Modelling complex ageing behavior of tensile modulus in CTPB-bonded propellant – parametric and kinetically based procedures

Abstract. Composite rocket propellants with CTPB-binder (carboxyl terminated polybutadiene) are in use besides the ones which are based on HTPB-binder (hydroxyl-terminated polybutadiene), which are cross-linked via polyisocyanates to a polyurethane network. CTPB binders are cross-linked with substances containing aziridine (aza-cyclopropane) groups or epoxide (oxirane, oxa-cyclopropane) groups. In each case polyaddition reactions occur. A system consisting of CTPB + MAPO is considered here. MAPO is called tris-[1-(2-methylaziridinyl)] phosphane oxide. Curing with MAPO works well, but problems can occur. One drawback is the splitting of the P-N-bond, because it is very susceptible to hydrolytic cleavage. The mechanical properties of samples of a CTPB-MAPO propellant with AP and plasticizer were aged at 45°C during nearly 3 years and investigated by uniaxial tensile measurements. The tensile modules (Young modulus, E modulus) showed a complex behavior: first increasing, then decreasing and further decreasing with lower rate. With rubber systems, the Young modulus is proportional to cross-link density. The cross-link density is a state property of the binder. Because of proportionality of the modulus with the chemical property cross-link density, it can be used directly in kinetic modelling. The task was with the available measurement data to develop a kinetically based description. The increase and following decrease of modulus are caused by post-curing and binder backbone splitting. The decrease can be distinguished in two parts, where the second one originates by diffusion-control of the water migration in the propellant. The kinetic modelling is performed in two ways: at first parametric descriptions are tried which are only formally based on kinetic concepts; secondly, kinetically based models are applied. With both types, complete data description can be achieved. But the character of these modellings are different. Post-curing and splitting of the binder could be modeled with combined reactions of second order.