Structural steel erection and connections field guide

Steel erection, and why the half-built frame is the dangerous part

Structural steel erection is the field process of raising a steel frame: picking each member off the lay-down, setting columns on their anchor bolts, connecting beams to the columns, plumbing the frame true, then bolting and welding the connections so the structure can carry its load. Done in order, it turns a pile of marked steel into a building skeleton.

The danger that governs the whole job is simple and it does not change. A finished steel frame is a stiff, stable box. A frame that is half up is neither. Until enough connections are made and enough temporary bracing or permanent diaphragm is in place, the partly-erected frame can rack, twist, or come down, and it can do it with people hanging off it. That is why steel erection is regulated as its own discipline under OSHA 1926 Subpart R, with hard rules for the anchor bolts under the columns, the stability of the frame as it goes up, and fall protection for the people putting it together.

The connections are the other half. A bolted or welded joint that has not been made to the structural requirements, and inspected, is not ready to take the load it was drawn for. So the work splits into two jobs that have to stay in step: keep the frame standing while it is incomplete, and prove every connection before the design load arrives. The columns start on the concrete the foundation crew left, so the anchor bolts, the base plates, and the grout tie this guide to the footings work, and every pick ties it to the crane and rigging work.

Why is a partly-erected steel frame unstable?

A partly-erected steel frame is unstable because the system that makes a finished building stiff, the completed network of connections, braces, and floor diaphragms, is not there yet. A single column standing on its anchor bolts is a tall lever with a small base. A beam dropped into its seat with a couple of bolts is held, not fixed. Until the bay is closed and braced, lateral load from wind, from the crane, or from the swing of the next pick has nothing solid to push against.

This is the truth that ranks above every other on a steel job, and it kills when it is ignored. Erection collapses happen during the gap between setting a member and securing the frame around it, when a crew has gone faster than the bracing or skipped the sequence the engineer drew. The members were strong enough. The incomplete frame was not.

Two things hold the frame up while it is incomplete: the erection sequence, which keeps completed, braced bays ahead of open ones, and the temporary bracing or guying that stiffens the steel until the permanent lateral system takes over. Both are engineered, not improvised. OSHA 1926 Subpart R, the AISC Code of Standard Practice (AISC 303), and the engineer of record set the requirements. The field's job is to follow the plan, not to decide on the iron how much frame can stand open.

What does OSHA Subpart R cover?

OSHA 1926 Subpart R is the steel erection standard, and it controls the specific hazards of raising a steel frame. It is the rule the compliance officer opens on a steel job, and it is written around the failures that have hurt and killed ironworkers: collapse of an unstable frame, falls, and being struck by loads or members.

The standard runs from the site conditions before steel arrives through the connections that finish the frame. Site layout and a firm path for the crane come first. Column anchorage and the readiness of the concrete are next, at 1926.755. Structural steel assembly and the minimum securing of members fall under 1926.754 and 1926.756. Fall protection for the steel trades is at 1926.760, with its own triggers and the controlled decking zone. Open web steel joists, systems-engineered metal buildings, and falling-object protection each have their own sections.

Treat the citations in this guide as a map, not the law. Section numbers and the exact wording shift between editions and interpretations, and a jurisdiction can adopt a state plan that is stricter than federal OSHA. Confirm the current standard, the AHJ's requirements, and the project's site-specific erection plan before you act on any single number. Subpart R sets the floor. The engineer of record and the contract documents can, and often do, demand more.

What is the OSHA 4-bolt rule?

The 4-bolt rule is the OSHA requirement that every column be anchored by a minimum of four anchor rods, commonly called anchor bolts, before the load is released from the crane. It lives in 1926.755, and it exists because a column on fewer than four anchors does not have the stability to stand against the eccentric load and the wind it will see before the frame is closed around it.

The rule does not stop at counting bolts. The standard ties the requirement to a strength criterion: the column, its anchor rod assembly, the column-to-base-plate weld, and the foundation have to be designed to resist a minimum eccentric gravity load of 300 lb placed 18 in from the outer face of the column, in each direction, at the top of the shaft. That is the engineer's calculation, not the field's, but the field is who confirms four bolts are actually there and sound before the column is set free.

