Rebar mechanical splices and couplers field guide

What is a rebar mechanical splice?

A rebar mechanical splice is a coupler that joins two reinforcing bars end to end and transfers force from one bar to the other through the device itself, not through the concrete around it. A lap splice does the opposite. It overlaps two bars and lets the concrete carry the load from one to the other by bond over the development length. The coupler skips that. The bars butt up inside or against a steel sleeve, the sleeve grips both ends, and the force walks straight across the joint.

That one difference drives everything else. With a lap, you need lap length, you need the extra bar that the overlap costs you, and you pile two bars side by side in a section that may already be tight. With a coupler, there is no lap length to find and no doubled-up steel at the splice. You pay for the hardware and the labor to install it instead.

You reach for a mechanical splice when lapping does not work or does not make sense. Large bars where the lap would be enormous. Cages so congested that a doubled bar will not fit or the concrete cannot get around it. Staged construction where there is no second bar to lap to yet, only a dowel waiting for the next pour. Seismic regions where a lap is not permitted in the part of the member that has to yield. For the lap-splice side of this, and how development length and bond actually work, see the rebar development length and lap splices guide. For how the cage gets inspected before the pour buries it, see the rebar placement and cover guide. This guide is about the coupler.

What is the difference between a mechanical splice and a lap splice?

A lap splice transfers force between two bars through the concrete by bond, so it needs overlap length and it adds a second bar through the splice zone. A mechanical splice transfers force through a coupler bar to bar, so it needs no lap length and adds no parallel steel. That is the whole trade in one sentence, and the rest is when each one wins.

The lap wins on cost and simplicity for ordinary bars in ordinary members. Nothing to buy, nothing to torque, no special inspection in most cases, and a rodbuster can build it from the same bar on the same reel. As long as there is room for the overlap and the lap is permitted where it lands, the lap is usually the cheaper call.

The coupler wins when the lap stops being practical. The overlap on a #11, #14, or #18 bar gets long enough that you are buying a lot of extra steel and stacking it in a tight section. Congested columns and mat foundations run out of room for doubled bars and clear spacing both. Staged pours and dowels have no bar to lap to. Seismic special moment frames do not allow ordinary laps in the plastic hinge zones. In each of those, the coupler is not a luxury, it is the only thing that fits or the only thing the code permits.

Do not frame this as one being better. They solve the same problem, continuing a bar across a finite length, by two different mechanisms. The drawings and the engineer pick which one applies where, and the splice schedule on the structural sheets tells you. When the drawing calls a coupler, you do not get to lap it instead because lapping is easier.

Why use a coupler instead of a lap

Four drivers put a coupler on the drawing, and they often stack on the same job. Large bars are the first. Lap length scales with bar size, so the overlap on the big bars in a high-rise column or a heavy transfer girder gets long and expensive fast, and on the largest bars a compliant lap may not be practical at all. The coupler removes the lap length from the equation.

Congestion is the second. In a column with a high steel ratio, or a mat with two heavy mats of bar, a lap means two bars where there was one, and the clear spacing and the room for the concrete to flow around the steel both suffer. The coupler keeps a single bar line through the splice, which opens the section back up for placement and consolidation.

No lap room and staged construction is the third. When you pour a footing or a slab and the columns or walls come later, you cast a coupler or a threaded dowel in the first pour and connect to it in the next. There is no second bar to lap to at the time of the first pour, so the coupler becomes the planned connection point for work that has not happened yet.

Seismic is the fourth, and it is the one with the hardest rule. In the regions of a member that are designed to yield and absorb energy, a plastic hinge in a special moment frame, ordinary lap splices are not permitted, and the splice that is allowed there has to develop the bar's full strength. The coupler, the right type of coupler, is how you continue the bar through a region where a lap cannot go. Continuity across the joint, under load reversal, without slip, is the whole point in that case.

