What Is a Backer Rod and Why Does It Matter?

A backer rod is a compressible foam rod pushed into a joint before the sealant goes in. It runs the full length of the slot and sits at a set depth below the surface. The sealant is then poured or gunned on top of it. Within a minute of that happening, nobody can see the rod at all. It still decides more about the finished joint than almost anything else in the slot.

At Shepherd and Sons we have been reforming and resealing joints on UK airfields, highways and ports for more than 40 years. We are members of the Extruded Sealant Association and work to the ESA Code of Practice. Most of the failed seals we are asked to look at come down to one of two things: preparation, or geometry. The backer rod sits in the middle of the geometry, because it is doing three separate jobs at once. This article takes those three in turn. It then covers closed cell against open cell, sizing, and what to do when a slot is too shallow to take a rod.

A sealant bead has a specified shape. The width comes from the saw. The depth comes from whatever the sealant is sitting on. Pour into an empty 40mm slot and you get a bead 40mm deep. Set a rod at the right height and you get the depth the data sheet asks for.

That relationship between width and depth is what governs how far the bead can stretch. A joint opens and closes by a fixed amount as the slabs move, and the bead has to absorb all of it. A bead that is too deep for its width behaves like a block rather than a strip. It cannot neck down across its section, so the strain concentrates at the slot faces. That is where it tears or peels away.

Every sealant carries its own width-to-depth figure. That figure belongs in the specification, not in somebody's head on site. It matters most with the high movement products. Specify a high extension N1 grade, then set the rod at the wrong depth, and you have paid for extension the joint will never use. The class describes what the material can do. The geometry decides whether it gets the chance. The difference between N1 and N2 is a movement question, and movement is exactly what the rod is there to protect.

A joint sealant should be bonded to two faces and free on the third. The two faces are the slot walls. The third is the floor of the slot.

Three-sided adhesion is the classic way to tear a seal. If the bead has stuck to the base as well as the walls, it is pinned along its bottom edge. It cannot deform freely when the joint opens. The strain gathers at the bottom corners and the bead splits from there. What you get is a failure that looks like a bad sealant and is nothing of the sort.

Closed-cell polyethylene does not bond to sealant, and that is the whole point of it. The rod is a physical break between the bead and the slot floor. That is why the ESA Code calls it a bond breaker rather than a filler. It is not there to fill anything. Some older joints have a compressible filler board sitting in the bottom from the original pour. That board is not a bond breaker and was never meant to be one. A proper bond breaker still goes in above it. That is why every joint sealing specification worth the name calls up the bond breaker as well as the sealant.

The third job is a plain commercial one. An empty slot takes as much sealant as it has room for, and hot applied material does not stop at the depth you had in mind.

Work the arithmetic on it. Take a 20mm wide slot on a resealing programme. Set the rod 5mm lower than the specification asks and every metre of joint carries an extra 100 square millimetres of sealant in section. That is a tenth of a litre per metre. Our M20 J8 to J9 coastbound programme reformed and resealed 23,304 linear metres of joint. At that length, 5mm of carelessness is over two cubic metres of sealant nobody budgeted for.

The material is the smaller half of the problem. That possession ran 10pm to 3:30am, twenty nights in a row. A crew that runs dry at one in the morning is not going back to the depot for more. Depth below the specified bead does no work either. It adds no movement capacity and no adhesion, and it costs exactly the same per kilogram as the part that does.

Two types of rod turn up on site and they are not interchangeable.

• Closed cell. Usually polyethylene. The cells are sealed, so the rod does not take up water and does not pass moisture through itself. That makes it the default for cold applied pavement and structural joints. It is not the one for a hot applied pour: standard closed cell polyethylene melts at pouring temperature, so hot applied work takes the heat resistant grade instead. Getting that wrong is not a subtle failure, and it shows up as soon as the material goes in.
• Open cell. Usually polyurethane. Softer, more forgiving on an irregular slot, and absorbent. It has its place indoors and in dry conditions.

Absorbent is the word that matters outdoors. A slot on a live carriageway is rarely bone dry. There has been rain, there is slurry water off the saw, and there is dew on a night shift in October. An open-cell rod takes that water up and holds it. It does not stay there once the sealant arrives. Under hot applied sealant the moisture flashes off as material at pouring temperature reaches it. Under a cold applied system it drives off during cure. Either way it works upward through the bead. You get bubbles, blowholes and a spongy top surface, and adhesion at the slot face suffers with it.

So closed cell is the sensible default on anything outdoors, anything cold applied, and anything sealed through a British winter. Two habits go with that choice. Handle the rod carefully, because a closed-cell rod that has been punctured or dragged over a sharp arris is not closed any more. Fit it with a blunt roller or a proper installation wheel, never a screwdriver. Then check the rod is rated for the temperature it is about to meet. Hot applied material goes in hot, and not every foam is made for it.

