Validation of Models for the Flow of Granular Media

TL;DR

This paper emphasizes the importance of using stochastic models and statistical inference for validating powder flow models, highlighting the need for proper assessment to ensure accurate physical representation.

Contribution

It introduces the application of spatial and multivariate statistical methods for fitting and validating models of granular media flow.

Findings

Proper model validation prevents failure to capture physics

Statistical inference improves model fitting accuracy

Assessment methods are crucial for model reliability

Abstract

Validation of models for powder flow requires that the models be stochastic and that they be fit by statistical inference. Methods from spatial and multivariate statistics can be used for model fitting and assessment. If the quality of the fitted model is not assessed, there is a significant risk the model will fail to represent the physics of powder flow.