When a bead splits, the sealant gets blamed. Usually it is not the sealant.
Quick answers
What is the most common reason a sealant joint fails?
Three-sided adhesion. The sealant bonds to the back of the joint as well as the two faces, and the first real movement tears it. Closed-cell backer rod or bond breaker tape prevents it.
What shape should a sealant bead be?
About half as deep as it is wide, with an hourglass profile, and no thinner than 1/4" at the bond line.
Can silicone be applied to a wet or frosty surface?
No. The silicone bonds to the frost or moisture film instead of the substrate, and the bead peels away later. Our weatherproofing grade is rated for application between 50 and 122 F.
1. Three-sided adhesion
The most common one. Sealant is supposed to bond to two opposite faces and stretch between them. If it also bonds to the back of the joint it is anchored on three sides, and the first real movement tears it — usually a cohesive split straight up the middle.
The cure is a bond breaker. Closed-cell backer rod is one in itself; sealant will not stick to it. Where the joint is too shallow for rod, use bond breaker tape. Never fill a joint solid.
2. Wrong shape
A weatherproofing bead wants to be about half as deep as it is wide, with an hourglass profile — thick where it bonds, thin in the middle. That thin middle is what does the stretching.
A bead as deep as it is wide has no thin section. It cannot elongate, so it tears. A bead thinner than 1/4" at the bond line has too little grip and peels instead. Rod size sets the depth.
3. The surface was not clean
Silicone bonds to the surface it touches. If that surface is a film of form release, a fingerprint, dust, frost or old sealant residue, the bond is to the film.
Solvent wipe with two cloths — one wet to lift, one dry to remove before it flashes off. One cloth just spreads the contamination around. Prime where the manufacturer says to prime.
4. Applied in the wrong conditions
Our weatherproofing grade wants 50 to 122 F at application. Below that, cure slows badly and the substrate may carry invisible frost, which is a guaranteed adhesion failure. A joint is also at its widest when the building is coldest — so a bead applied cold gets compressed for the rest of its life rather than stretched, which is the better way round if you have the choice.
Do not seal a wet or frost-laden surface. Do not seal a joint that is about to be rained on before it skins.
5. It really was the wrong product
Sometimes it is the material. Acetoxy silicone against coated aluminium corrodes the frame and bonds poorly — that is a known failure mode. A weatherproofing sealant used where the glass is structurally bonded is not a failure mode, it is a hazard. The two grades are not interchangeable.
Test it before you commit a facade
Adhesion depends on your substrate, not on the data sheet. Coatings vary between batches; the mill finish on this job is not the finish on the last one. Ask us for a sample and the full test report, run it on your own detail, and pull it. We would rather post a sample than be part of a claim.