The concrete under the column has to be ready: poured, cured enough to take the load, and surveyed so the anchor bolts are in the right place at the right elevation. Anchor bolts that are mislocated, bent, or short are common, and the fix is not the erector's call. OSHA is explicit that anchor rods shall not be repaired, replaced, or field-modified without the approval of the project structural engineer of record. Bending a bolt back or welding on an extension without that approval is exactly the shortcut that puts a column on a base it cannot trust. The anchor bolts, the base plates, and the foundation belong to the footings work, so read this alongside the foundation and footings guide, and hedge any anchor-bolt repair to the EOR every time.

Shake-out: sorting the steel before it flies

Shake-out is the work of receiving, sorting, and laying down the steel so the right piece is ready when the crane calls for it. It happens before erection and it sets the pace of everything after. Steel arrives on trucks bundled for shipping, not for the order it goes up, so the crew breaks the bundles down and stages members by piece mark and by sequence.

Every structural member carries a piece mark that matches the erection drawings. Reading those marks and laying the steel out so columns, beams, and braces come to hand in the order the sequence calls for is what keeps the crane working instead of waiting. A piece in the wrong place is a pick wasted, and a wrong piece flown up and found wrong at the connection is worse.

The lay-down also gets the steel ready to rig safely. Members go on cribbing, dunnage, or blocking, off the mud, oriented so the connector can reach the picking point and the crew can hook it without working under it or rolling it by hand. Shake-out is not glamorous and it is where a smart foreman buys back hours. Sort it once on the ground, or sort it again in the air at four times the cost and twice the risk.

The crane pick and rigging the member

Every member goes up on a crane, and the pick is governed by the load chart, the rigging, and the lift plan, not by the operator's read of the day. The weight of the member, the radius it has to reach, and the configuration of the crane decide whether the pick is inside the chart. A piece that is fine at a short radius can be over capacity swung out to the far corner of the building, and that is where erection picks go wrong.

Rigging the steel is its own discipline. The member is hooked at its rigging or picking points so it flies level and lands oriented the way the connection needs it, with slings and shackles rated for the load and the angle. A tag line keeps the piece from spinning and lets the crew control it from the ground without anyone putting hands on a suspended load. The connector takes it at the column or the seat, not the rigger at the back of the truck.

This is the heart of the crane and rigging discipline, and it has its own standard, OSHA 1926 Subpart CC, plus ASME B30 for the rigging. The full treatment of the load chart, the power-line clearance, sling angle, and the exclusion zone lives in the crane, rigging, and signaling guide. Read it as the companion to this one. On a steel job the crane and the iron are the same operation, and the rules for both apply to every pick.

Setting the columns on the base plates and anchor bolts

Columns go up first, because everything else hangs off them, and setting a column is the move where the foundation work and the steel work meet. The column lands on its base plate over the anchor bolts, the leveling nuts or shim packs take the elevation, the bolts are run up, and the column is roughly plumbed before the crane is released, with the four-bolt minimum and its strength criterion satisfied.

How the base bears matters as much as where it lands. OSHA recognizes a column set on a level finished floor, on pre-grouted leveling plates, on leveling nuts, or on shim packs, as long as the arrangement can carry the construction loads. Leveling nuts under the plate, adjusted up or down, are the common way to set elevation and get the column plumb on a heavy frame; shim packs do the same job on lighter steel. Either way the load path from column to plate to anchorage to concrete has to be real before the hook comes off.

The first columns set in a frame are the most exposed, standing alone before any beam ties them together, so they get guyed or braced as the competent person determines. A column is not done when it is bolted down. It is done when it is plumbed, secured, and either braced or tied into a beam that is itself secured. Set it, level it, and do not turn the crane loose until the base is right and the column will stand. The anchor bolts and base plates trace back to the footings work, so confirm the foundation was built and surveyed to the drawings before the first column flies.

Leveling the base plate and grouting under it

Leveling sets the column to the right elevation and plumb; grouting fills the space under the base plate so the load transfers into the concrete in full bearing. The two are separate steps and the grout comes later, after the column is set, plumbed, and the frame around it is far enough along that the base is in its final position.

Leveling nuts and shim packs hold the plate off the concrete during erection, which leaves a gap between the underside of the plate and the top of the footing or pier. That gap gets packed with non-shrink grout so the plate bears on a solid, continuous surface instead of on a few points of steel. Non-shrink is the word that matters: ordinary grout shrinks as it cures and pulls away from the plate, leaving voids that concentrate the load and defeat the point of grouting at all.

