As multiscale structures are inherent in multiphase flows, constitutive models employed in conjunction with transport equations for momentum, species, and energy are scale dependent. We suggest that this scale dependency can be better quantified through deep learning techniques and formulation of transport equations for additional quantities such as drift velocity and analogies for species, energy, and momentum transfer. How one should incorporate interparticle forces, which arise through van der Waals interaction, dynamic liquid bridges between wet particles, and tribocharging, in multiscale models warrants further study. Development of multiscale models that account for all the known interactions would improve confidence in the use of simulations to explore design options, decrease the number of pilot-scale experiments, and accelerate commercialization of new technologies.

中文翻译：

---

建立气体颗粒流中动量，种类和能量传输的本构模型。

由于多相结构是多相流中固有的，因此与动量，物种和能量的传输方程结合使用的本构模型与尺度有关。我们建议可以通过深度学习技术和运输方程式（例如漂移速度以及物种，能量和动量传递的类比）的附加量公式化来更好地量化这种规模依赖性。在多尺度模型中，如何将范德华相互作用，湿颗粒之间的动态液体桥以及摩擦带电产生的颗粒间力纳入其中，值得进一步研究。考虑到所有已知相互作用的多尺度模型的开发，将提高使用仿真探索设计方案的信心，减少试点规模实验的数量，