Threaded couplers

A threaded coupler is the most common mechanical splice. The bar ends are threaded, the coupler is an internally threaded steel sleeve, and the bars screw into the two ends like two bolts into one long nut. Force crosses the joint through the threads and the body of the sleeve. Installation is fast, the connection is checkable, and the hardware is compact, which is why threaded systems show up on most jobs that use couplers at all.

Two thread forms dominate. Parallel thread, where the bar end is enlarged or the thread is rolled so the threaded section is at least as strong as the bar, then a straight thread runs the length of engagement. And taper thread, where a cone-shaped thread is cut into the bar end and the coupler tapers to match, so the joint draws tight and self-seats as you torque it. Parallel thread keeps the bar's section at the root by upsetting or forging the end first. Taper thread cuts into the bar, so the system is engineered to keep the threaded net section strong enough.

The detail that decides whether a threaded splice performs is engagement. Both bars have to thread fully into the coupler and meet the manufacturer's torque and engagement marks. A bar that bottoms out short, or threads that are damaged or dirty, gives you a joint that looks done from the outside and slips under load. The position mark painted on the bar at the correct depth is there so the inspector can see, without a wrench, whether the bar went in far enough.

Bar-end preparation for threaded systems

Threaded couplers live or die on the bar-end prep, and that prep happens with the manufacturer's equipment, not with whatever is in the gang box. The bar is cut square first. A skewed cut throws off the thread and the seating, so the saw cut squareness is the first thing that matters and the first thing to go wrong when a crew rushes.

Parallel-thread systems usually upset or forge the bar end before threading, enlarging it so that cutting the thread does not reduce the bar below its full strength. The forged or upset head is formed cold or hot on a dedicated machine, then the straight thread is rolled or cut. Done right, the threaded section is stronger than the bar body, so the bar yields before the thread strips.

Taper-thread systems cut a tapered thread directly into the squared bar end with a die set and a threading machine matched to the bar size. Because the thread is cut into the bar rather than into an enlarged end, the system geometry is engineered so the net section at the thread still carries the required force. Either way, the threading is checked with a thread gauge. A go gauge has to run on to the required length and a no-go gauge has to stop, and a thread that fails the gauge is rejected and recut, not installed and hoped over.

This is the step the field tries to skip when the schedule gets tight, and it is the step that the whole splice depends on. A coupler torqued onto a bad thread is a bad splice with a good-looking outside.

Swaged and cold-pressed couplers

A swaged coupler is a steel sleeve pressed cold onto the bar ends with a hydraulic press. The sleeve slides over the ribbed bar, the press squeezes it down in a series of bites, and the steel of the sleeve flows into and around the deformations on the bar. When the joint is loaded, the dents pressed into the sleeve interlock with the bar ribs, and force crosses the joint through that mechanical grip. There is no thread and no grout, so there is no bar-end machining beyond a clean cut.

The advantage is that it tolerates field conditions well and works as a repair or a tie-in where threading is awkward. The press and the dies are sized to the bar, and the number of presses per side is set by the manufacturer, so a half-swaged sleeve is a rejectable splice. Crews like swaged couplers for retrofit and for joining to existing bar where you cannot spin a threaded sleeve on.

The catch is the equipment. The hydraulic press is heavy, the dies are bar-specific, and in a tight cage there has to be room to get the press around the sleeve and stroke it. Swaging in a congested column corner is harder than it looks on the submittal. Inspection checks the swage count and the sleeve dimensions, and some systems use a saw-cut sample or marked stops to confirm the bar seated fully before the swage.

Grout-filled sleeve couplers

A grout-filled coupler is a steel sleeve, larger than the bar, that the bar ends slide into from each side, with the gap between bar and sleeve then packed with high-strength non-shrink grout. The cured grout keys into the bar deformations and the inside of the sleeve, and force crosses the joint through that bond. It is the splice that does not need the two bars to meet on a thread or a swage, which is exactly why precast uses it.

