Concrete Slab for a Metal Building

Why it is not just a thicker garage slab

Alexander Newman, writing in STRUCTURE, puts the weight of a single-story metal building system at 2 to 5 pounds per square foot, and notes that gable rigid frames "exert significant horizontal column reactions on the foundations." He also points out the rule that makes the uplift problem bite: codes permit no more than 60 percent of the dead load likely to be in place to be counted against wind uplift. Sixty percent of 2 to 5 psf is 1.2 to 3 psf, which is almost nothing. A building this light is in net uplift under design wind in much of the country while simultaneously pushing its column bases apart.

Compare that with what the site's other slab pages are about. A garage slab's problem is a downward wheel load, and ACI 330R Table 2.4 says 3-1/2 to 4 inches handles passenger vehicles across the whole range of subgrade support. A hot tub pad's problem is a concentrated static load. A metal building slab's problem is horizontal, and thickness does not answer a horizontal problem.

On the magnitude, we are going to decline to give you a number, and here is why. No manufacturer publishes a reaction schedule for a homeowner-scale building keyed to width. Nucor publishes endwall column base reactions on a stated basis of 25 foot bays, 40 to 60 foot spans and a 1:12 roof slope, and on those tables a single endwall interior column at 30 foot spacing reaches 15,000 lb down at 40 psf live load and 11,200 lb of uplift. Those are endwall columns, not main rigid frames, under an older generation of wind provisions, and reading them as general values would be wrong. The horizontal figure for your building is on your own reaction schedule, and Newman's own assessment of the whole subject is that the design procedures "are often not well understood, because they are not specified in the building codes and technical design guides."

For scale rather than for design: 11,200 lb of uplift at one bolt group, held down by nothing but the self-weight of a 4 inch slab at 50 lb per square foot, would need 224 square feet of slab per corner. That is our arithmetic on Nucor's published figure, not a design method, and it is here only to show why the anchorage is not a detail.

How the outward thrust gets resisted, cheapest first

Newman ranks the options, and the ranking runs opposite to the price:

1. Hairpins tied into the slab on ground. The "least expensive method of resisting horizontal column reactions," and the one that "suffers from total reliance on floor slab."
2. Tie rods between opposing columns, "extinguishing" both horizontal reactions. These run "from the cheapest and least reliable (reinforcing bars placed in the slab) to the relatively expensive and much more reliable (concrete grade beams)."
3. Moment-resisting foundations, which behave like cantilevered retaining walls. "One of the most reliable systems available."
4. Trench footings.

One practical warning travels with the hairpin and friction options, from an engineering practice that designs these: if the slab is later "modified in the future by cutting or removing, the opposing friction forces can be unknowingly and detrimentally reduced." Someone saw-cutting a trench for a drain in ten years will not know the slab was holding the building together.

And here is the finding that should change what you pour. In a second STRUCTURE article Newman sets out what the slab actually becomes if hairpins are used. Under ACI 318-19 section 18.13.3.2, where those columns form part of the seismic-force-resisting system in Seismic Design Categories C through F, the slab-on-ground "must be designed and detailed as concrete diaphragms with a load path." That means a minimum steel area of 0.0018 times the slab's gross area, a maximum bar spacing of 12 inches, deformed bars rather than smooth welded wire fabric, reinforcement continuous through the joint lines, and a positive connection to the perimeter foundation.

Now put that next to standard flatwork practice, which is also on this site. NRMCA CIP 6 puts control joints at 24 to 36 times the slab thickness with a 15 foot cap, which on a 4 inch slab is every 8 to 12 feet, and calls for isolation joints where slabs meet walls, footings or columns. The diaphragm case wants reinforcement running straight through those joints and the slab tied to the perimeter. Newman is explicit that a slab "without isolation and control joints or where the deformed reinforcement continues through the joints (as in a structural slab) is vulnerable to random cracking."

Those two requirements cannot both be satisfied. So there are two honest ways to build this, and you have to pick one on purpose:

• A non-structural floor, jointed and isolated the normal way, with the thrust resisted entirely outside the slab by tie rods designed as grade beams or by moment-resisting footings. Newman's own recommendation.
• A structural diaphragm, with #4 bar at 12 inches each way satisfying both the 0.0018 area and the 12 inch spacing cap, continuous through the joint lines, tied to the perimeter, and accepting that it will crack where it wants to.

