Q

0 20 40 60 80 100 120 160 180 200 0 20 40 60 80 100 120 160 180 200 Time (min) Time (min)

Fig. 7.39. IPB Maintains Resin Pressure17

bagging procedure resulted in a severely overbled and porous laminate in the previous test, the addition of internal bag pressure prevented both overbleeding and porosity. The absence of overbleeding was a result of the lower membrane pressure (30psi for the IPB cure vs. 100 psi for the normal cure).

7.9.4 Resin and Prepreg Variables

The resin mixing and prepregging operations can also influence the process-ability of the final prepreg. During normal mixing operations, air can easily be mixed into the resin. This entrained air can later serve as nucleation sites for voids and porosity. However, some mixing vessels are equipped with seals that allow vacuum degassing during the mixing operation, a practice that has been found to be effective in removing entrained air and may be beneficial in producing superior quality laminates.

Prepreg physical properties can also influence final laminate quality. Prepreg tack is one such property. Prepreg tack is a measure of the stickiness, or self-adhesive nature, of the prepreg plies. Many times, prepregs with a high tack level have resulted in laminates with severe voids and porosity. This could be due to the potential difficulty of removing entrapped air pockets during collation with tacky prepreg. Again, moisture can be a factor. Prepregs with a high moisture content have been found to be inherently tackier than low moisture content material. Previous work18 has indicated a possible correlation between prepreg tack and resin viscosity, i.e. prepregs that are extremely tacky also have high initial resin viscosities. Resins with such high viscosities will be less likely to cold flow and eliminate voids at ply terminations.

AS-4/3501-6 Overbleed Normal Pressurization

AS-4/3501-6 Overbleed Internally Pressurized Bag

AS-4/3501-6 Overbleed Normal Pressurization

AS-4/3501-6 Overbleed Internally Pressurized Bag

0 20 40 60 80 100 120 160 180 200 0 20 40 60 80 100 120 160 180 200 Time (min) Time (min)

Fig. 7.39. IPB Maintains Resin Pressure17

Prepreg physical quality can greatly influence final laminate quality. Ironically, prepreg that appears "good" (i.e., smooth and well impregnated) may not necessarily produce the best laminates. Several material suppliers have determined that only partially impregnating the fibers during prepregging results in a prepreg that consistently yields high quality parts, whereas a "good" (i.e., smooth and well impregnated) prepreg can result in laminates with voids and porosity. Partially impregnated prepregs have the same resin content and fiber areal weight as the fully impregnated material. The only difference is the placement of the resin with respect to the fibers. The partial impregnation process provides an evacuation path for air, and low temperature volatiles, entrapped in the lay-up. As the resin melts and flows, full impregnation occurs during cure. A closely related phenomenon is the surface condition of the prepreg. A fully impregnated prepreg will not cause a problem if a surface has the impressions of the fibers (sometimes called a corduroy texture), again providing an evacuation path. These three prepreg conditions are summarized in the highly idealized schematic shown in Fig. 7.40.

7.9.5 Condensation Curing Systems

The chemical composition of a thermoset resin system can dramatically affect volatile evolution, resin flow, and reaction kinetics. Addition curing polymers, in which no reaction by-products are given off during crosslinking, are, in general, much easier to process than condensation systems. Because of the reaction byproducts and solvents that evolve during processing, condensation systems, such as phenolics and polyimides, are extremely difficult to process without voids and porosity.

Condensation curing systems, such a polyimides and phenolics, give off water and alcohols as part of their chemical crosslinking reactions. In addition, to allow prepregging, the polymer reactants are often dissolved in high temperature boiling point solvents, such as DMF (dimethylformamide), DMAC (dimethylac-tamide), NMP (N-methylpyrrolidone), or DMSO (dimethylsufoxide). Even the addition curing polyimide PMR-15 uses methanol as a solvent for prepregging.

Fully Impregnated Fully Impregnated Partially Impregnated

Smooth Surface Rough Surface Good

Bad Good

Fig. 7.40. Effects of Prepreg Physical Quality17

The eventual evolution of these volátiles during cure creates a major volátiles management problem, which can result in high void and porosity percentages in the cured part. Unless a heated platen press or a hydroclave with extremely high pressures (e.g., 1000psi) is used to keep the volatiles in solution until gellation, they must be removed either before the cure cycle, or during cure heat-up when the resin viscosity is low. In addition, since these materials boil or condensate at different temperatures during heat-up, it is important to know the point(s) during the cycle when the different species will evolve.

