Winter steel in New England: what the cold actually does to a job

Six degrees on the shop thermometer, and the first pick of the day is a 40 ft. W24x84 that spent the night on the yard rack. It comes in with frost on the top flange, goes down on the sawhorses, and the fitter puts a tape on it before anybody lights a torch. What follows is what the instruments read on Note S-114, the cold weather sheet kept for a school gym job.

Does cold steel actually get brittle?

The sentence you hear in a trailer in January is that steel goes brittle and shatters. On ordinary structural steel at the temperatures a New England winter delivers, that is mostly not what fails a job. Toughness is settled when the material is ordered: the AISC 360 specification and the ASTM A6 supplementary requirements put Charpy V-notch testing on specific shapes and temperatures, and the mill ships the material with the impact test done or it is not what the drawings call for. A fitter in a yard cannot test for it. What changes in the cold is everything around the steel: how it is measured, cleaned and welded, and how long it takes with gloves on.

The tape, the beam, and fifty degrees of difference

The instrument was a 100 ft. steel tape, checked against the shop's reference tape at 68 F. Steel moves about six and a half millionths of an inch per inch per degree F, roughly 0.008 in. over the full blade, so the 12 F yard put the tape about 3/8 in. short. Across the 40 ft. beam the same arithmetic gives 5/32 in., and two fitters read exactly that: one with the tape off the cold rack, one with a tape that had ridden in the truck cab, the reads 5/32 in. apart, more than the 1/16 in. anybody writes down. Nobody fails a job over 5/32 in., but a fitter who does not know it is there will chase it all afternoon and blame the iron.

Condensation and flash rust at the shop door

Protocol: surface temperature read with a magnetic contact thermometer, cross-checked with an infrared gun at three points along the web, dew point off the shop hygrometer every two hours. The rule the coating maker prints is that a surface sits at least 5 F above the dew point before anything is blasted or primed, and a beam off a flatbed at 15 F wheeled into a 65 F shop is below that line for hours. Three beams brought in on a Tuesday morning had a full flash rust bloom on freshly blasted webs by the second shift and had to be re-blasted. The beams that waited on the truck under a tarp came through in one pass. The AISC Code of Standard Practice puts shop painting and surface preparation on the fabricator, and a surface that flashed rust is not a prepared surface.

What does preheat do that more amps cannot?

Protocol: base metal temperature read with a contact pyrometer and confirmed with temperature-indicating crayons at four points, 12 in. apart, on both sides of the joint. The procedure, written to AWS D1.1, called for 32 F minimum base metal temperature before any arc and 70 F minimum preheat on the 1 in. flange of the built-up column. That flange read 38 F, which satisfies one line and not the other, so the rosebud came out and the arc did not start until the crayons were refused at all four points. Turning the machine up does not do this job: more current on cold steel drives hardness into the heat affected zone and does nothing about moisture on the surface or hydrogen off a damp electrode. This note read interpass as well as preheat, because a welder who preheats well and then runs five passes without looking can walk the interpass past the top of the range as easily as under the bottom.

The fit-up that closed in June

A bay that squared easily in August is a different animal in January. The tolerance does not change with the season: the AISC Code of Standard Practice puts the acceptable fit-up gap at plus or minus 1/8 in., and a 40 ft. member that grew 3/16 in. in the afternoon sun gets pulled into place with a come-along instead of a drift pin. A connection that started two bolts by hand in July needs a spud wrench on every one of them with gloves on. A gap forced past 1/8 in. comes back as a weld that will not pass.

What the calendar actually costs a schedule

The honest numbers on this job, not a rule for anybody else's: a mid-December day in southern New Hampshire runs about nine hours between sunrise and sunset, so a crew that worked ten in June works eight in December with the same start. Handling slows, not because anybody is slow, but because every bolt needs a second attempt. Grout under the base plates needs 40 F and rising for a full day, so plates set in January wait for a window, and touch-up paint needs a surface above 50 F, which pushes a punch list meant to close in December into April. The AISC Code of Standard Practice does not set the calendar; the contract documents do, and if they do not say who pays for the winter, the answer gets invented later by somebody with a lawyer.

What this note does not cover

It does not evaluate toughness, or say whether the material ordered was right for the service temperature. It does not qualify a welding procedure: that procedure was qualified somewhere else by somebody else, and reading a preheat number in the field is not the same as testing it. It does not cover fall protection, ice on walking surfaces, or the winter requirements OSHA carries, which belong to a safety plan this note is not. It does not cover coatings beyond the shop door temperature and dew point reading, and it certifies nothing about a coating warranty. It does not cover the hospital wing addition that ran the same winter, which had its own sheet. It does not commit anybody to a date, and it says nothing about the members outside the grid lines named at the top of this sheet.

What a GC or an owner can ask for

Ask for the cold weather plan in October, not January, when the answer can still change. Four questions do most of the work. Who reads the surface temperature and the dew point before blasting and priming, and where is it written down? What is the minimum preheat on the procedure, how is it measured, and at how many points? Where are the tapes verified against the reference? And which base plates will be grouted this winter and which will wait for spring, with a date attached to the wait. Then on the first cold Monday, walk the shop and watch whether anybody puts a thermometer on the steel before the primer goes on. If they do, the plan is probably real. If they do not, nothing else in the binder matters.