Here is the part that gets missed most often. A backer rod that drops into the slot easily is the wrong rod.

Nothing holds a rod in place except its own compression. There is no adhesive and no mechanical fixing. So it is chosen oversize against the joint width and squeezed in. The ESA Code of Practice puts it plainly. The bond breaking material should be slightly wider in diameter than the joint, so that it stays firmly located. A rod cut to slot width will drift, ride up, or float on the sealant.

Two things follow. Joint widths vary along a pavement, particularly on older concrete and anywhere a slot has been widened at some point in its life. A programme therefore needs a range of rod sizes on the van, and the Code says as much. The second point is the reverse of the first. A rod forced into a slot far too narrow for it buckles instead of sitting level. You then get a wavy sealant depth running down the joint.

Once it is in, the depth is measured rather than eyeballed. Readings are taken from the top of the bond breaker and from the top of the sealant. They are taken at three points along a transverse joint, or at one metre spacing, to an accuracy of ±0.5mm. A metal ruler and a 150mm straightedge laid across the joint is all it takes. In our experience that check is the quickest way to catch a rod running high or low, and it costs seconds per joint.

Some slots have no room for a rod. Narrow saw cuts, shallow chases and thin details all fall into that bracket. Push a rod into a slot that shallow and the sealant ends up thin, proud of the surface, or both.

The answer there is a bond breaker tape. It is a thin polyethylene or similar strip, usually self-adhesive, laid along the base of the slot before sealing. It does the second of the three jobs above and only the second. It breaks the bond on the floor of the slot. It sets no depth and it saves no material. The depth therefore has to come off the saw, and be cut right the first time.

Tape width matters as much as rod diameter does. Tape narrower than the slot leaves a strip of exposed concrete either side of it, and you are back to three-sided adhesion in patches. Tape wider than the slot creeps up the walls and takes bond area off the faces you want the sealant gripping. Crack chasing work often lands in exactly this territory, because a chased slot is cut to suit the crack rather than to a standard joint profile. Silicone joint seals are frequently shallow by design as well.

One thing not to confuse with either. Overbanding is applied over the surface of a crack rather than poured into a slot. There is no third face, so there is no bond breaker involved at all.

A backer rod set 5mm wrong is not one defect. It is the same defect, made identically, for as long as the shift lasts. That is what makes it worth briefing properly at the start rather than deciding joint by joint.

Nobody sees it either. The rod is under the sealant within minutes and the finished joint looks correct. The evidence turns up two or three winters later. It appears as tearing along a continuous run of joints that were all sealed by one crew on one night. Cohesive tearing that follows a shift pattern rather than a traffic pattern is usually a geometry problem, not a product problem.

On highway resealing programmes running to tens of thousands of linear metres, the rod size and the setting depth are fixed and briefed before the first joint. The same discipline came out of airside work. We have been sole contractor for the British Airport Authority since 2008, across all seven former BAA airports. A taxiway hands back at a stated time whether the detail was right or not, so the checks have to sit inside the shift.

The rod goes in near the end of the preparation, not at the start of it. On a reseal the order runs like this.

• Cut or widen the slot to the specified dimensions.
• Remove the old sealant completely, without damaging the slot faces.
• Wash the slurry out before it dries in the slot.
• Roughen the sawn faces so the primer and sealant can key in.
• Blow the slot out with oil-free compressed air.
• Fit the backer rod to the setting depth and check it.
• Prime where the manufacturer requires it, then seal.

Two things trip people up around the rod. The slot has to be dry when the rod goes in, and still dry when the sealant follows. Wet weather therefore stops the sealing even where it does not stop the cutting. The rod also goes in before the primer on this sequence. If the slot gets wet after priming and the preparation has to be redone, the rod comes out with it.

A backer rod is a line item and it should read like one. A specification that covers it names the rod type and the diameter against the slot width. It names the setting depth below the surface, the sealant depth that results, and the width-to-depth ratio those two produce. It names the recess and the depth check regime. Where the slot is too shallow for a rod, it names the tape and the tape width instead.

A specification that says insert backer rod and stops there has left out the dimension that matters most. It hands that decision to whoever happens to be holding the roller at two in the morning.

We've found the quickest way to settle it is to work backwards from the sealant. Take the product data sheet, take the actual slot width measured on site, and the rod size and setting depth fall out of the two together. If you are planning a resealing programme and want that worked through against real joint dimensions, get in touch. We will look at the slot, the movement and the sealant as one decision rather than three.