There are three strategies for volatile management: (1) use a press or a hydro-clave with an applied hydrostatic resin pressure greater than the volatile vapor pressure to keep the volatiles in solution until the resin gels; (2) remove the volatiles by laying-up only a few plies at a time and hot debulking under vacuum bag pressure at a temperature higher than the volatile boiling point; or (3) use slow heat-up rates and vacuum pressure during cure, with intermediate holds, to remove the volatiles before resin gellation. It should be noted that more than one of these strategies can be used at the same time. The advantage of a heated platen press or hydroclave is that high pressures can be used to suppress volatile evolution. However, the tooling must be designed to withstand the higher pressures, and special damming systems must be incorporated to prevent excessive resin squeeze out. The second method, intermediate hot debulks under vacuum pressure, is effective but is very costly and labor intensive, since the ply collation operation has to be interrupted every several plies; the part bagged and moved to an oven; hot debulked; and then cooled before further collation. The last method, as shown in Fig. 7.41 for a typical autoclave cure cycle for PMR-15, incorporates multiple holds under vacuum during heat-up to evacuate the volatiles during various points in the cure cycle. It should be noted that some manufacturers use a 600° F cure and post-cure rather than the 575° F shown in the figure. In addition, some use only a partial vacuum during the early stages of cure, and apply a full vacuum in the latter stages. Although the final cure of PMR-15 is an addition reaction, it undergoes condensation reactions early in the cure cycle during the imidization stage that creates a volatile management problem. The tricky part to this approach is determining the optimum times and temperatures for the hold periods, and the heat-up rates to use. Physiochemical test methods can be used in helping to design these cure cycles. To obtain full crosslinking, polyimides often require extended post-cure cycles. Note that even the post-cure cycle incorporates multiple hold periods during heat-up to help minimize residual stress build-up and thus reduce the likelihood of matrix microcracking.

7.9.6 Residual Curing Stresses

Residual stresses develop during the elevated temperature cure of composite parts. They can result either in physical warpage, or distortion, of the part

Fig. 7.41. Typical PMR-15 Cure Cycle19

(particularly thin parts) or in matrix microcracking either immediately after cure or during service. Distortion and warpage causes problems during assembly and is more troublesome for composite parts than metallic ones. While the distortion in thin sheet metal parts can often be pulled-out during assembly, composite parts run the danger of cracking, and even delamination, if they are stressed during assembly. Microcracking is known to result in degradation of the mechanical properties of the laminate, including the moduli, Poisson's ratio, and the CTE.20 Microcracking (Fig. 7.42) can also induce secondary forms of damage, such as delaminations, fiber breakage, and the creation of pathways for the ingression of moisture and other fluids. Such damage modes have been known to result in premature laminate failure.21

The major cause of residual stresses in composite parts is due to the thermal mismatch between the fibers and the resin matrix. Recalling that the residual stress on a simple constrained bar is:

where a = Residual stress a = Coefficient of thermal expansion (CTE)

Fig. 7.42. Matrix Microcracking

E = Modulus of elasticity

AT = Temperature change

A rather simplified analogy for a composite part is that the CTE difference between the fibers (~0 for carbon fiber) and the resin is large (~20-35 x 10-6/° F for thermoset resins). The modulus difference between the fibers (30140 msi) and the resin (0.5 msi) is also large. The temperature difference (AT) is the difference from when the resin becomes a solid gel during cure and the use temperature. The so-called "stress free temperature" is somewhere between the gel temperature and the final cure temperature, as the crosslinking structure develops strength and rigidity. The use temperature for epoxy composites usually ranges anywhere from -67 to 250° F.

There are several observations we can make from this simplified analogy. High modulus carbon, graphite, and aramid fibers have negative CTEs. Normally, the higher the fiber modulus, the more negative the CTE becomes, which leads to increases in residual stresses and helps to explain why more matrix microcracking is observed with high modulus graphite fibers than with high strength carbon fibers. Carbon/epoxy resin systems are usually cured at either 250 or 350° F. Since there will be a smaller AT for the systems cured at 250° F, they should experience less microcracking than the systems cured at 350° F. Very high temperature polyimides, and many thermoplastic resins, that are often cured or processed at temperatures in the range of 600-700° F develop very high residual stresses and are very susceptible to microcracking. Since the AT differential becomes larger when the use temperature is lowered, for example, when the temperature is -40 to -67° F for a cruising airliner at 30 000-40 000 ft, more microcracking is normally observed after cold exposures than elevated temperature exposures. The analogy presented above greatly oversimplifies the residual stress problem in composite structures. In fact, analysis of residual stresses in composites is probably one of the most complex problems analysts have tried to address. There is quite a bit of conflicting data in the literature over the various causes of residual stresses and the effects of material, lay-up, tooling, and processing variables on residual stresses.

While residual stresses in composites are extremely complicated and there is considerable conflicting data on the effects of different variables, the following guidelines are offered for minimizing their effects:

• Use only balanced and symmetric laminates. Minimize ply lay-up misori-entation or distortion whenever possible.

• Design tools with compensation factors to account for thermal growth and angular spring-in. The use of low CTE tools will probably help to minimize residual stresses when curing carbon fiber composites.

• The use of lower modulus fibers and tougher resin systems helps to minimize residual stresses and microcracking.

• Slow heat-up rates during cure with intermediate holds and lower curing temperatures probably helps in minimizing residual stresses by balancing the rate of chemical resin shrinkage with the rate of thermal expansion. Likewise, there is some evidence that slow cool-down rates help.

Was this article helpful?

0 0

Post a comment