Uncertainty Quantification of Time-Dependent Quantities in a System with Adjustable Level of Smoothness

Abstract We summarise the results of a computational study involved with Uncertainty Quantification (UQ) in benchmark turbulent burner flame simulation. UQ analysis this simulation enables one to analyse convergence performance most widely-used uncertainty propagation techniques, Polynomial Chaos Expansion (PCE) at varying levels system smoothness. This is possible because simulations, smoothness time-dependent temperature, which study's QoI found evolve development state. deemed important as it known that PCE cannot accurately surrogate non-smooth QoIs and thus perform convergent UQ. While restriction gets accounted for, there no understanding whether quantifiable scaling relationship between PCE's metrics level QoI's It QoI-smoothness can be quantified by its standard deviation allowing observe effect on performance. for our flow scenario, exists power-law comparative parameter, defined measure relative Monte Carlo sampling, deviation, allows us make more weighted decision choice technique.