Grout placed too soon, before the column is plumbed and the bay is set, locks in whatever is wrong. Grout skipped or done badly leaves the base bearing on leveling nuts that were never meant to be the permanent load path. The bearing detail, the grout type, and the timing come from the structural drawings and the EOR. Build the base the way the detail draws it, and read the footings guide for how the concrete under the plate was meant to behave.

The raising gang: connecting the beams

The raising gang is the crew that takes the steel from the crane and connects it into the frame: the connectors up in the iron, the operator, and the signal person and riggers feeding it. They set the beams between the columns, drop them into their seats or up to their clip connections, and put bolts in to secure each member before the load comes off the hook.

OSHA sets the floor on how secure that is. During the final placing of solid web structural members, the load is not released from the hoisting line until the member is secured with at least two bolts per connection, of the same size and strength shown on the erection drawings, drawn up wrench-tight, or the equivalent specified by the engineer of record. Two bolts wrench-tight is the minimum to let go, not the finished connection. Diagonal bracing members get at least one bolt wrench-tight before release.

The two-bolt minimum is a stability rule, not a completion rule, and a competent person decides when more is needed. Cantilevered members, and any member whose stability depends on it, can require more than two bolts before the crane is released, and the competent person determines that and gets them in. The connectors put the steel where it goes and make it safe to release. The full pretension or the weld that finishes the joint comes after, on the ground crew's schedule, and it has to be inspected before the frame is loaded.

Plumbing up the frame

Plumbing up is the work of pulling the frame true, vertical columns plumb and bays square, before the connections are finalized. A frame goes up with slack and tolerance in every joint, and left alone it leans. Plumbing-up takes that lean out and holds the steel in position while the bolts get pretensioned and the welds get made, so the connections lock in a frame that is straight instead of one that is racked.

The crew pulls the frame with plumbing guys, turnbuckles, and come-alongs, working against a survey. A theodolite, a total station, or plumb readings tell the crew which way the column is out and how far, and the rigging pulls it back inside tolerance. AISC 303, the Code of Standard Practice, sets the erection tolerances, with column plumbness commonly held to 1/500 of the height between working points. The crew tunes the frame to that, then holds it there.

The order is the part rookies get backward. You plumb the frame, then you finalize the connections, not the other way around. Final-bolt or weld a bay before it is plumbed and you have permanently fixed whatever was crooked, and now the fix is loosening or cutting a finished connection. Hold the plumbing guys and the bracing until the connections that lock the frame are made and accepted. The survey, the tolerances, and when the temporary supports can come off are the engineer's call, not a judgment made on the iron.

Temporary bracing and guying for stability

Temporary bracing and guying keep the frame standing during erection, in the gap between setting the steel and finishing the permanent lateral system. This is the most direct answer to the unstable-frame hazard, and it is engineered work, not field improvisation. Guy cables, temporary diagonal bracing, falsework, and cribbing stiffen the open frame against wind and erection loads until the permanent bracing, moment connections, and floor diaphragms can do the job.

AISC 303 puts the responsibility on the erector to determine, furnish, and install the temporary supports needed for the operation, working from information the owner's designated representative provides about how the finished structure resists load. Steel frames are normally braced with temporary guys that, together with the steel deck floor and roof diaphragms, hold the frame stable as it goes up. The same guys are what the crew pulls against to plumb the frame.

The rule that gets people killed when it is broken: temporary bracing stays until the permanent system that replaces it is in and capable. Pulling a guy or a brace early, to get it out of the way of the next pick or because the frame looks solid, removes the support the frame is leaning on before its replacement exists. That is a collapse mechanism, plain and direct. When the bracing comes off, in what order, and what has to be complete first is engineered by the EOR and the erection engineer. Do not remove temporary support early, and do not decide on the iron that the frame no longer needs it.

What is the difference between snug-tight and slip-critical bolts?

A snug-tight bolt is brought up only until the connected plies are in firm contact; a slip-critical bolt is pretensioned to a high, specified clamping force so the joint carries load by friction between the faying surfaces and does not slip. Both are made with high-strength structural bolts, commonly ASTM F3125 Grade A325 (the spec that absorbed the old A325 designation, with a minimum tensile strength around 120 ksi for the usual diameters). The difference is how hard they are tightened and what the joint is counting on.

