The one number that is not an estimate
Every figure that goes into planning a layup is a guess dressed up: the ratio, the waste, the area of a curved surface. The finished part is not a guess. Put it on a scale, subtract the glass, and what is left is resin, exactly, whatever anybody intended. That subtraction is the whole page.
Worked through for the default numbers: a part that weighs 35.1 pounds with 17.08 pounds of glass in it holds 18.02 pounds of resin. That is 1.05 parts resin per 1 part glass by weight, or 51.3 percent resin by weight. At 2.55 and 1.15 g/cm3 the volumes are 3039 and 7106 cubic centimetres, so the fibre volume fraction is 30.0 percent — a number that looks nothing like the weight fraction and is not supposed to.
Where the ratio goes wrong before you weigh anything
The glass weight is the input most likely to be wrong, and the cloth label is the usual culprit. Areal weight is quoted per square yard in the United States and per square metre almost everywhere else, and the two differ by a factor of about 34. A stack of fabrics entered as one number, when it was really the weight of one ply out of five, is the other common one. And a curved part measured by its footprint rather than its developed surface understates the area and therefore the glass.
Any of those three shows up the same way: a resin ratio that is wildly out, or in the worst case a glass weight larger than the part, which the page catches and refuses to work with. When the number looks impossible, check the cloth label before anything else.
What the thickness gap means and what it does not
The glass and the resin have known volumes. Spread over the laminated area they give a thickness the part would have if it were solid all the way through. Measure the real thickness and the difference is space with nothing in it. On the default numbers the implied thickness is 0.107 inches and a measured 0.118 leaves 9 percent unaccounted for.
What that percentage is not is a quality grade. Some of it is genuine porosity between fibre bundles. Some of it is the caliper riding on the texture that peel ply leaves, or on a rough back face, or on a slightly tapered edge where the readings got taken. A part measured with a micrometer on a machined coupon and a part measured with a plastic caliper on the mould flange are not measuring the same thing. This page reports the gap and leaves the interpretation alone, because how much porosity matters is a question about the part and its service, and nobody can answer it from a number.
Recording it is the point
A resin ratio measured off a finished part is worth writing on the fabric roll in marker. It is specific to that fabric, that resin, that technique and that pair of hands, and it will estimate the next part of the same kind far better than any published figure. Change the fabric and it does not carry over. Switch from a brush to a squeegee, or start bagging what you used to lay open, and it does not carry over either. Which is a reason to do this on every part until the numbers stop moving, not a reason to skip it.
Questions people ask
How do I find out what resin to glass ratio I actually got?
Weigh the finished part, trimmed and with nothing bonded to it, then subtract the weight of the glass that went in. What remains is resin, and dividing it by the glass weight gives the ratio by weight. For the numbers in the form — a 35.1 lb part with 17.08 lb of glass — the resin is 18.02 lb and the ratio is 1.05 parts resin per 1 part glass. That is 51.3 percent resin by weight.
Why is the fibre volume fraction so different from the glass weight percentage?
Because glass is about twice the density of cured resin, so a given weight of glass takes up much less room than the same weight of resin. In the default case the laminate is 48.7 percent glass by weight and only 30.0 percent glass by volume. Both numbers are correct and they describe the same laminate; they are just answering different questions. Quoting one and calling it the other is the most common way composite figures get compared to nothing.
The calculator says my glass weighs more than my part. What went wrong?
One of the inputs, and the cloth label is the first place to look. Areal weight quoted in grams per square metre entered into a field expecting ounces per square yard is out by a factor of about 34, which is enough to make the glass heavier than the part. After that: a ply count that covered one layer of a stack rather than all of them, an area taken from the footprint of a curved part rather than its developed surface, or a part weight taken before something was trimmed off it. A cured laminate always weighs more than the glass in it.
What does the gap between measured and calculated thickness tell me?
That there is space in the laminate with nothing solid in it, and roughly how much. On the default numbers the weights imply 0.107 inches and a measured 0.118 leaves 9 percent unaccounted for. Part of that is real porosity and part of it is the caliper reading across surface texture. The page reports the gap and does not grade it, because whether a given amount of porosity matters depends entirely on the part and what it does.
Can this tell me if my laminate is any good?
No. It tells you what is in it by weight and by volume and how thick that ought to make it. Whether a laminate is fit for a job depends on the loads, the fabric orientations, the cure, the bond lines, the geometry and a great deal else, and it is an engineering judgement rather than an arithmetic one. What this page is good for is estimating: the ratio you measure here is the number that makes your next quote right.
Do I need to remove the core before weighing?
You need to get it out of the number one way or another. The subtraction assumes the part is nothing but glass and cured resin, so any core, filler, fairing compound, paint, bonded hardware or leftover release agent lands in the resin figure and inflates the ratio. If it cannot physically come off, weigh it separately and subtract it from the part weight before entering it. A cored panel weighed whole will give a resin ratio that is nonsense.