How to Calculate Rebar Welding Capacity: Butt Welds & Fillet Splices

Two #6 bars meet in the middle of a column cage, the lap zone is too congested to tie, and the detailer has written "weld splice — see AWS D1.4" on the drawing. What does that weld actually have to carry, how long must it be, and which electrode goes in the stinger? Rebar welding calculation answers all three with two short formulas: the weld's working capacity F = 0.6 × E × (throat geometry) and the bar's yield capacity T = A_bar × fy. Put the two side by side and the governing side of the splice falls out immediately — usually the weld on a simple butt joint, almost never on a two-sided fillet. The calculator above runs both checks live for bars #3 through #8, Grades 40–75, E60 or E70 electrodes; the sections below walk every number, every minimum, and the AWS D1.4 rules that keep the weld from cracking six months after the pour.

The Welding vs Lap Splice Decision

A lap splice is the default splice in rebar work because it needs no skill and no equipment: lay the two bars alongside each other for the required lap length and tie them, and concrete bond does the rest. The price is steel. A lap runs 30–60 bar diameters depending on the code path, class of splice, and concrete strength, which on a #6 means roughly 22–36 inches of doubled bar — steel that occupies space, congests the joint, and carries load in only one of its two layers. The full arithmetic is in how to calculate rebar lap length, and it is the same tension-development math that governs every other bond issue on the site.

Welding attacks the problem from the other direction: the force crosses through weld metal instead of concrete bond, so the bars need no overlap at all. That saves the lap steel and the congestion, and it buys a splice whose capacity you can compute on a napkin — which is exactly what the calculator at the top of this page does. What welding costs is skill, procedure, and verification, and the comparison table below is where every contractor decides which currency they would rather pay.

Read the last row first, because it settles most jobs: if the bars are #6 or smaller and the joint is not congested, the lap wins on pure economics and the welded splice is a specialty tool. Where welding earns its keep is the set of problems laps cannot solve — a #8 column cage where a 40-inch lap doubles the steel in a 24-inch-wide column, a splice to a structural steel plate where there is no second bar to lap with, and repair work where the original bar is already cast in place.

Three Weld Types Used on Rebar

AWS D1.4, the structural welding code for reinforcing steel, recognizes several joint details, but three of them do essentially all the work on buildings and bridges. Knowing which one the drawing means matters, because each has a different capacity basis — and only one of them involves a weld that runs along the rebar rather than across it.

The Weld Is a Structural Member

Every one of these details makes the weld itself a load-carrying member with a computable capacity — which is the whole point of a rebar welding calculation. The butt weld carries tension through its full cross-section, the two-sided fillet carries it through two shear throats, and the slot weld carries it through the slot. Once the geometry is known, the capacity is arithmetic; the craft is in making the deposited metal actually match the assumption.

Butt Weld Capacity: F = 0.6 × E70 × A_groove

The 0.6 factor is the same derating structural welding has used for decades: an electrode rated 70 ksi in tension is trusted to only 0.6 × 70 = 42 ksi as a working stress through the weld throat, a margin that covers shear-on-the-throat behavior, residual stresses, and the reality of field-deposited metal. For E70 filler the effective design strength is therefore 42 ksi, and the butt weld's working capacity is simply 42 ksi times the area it actually fuses.

Run the defaults on the calculator: a #6 bar (0.44 sq in), Grade 60, E70 electrode. The weld works out to F = 0.6 × 70 × 0.44 = 18.48 kips, while the bar itself yields at T = 0.44 × 60 = 26.4 kips. The weld governs — it carries only about 70% of what the bar can deliver. That is the honest teaching point this guide keeps repeating: a simple butt check on the 0.6Fe_xx basis does not develop the bar. This is exactly why practical full-strength welded splices are not flush beads on square-cut ends — they are complete-joint-penetration bevel groove welds, ends prepared with an angle, rooted, and filled with matching-strength filler, detailed per AWS D1.4 so the splice develops at least 125% of the bar's yield as ACI requires for welded splices in tension. The 18.48-kip number is the working check that tells you the direction of the problem; the CJP detail is the solution.

Fillet Lap Weld Capacity: Two Sides of Support

The 0.707 is geometry, not empiricism: a fillet's strongest plane is its 45-degree throat, and the throat of a right-triangle fillet equals the leg times cos 45° = 0.707. So a 3/8-in fillet has an effective throat of 0.375 × 0.707 = 0.265 in, and every inch of weld length puts 0.265 sq in of weld metal into shear at the 42-ksi effective stress of E70 filler. Two bars lapped side by side get welds on both flanks — a flare-bevel detail lets the bead ride up around the rebar circumference — and both throats count, hence the × 2.