What you should not do is pour a garage slab, let the erector grout hairpins into it, and assume the two things add up. No source we could find publishes a hairpin bar size, leg length or spacing, and we are not going to invent one. If your design relies on hairpins, that schedule comes from an engineer.

Anchor bolts, the part you cannot fix afterwards

Everything else on this page is recoverable. Anchor bolts in the wrong place are not, once the concrete sets.

The tolerance nobody reconciles for you

The steel vendors ask for tighter than the concrete standard allows, and the gap is a factor of four to eight.

ACI 117-10 is the standard your concrete contractor works to, and on a 2 inch rod it permits eight times the deviation the kit seller asks for. The American Society of Concrete Contractors published a position statement precisely because of this and offers to mediate "tolerance conflicts between Division 3 and Division 5." On a commercial job there is a Division 3 and a Division 5 to mediate between. On your job there is one concrete contractor and a steel crew arriving in three weeks.

Two notes on the history, because the numbers circulating online are stale. ACI 117-90 allowed a flat 1 inch, which is why older advice is so relaxed. And in 2016 AISC 303 is reported to have adopted the diameter-based ACI 117 figures, so the widely quoted 1/8 inch is the pre-2016 AISC number rather than the current one. We could not read AISC 303-16 section 7.5.1 directly, so treat the current AISC figure as the harmonized ACI one rather than 1/8 inch.

The practical answer: agree the tolerance in writing before the pour, and tell your concrete contractor it is tighter than his standard. The published price of getting it wrong is a $500 to $1,500 drilled-anchor retrofit.

Embedment, and why it forces a thickened edge

AISC Design Guide 1 gives the minimum embedded length of an anchor rod as 17 times the rod diameter, and that is one term in a formula rather than the whole answer. Rhino publishes 20 times the diameter as a rule of thumb; one vendor content site says 25 times. For a 3/4 inch rod those give 12.75, 15 and 18.75 inches.

Every one of those exceeds a 4 inch slab, and a 6 inch slab, by a wide margin. An anchor rod for a rigid frame column cannot be developed in a slab. It needs a pier, a thickened edge or a footing under it. That single piece of arithmetic is the reason a metal building slab has a perimeter haunch, and AISC's own minimum establishes it without needing a vendor to say so.

The real embedment is a calculation under ACI 318 Chapter 17 against your own reaction schedule, which is exactly what every manufacturer says is not theirs to do. Treat 17d as the floor below which nothing can be right, not as the answer.

What the standards and the vendors publish

• Material. ASTM F1554 Grade 36 is the default at Nucor and Miller; Valor and Northern State allow A307. AISC Design Guide 1 prefers 1/2 inch diameter F1554 Grade 36 rod wherever possible and suggests staying with Grade 36 up to about 2 inches before reaching for higher strength.
• Count. OSHA's steel erection standard requires a minimum of four anchor rods in a column base plate connection, excluding post-type columns under 300 lb. Rhino states the same.
• Shape and angle. Rhino: a curved L or J shape, and "perpendicular to the concrete surface at exactly 90 degrees."
• Projection. Nucor publishes 3 inches above the bottom of the base plate, before grout. Mueller says 2-1/2 inches of thread above the concrete, Valor at least 2 inches. One small kit ships 5/8 by 14 inch J-bolts with a 2 inch projection specified.
• Washers. Design Guide 1 Table 2.3 pairs each rod with an oversized hole and a minimum plate washer: a 3/4 inch rod takes a 1-5/8 inch hole and a 3-3/8 inch washer. Plate washer thickness is about a third of the rod diameter, and they do not need to be hardened.

That washer figure is worth pausing on. A 3/4 inch rod set near a slab edge has a 3-3/8 inch washer bearing on the plate, an oversized hole, and the concrete breakout cone that ACI 318 Chapter 17 governs, all competing for the same few inches of edge. We found no published minimum concrete edge distance for a metal building anchor rod, and Design Guide 1's "2 inch" figure is a plate dimension rather than a concrete one, so we are not going to print a number. The mechanism is clear enough without one.

Templates and bolts: nobody supplies either

Every erection manual we read tells the contractor to make a template. None says one ships with the kit. Buck Steel: "all anchor bolts should be held in place with a template or similar means, so that they will remain plumb and in the correct location during placing of the concrete." Simpson Steel, in capitals: "Never place anchor bolts by hand, use template MADE TO SCALE to assure accuracy." Valor and Northern State both specify drilling 1/8 inch air relief holes in the template so trapped air can escape.