Snug-tight is the right answer for most bearing-type connections, where the bolt bears against the hole and the plies do not need to be clamped against slip. Snug is reached with a few impacts of an impact wrench or the full effort of a spud wrench, bringing the plies into firm contact. Pretensioned connections are tightened well beyond snug, and slip-critical connections add a requirement on the faying surfaces (the Class A or B condition of the steel that the friction depends on). Slip-critical is specified where slip cannot be tolerated, such as joints with load reversal, fatigue, or oversized and slotted holes.

What governs which one you make is the drawing and the connection design, full stop. The RCSC specification (the Research Council on Structural Connections), referenced through AISC 360 and AISC 303, defines snug-tight, pretensioned, and slip-critical and the methods for each. The erection drawings call out the joint type bolt by bolt. Do not decide in the field that a snug joint is good enough where the design called for pretension, and do not pretension everything because more feels safer. Make the connection the design specifies, and hedge the call to the engineer of record and the RCSC.

Installing and verifying the pretension

Pretensioned and slip-critical joints are installed in two stages: bring the whole connection to snug first, then pretension by an approved method. Snugging first matters because it pulls the plies together; if you pretension one bolt while the plies are still gapped, you lose the tension as the joint closes under the next bolt. Snug the pattern, then tension the pattern.

The RCSC recognizes several pretensioning methods, and the project specifies which. Turn-of-nut applies a defined additional rotation past snug, matched to the bolt length and grade, so the stretch in the bolt develops the tension. Direct tension indicators, the DTI washers with bumps that flatten under load, give a visible go or no-go you read with a feeler gauge. Twist-off-type tension-control bolts shear off a spline at the design tension, installed with a special wrench. Calibrated-wrench tightening sets a torque verified on a tension-measuring device that day. Each method has a pre-installation verification, run on a Skidmore or equivalent, to prove the lot and the crew's setup actually reach tension before any of it goes into the structure.

Verification is what the inspector checks, and it follows the method. For turn-of-nut, the match-marks on the nut and the bolt show the rotation was applied. For DTIs, the gaps closed to the specified condition. For TC bolts, the splines are sheared. The inspector confirms snug was reached first and that the right method was applied and arbitrated correctly. A bolt that was torqued to a number with no verification of actual tension is a bolt nobody can prove is pretensioned, and on a slip-critical joint that is the whole ballgame.

Field welding the connections

Field welding makes the connections that are welded rather than bolted, often the moment connections that resist the frame's lateral load, and it is qualified, inspected work. The weld is made by a welder qualified to a welding procedure specification (WPS) that defines the process, the filler metal, the joint detail, the position, and the parameters. Structural welding on buildings follows AWS D1.1, the structural welding code for steel, and the connection is only as good as the procedure and the welder behind it.

Field conditions make field welds harder than shop welds, and the inspection accounts for it. Wind, position, access, and weather all work against a clean weld up in the iron, where the shop had the piece flat on a table. Preheat where the detail calls for it, weather protection, and clean joints are the difference between a sound weld and porosity or a lack of fusion that nobody can see from the surface.

Inspection of structural welds is the job of a certified welding inspector, a CWI, and it runs from visual inspection on every weld to volumetric testing where the code and the connection require it. Visual inspection catches the surface defects: undersized welds, undercut, cracks, profile, the things a trained eye reads. Complete-joint-penetration groove welds, common in moment connections, get ultrasonic testing (UT) or another nondestructive method to find the flaws below the surface that vision cannot reach. What gets which inspection, and the acceptance criteria, come from AWS D1.1, the project specifications, and the building code's special-inspection requirements. The CWI signs it off; the field does not self-certify a structural weld.

How are steel connections inspected before the load comes on?

Steel connections are inspected before the structure is loaded so that every bolted and welded joint is proven to the structural requirements while there is still time and access to fix it. This is special inspection, required by the building code for structural steel, and it is independent of the erector's own quality control. The premise is plain: the frame is only as strong as its connections, and a connection nobody checked is a connection nobody can stand behind.

Bolted connections are inspected by joint type. Snug-tight connections are confirmed to be in firm contact and complete. Pretensioned and slip-critical connections are verified by the method used, match-marks for turn-of-nut, gaps for DTIs, sheared splines for TC bolts, with the pre-installation verification on record. Welded connections are inspected by the CWI: visual on all welds, plus UT or other nondestructive testing on the complete-penetration welds the code requires. The faying surfaces, the bolt and washer arrangement, and the weld profiles all get looked at against the drawings.

