Identify contaminant sources in airliner cabins by inverse modeling of CFD with information from a sensor

In case contaminants are found in an aircraft cabin, it is useful to identify the contaminant source location and strength. This can be done through inverse Computational Fluid Dynamics (CFD) modeling. Since inverse CFD equations are ill-posed, this study proposes to solve a quasi-reversibility (QR) equation and a pseudo-reversibility (PR) equation. The QR equation improves the numerical stability by replacing the second-order diffusion term with a  fourth-order stabilization term in the governing equation of contaminant transport. The PR equation identifies contaminant sources by reversing airflow. Both methods solve for backward location probability density function (PDF) of contaminant, which can be linked to the contaminant source location. The source strength can be further determined by scaling the nominal contaminant concentration computed by CFD with the concentration measured by a sensor. Our study shows that the two methods can identify the contaminant source location and strength. The QR method performs slightly better than the PR method but with a longer computing time.