Recognizing the importance of collaboration between engineering disciplines is crucial to improving our understanding of PMCs and extending their use range. Just as in polycrystalline metals, materials science without mechanics or mechanics without materials science is likely to lead to false conclusions, invalid models, and the ultimate failure of long-term predictions of lifetime or degradation in any extreme environment.

Considering the past 50 years of research by the multitude of materials scientists, solid state physicists, and mechanical and materials engineers on the structure-property relationships and processing-properties relationships for metals, and the remaining unknown issues within the engineering metals field (fatigue crack initiation or stress-corrosion cracking are good examples), it is no surprise that interdisciplinary work remains to be done on composite materials. To amass the knowledge necessary to push the envelope for use of PMCs in extreme environments, many fundamental studies are required in an array of disciplines (remember that the solid state physics and mechanics communities supplied much of the dislocation theory for the metallurgists). This need is not only a challenge but also an opportunity.

However, the problems are more difficult in PMCs for the reasons cited. In addition, the disciplines required are even broader. They include organic polymer chemistry, polymer physical chemistry, materials science, solid-state physics, and materials and mechanical (includes mechanics and manufacturing) engineering. It is highly unlikely that any individual in a single discipline can address all of the relevant topics with any depth of understanding. Thus a key challenge is to develop a truly multidisciplinary approach to the understanding of composite performance in different environments.

It is essential that the causes of environmental interactions be identified rather than just the symptoms. For example, microcracking of the matrix may be a symptom of a hostile environment, but absorption of water followed by a particular thermal history may be the root cause of the cracking. Is the moisture absorption caused by the wrong choice of a matrix or by improper postcuring of the matrix? Is the interface being degraded by the environment, leading to a loss of interfacial bond strength and the subsequent wicking of the moisture along the fibers? These well-known examples illustrate the need to understand the chemical-physical root mechanisms before undertaking any sustained effort to model and then predict long-term performance. The challenges are enormous and will require a sustained and interactive effort by many members from many disciplines within the technical community.