The framework is set by AISC 360 and the building code's special-inspection provisions (commonly Chapter 17 of the IBC), with the inspection scope and the acceptance criteria in the project specifications and the structural drawings. The exact requirements, who inspects, and what counts as a passing joint are the engineer's and the AHJ's to define, so verify them against the contract documents and the adopted code edition. Inspect the connections before the load comes on, not after the deck is poured and the frame is buried in the building.

The erection sequence

The erection sequence is the engineered order in which the frame goes up, and it is built to keep the structure stable at every stage, not to suit the crane's convenience. Steel is erected bay by bay, completing and bracing a section before opening the next, so that a stable, braced frame always stands ahead of the open leading edge. Random erection, grabbing whatever piece is handy, is how a frame ends up with an unstable run of columns and beams that cannot hold themselves up.

The sequence ties together everything else in this guide. It sets which columns get set and braced first, when the beams close each bay, when the frame gets plumbed, when the temporary bracing goes in and comes out, and when the decking gets laid to give the floors their diaphragm action. The metal deck is part of the stability plan, not just a floor, because the diaphragm it forms is often what the permanent lateral system relies on to brace the frame.

The sequence is the engineer's plan, developed by the erection engineer with the EOR, and it is in the erection drawings and the site-specific erection plan. The field follows it. Changing the order on the iron, to dodge a conflict or pick up the pace, can leave a bay unbraced longer than the plan allowed or load a connection before it was meant to carry. If the sequence has to change, it goes back to the engineer, not around them.

What is a controlled decking zone?

A controlled decking zone, the CDZ, is an area where metal decking is being initially placed and forms the leading edge, in which OSHA allows the decking crew to work under specific controls instead of conventional fall protection, between 15 and 30 ft above the level below. It is the standard's answer to a real problem: the deck is what a crew would tie off to, and at the leading edge it does not exist yet, so the rule trades conventional tie-off for a tightly bounded, access-controlled zone with trained deckers.

The metal deck itself is the floor and roof surface, and structurally it is the diaphragm that ties the frame together and braces it against lateral load once it is attached. That is why decking is part of the erection sequence and the stability plan, not a finishing trade that follows behind. The deck gets laid, then placed and finally fastened, and only fully-attached deck counts as a working surface.

OSHA bounds the CDZ hard. It applies only over 15 and up to 30 ft, only where decking is initially being installed at the leading edge, and access is limited to the deckers placing and securing the deck. The zone is no more than 90 ft wide and 90 ft deep from any leading edge, and no more than 3,000 sq ft of unsecured decking is allowed at one time. Final deck attachment and the installation of shear connectors are not done in the CDZ, because that work is no longer leading-edge placement. The fall hazard is the reason for every one of those limits. Read the CDZ rules in 1926.760 and confirm them against the current standard.

What triggers fall protection in steel erection?

Fall protection in steel erection is triggered at 15 ft above a lower level, under OSHA 1926.760, which is higher than the 6 ft trigger that applies to most other construction work and is specific to the steel trades. Above 15 ft on a walking or working surface with an unprotected edge, the worker is protected by guardrails, safety nets, a personal fall arrest system, a positioning device, or a fall restraint system. Falls are the number-one killer in steel erection, so this is the rule that is never relaxed by field judgment.

Two roles get specific treatment because of how they work. Connectors, the ironworkers receiving and securing members up in the frame, between 15 and 30 ft are required to be provided with a personal fall arrest or positioning system and to wear the equipment needed to tie off, or be given other means of protection. Deckers in a controlled decking zone, also between 15 and 30 ft, work under the CDZ controls. Above 30 ft, or two stories, conventional fall protection applies to both.

The perimeter of the working deck gets a perimeter safety cable system, the wire-rope guardrail strung around the open edges of the floor as the steel goes up, so the edge is protected before the trades behind the iron arrive. Fall protection in steel erection is non-negotiable, and the specifics, the trigger heights, the connector and CDZ provisions, and the perimeter cable, are set by 1926.760 and the site-specific fall-protection plan. Confirm them against the current standard and the AHJ. The hazard does not get hedged.