And the bolts themselves are not in the box. Nucor: "anchor bolts and nuts are not supplied by Nucor as part of the building order." Miller: "anchor rod will not be supplied by Miller." Northern State, again in capitals: "does not provide anchor bolts or engineering for foundations." Rhino notes the concrete contractor usually buys them locally. Budget for them and order them against the setting plan, not against the general arrangement drawing. Nucor is explicit that its own anchor bolt setting plan "MUST be used in lieu of the Architect's."

One more sequencing item that catches people: Buck Steel requires that normal Portland cement concrete cure at least seven days, or high-early-strength at least three, before the structural columns are erected. If the steel crew is booked for the Monday after the Friday pour, move one of them. Concrete cure time covers what is actually happening in those seven days.

Thickness, strength and steel, and who is actually telling you

Be clear about the sourcing here, because it is weaker than anywhere else on this site. No code, ACI, NRMCA or MBMA document publishes a slab thickness or a concrete strength for a metal building. Every figure below comes from a company selling steel buildings.

If the building is a residence rather than a shop, much of the above changes: a barndominium slab is permitted as a dwelling, which brings the vapor retarder and the full-area insulation requirements with it, and most barndominiums are post-frame rather than steel so the thrust problem does not arise at all.

Two of those need correcting against standards the site already carries.

First, the strength figure is low and it is missing the thing that matters. IRC Table R402.2 puts a weather-exposed slab at 2,500, 3,000 or 3,500 psi by weathering potential, and the industry sits above the code floor: the Michigan Concrete Association specifies 4,000 psi with 6.5 percent air and a water-cement ratio of 0.45 or less, Ohio Concrete 4,500 psi with 6 percent air, Iowa DOT 4,000 psi with a 7 percent air target. The vendors' "3,000 to 4,000 psi" is therefore at or below consensus for exterior flatwork, and not one metal building source mentions air entrainment at all, which in freeze-thaw country matters more than the psi. What psi concrete you need has the full comparison. Specify the air content yourself, because the vendor will not.

Second, one source puts the reinforcement "2 to 3 inches from the bottom surface," and on the lower reading that is wrong. IRC R506.2.4 requires steel to sit from the center to the upper third of the slab for the duration of the placement, and ACI 302.1R puts shrinkage and temperature steel in the upper third because cracks originate at the surface and are widest there. In a 4 inch slab, steel near the bottom does nothing useful. Rebar, mesh and vapor barrier has the chair spacing and the cover rules.

What the site's own benchmarks give you, which are not written about metal buildings but are the right yardsticks: the IRC minimum slab thickness is 3-1/2 inches (R506.1), ACI 330R says there is "no benefit in building slabs less than 3-1/2 in. thick," and bending capacity rises with the square of thickness so going 4 to 6 inches buys 1.5 times the concrete and about 2.25 times the bending resistance. How thick should a concrete slab be works through it. And if the steel comes with a reaction schedule, the thickness in the field is the easy question; the detail under the columns is the hard one.

On vapor barriers, the residential rule still applies: IRC R506.2.3 exception 1 covers "garages, utility buildings and other unheated accessory structures," so an unheated steel storage building needs no retarder under the code. Where one is required, since the 2021 IRC it is 10 mil meeting ASTM E1745 Class A with 6 inch laps, not the 6 mil poly older editions allowed. One thing changes if your building is permitted as commercial rather than residential: under the 2018 IECC, an unheated slab-on-grade in climate zones 4 and 5 needs nothing residentially but picks up R-10 perimeter insulation to 24 inches under the commercial table C402.2.4. A conditioned shop permitted commercially therefore carries a slab edge insulation requirement the identical slab under a detached residential garage does not. Under-slab insulation has the residential tables.

The notch at the edge, and sealing the panel to the slab

This is the part that exists only on a metal building, and it is the part that decides whether water and mice get in.

A corrugated wall panel sits on the slab as a row of open flutes. Three published things close it. A closure strip, which Miller Buildings specifies as "solid or closed-cell, preformed rubber, neoprene or polyethylene," fills each flute. Tape sealant seals the lap, though note that the 1/2 by 3/32 inch figure Miller publishes is for roof panels and we found no published dimension for the base of the wall. And a base angle or base trim holds the panel off the concrete, which Buck Steel says plainly is to prevent "the bottom of the wall panel from coming in direct contact with the concrete, reducing the risk of wall panel corrosion."