Now the calculator's fillet default: #6 bar, E70, 3/8-in legs, 4 inches per side. F = 0.6 × 70 × 0.707 × 0.375 × 4 × 2 = 89.08 ≈ 89.1 kips, against the bar's 26.4 kips of yield capacity. The bar governs by a factor of more than three — the two-sided fillet is enormously strong for its size, and the tile flips to BAR GOVERNS the moment you switch weld types. This is the field reality of the weld along rebar calculation capacity question: fillet splices almost never fail on capacity, so their design is governed by minimum length rules, bar spacing for electrode access, and the geometry of the joint — not by the weld's strength, which you effectively cannot exhaust on a single bar.

Minimum Weld Length: 5 Bar Diameters and 1 Inch

AWS D1.4 sets a floor on how short a rebar weld may be: the effective length must be at least 5 bar diameters and at least 1 inch, whichever is greater. The reason is stress concentration rather than average stress — load entering a weld funnels through its ends, where the stiffness of the bar meets the compliance of the bead, and a short weld concentrates that transfer into a fraction of an inch of metal with a crater at each end. Short welds crack at the craters; it is the signature failure of undersized splices, and it is why the code writes the minimum as a length requirement instead of trusting capacity math alone.

Notice that 5 diameters beats 1 inch at every bar size from #3 to #8, so in everyday work the 5d rule is the rule. The calculator carries it underneath the fillet slider for exactly this reason: if you drag a #6 fillet weld down to 3 inches, the panel still flags the AWS minimum of 3.75 in. Capacity says the weld is fine; the code says the ends will crack — and the code has the crash history on its side.

Electrode Selection: Matching Filler to Bar Grade

The electrode is not an accessory — it is the filler metal that becomes the splice, and choosing it wrong quietly downgrades the joint. E6013, a 60-ksi general-purpose rutile rod, is acceptable for older Grade 40–50 bars on non-critical work where the weld is not asked to develop the bar. E7018 low-hydrogen is the standard choice for Grade 60 rebar and the required class for A706 weldable-grade bars: the 70-ksi filler overmatches the 60-ksi stock, and the low-hydrogen coating is what keeps the joint from cracking.

Why does hydrogen matter so much on rebar specifically? Because reinforcing steel lives outside, drinks humidity, and often gets welded damp at the restart of a shift. Hydrogen dissolved into hot weld metal escapes as it cools — except where it is trapped in the hardened heat-affected zone, where it collects at inclusions and grows cracks under residual stress, days after the welder has packed up. Low-hydrogen electrodes have baked coatings that emit almost no hydrogen; keep opened cans in a heated rod oven and discard rods that have gone soft, because a damp E7018 is worse than a dry E6013.

Preheat and the AWS D1.4 Essentials

Large bars — roughly #9 and up — and cold-weather welding need preheat, commonly in the 250°F class for big sections, so the heat-affected zone cools slowly enough to stay ductile. Beyond temperature, AWS D1.4 asks for four things on every job: a qualified weld procedure specification (WPS), welders qualified to it, 100% visual inspection of every rebar weld, and chemistry verification before welding standard A615 bars — the rolled and rail-steel grades whose carbon equivalent is unknown unless the mill cert says otherwise. A706 is the weldable grade precisely because its chemistry is capped at the mill, which is why the smart move is to specify A706 up front whenever the design anticipates welded splices.

One more grade note for completeness: when the design mixes rebar with structural steel — dowels welded to beam flanges, bars to embed plates — the electrode must satisfy both materials, which usually means the E70XX class and a joint detail the fabricator signs off on. The capacity math on this page still applies; the weld around rebar calculation simply uses the weakest of the connected parts as its target capacity.

Worked Verification: Two #6 Grade 60 Bars, Butt Weld

Set the calculator to its defaults and follow the same numbers on paper — this is the check an inspector or a detailer runs before approving the splice detail.

When Does Welding Beat a Lap Splice? The Break-Even

Put dollars on the two methods and the decision stops being philosophical. A lap splice on a #6 runs roughly 30 inches of doubled bar: 2.5 ft × 1.502 lb/ft ≈ 3.76 lb of extra steel, about $4–5 of material per splice at everyday bar prices — and it consumes zero skill, because any rodman can tie it. A butt weld needs an AWS D1.4-qualified welder, electrode time, and inspection: call it $12–20 per splice in labor and consumables on a production basis. Head to head on a single small-bar splice, the lap is cheaper by a wide margin.

The economics flip as bars grow and counts climb. Lap length scales with bar diameter, so the extra steel per splice grows roughly with the square of the bar number — a #8 lap burns more than four times the steel of a #5 lap at the same code minimum. Meanwhile the welder's time barely changes: a #8 butt weld takes minutes more than a #5, while the lap it replaces costs over 13 lb of steel. The break-even sits around #7 and larger bars, and welding pulls clearly ahead where splices repeat by the hundred — walls and columns with hundreds of splices, where saved lap steel compounds across the floor plate and the welder's productivity amortizes over the run. Small bars and one-off splices: lap. Big bars, tight cages, congested joints, and anything that must splice to structural steel: weld.