What the sleeve saves, what it never changes, and the 0.20 in that the spacing table assumes
← Back to the calculatorA 60 ft joint at 12 in on centre takes 60 dowels — one per foot, with the end dowels 6 in in from each end. In 3/4 in by 14 in that is 105.1 lb of bar net, and bought as 20 ft sticks it is four sticks, 120.2 lb, because 14 in does not divide 240 in.
What the sleeve buys is sequence: it is cast into the pour you have already finished, so the bar goes in last and the alignment is set while the form is still open. What it does not buy is capacity. Load still crosses the joint by bearing on the concrete under the bar, and the bar still has to be smooth and still has to be half in each pour.
Three parts, and only one of them is concrete work. A moulded polypropylene sleeve — 0.083 in wall on the lateral-movement version, 0.125 in on the standard one — a one-piece base with alignment marks and pre-installed nails, and a plain smooth round bar that you supply separately. The sleeve is 100% polypropylene, rated 5,500–8,000 psi in compression to ASTM D-695, and the bar never touches it until the form is off.
The sequence is five steps, and the order is the whole point:
Read step 5 again, because it is where the money is. The bar is the last thing placed, into a pocket that is already cast, in a joint face that is already straight. A conventional slip dowel does the opposite: the bar goes through a drilled hole in the form before the pour, has to be greased so it will release, and has to be spun or backed off afterwards. Same joint, same bar, different order.
The selection is driven by slab depth, not by load, and every one of the three sits at 12 in on centre. That is the first thing a “speed” page should say plainly: the speed is in the installation, so the layout is exactly the layout you already had.
| Slab depth | Sleeve | Smooth dowel | Spacing | Bar weight | Per carton |
|---|---|---|---|---|---|
| 5–6 in | Red | 3/4 in × 14 in | 12 in | 1.752 lb | 140 |
| 7–8 in | Yellow | 1 in × 14 in | 12 in | 3.115 lb | 100 |
| 9–11 in | Green | 1-1/4 in × 16 in | 12 in | 5.737 lb | 80 |
Swipe the table sideways for more columns →
One column is worth pausing on. The bar weights are 1.752, 3.115 and 5.737 lb per dowel — a 3.3× spread across a table where the spacing never moves. That spread is why “how much rebar does a joint need” has no answer without the slab depth, and why the same 60 ft joint can be 105 lb or 344 lb of steel.
The manufacturer lists what the system eliminates: form drilling, greasing, spinning, removing and replacing. That is five operations. It also adds one — nailing the base to the form. So the honest net is four fewer operations per dowel, and on a 60 ft joint that is four fewer times 60.
| Operation | Conventional slip dowel | Speed dowel | Difference × 60 |
|---|---|---|---|
| Drill the form | One hole per dowel, on layout | None | −60 holes |
| Grease the bar | One bar per dowel | None — the sleeve is the slip surface | −60 bars |
| Set the bar | Through the drilled hole | Into the cast sleeve, after stripping | same count, later in the job |
| Spin / back off | Every bar, before the second pour | None | −60 operations |
| Nail the base | None | One sleeve per dowel | +60 |
| Net | 4 operations per dowel | 2 operations per dowel | −120 on one joint |
Swipe the table sideways for more columns →
Two things follow that the sales copy does not say. First, the saving is counted in operations, not in hours, and an operation on a form line is cheap while an operation on a stripped joint face at 6 a.m. is not — which is why the same four operations are worth different amounts on different jobs. Second, the one added operation happens while the form is open and the layout is still wet, which is the cheapest moment in the whole pour to fix a mistake.
Take a 60 ft construction joint at 12 in on centre. The count is the rule this site uses for any dowelled joint: floor((length − 12 in) ÷ spacing) + 1, because the first and last dowel sit 6 in in from the ends. That is floor(708 ÷ 12) + 1 = 60, and the check is that 6 in + 59 spaces of 12 in + 6 in = 720 in exactly. One dowel per foot — which is why 60 ft is 60 dowels and not 61. From there the arithmetic is the same three lines this site uses everywhere: bars, total length, total weight — then the order, which is a different number.
| Dowel | Each | 60 dowels | Per 20 ft stick | Offcut | Sticks | Order weight | Allowance |
|---|---|---|---|---|---|---|---|
| 3/4 × 14 | 1.752 lb | 105.1 lb | 17 | 2 in | 4 | 120.2 lb | +14.3% |
| 1 × 14 | 3.115 lb | 186.9 lb | 17 | 2 in | 4 | 213.6 lb | +14.3% |
| 1-1/4 × 16 | 5.737 lb | 344.2 lb | 15 | 0 in | 4 | 344.2 lb | 0.0% |
Swipe the table sideways for more columns →
The 16 in system is the row to look at. It cuts with zero offcut — 15 pieces out of 240 in exactly — and 60 dowels is exactly four sticks, so the order weight equals the net weight: 344.2 lb either way, a 0.0% allowance. The 14 in cut looks tighter and is not. It leaves 2 in on every stick, so 60 dowels is 3.53 sticks, you buy four, and you pay 14.3% over the net weight for the privilege. A clean cut length and a clean order are two different things, and only the 16 in system gets both.
Sleeves come in cartons, and the carton sizes do not line up with a joint either: 140, 100 and 80 per carton against 60 dowels means 2.33, 1.67 and 1.33 joints per carton — 0.43, 0.60 and 0.75 of a carton per joint. One carton of the small size does two joints and a bit; one carton of the large size does one joint and not the next. On a slab with a dozen joints, that is the difference between a tidy order and a shelf of leftovers.
