Tensile strength gets the attention because it is the number engineers design with. But a tensile test breaks a bar and reports a result without saying why it came out that way. Two other numbers, degree of cure and glass transition temperature, go one step further back. They report what happened to the resin inside the die, which is where most of a bar's long-term properties are decided, and they can be measured on a sliver of bar a few milligrams in weight.
What curing actually does to the resin
Resin enters the die as a liquid mixture of molecules that have not yet joined up. Heat in the die starts a reaction that links them into a rigid three-dimensional network, and that network is what holds the glass fibres in place and transfers load between them. The reaction releases heat as it goes. Degree of cure is simply how far that reaction got before the bar left the heated zone, expressed as a percentage of the reaction that was chemically available.
A bar that leaves the die at 85 percent cure still has a share of its resin sitting unreacted inside it. It looks finished. It is the right diameter and the right colour. But the network holding the fibres is looser than it should be, which is why an undercured bar underperforms on the properties that depend on the resin rather than the glass: transverse shear, bond, moisture uptake and resistance to alkali.
How DSC measures it
Differential scanning calorimetry heats a small sample at a controlled rate and records how much heat flows in or out of it. If any resin in the sample is still unreacted, it finishes curing inside the instrument and gives off heat while doing so. Compare that leftover heat with the total heat released by fully uncured resin of the same system, and the difference is the degree of cure. ASTM E2160 is the test method for this, and ASTM D7957, the American specification for GFRP bars, requires a degree of cure of at least 95 percent.
The same DSC run also shows the glass transition temperature, or Tg: the temperature at which the cured resin moves from a rigid, glassy state to a softer, rubbery one. Tg rises as cure progresses, because a more completely linked network needs more heat to loosen. Specifications built on ASTM D7957 commonly require a Tg of at least 100 °C by DSC under ASTM E1356. A low Tg is often the first visible sign of undercure, and it also tells an engineer how much margin the bar has before heat starts to soften the matrix.
Why the thick bars are the ones to test
Heat reaches the centre of a bar from the die wall, and the resin at the core cures last. A 16 mm or 20 mm bar has far more material between the surface and the centre than an 8 mm bar, so at the same line speed and die temperature the core of a thick bar spends less effective time at cure temperature. That is why published test programmes often pick their largest diameter as the worst case for degree of cure. A line that passes comfortably on 8 mm can fall short on 20 mm, and the only fix is slowing the line or adding cure length, both of which change the plant's output per shift.
What this says about the line, not just the batch
Degree of cure is the most direct evidence of whether a die and curing section hold their temperature under real running conditions. A heater zone that sags during a long run, a controller that overshoots and then undershoots, or a curing section too short for the speed the plant wants to run — each shows up as cure that drifts from one batch to the next. A single good DSC result proves one sample was cured. Consistent results across diameters, shifts and the start and end of a run prove the line can do it every time.
This is also why cure is a buying question and not only a quality-control question. When evaluating a line, ask what degree of cure and Tg it achieves on its largest rated diameter at its rated speed, measured on bar taken from the end of a long run rather than the first metres after a warm start. A supplier who can show that data has checked the part of the process that matters most. A supplier who can only show a tensile certificate has shown you the result, not the process that produced it.
A cheap test to run often
A DSC test needs milligrams of material and a few hours of instrument time. That is far less than a full tensile, bond and alkali programme. It will not replace those tests, and it is not a substitute for certification. But it can be run on every batch, and it catches undercure before a bad batch reaches a site. For a plant, it is the quickest way to know whether this week's bar was cured the same way as the bar that passed certification.