Research On Algorithms For Dynamic Nonuniform Regression Of Solid Rocket Motors And Its Application

In order to meet the increasing performance and dimensional requirements of solid rocket motors,motors with large aspect ratios,high modulus number,embedded wire,and non-axial overload are recently widely used.However,motors having the above characteristics are often affected by the non-uniform combustion effect,which is hard to simulate via ordinary methods.With an object to improve the accuracy of the regression simulation and the internal ballistic prediction of such motors,and eventually boost the design process,this thesis had proposed a group of regression algorithms and regressionburn rate coupling simulation methods,which can be used to perform regression and performance prediction for such motors.The proposed level set-based regression algorithms work well with non-uniform dynamic burning rate distribution and fits with arbitrary grain coating and geometry.Therefore,theoretically,the methods are capable of any motor.In order to handle the large amount of computation required by the level set-based method,this paper has integrated various simplifications and acceleration technologies,which reduces the absolute computation as well as increases computing performance and significantly boosts the entire simulation.In addition,in order to evade the problem that Cartesian grid-based simulations are(although fast)difficult to couple with burn rate simulation,this paper transplants the level set method to the unstructured grid and developed a group of innovative methods including a spatial discrete format,a gradient estimation method,etc.to ensure accurate and robust regression of the simulation of motors on the unstructured grid.Besides,the paper also studies the coupling numerical solving of the regression algorithm and burn rate models.An integrated flow,which extracts necessary geometry and boundary conditions and feeds the burn rate field back to the regression module,is implemented.In terms of burn rate models,this paper mainly focuses on several representative non-uniform combustion effects including erosive burning,embedded wire and rotating overload effect.On the base of the above work,this article finally managed to accurately predict the performance of the motors under non-uniform burning effects.A serial of testing problems,of which clear analytical solutions or experimental data are available,are utilized to test the proposed methods.Performance tests show that the proposed algorithms are significantly faster than their ordinary equivalents.A highprecision regression simulation can be accomplished within only 25 minutes.Other validations show that the proposed methods accurately predicts the non-uniform regression procedure and the corresponding internal ballistic curve.