The reason it dominates precast is tolerance. When you set a precast column on a footing, or stack one precast element on the next, the dowels coming up never line up perfectly with the sleeves cast into the piece above. The grout-filled sleeve is sized to swallow that misalignment, accept the bar over a range of position, and still cure to a full-strength connection. Threaded and swaged couplers want the bars where they are supposed to be. The grout sleeve forgives where they actually are.

Grout-filled sleeves also see heavy use in seismic precast, because the right grout-filled systems are tested to develop the bar's full tensile strength and behave under load reversal. The cost is the grout and the wait. You mix and pump the grout to the manufacturer's spec, you fill until grout flows clean from the outlet port so you know the sleeve is full and not voided, and you do not load the connection until the grout has reached its specified strength. Skip the cure and you have a connection that has not become a connection yet.

Shear-screw and other coupler systems

Shear-screw, or bolted, couplers clamp the bar with a row of hardened set screws or serrated rails instead of a thread, a swage, or grout. The screws bite into the bar surface, and the heads are designed to shear off at a set torque, which gives the installer a built-in signal that the joint reached the right clamping force. The big draw is that there is no bar-end preparation at all. Cut the bar, slide the coupler, run the screws until the heads pop. That makes it quick on site and forgiving of mixed bar sizes within a range.

The trade-off is that performance and seismic suitability depend heavily on the system and the bar, so the evaluation report is doing real work here. Some bolted systems qualify for demanding applications and some do not, and that is a question for the engineer and the listing, not a field assumption.

Welded splices exist as a separate path, joining bars by welding rather than a coupler, and they carry their own rules for weldability of the bar, electrode, procedure, and inspection. Welding reinforcing steel is governed separately and is not interchangeable with screwing on a coupler. Whether a bar can even be welded depends on its chemistry, so welded splices are an engineering and procedure question on their own, outside the scope of this guide.

What is a Type 2 mechanical splice?

ACI 318 sorts mechanical splices into two performance classes, and the difference decides where a given coupler is allowed to go. A Type 1 mechanical splice is required to develop at least 125 percent of the bar's specified yield strength, fy, in tension or compression. A Type 2 mechanical splice meets that Type 1 requirement and, on top of it, develops the specified tensile strength, fu, of the spliced bar. Treat those figures as the code's framework and confirm the exact wording, percentages, and the section numbers against the ACI 318 edition in force and the coupler's evaluation report, because the definitions are the kind of thing that gets refined between code cycles.

Why the second class exists is the seismic reason. A Type 1 splice is fine where the bar is not expected to yield. But in a region designed to undergo inelastic deformation under an earthquake, the splice has to survive the bar going past yield and into the strain-hardening range without failing first. A splice that only reaches 125 percent of fy could become the weak link there. The Type 2, developing the bar's full tensile strength, is what is required in those yielding regions of special seismic systems.

On the job this is a hard line, not a preference. Specifying or installing a Type 1 where the design calls for a Type 2 is one of the more serious mistakes you can make on a coupler job, because the splice will look identical and pass a casual look while being the wrong device for a region that has to perform in a quake. Read the splice schedule and match the type called out. When the type is not clear, it is an RFI, not a guess.

Couplers in seismic regions

In structures detailed for high seismic demand, the rules on splices tighten and the coupler often becomes the required tool. Ordinary lap splices are restricted or not permitted in the regions of beams and columns that are expected to form plastic hinges, because those regions have to yield and dissipate energy under load reversal, and a lap can degrade there. The splice allowed in those regions has to develop the bar's full strength, which points to a Type 2 mechanical splice or the equivalent the code recognizes.

Location and staggering still matter even with the right type. The detailing rules push splices away from the worst of the hinge region and out toward where the demand is lower, and they limit how many bars you splice at one section. The exact zones, the offsets from the joint face, and the stagger come from ACI 318 and the structural drawings for the specific system, so build to the splice schedule and the seismic detailing notes rather than a rule of thumb carried from a non-seismic job.