Struck-by hazards and working under loads

After falls, the steel job kills by struck-by: a member or a load coming down on someone, or swinging into them. The hard rule is that nobody works or stands under a suspended load. The load path the crane swings, from the lay-down to the connection, is kept clear of people, and the connector takes the piece at the point of connection rather than anyone reaching under it on the way.

Loose material and tools fall too. A bolt, a drift pin, a spud wrench dropped from up in the iron is a projectile by the time it reaches grade, so falling-object protection and a hard hat are not optional, and the area below active erection is controlled. Subpart R addresses falling-object protection directly, and the exclusion zones for the crane and the overhead work keep the people out from under the hazard.

The connection between this and the crane work is total: most struck-by exposure on a steel job is a crane and rigging exposure, the dropped or swinging load, the load over people. The exclusion zone, the tag line that controls the piece, and the discipline of keeping clear of the swing are covered in the crane, rigging, and signaling guide. On a steel job the rule is short. Do not put yourself, or let anyone, under the load.

The engineer of record and the erection plan

The engineer of record (the EOR) designs the structure and its connections; the erection engineer designs how it gets safely raised, including the stability of the partly-built frame, the temporary bracing, and the erection sequence. The site-specific erection plan pulls it together: the sequence, the bracing scheme, the crane and rigging plan, the fall-protection plan, and the connection requirements. These are engineering decisions, and the field executes them rather than originates them.

This is where the line between the field and the engineer has to be sharp, because the temptations to cross it are constant and the consequences are structural. How many anchor bolts, whether a column needs guying, how long a bay can stand unbraced, when temporary supports can be removed, whether a mislocated anchor bolt can be repaired, what makes a connection complete: every one of those is engineered, and every one of them has been the cause of a collapse when someone in the field decided it on their own.

The stability calculation, the connection design, and the sequence are not the erector's to change. When the steel does not match the drawings, when an anchor bolt is wrong, when a connection cannot be made as detailed, the answer is to stop and go to the EOR and the erection engineer, not to engineer a fix on the iron. Hedge the structural calls to the engineer, every time, and keep the plan on site where the crew can work to it.

Survey and erection tolerances

Erection tolerances are the allowable limits on how far the finished frame can be out of plumb, out of level, and out of line, and they are set by AISC 303, the Code of Standard Practice. They exist because no frame goes up perfect, and the structure was designed to work within a defined envelope of position. Survey is how the crew proves the frame landed inside that envelope.

The numbers a crew carries are the plumbness limit and the alignment limits. Column plumbness is commonly held to 1/500 of the distance between working points, which is the figure the plumbing-up crew tunes to with the guys and turnbuckles. Anchor-bolt placement, member length, and frame line all have their own tolerances, some from AISC 303 for erection and some from ASTM A6 for the mill material. The crew shoots the frame with a total station or theodolite, compares it to the working points, and pulls anything outside the limit back in before the connections are finalized.

The as-built survey is the record that the frame is where it is supposed to be. It also catches the foundation problems that surface as steel goes up, the anchor bolts set off location, the pier a half-inch low, the things that came from the concrete and now show in the steel. The specific tolerances, and what happens when the frame is outside them, come from AISC 303 and the engineer, so verify against the project documents rather than a remembered number.

Weather and site conditions

Weather stops steel work, and the call is not optional bravado. Wind is the first limit: the crane has a wind speed at which the pick stops, set by the manufacturer's chart and the load, and a member flying in gusty wind is a load nobody on the ground can control with a tag line. When the wind hits the limit, the pick is done for the day, not muscled through.

Ice and rain change the iron from a working surface to a skating rink. Steel beams are narrow and unforgiving dry, and slick with ice or rain the fall hazard climbs even with fall protection on. Cold affects the welding, where preheat and procedure assume conditions the weather may not give, and lightning clears the iron entirely. The connector working at the leading edge is the most exposed person on the job, and the conditions that connector works in are part of the safety plan, not a personal toughness test.

The limits come from the crane manufacturer's chart, the welding procedure, the fall-protection plan, and the site-specific erection plan, with the competent person making the call on the ground. When the weather says stop, the frame waits. It is more stable braced and left alone than it is with a crew pushing picks through conditions the equipment was never rated for.

What to document

The erection record is what proves the frame was built and inspected the way it was designed, and it is what answers the question years later when someone opens up a connection or adds load. The pieces that matter are the erection plan and sequence that were followed, the foundation and anchor-bolt survey, the bolt and weld inspection reports, and the as-built survey showing the frame inside tolerance.

