cheaper than a field fix that has to be cut out and re-welded overhead.
The AISC Code of Standard Practice draws the line between what the fabricator owns and what the erector owns. It does not redraw a weld symbol for you, and it will not settle an argument about a flag your own shop set.
Can you read a callout without the general notes?
No, and this is where most misreads begin. The symbol carries size, length, spacing, side, and whether the weld is field or shop. It does not carry the acceptance criteria, the minimum fillet size, the preheat, or the electrode unless the tail names one. The minimum fillet size floor for a given plate thickness lives in the tables of AWS D1.1, not on the detail. On the 3/8 in. tab on this job that floor is 3/16 in., so the 5/16 in. called out sits above it and nothing about the callout is undersized. Had the same detail read 1/8 in. on that plate, the number would have looked perfectly reasonable on the sheet and been wrong at the bench, and only the general notes and the standard behind them would have caught it.
The values as they were read on this check
Written down as read, not as law. Reference line: fillet, arrow side. Size 5/16 in., length 12 in. at 12 in. centers. Tail: E70XX, WPS on file. Flag: absent, so shop weld. General notes: AWS D1.1, electrodes per the WPS, thermally cut edges held to the workmanship requirements of that same standard. Fit-up on the two tabs, checked with a tape and a square: 1/16 in. and 3/32 in. gap, both inside the 1/8 in. the shop works to on fit-up. Material: A992 for the beam at 50 ksi yield, A36 for the tab at 36 ksi yield, both heats with mill reports on file. The companion connection on the same detail is bolted, snug tight per the RCSC specification, with no pretension called out, which is worth knowing before anyone decides the weld is the only thing on that joint doing any work.
What this note does not cover
Several things, and they matter more than what it does cover. This note is not a WPS and does not qualify one. It is not a procedure qualification record. It is not a welder qualification, and no continuity record was reviewed. It is not an inspection record: no visual or nondestructive examination was performed for this check, no dye penetrant, no ultrasonic, no magnetic particle. It does not cover preheat, interpass temperature, or heat input, because neither member here is thick enough to put those questions in play. It does not cover coating or galvanizing to ASTM A123, or how either behaves at a field weld on a coated surface. It says nothing about the engineer's intent beyond the revision of the sheet that was on the table, and it covers no field condition, because the field was never part of it.
The gesture worth copying
If you run a shop or you run a job, take a highlighter to the sheets in the fit-up area. Mark every symbol whose tail points somewhere other than your own WPS binder, and every symbol where the habit on your floor and the line on the drawing disagree. Do it on the day the sheets arrive, before the pick goes up. A field weld that nobody named in advance costs a lift, a permit, and a welder's afternoon. A highlighter costs about a minute.

