Senior Laboratory -- Composites

Frequently Asked Questions

Extent of Cure Crossover

What is the extent of cure cross-over?

We define the point at which the extent of cure (the fractional portion of the reaction completed) for the surface and the center (1/2" through thickness) locations are identical to be the extent of cure cross-over point. This is a very important piece of information, since it tells us whether the part is curing inside/out or not.
At the beginning of the cure cycle, the surface temperature of the composite will be greater than that of the surface. This is because the surface is much closer to the heating source and the fact that the thermal conductivity of the composite is rather low, resulting in the center taking a while to heat up. Since the curing reaction of the resin is heavily influenced by the temperature through the Arrhenius relationship, the surface cures more than the center during this stage. Unfortunately, if the surface cures extensively before the center, large stresses are generated within the part when the center finally cures, possibly cracking the part. Because the resin has a high shrinkage rate (8% by volume), as the center cures and shrinks it is pulled away from the center toward the hardened (already cured) surfaces. However, if composite cures from the center out, these stresses will not be generated and no cracking will occur.
Thankfully, we are helped out by the fact that the center is more thermally insulated than the surface. Once the center begins to react, some of the heat released by the reaction cannot be transferred out, which raises the temperature of the resin. Thus the resin at the center cures at a faster rate due to the higher temperatures, resulting in even higher temperatures and a faster curing reaction. At this point, the extent of cure at the center of the part rises quickly and surpasses the reaction at the surface of the composite. The earlier the extent of cure cross-over, the better the inside/out cure.
For the purposes of this lab, we estimate that a part of reasonable quality can be produced if the extent of cure cross-over occurs before 0.30 (or 30% of the reaction is completed). This means that the extent of cure for the center node of the simulation must exceed the value for the surface node before 30% of the reaction is completed in the center.

Insulating Boundary Conditions at Mold Edges

In your input data file, you have the initial temperatures at 25 degrees C throughout the process. Are the left and right sides going to really stay at ambient temperature, or will they be heated up by the water bath to 70 degrees like the rest of the surface (or whatever surface temperature we choose)?

Well, they will change temperatures, but if you will notice the boundary conditions for the two sides, we assume that they are insulated. We do this because it is much much faster to solve the problem in one dimension. This is only an approximation, and as you point out, the sides should change temperature. Because the sides are assumed to be insulated, it does not matter what temperature is given, the simulation does not allow any heat transfer out the sides. I just used 25, but it is pretty arbitary. The program needs to have temperature values, even if it does not use them. If you increase the number of temperature points, just add the required number of extra temps to keep the program happy.

Sensitivity Analysis

We have been varying parameters, i.e. thermal conductivity, density, etc.), but these parameters can't be varied in the laboratory by experimental conditions--can they??? I guess the question is which parameters can we vary or are expected to vary by controlling the experimental conditions?

First off, you should only change these parameters once you have found a good cure cycle (one with an early extent of cure cross-over point). The idea behind changing the parameters for a sensitivity analysis is that you understand how each parameter effects the reactions within the composite. I can almost guarantee that the values I gave in the input file are not what you will see during the actual experiment. This is not meant to be a trick. It is just that some of the values are difficult/impossible to measure experimentally (some of the kinetic parameters). Others can change from experiment to experiment, so the numbers given are only an observed average (boundary conditions, resin volume fraction, etc). You will be able to use the results of the sensitivity analysis to guide what you do in the actual experiment. For example, if you see that the center is taking longer to heat up, what are the possible reasons. Give up?, you might want to look at boundary conditions, thermal conductivity, heat capacity, and density. The sensitivity analysis would tell you which one may be the most likely culprit. For the same reasons, you want to look at changes in kinetic parameters to see how the temperature peaks and cure are effected. Lastly, you will use the results of the sensitivity analysis to help fit the simulation to the actual experimental run.

[*] Back to Composites Lab Page