The notch is the near-consensus detail. Buck Steel publishes a 1-1/2 by 1-1/2 inch notch with a 2 by 4 inch base angle, and has the slab 3 inches wider and 3 inches longer than the steel building so the notch has somewhere to be. Valor Steel publishes an identical 1-1/2 by 1-1/2 inch notch mold. Wolf Steel calls it a 1-1/2 inch drop that "overlaps with the metal building panel," gives the reasons as water, pests and appearance, says the slab should slope away from the building, and adds the warning worth repeating: a notch detail "is not a one-size-fits-all solution."

On the overall footprint the sources disagree and you should go by your own anchor bolt plan. Buck Steel wants the slab 3 inches larger each way. Worldwide Steel says the width usually aligns with the building but "the slab length often stops short of the roofline." Wolf Steel says it varies with the frame type and is settled in planning.

Note this is a deliberate departure from what we tell people building a shed. The shed slab page argues for pouring exactly to the footprint, so roof runoff lands on soil rather than on a splash ledge. A metal building reverses that logic, because the panel sits in a rebate at the slab edge and the slab has to be slightly larger than the steel line for the rebate to exist. Then you slope away from the building to deal with the runoff. The slope calculator has how much fall and what it costs in concrete.

Simpson Steel fixes its 4 by 2 inch base angle to the slab with nails or expansion anchors at 24 inches on center. Ceco's list of what has to be embedded in the foundation is worth reading before the pour: "bolt embedment, bearing angles, tie rods and/or other associated items." Bearing angles and tie rods go in during the pour or not at all.

Control joints

Nobody publishes a joint layout for a metal building slab, so the answer depends on which of the two cases above you are in.

If the thrust is handled outside the slab, use the normal flatwork rules the site already carries from NRMCA CIP 6: joints at 24 to 36 times the slab thickness with a 15 foot cap, which is every 8 to 12 feet on a 4 inch slab and 12 to 15 feet on a 6 inch slab; panels no longer than 1.5 times their width; the groove at least a quarter of the thickness and never under an inch; and saw cutting inside 4 to 12 hours. Isolation joints where the slab meets the perimeter foundation and around any column pier.

If the slab is doing the structural work, those rules do not apply and you need the reinforcement running through. That is a designed slab, not a jointed one.

One detail from the garage page that transfers directly: BendPak requires lift anchors to sit at least 6 inches from any joint. The same logic applies to an anchor bolt group near a saw cut, and it is a reason to lay out the joints against the anchor bolt plan rather than after the fact.

What it costs

No source publishes a measured premium for a metal building slab over a plain one. The ranges vendors print, $4 to $8, $6 to $12 and $7 to $14 per square foot, are the same ranges published for ordinary slabs. The only route to a premium is to add up the published adders, all from one publisher: rebar instead of mesh at $0.50 to $1.20 per square foot, each extra inch of thickness at $1.25 to $1.50, and frost-depth perimeter footings at $1.00 to $2.00 per square foot equivalent. Taking all three is roughly $2.75 to $4.70, and that is one publisher's arithmetic rather than a measurement.

The more useful figure is what the thickened edges do to the concrete order. One vendor puts a 30 by 40 slab's field at about 15 cubic yards and the same slab with thickened edges at 19 to 22. We can confirm the base: 1,200 square feet at 4 inches is 14.81 cubic yards. So on that source's numbers the perimeter haunch adds 28 to 49 percent to the concrete, which matches the warning on the thickness page that a turned-down edge takes noticeably more concrete than area times field thickness suggests. The slab calculator sizes the field and the footing calculator handles the perimeter separately.

Separating the slab from the kit

Most people searching for a cost here are pricing the building and the slab together. One vendor breaks a 30 by 40 down:

On those numbers the slab is 24 to 32 percent of the whole job. Treat the kit figure with caution: another vendor quotes a 30 by 40 "from $28,500," which is 2.6 times the low end above. Almost certainly bare kit against something more inclusive, but neither says so, and we are not going to average them.

An itemized slab breakdown from the same part of the market, for a 1,200 square foot slab: fine grading $800 to $1,600, gravel base $800 to $1,400, vapor barrier $250 to $600, reinforcement $600 to $1,700, ready-mix $2,800 to $5,200, labor $1,700 to $3,200, anchor bolts $250 to $700, and a drilled-anchor retrofit at $500 to $1,500 if the bolts end up in the wrong place.