Grout-filled and high-performance threaded systems both show up in seismic work, in cast-in-place and precast, because the qualified versions are tested under cyclic and tensile load to behave the way the design assumes. The seismic detailing of a structure is its own deep subject and is engineered member by member. The field job is to install the type the drawings call out, in the location they call out, and to not move a splice into a region the detailing kept it out of.

Tension and compression splices

A coupler can be carrying tension, compression, or load that reverses between the two, and the demand on the splice is not the same in each case. In tension the coupler has to hold the bars together against a pull, so the thread engagement, the swage grip, or the grout bond is carrying the full force across the joint. The performance classes, Type 1 and Type 2, are written around developing strength in tension because that is the harder case for most splices.

In pure compression a different path is available. Bars in compression can transfer part of the force by end bearing, the squared and cut ends butting against each other or against a bearing surface inside the coupler, with the device holding the bars in concentric alignment. End-bearing splices have their own rules for the squareness of the cut and the alignment, since a force that crosses by bearing depends on the ends actually meeting flat and true.

Most structural splices have to work in both directions or under reversal, especially in seismic frames where the load swings tension to compression cycle after cycle. When the demand reverses, you cannot lean on end bearing for the compression half and ignore the tension half. The splice has to develop the bar both ways, which is why the qualified couplers are tested in tension, in compression, and under cyclic load. Match the splice to the demand the drawing assigns it, and do not treat a compression-only device as a general-purpose splice.

Staggering mechanical splices

Splices get staggered so that not all of the bars are spliced at the same section, and the same logic applies to couplers that applies to laps. A section where every bar is spliced at once is a built-in plane of relative weakness and congestion, even with full-strength couplers, so the detailing spreads the splices out along the member. How far apart, and what percentage of bars you may splice at one section, comes from ACI 318 and the structural drawings, and it depends on the splice type and the application. For the way stagger works on lap splices, see the rebar development length and lap splices guide.

One thing couplers change is the congestion math at the splice plane. Because a coupler does not double the bar the way a lap does, a section full of couplers is far less congested than the same section full of laps. That can relax how aggressively you need to stagger for placement reasons, but it does not erase the structural reasons the code staggers splices, and the higher-performance applications still call for stagger.

Build the stagger to the schedule. The drawings will show the offset, and the rodbusters set the coupler elevations to match. Bunching the couplers at one elevation because it is easier to set them all at the same height is exactly the move the stagger requirement exists to prevent.

Installing each coupler type

Installation is specific to the system, and the manufacturer's instructions are the procedure of record, not a starting point you adjust. Get the type-specific steps wrong and the splice fails the way it was never meant to.

For threaded couplers, the bars are threaded and gauged, the coupler is spun on, and the joint is torqued to the manufacturer's value with the bars seated to the position marks. The mark painted on the bar at the correct insertion depth is the field check that the bar went in far enough. Under-torque and short engagement are the two failure modes, and both hide behind a coupler that looks fully made up.

For swaged couplers, the sleeve goes over the squared bar ends and the hydraulic press strokes it the specified number of times per side with the correct dies. The swage count and the finished sleeve dimensions are the proof. A sleeve that did not get the full set of presses, or got pressed with the wrong die, is rejectable even though it is on the bar.

For grout-filled sleeves, the bars are inserted to the marked depth, the grout is mixed to the water ratio on the bag and pumped in, and you fill until clean grout flows from the outlet port so you know the sleeve is solid and not voided. Then the connection waits. It does not carry load until the grout reaches its specified strength, and on a cold day that wait gets longer, so plan the schedule around the cure, not against it.

QC and special inspection

Mechanical splices are the kind of work that gets covered by concrete and never seen again, so the inspection happens while the splice is still exposed and it tends to be a special-inspection item. The International Building Code, in its special-inspection chapter, calls for special inspection of mechanical reinforcing-bar splices, and the building official, the structural drawings, and the statement of special inspections set the exact scope and frequency for the job. Confirm what is required against the adopted IBC edition and the project documents rather than assuming.

