Behavior and variability of solid propellants and criteria for failure and for rejection.

Abstract

T rapid development of larger and larger solid propellant propulsion units has been accompanied by an increase in capability of stress analysis and mechanical property measurement which is significantly more advanced than other fields of plastics or elastomeric application. Along with the increased capability has come a growing realization of the large variability inherent in solid propellants and of the necessity of including the variability characteristics in the stress analysis and failure criteria development. The most serious area of theoretical and experimental difficulty still remains as that of variability associated with failure behavior. This was implied in the early failure work of Griffith based on failure initiating from a defect of some magnitude, but the variability was directly recognized in the statistical approaches developed by Weibull. The basic approaches, originally applied to both simple failure and fatigue failure of metals, were extended first into the field of highly brittle materials such as glass, then to the more brittle plastics such as polystyrene, and finally to the rubbery materials by such workers as Rivlin and Thomas, who worked on the tearing of rubber, and by Bueche and Halpin,'who have examined the implications of timedependent failure and further generalized the concept of a failure envelope originated by Smith and Stedry. The relations of these theories to structural analysis has been well discussed by Williams,who also introduced the cumulative damage concepts of Miner to the problem of solid propellants. Further generalizations were offered by Landel and Fedors, who pointed out the effect of the distribution of molecular weights in the polymer through use of crosslink density as a correlating value, and who also pointed out the applicability to failure analysis of the empirical generalization of the total mechanical property curve proposed by Martin, Roth, and Stiehler. The application of such failure studies to the prediction of failure in motors and the establishment of failure criteria has been dealt with by studies of small motors by Milloway and Wiegand and the correlation of the viscoelastic behavior in motor studies by Kruse and Mahaffey.The studies of motor and specimen fatigue failure has been explored by Bills et al., whose data indicate that a cumulative damage approach is possible, but that the sequence of