Each of those carries its own detail. The bolt inspection records the joint type, the method, and the verification for each connection. The weld inspection carries the WPS, the welder qualification, and the CWI's visual and NDT results. The anchor-bolt and as-built surveys carry the positions against the working points. Capturing these as the work happens, on a field tool like FieldOS, keeps the record tied to the bay and the day instead of reconstructed from memory at closeout. The inspection that nobody can find is the inspection that did not happen, as far as the next engineer is concerned.

Common mistakes

• Erecting without temporary bracing or a stability plan, and letting the open frame run ahead of what is braced.
• Setting a column on fewer than four anchor bolts, or on a foundation that is not cured, surveyed, and ready.
• Repairing, bending, or field-modifying an anchor bolt without the engineer of record's approval.
• Finalizing connections before the frame is plumbed, locking in a racked frame.
• Treating a snug-tight joint as good enough where the design called for pretensioned or slip-critical.
• Pretensioning by torque alone with no verification of actual bolt tension.
• Loading the frame before the bolt and weld connections have been inspected and accepted.
• Working or standing under a suspended load, or letting anyone into the swing.
• Removing temporary guys or bracing early, before the permanent lateral system is in and capable.
• Changing the engineer's erection sequence on the iron to save time.

Field checklist

Standards and references

OSHA 1926 Subpart R is the steel erection standard and the controlling safety document on a steel job. Column anchorage and the four-bolt minimum are in 1926.755, with the requirement that anchor rods not be repaired, replaced, or field-modified without the engineer of record's approval. Structural steel assembly and the securing of members, including the two-bolt minimum before releasing the load, are in 1926.754 and 1926.756. Fall protection, the 15 ft trigger, the connector provisions, and the controlled decking zone are in 1926.760. Crane work on the same site falls under Subpart CC.

The structural side comes from AISC and RCSC. AISC 360 is the specification for structural steel buildings, and AISC 303, the Code of Standard Practice, sets the erection tolerances and puts the temporary support of the frame on the erector working from the owner's designated representative's information. The RCSC specification, referenced through AISC, defines snug-tight, pretensioned, and slip-critical joints and the installation and verification methods. Field welding follows AWS D1.1, the structural welding code, with inspection by a certified welding inspector. Connection inspection is special inspection under the building code, commonly Chapter 17 of the IBC.

Three things hold above all the citations. Stability during erection governs the whole job, so brace it and sequence it to the engineer's plan. The column starts on four anchor bolts and a ready, surveyed foundation, and every connection gets inspected before the load comes on. Fall protection and crane safety are not negotiable. The section numbers and the exact requirements shift between code editions and interpretations, so confirm them against the adopted standards, the AHJ, the project specifications, and the engineer of record before acting on any one of them.

Units, terms, and conversions

Steel erection carries its own vocabulary, and the same idea reads differently across a fabrication drawing, an erection drawing, and the OSHA standard. The terms below are the ones a crew and an inspector use on the iron.

Steel erection

The field process of picking, setting, connecting, and plumbing steel members into a frame, regulated under OSHA 1926 Subpart R

OSHA Subpart R

29 CFR 1926 Subpart R, the steel erection standard covering anchorage, stability, connections, and fall protection

4-bolt rule

1926.755 requirement that each column be anchored by a minimum of four anchor rods (anchor bolts), with a 300 lb eccentric-load strength criterion

Base plate, leveling, grout

The steel plate at the column base, set to elevation on leveling nuts or shims, with non-shrink grout packed under it for full bearing

Plumbing-up

Pulling the frame true with guys and come-alongs to within AISC 303 tolerance before connections are finalized

Snug-tight vs pretensioned / slip-critical

Snug brings plies into firm contact; pretensioned and slip-critical bolts are tightened to a high specified tension, with slip-critical adding a faying-surface requirement

Controlled decking zone (CDZ)

A bounded leading-edge area, 15 to 30 ft up and no more than 90 by 90 ft with under 3,000 sq ft unsecured, where trained deckers place metal deck under specific controls

Erection sequence / EOR

The engineered order of raising the frame to keep it stable; the engineer of record designs the structure and connections, the erection engineer the means of raising it

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is a footing?
• What is a qualified rigger?
• What is precast concrete?
• What is a special inspection on a construction project?
• What is the difference between a snug-tight and a pretensioned joint?