The visual checks are where most problems get caught. On threaded couplers the inspector looks for full thread engagement and the bar seated to its position mark, and confirms the torque was reached. On swaged couplers, the swage count and the sleeve dimensions. On grout-filled sleeves, that the grout flowed from the outlet port and that the cure was met before loading. A common verification on threaded systems is to saw-cut a sample coupler and confirm the threads engaged fully inside, since the outside of a fully made-up and a short-seated coupler can look the same.

The inspector is also checking that the type installed matches the type specified, Type 1 versus Type 2, because that substitution is invisible once the section is poured and it is exactly the kind of thing a special inspection exists to catch. Record the splice locations, the type, the lot or heat data where required, and the torque or swage or grout verification, because that record is the only proof the splice was right once the concrete hides it for good.

Testing mechanical splices

Couplers get proven two ways, and both lean on the test methods in ASTM A1034 for testing mechanical splices for steel reinforcing bars. That standard covers tension, compression, slip, cyclic, and fatigue testing of a bar-splice assembly, which is two bars joined by the actual splice, loaded the way it would be loaded in the structure. The acceptance criteria the splice has to meet come from the design and the code class it is being qualified to, Type 1 or Type 2.

The first way is pre-qualification. The coupler system is tested to demonstrate it develops the required strength, and that data is what supports the evaluation report the engineer relies on to allow the product. You are not re-proving the system on every job. The manufacturer proved the design, and the report carries it.

The second way is production or job sampling. On many projects, especially seismic and critical work, the specification calls for pulling sample assemblies made up by the same crews with the same equipment on the same job and tension-testing them to confirm the field installation matches the qualified performance. The number of samples, the frequency, and the pass criteria come from the project specification and the engineer. The point of the production sample is simple. The system being good in a lab does not prove the splice your crew made this morning is good. The sample test is what ties the field work back to the qualification.

Couplers as dowels and form-savers

One of the most common uses of a threaded coupler has nothing to do with congestion or large bars. It is the planned connection between one pour and the next. You cast a coupler into the first pour with its open end at the construction joint, strip the forms, and later thread a dowel or a continuing bar into it for the second pour. The coupler is the future connection, set and protected while the next phase waits.

The form-saver is the version of this at the edge of a form. A threaded coupler is mounted to the inside face of the formwork so its open end finishes flush at the concrete face when the form comes off. After stripping, you thread the dowel-out bar into the exposed coupler and continue into the adjacent pour, slab to wall, wall to slab, or stage to stage. It saves you from drilling and epoxying dowels later, and it gives a positive mechanical connection instead of a field-drilled one.

The detail that gets fouled is keeping the coupler clean and capped until the next bar goes in. Concrete slurry, rust, or a bent thread in an exposed coupler means the dowel will not seat, and now you are chasing a fix at the worst time, with the next pour bearing down. Cap the couplers, protect them, and verify the thread is clean before you thread the dowel. The staged-construction use is also why couplers show up so often on jobs that are not congested at all. The driver there is sequence, not steel ratio.

Alignment and tolerance

Couplers vary a lot in how much bar misalignment they will take, and that tolerance is a real selection factor, not a footnote. Threaded and swaged couplers want the two bars close to coaxial, because the thread or the swage assumes the bars line up. Push them out of alignment and you stress the device in a way it was not qualified for. Grout-filled sleeves are the forgiving end of the range, sized to accept the bar over a window of position, which is why precast erection, where the bars never land perfectly, leans on them.

The misalignment a given coupler tolerates is set by the manufacturer and the evaluation report, so design the connection and the bar placement to land inside that window. On staged and precast work, this is where the planning pays off. The dowels that come up out of the first pour have to fall within the coupler's acceptance for the piece that lands on top, and that means the survey and the templating have to be right before the concrete sets the dowels in place.

