"Twenty four hours" is the answer to a question nobody on a glazing job is actually asking. A silicone joint passes through three separate states, days apart, and the number that matters depends on which one you need. Skin over. Through-cured. Able to carry load. They are not the same day, and on a facade the difference between them is the whole job.
Quick answers
How long does silicone sealant take to cure?
A one-part neutral cure silicone skins over in under an hour, and cures through at roughly 2 to 3 mm of depth per twenty four hours. A 12 mm deep joint is therefore a four to five day proposition, not a one day one. Our own neutral cure grade tested tack free at 0.7 hours.
What is the difference between dry and cured?
Silicone does not dry, it cures. Drying is a solvent leaving. Curing is a chemical reaction with moisture in the air that turns a paste into rubber, and it works inward from the surface. Touch dry means the outside has reacted. It says nothing about the middle.
Does it cure slower on glass and aluminium?
It depends entirely on the air, which is the same thing. Glass and metal are not porous, so they contribute no moisture at all. Everything the reaction needs comes out of the atmosphere, which is why a sealed-up, still, dry space is the worst place to leave a fresh bead and why the same product behaves differently in August and in February.
When can the joint take structural load?
Not on a rule of thumb, ever. That figure comes off the manufacturer's technical data sheet for the specific grade, and it is measured in weeks, not days. Ask us for the sheet and we send it with the sample.
The three states, and who each one is for
| State | Roughly when | What it means you can do |
|---|---|---|
| Skinned, tack free | Under an hour on our neutral cure grade, at room temperature | Stop tooling. Dust will no longer stick to it. Nothing else. |
| Cured through | About 2 to 3 mm of depth per day, trade rule of thumb | The joint moves as rubber, sheds water, and can be adhesion tested. |
| Full structural strength | Manufacturer's data sheet, measured in weeks | The bond may carry the load it was designed for. |
The trap is that the first number is the one printed largest on most packaging, and the third is the one the facade depends on.
Depth is the variable nobody accounts for
Cure marches inward from the face at a roughly fixed rate. So the depth you tool is the depth you wait for:
- 6 mm of depth, about two to three days.
- 12 mm of depth, about four to five days.
- A joint filled solid because no rod went in, weeks, and it will still be soft in the middle when the building moves.
This is the quiet argument for backer rod that has nothing to do with saving sealant. A rod sets the depth, which sets the cure, which sets when the joint is actually a joint. Rod size by joint width. · What a correctly shaped joint uses.
Deeper is not safer. A thick bead is not insurance, it is a slower cure and a stiffer joint, and the joint was designed to stretch.
What actually changes the timeline
- Humidity. The reaction consumes atmospheric moisture. Dry air slows it, humid air speeds it. A heated, sealed interior in January is dry air.
- Temperature. Cold slows the chemistry down hard. Below 40°F a bead that normally cures in a day can still be soft after two. What else goes wrong below 40°F.
- Air movement. Still air near the bead goes stale. A little airflow brings fresh moisture in and carries the cure byproducts away.
- Chemistry. Acetoxy cures faster and tells you so by smell. That speed is not a reason to use it on coated or anodised aluminium, where it attacks the finish. Why neutral cure is the only answer there.
Three mistakes that cost a week
- Pull testing the next day. A part cured bead fails a test the same bead passes once it is through, and the crew then goes looking for a primer it never needed. Cure first, then test. The field adhesion test, and how to read the break.
- Water on it too early. A skin is not a seal. Rain or a wash down before the joint is through leaves a bead that looks right and is not.
- Taking the temporary supports out to the same schedule as last summer. Panels need their temporary stops until the bond has the strength the engineer assumed, and that date moves with the weather.
The one number we will not print
We will not publish a structural cure time for our sealants, because a structural bond is not a thing to take on somebody's blog post. It is on the manufacturer's data sheet for the exact grade, and where a specification names ASTM C1184 by title that sheet is part of the submittal anyway.
Tell us the grade and we send the sheet, the laboratory report and a sausage to test on your own substrate, before you order anything.
