Much of the research that I undertake has its primary focus on the aerospace industry. I do however try as much to take an equal look at the aerospace and automotive sectors to see what things can be learnt from both. For example, in the automotive sector the best-practice for meshing is typically unstructured grids with many people saying that structured grids are purely for academic studies. However during my time at NASA I saw first hand the way that overset structured meshes can be used for extremely complex geometry such as an entire aircraft. In fact, the standard best-practice for Boeing is overset structured meshes. The mesh quality is very high but compared to unstructured meshes there is a mesh generation penalty particularly if you are investigating new geometry on a weekly basis.
I was a member of the organising committee of the 3rd AIAA High-Lift workshop (https://hiliftpw.larc.nasa.gov) to generate grids and benchmark different turbulence modelling options. A paper on this has been published in Journal of Aircraft which you can see below. The main conclusions were that OpenFOAM showed very similar levels of accuracy to the commercial code STAR-CCM+ using the Spalart-Allmaras RANS model but both were still not able to predict the correct stall behaviour. In addition it was found that OpenFOAM was less robust and computationally efficient than STAR-CCM+ and thus took longer to achieve the same result (with more manual changes in CFL etc to keep stability).Ashton, Skaperdas - 2019 - Verification and Validation of OpenFOAM for High-Lift Aircraft Flows
In another paper below I looked at how structured and unstructured compared for a multi-element airfoil from the recent BANC aeroacoustics workshop. The paper ended up focusing on a problem with encountered with the IDDES model (sensitivity to the boundary layer meshing) but ignoring this the results were quite similar. The mesh generation process only took a few days and could easily adapted for different geometry. We did however found that the regions between the prismatic boundary layer and the hexahedral cut-cell off-body region caused poor mesh quality which did cause some problems for the acoustics. I am currently working on developing improved unstructured meshing e.g advancing front method to remove this problem.Ashton, West, Mendonça - 2016 - Flow Dynamics Past a 30P30N Three-Element Airfoil Using Improved Delayed Detached-Eddy Simulation