Here is the number that makes people distrust a slip dowel. The bonded half of a 3/4 in dowel is 7 in long. Run the same bar through the development length table as a deformed bar in tension and a #6 needs 23 in. Seven against twenty-three is a factor of 3.3, and it looks like a dowel that is three times too short.
Those two failures are not the same failure. An adhesive anchor pulls on the bar, so the bar has to develop its yield force through bond, and bond is what sets the 23 in. A dowel in a joint is pushed down by one slab and held up by the other: the bar is in double shear across the joint face, and what has to be adequate is the bearing stress on the concrete beneath it, plus enough embedment that the concrete does not spall at the face. That is a different calculation with a different answer, and it is why 7 in is normal in a joint and 16 in is normal in a wall.
Which also means: if someone quotes you a speed dowel as “stronger than a drilled and epoxied dowel”, the comparison has no basis. They are two products for two load paths. The sleeve's contribution is that the sliding half actually slides — a bar bonded on both sides of a construction joint locks the joint, and the crack moves somewhere you did not plan for.
And the word “speed” sits on two different products. On the round system, the spacing stays at 12 in for all three sizes. On the plate system from the same manufacturer, the published claim is the opposite kind: the plate reduces the number of dowels required and lets the installer increase the spacing. If your bar count is what you are trying to move, the plate is the product that moves it — the sleeve does not.
The selection table carries a line most people skip: the values assume a maximum joint opening of 0.20 in. That is a budget, and a joint spends it on temperature and shrinkage. With the usual coefficient of thermal expansion for concrete, 5.5 × 10−6 per °F, a joint's movement per degree is its length times that number — and dividing 0.20 in by it says how much temperature swing the table can absorb.
| Joint length | Movement per °F | Temperature swing that uses 0.20 in | At 100 ft of joint |
|---|---|---|---|
| 40 ft | 0.00264 in | 75.8 °F | — |
| 60 ft | 0.00396 in | 50.5 °F | — |
| 100 ft | 0.00660 in | 30.3 °F | one season |
Swipe the table sideways for more columns →
The sliding half has to pass through the pocket. Deformations are exactly what the sleeve exists to avoid, and the published guidance is a smooth round dowel. A deformed bar keys against the pocket instead of sliding, which turns a slip joint into a restrained one — the joint stops moving and the crack shows up somewhere with no detail in it.
Grease is for a bar cast bare into concrete. Here the sleeve is the slip surface, and greasing adds back one of the five operations the system removed. Worse, the half that is supposed to bond is the half that gets handled, and a bond breaker on it is a dowel that pulls out of the second pour.
The manufacturer's own application note flags internal vibration to consolidate concrete around the sleeve. A pocket with a void under it is a dowel bearing on air, and it will not look wrong from the surface: the sleeve is still in line, the bar still slides, the joint still opens. It fails later, under load.
This is a construction-joint product: sleeve in the first pour, bar bridging into the second. An expansion joint has a compressible board and a movement several times larger, and it takes a sleeve that passes through that board. Put the construction-joint sleeve in an expansion joint and the joint has nowhere to go.
No, and they do not do the same job. A sleeved slip dowel transfers shear by bearing on the concrete under the bar, which is why 7 in of embedment each side is normal. A post-installed adhesive dowel is an anchor that develops the bar in tension, which is why the same 3/4 in bar needs about 16 in of bonded length in a #4-sized hole and 23 in as a #6. The sleeve removes work; it does not add capacity.
It follows slab depth, not load: 5 to 6 in slab takes a 3/4 in by 14 in smooth dowel, 7 to 8 in takes 1 in by 14 in, and 9 to 11 in takes 1-1/4 in by 16 in. All three sit at 12 in on centre. Those three bands are the manufacturer's selection table, and the spacing is tied to a maximum joint opening of 0.20 in.
60 at 12 in on centre — one dowel per foot. The rule is floor((length − 12 in) ÷ spacing) + 1 because the end dowels sit 6 in in from each end, so 60 ft gives floor(708 ÷ 12) + 1 = 60. In 3/4 in by 14 in that is 105.1 lb net; bought as 20 ft sticks it is four sticks, 120.2 lb, a 14.3% allowance for the 2 in offcut on every stick.
Not in a sleeve. The sliding half has to pass through the pocket, and deformations are what the sleeve exists to avoid — the published guidance is a smooth round dowel. A deformed bar in a sleeved joint keys against the pocket instead of sliding, which turns a slip joint into a restrained one and moves the crack somewhere else.
No. Greasing exists to let a bar slip in bare concrete; the sleeve already provides the slip surface, and the manufacturer lists greasing, spinning, removing and replacing as operations the system eliminates. Greasing a sleeved bar adds a step back and buys nothing, and it puts a bond breaker on the half that is supposed to bond.
14 in for the 3/4 in and 1 in systems, 16 in for the 1-1/4 in system, set so the bar is roughly half in each pour: 7 in sliding in the sleeve and 7 in bonded in the second pour, or 8 and 8 on the largest. The sleeve length is the sliding length, not a cover dimension.
The selection table is based on 0.20 in. On a 60 ft joint that is used up by roughly a 50 °F drop at the usual 5.5 × 10−6 per °F coefficient of thermal expansion, and shrinkage spends part of the same budget. A 100 ft joint reaches 0.20 in at about 30 °F. Past that, the spacing in the table is no longer the spacing you designed.
No. It is a construction-joint product: the sleeve is cast into the first pour and the bar bridges into the second. An expansion joint has a compressible board and a much larger movement, which takes a different sleeve — one that passes through the board. Using the construction-joint sleeve there gives the joint nowhere to go.