When the field finds bars that will not line up with their couplers, the answer is not to force them or to bend them cold into the sleeve. It is to stop and get the engineer's direction, because a forced or mis-seated coupler is a splice that did not get the load path it was designed for.

Congestion relief and cost versus a lap

The single biggest placement reason to choose a coupler is that it does not double the steel at the splice. A lap puts two bars side by side through the overlap, and in a heavily reinforced column or a thick mat that doubled steel eats the clear spacing the concrete needs to flow and consolidate. The coupler keeps one bar line through the splice, so the section stays open enough to place and vibrate properly. Congestion is not just a fit problem. When bars crowd together the coarse aggregate bridges across them and the concrete below the splice goes unconsolidated, leaving honeycomb and voids right where the structure is most loaded. For how clear spacing and consolidation interact at the cage, see the rebar placement and cover guide.

On a single splice, though, a coupler costs more than a lap. You buy the coupler, you pay for the bar-end preparation or the press time or the grout, and on many jobs you pay for special inspection and sample testing on top. A lap costs the extra bar in the overlap and the labor to tie it, and usually nothing else. For an ordinary bar in a member with room, the lap is the cheaper line item, which is why most splices are still laps.

The comparison flips as the bar gets bigger and the section gets tighter. The overlap on a large bar is a lot of expensive steel, and that lap length grows with the bar while a coupler is roughly a fixed cost per splice. At some bar size the coupler is simply cheaper than the steel the lap would burn, and once you add the slower placement and the consolidation risk a doubled bar creates, the coupler can win on total cost even where the lap looks cheaper on the takeoff. Run the comparison on the real job, bar by bar and member by member. The estimate that priced everything as a lap and then meets a column full of #14s detailed with couplers is the estimate that turns into a change order, so catch the splice method at takeoff.

Submittal, approval, and proprietary systems

A coupler is a proprietary product carrying load in a structure, so it goes through approval before it goes in the wall. The system is submitted with its evaluation report, the third-party document that records what the product was tested to and which code provisions it satisfies, commonly an ICC-ES report or an IAPMO-UES report. That report is what lets the engineer of record accept the product as meeting the Type 1 or Type 2 requirement and the application it is being used for.

Read the report for the specifics, because they bound what you are allowed to do. It states the bar sizes and grades the system covers, the type classification, the installation requirements, and any limits on use. A coupler qualified for one bar grade or one application is not automatically good for another, and the report is where that line is drawn. Match the product on site to the product that was approved. Substituting a different coupler system, or a different size or grade than the submittal covered, because it was what the supplier had, undoes the approval. A substitution goes back through the engineer with the substitute's own evaluation report, not in on a verbal.

The systems themselves are not interchangeable, and that is the field consequence of a proprietary market. Several manufacturers make threaded, swaged, grout-filled, and bolted couplers, each with its own threading dies, swage press and dies, torque values, grout, and inspection method, all belonging to the system on the submittal. Couplers concentrate at the heavy end of the work: large-bar high-rise columns and core walls, deep transfer girders, mat foundations with two heavy mats of bar, and the big foundation and steel-to-concrete work on data centers and industrial plants. Those jobs have the largest bars, the worst congestion, and often the tightest seismic detailing, so the coupler program gets planned up front, with systems selected, submittals approved, and crews trained before the steel arrives.

What to document

Concrete seals a coupler in for good, and once it does there is no inspecting the joint again, which leaves the record as the only evidence the splice was ever made right. The record is what answers the inspector now and the question years out about whether the joint was ever the right device, installed right.

Capture the coupler system and its evaluation report number, the type class installed, the bar size and grade, the splice locations, and the installation verification appropriate to the type. For threaded, the torque and that the bars seated to the position marks. For swaged, the swage count and sleeve check. For grout-filled, the grout flow from the outlet port and the cure confirmation. Add the special-inspection sign-off and any sample-test results. If a type was changed or a substitution approved, record the approval. The table below is the short version of what belongs in the record.

