Simulation of reactive nanolaminates using reduced models: I. Basic formulation
Simulation of reactive nanolaminates using reduced models: I. Basic formulation
复制标题
使用简化模型模拟反应性纳米层压板：I. 基本配方
DOI：
10.1016/j.combustflame.2009.06.019
复制
发表时间：
2010
影响因子：
4.4
通讯作者：
O. Knio
中科院分区：
文献类型：
--
作者：
M. Salloum;O. Knio
建立了一个用于模拟Ni/Al多层膜反应前沿的多维瞬态简化模型。该配方是基于早期开发的准一维轴向和正常传播的方法的推广，特别是通过调整减少原子混合和热释放的形式主义。这种方法使我们能够专注于解决热锋结构，其演变是由热扩散和热释放。一个混合积分方案用于此目的，结合扩展稳定性，龙格-库塔-切比雪夫（RKC）积分的扩散项与化学源项的精确治疗。因此，避免了对各个层内的原子混合的详细描述，这使得能够对多维中的简化运动方程进行瞬态建模。首先进行二维模拟的前传播的复合材料结合两个双层周期。结果与Knepper等人的实验测量结果进行了比较。[22]，结果表明反应速度可能显著依赖于分层频率。比较表明，使用浓度依赖的电导率模型，瞬态二维计算可以合理地再现实验行为。基于表面引发反应的3D计算进行额外的测试。计算预测与激光点火测量的比较表明，计算提供了合理的估计点火阈值。最后提供了一个详细的讨论潜在的概括和相关的障碍。
A transient multidimensional reduced model is constructed for the simulation of reaction fronts in Ni/Al multilayers. The formulation is based on the generalization of earlier methodologies developed for quasi-1D axial and normal propagation, specifically by adapting the reduced formalism for atomic mixing and heat release. This approach enables us to focus on resolving the thermal front structure, whose evolution is governed by thermal diffusion and heat release. A mixed integration scheme is used for this purpose, combining an extended-stability, Runge–Kutta–Chebychev (RKC) integration of the diffusion term with exact treatment of the chemical source term. Thus, a detailed description of atomic mixing within individual layers is avoided, which enables transient modeling of the reduced equations of motion in multiple dimensions. Two-dimensional simulations are first conducted of front propagation in composites combining two bilayer periods. Results are compared with the experimental measurements of Knepper et al. [22], which reveal that the reaction velocity can depend significantly on layering frequency. The comparison indicates that, using a concentration-dependent conductivity model, the transient 2D computations can reasonably reproduce the experimental behavior. Additional tests are performed based on 3D computations of surface initiated reactions. Comparison of computed predictions with laser ignition measurements indicates that the computations provide reasonable estimates of ignition thresholds. A detailed discussion is finally provided of potential generalizations and associated hurdles.