Effects of Initial Conditions and Chemical Kinetics on Cellular Multiplicity of Regular Channel Hydrogen Detonations

Detonation cellular structures are often treated as uniquely determined by a mixture’s thermochemical state, yet recent studies have shown that confinement, numerical choices such as initial conditions (ICs) and the chemical kinetic model can induce cellular multiplicity. This work investigates how sinusoidally perturbed ZND ICs and the choice of the chemical kinetic model influences the long-time evolution of regular hydrogen detonations using detailed chemistry and shock-frame simulations. Eighteen cases spanning three stable/weakly unstable hydrogen mixtures were examined while varying the number of imposed sinusoidal waves and the channel height. We introduce the concept of self-regularization, wherein detonations initialized with different perturbations relax to the same equilibrium cellular structure through a sequence of transverse-wave creation, interaction, and merging events. Sixteen cases displayed this behavior, ultimately recovering the mixture-determined cell count despite initially irregular or mismatched patterns. Two cases exhibited persistent multiplicity, converging instead to a stable half-cell–shifted structure caused by inhibited transverse-wave formation. A relaxation-length analysis revealed (i) a U-shaped dependence of normalized relaxation length on the number of imposed waves, suggesting the existence of rapidly relaxing ICs; (ii) a strong increase in relaxation distance with channel height; and (iii) faster, less IC-sensitive convergence for more reactive mixtures. The influence of the chemical kinetic model is found to be highly mixture dependent, with the argon-diluted hydrogen mixture exhibiting lesser sensitivity to kinetics compared to the stoichiometric hydrogen-oxygen mixture. These results outline the importance of the choice of the chemical kinetic model and the conditions under which sinusoidal ICs yield the same converged cellular structures or lead to multiplicity and provide practical guidance for selecting initial conditions and computational domain sizes in detailed chemistry detonation simulations.