Coupler types at a glance

The systems differ in how they grip the bar, what prep they need, and what they are best at. This is the field summary, with the evaluation report and the engineer controlling which one is allowed where.

Common mistakes

• Installing a Type 1 splice where the design calls for a Type 2 in a seismic yielding region.
• Short thread engagement or under-torque on a threaded coupler that looks fully made up from outside.
• Skipping or rushing the bar-end preparation, then threading a coupler onto a bad thread.
• Bunching couplers at one section instead of staggering to the splice schedule.
• Loading a grout-filled sleeve before the grout reached its specified strength.
• Using a coupler outside the bar size, grade, or application its evaluation report covers.
• Substituting a different coupler system than the approved submittal because it was on hand.
• Treating mechanical splices as ordinary work and skipping the required special inspection.
• Leaving cast-in or form-saver couplers uncapped, so slurry or rust fouls the thread before the dowel goes in.
• Forcing misaligned bars into a coupler instead of stopping for the engineer.

Field checklist

Standards and references

ACI 318 is the structural framework for mechanical splices in reinforced concrete. It defines the Type 1 and Type 2 performance classes, sets where each is permitted, and ties the seismic requirements to the type that develops the bar's full strength in the regions that yield. The specific percentages, the wording of the definitions, and the section numbers are refined between editions, so confirm them against the ACI 318 edition the project is built to and against the coupler's evaluation report before citing them on a submittal.

The International Building Code, in its special-inspection provisions, is where the requirement to specially inspect mechanical reinforcing splices lives. The adopted IBC edition, the building official, and the project's statement of special inspections set the scope and frequency. ASTM A1034 is the test-method standard for mechanical splices, covering tension, compression, slip, cyclic, and fatigue testing of the bar-splice assembly, and it is what backs both the product qualification and any job sample testing.

The proprietary product itself is governed by its evaluation report, commonly ICC-ES or IAPMO-UES, which states the qualified bar sizes and grades, the type class, the installation requirements, and the limits on use. CRSI and the coupler manufacturers publish installation and inspection guidance for the specific systems. Cite the standard that controls the point, and let the engineer of record and the evaluation report govern which coupler is allowed in which member.

Units and terms

Mechanical splice work carries its own vocabulary, and the same device goes by a few names across a submittal, a spec, and a shop drawing.

A mechanical splice is also called a coupler or a mechanical connection. Bar sizes run in the imperial system as #3 through #18, where the number is roughly the bar diameter in eighths of an inch, and in metric as 10M through 55M or by millimeter diameter. Strength terms matter here: fy is the specified yield strength of the bar, fu is its specified tensile strength, and the Type classes are written against those. Torque on threaded couplers is given in foot-pounds or newton-meters. Grout strength is specified in psi or MPa, and the connection waits until that strength is reached.

Mechanical splice / coupler

A device that joins two reinforcing bars end to end and transfers force bar to bar, not through concrete bond

Type 1 splice

A mechanical splice required to develop at least 125 percent of the bar's specified yield strength, per ACI 318

Type 2 splice

A splice that meets Type 1 and also develops the bar's specified tensile strength, required in seismic yielding regions

fy / fu

Specified yield strength and specified tensile strength of the bar, the figures the splice classes are written against

Form-saver

A threaded coupler mounted at the form face so a dowel can be threaded in for the next pour

Grout-filled sleeve

A coupler where bars set in an oversized sleeve are bonded by high-strength grout, common in precast

Evaluation report

The third-party report (such as ICC-ES or IAPMO-UES) that qualifies the coupler system and bounds its use

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is rebar development length?
• What is the minimum concrete cover for rebar?
• What is post-tensioning and how is it done?
• Why does rebar corrode in concrete?
• What is a bar bending schedule?