The number that should settle the argument. The same source prices a PE-stamped foundation plan set, with full ACI 318 anchor bolt design, at $0.40 per square foot for buildings up to 5,000 square feet. On a 30 by 40 that is about $480, against an $8,400 to $16,800 slab and a $26,400 to $56,400 project. One to two percent of the build to have the thing designed by someone whose job it is. Phoenix, for one, requires it in writing: all drawings and calculations "must be signed and sealed by an Arizona registered professional engineer or architect," with the manufacturer's sheet permitted only for the anchor bolt locations and column reactions.

One cost trap this page does not need, unusually for this site: the short load penalty. The smallest building anyone searches for, a 20 by 30, is 7.4 cubic yards at 4 inches, comfortably over every published ready-mix minimum. Pad cost is where that problem lives, and it does not reach you here. Concrete slab cost by size has the per-square-foot curve, and the slab cost calculator prices your own dimensions.

Before you pour

1. Get the reaction schedule and the anchor bolt setting plan from the manufacturer, in writing. Nucor supplies three sets of anchor bolt drawings and reactions with every order. These are inputs, not a foundation design.
2. Hire a local engineer to design the foundation against those reactions, and expect around $0.40 per square foot. Ask specifically how the horizontal thrust is resisted and whether the slab is part of the load path, because that answer decides your reinforcement and your joints.
3. Check whether a stamped plan is required by your jurisdiction, and whether the building is permitted as residential or commercial, because the energy code differs.
4. Agree the anchor bolt tolerance in writing with the concrete contractor, and tell him it is tighter than ACI 117.
5. Buy the anchor bolts, because they are not in the kit, and order them against the setting plan.
6. Build templates to scale with air relief holes. Never place bolts by hand.
7. Confirm what else gets embedded: bearing angles, tie rods, hairpins. These go in during the pour or not at all.
8. Specify the air content, not just the psi, if you are anywhere with a freeze-thaw winter.
9. Decide the edge detail, notch dimension and slab footprint against the anchor bolt plan, not after it.
10. Check the bolts after the pour and before the steel arrives, which is what Buck Steel's manual asks for, and allow seven days of cure before columns go up.

Sources

• Alexander Newman, P.E., in STRUCTURE: Foundations for Metal Building Systems (July 2013) for the building weight, the horizontal reactions, the 60 percent dead load limit and the reliability ranking; and Avoiding Problematic Uses of Slabs on Ground (January 2021) for the diaphragm requirements and the four flawed assumptions behind hairpins.
• ACI 318-19 section 18.13.3.2 and section 24.3.2, for the diaphragm design requirement and the 12 inch maximum bar spacing, quoted via Newman. The 0.0018 times gross area benchmark is on our rebar page from ACI 318 directly.
• ACI 117-10 section 2.3.4 anchor bolt tolerances, and the superseded ACI 117-90 figure, via the American Society of Concrete Contractors Position Statement 14.
• AISC Design Guide 1, Base Plate and Anchor Rod Design, 2nd edition, for the F1554 Grade 36 preference, the 17 times diameter minimum embedment, Table 2.3 hole and washer sizes, and the OSHA four-rod minimum.
• Manufacturer engineering documents, all quoted verbatim: Nucor Building Systems anchor bolts and foundation reactions, Nucor Buildings Group specification guide, Ceco Building Systems, Miller Buildings, Valor Steel, Northern State Steel, Simpson Steel, Buck Steel and its base conditions page, and Rhino Steel.
• City of Phoenix metal buildings structural technical guideline, for the stamped-drawings requirement and what the manufacturer's sheet may cover.
• 2018 IECC section C402.2.4 commercial slab perimeter insulation, via the University of Illinois Smart Energy Design Assistance Center. A reproduction rather than the code itself, which is paywalled.
• Cost figures from SteelReady, SteelBuildingKit, MBMI, AAIRE and Worldwide Steel, all commercial publishers selling into this market.

What we deliberately will not print. A horizontal thrust figure for your building, because no manufacturer publishes one at this scale and the only numbers in circulation are forum posts. A hairpin bar size, leg length or spacing, because no standard or vendor publishes one. A tie rod diameter or spacing, for the same reason. A minimum concrete edge distance for an anchor rod, because ACI 318 Chapter 17 governs it as a calculation and gives no single number. We also could not read the MBMA Metal Building Systems Manual, which is paid, so nothing on this page is attributed to MBMA. The 224 square feet of slab per bolt group, the thickened edge percentage and the #4 at 12 inches recommendation are our arithmetic on the published figures above.