The torque check on an A325 connection: what it verified

The bolts came out of the keg onto the bench in a school gym job, 7/8 in diameter by 2-1/4 in long, with the connection schedule off sheet S-4 open on the table beside them. One bolt took a hardened washer under the nut, a match mark got struck across the nut flat and onto the threads with a paint pen, and the impact wrench went on. That is the whole of the operation. What follows is what the instruments read on Note B-022, in the order they were read.

What the drawings asked for before the wrench came out

Bolted connections do not get checked against a feeling. They get checked against the general notes and the connection schedule, so the first thing on the table was Note 14, not a tool. This frame had two kinds. Most of the shear connections carried the words snug tight, which the RCSC specification defines as the tightness from a few impacts of an impact wrench or the full effort of an ironworker on an ordinary spud wrench, enough to bring the plies into firm contact. The moment and brace connections carried the word pretensioned, a different job with a different record. The bolts were marked F3125 Grade A325 Type 1, which is where A325 lives now that ASTM folded it into F3125. Holes were standard, 15/16 in for a 7/8 in bolt, and the detail called for a hardened washer under the turned element.

Which joints actually needed pretension?

This is the question that decides how much of the check is even required, and it is answered on the drawings, not on the frame. The AISC Code of Standard Practice and AISC 360 both leave the choice to the connection design, and the RCSC specification treats snug tight as a complete and legitimate installation for a bearing connection where slip is not a design consideration. A crew that torques every bolt on the job to some remembered value is not being careful. It is spending money, putting washers where nobody asked for them, and producing a record that means nothing because no requirement stood behind it. On this job the count was 62 bolts in pretensioned joints and roughly 900 in snug tight ones. The 62 got the record and the walk. The 900 got a wrench, a look at the faying surfaces and nothing else.

The turn-of-nut marks, read against the chart

Turn of nut is the pretensioning method that leaves a mark somebody can read the next morning. Section 8 of the RCSC specification gives the rotation: for a bolt whose length is four diameters or less, one half turn from the snug condition; over four and up to eight diameters, two thirds of a turn; past eight diameters, a full turn. Our bolt is 2-1/4 in long on a 7/8 in diameter, which is 2.6 diameters, so it sat in the first band and the number was 180 degrees. The ironworker struck the mark after snugging, ran the wrench until the marks stood half a turn apart, and that was the acceptance. Reading it back the following morning is the check: 62 bolts, marks at 180 degrees on every one, none short. One bolt failed the look because a sleeve had rubbed the paint flat on the nut flat, so the mark was re-struck and the bolt re-read rather than assumed.

What does a torque wrench prove on a bolt that is already tight?

The torque wrench came out for six of the 62, and its role needs stating plainly. Torque is a proxy for tension, tied to it by a coefficient that depends on thread condition, lubrication and plating, and the scatter around it is wide enough that one torque can produce different tension from one lot to the next. That is why the RCSC specification makes the calibrated wrench prove itself against a tension calibrator, a Skidmore-Wilhelm, instead of trusting a published table. Our wrench was set at 410 lb-ft, the reference the tool room keeps for this size and finish. On the six bolts pretensioned by turn of nut the day before, the nut did not move on any of them, which says those bolts sit at or above the tension that torque corresponds to. It does not say the number. Then the wrench was walked onto four bolts that had only been snugged: those nuts turned at 180 to 260 lb-ft. Same wrench, same setting, two populations, and the reading separates them without proving the tension in either.

The DTI squirt on the hospital wing addition

Second job, different detail, same question. These connections used direct tension indicator washers under the nut, and this is where the visible part of the method gets mistaken for the test. The washer carries raised protrusions, and the installation is complete when they have flattened enough that a 0.005 in feeler gauge is refused at the required number of gaps, which depends on how many protrusions the washer has. Some versions hold silicone in the protrusions, and when the gap closes the silicone extrudes at each bump. It is a satisfying thing to watch and it is not the criterion. On the hospital wing addition the silicone squirted at all four protrusions on 22 of 24 bolts, and on those the gauge was refused at three gaps of four. On the other two the silicone showed at three protrusions only, and the gauge agreed: it entered at two gaps. Those two washers went on the scrap pile with their bolts: a DTI washer compressed past its window is spent, and adding rotation to it proves nothing.

What this note does not cover

This is a bolt installation note on two jobs. It does not certify the connection design and it does not say the joint is the right joint for the load. It does not match the bolt lot against the mill certificates for the heat, which is a receiving check done when the kegs arrived. It does not cover faying surface condition, which is cleaned and inspected before assembly under its own note. It does not cover slip-critical joints, because there were none on either job, and a slip-critical connection carries a surface preparation requirement and a rotation check this note never touched. It does not cover galvanized or Type 3 weathering bolts, which behave differently under the wrench. It does not cover the wrench's own calibration beyond the daily verification sheet, and it says nothing about the bolts outside the grid lines named at the top. None of this replaces the special inspection a project's statement of special inspections may require.

What a GC or an owner can ask for

Ask for the installation method by name on the drawings, and ask the erector for a written count of which joints were pretensioned and which were snug tight. Ask for the match marks or the feeler gauge readings as the record rather than a torque number, because a mark on a nut and a gauge that refuses to enter can be read after the fact by somebody who was not there, and a torque figure on a clipboard cannot. Then ask what happened to the bolts that failed the look, since a replaced bolt and a re-tightened bolt leave different evidence behind. The check in this note took about fifteen minutes at the bench and a walk of the frame the next morning. It is the cheapest record on the job, and the only one saying the connection was installed as designed.