quality · procurement
How to read a sealant data sheet
Updated 2026-11-10
Every model page on this site has the same four-column table. Most buyers read the third column and ignore the other three, which is where the useful information is.
Column 1: the property
Tensile strength, elongation at break, skin time, Shore A, density, extrusion rate, volatile content, cure rate, elastic recovery. Each one predicts a different failure mode. A data sheet that lists only tensile strength is telling you what the supplier thinks you want to hear.
Column 2: the standard
A number without a standard is an opinion. GB/T 528 defines the tensile test; GB/T 13477.5 defines skin time; HG/T 4363 defines cure rate. If a supplier quotes a tensile figure without a method, you cannot compare it with anyone else’s.
Where a catalogue cites a method that does not match the property — our own catalogue once cited a shear standard for elastic recovery — treat it as a documentation error and ask for the correct reference.
Column 3: the typical value
This is a design target, not a guarantee. Two batches can differ by several percent and both be in specification. If you write the typical value into a purchase contract as a minimum, you have bought a dispute.
The practical approach: specify a minimum that the process can actually hold, and ask for the batch record to verify it lot by lot.
Column 4: the method
Sometimes a standard number (GB/T 528), sometimes a note. The method tells you what the number would look like if you tested it yourself — useful when a competitor quotes a different figure for a similar product.
What a data sheet will not tell you
- How the product behaves on your substrate after your preparation method.
- How it cures in your bay at your temperature and humidity.
- Whether the batch you receive matches the catalogue.
Only a coupon test and a batch record answer those.
One row, read properly
Take a single row from a windscreen adhesive table:
Tensile strength | 3.0–4.0 MPa (spec) | 3.5 (typical) | GB/T 528What each cell is telling you:
- Property — the peak stress of a standard dumbbell specimen pulled at a fixed rate. It says nothing about peel, fatigue or adhesion.
- Standard — GB/T 528 defines specimen geometry and pulling speed. A number quoted to a different standard is not comparable, even if the units match.
- Typical — the centre of the process window, not a guaranteed minimum. A batch at 3.2 MPa is inside the spec and 9% below typical.
- Method — how you would reproduce it. If this cell is blank or says ”/”, the row is decoration.
Read one row this way and the whole table becomes a set of claims you can test rather than a set of numbers you compare.
The rows that predict failure
Most buyers look at strength. The rows that actually predict a comeback are:
| Row | What it predicts |
|---|---|
| Skin time | Whether the bead can be tooled before it skins |
| Cure rate | Whether the vehicle or joint can be released on schedule |
| Elastic recovery | Whether the bead accumulates strain over cycles |
| Volatile content | Whether the bead shrinks and pulls away from the edge |
| Extrusion rate | Whether the gun can lay the bead at the speed the job needs |
| Sag resistance | Whether a vertical bead stays where it was put |
A product can have an excellent tensile figure and a poor recovery figure, and it will fail in a moving joint. That is the mismatch the four-column table is designed to expose.
What this means for buyers
- Compare numbers only when the method matches.
- Treat typical values as design targets, not guarantees.
- Ask for the batch record on the two properties that matter for your joint.
- A data sheet without methods is marketing, not engineering.