Why a Sealed Joint Fails

A sealed joint has more parts than it looks. There are two joint faces the sealant bonds to, a backing behind the bead, and the bead itself with a profile determined by how deep it is relative to how wide the joint is. Each of those is a decision, and only one of them is the cartridge.

Setting the backing rod is what fixes the depth of the bead in relation to the width of the joint, and that proportion is what decides whether the sealant can stretch as the joint opens or simply sits in it as a thick plug. A joint too shallow to take a rod uses a bond breaker tape instead, which does the bond-breaking part without controlling depth.

A bead stuck to the two faces and the back of the joint has nowhere to stretch. When the joint opens it tears or pulls away rather than extending, and the failure looks like a poor sealant when it is a poor joint. That is the reason the backing rod is described as a bond breaker as well as a depth control.

Omitting the backing is invisible on completion and is one of the most common reasons sealed joints fail early. Renewal work frequently goes over an old compressed backing that was never taken out with the bead, and nobody looks at it again once new sealant covers it.

Backer rod comes in closed-cell, open-cell and bi-cellular forms, and the sealant manufacturer states which type its product is meant to be used with. A closed-cell rod that has been cut, punctured or over-compressed can release gas into fresh sealant and leave bubbles or blistering in a bead that was otherwise fine.

Open-cell products take up water, which matters in joints that are wetted or formed before the construction is weathertight. The rod is frequently bought on diameter alone, with the cell structure recorded nowhere, so the one characteristic that has to match the sealant is the one least likely to appear on the paperwork.

Glass, glazed tile, painted surfaces, plastics, powder-coated metal, concrete, render and masonry all behave differently, and priming requirements are product-specific. What matters is the surface actually present, including any coating, curing residue or laitance, rather than what the drawing says the material is.

There is a specific renewal trap: fresh silicone bonds poorly to cured silicone residue, which makes thorough removal part of every renewal rather than an optional tidiness. A top-up over an old bead is a strip-and-redo job pretending to be a quick fix.

The convenience of a single product for every joint is exactly how wet-area and movement joints end up with the wrong material. An acrylic frame sealant is a decorator's material for small static gaps: it is firm rather than elastic, it is not formulated for continuous wetting, and a junction that keeps moving shows a crack through both the caulk and the paint.

The reverse error is equally common. A silicone at a skirting cannot be painted, a polyurethane around sanitaryware is not the usual conversation, and a hybrid chosen for paintability and smell generally accommodates less movement than a low-modulus product made for a designed movement joint. Movement capability is a property of the specific product.

Not every seal is gun-applied. A pre-compressed expanding tape is set into the joint before the component that closes it, and it seals by the pressure of the foam against two faces rather than by bonding or curing. A joint that has already been closed cannot receive it, so a forgotten tape is a removal job rather than a small omission.

It also has a right way round, being made to resist driving rain from outside while remaining relatively open to vapour passing outwards. Fitted reversed it is doing a different job from the one specified, and a gun foam is not a substitute for it even though a gun foam is what is most often used instead.

Joint sealants have a finite life and are among the components most likely to need replacing during a building's life. Beads at wet junctions are commonly inspected as part of routine upkeep, because discolouration, lifting edges or splitting are visible signals that something at that junction has changed.

A neat bead is not evidence of performance. It can conceal a junction that is not draining, and a sanitary grade slows surface mould growth without doing anything about why a surface is staying wet. That distinction is worth holding on to before renewing the same bead for the third time.