Automatic Gap Detection and Patching

Introduction

If you look at your car closely, you will find many gaps. For example, look between the fender and the hood. There is about 4-6 mm gap. Such gaps exist all over the car.  

Challenges

For performing external aerodynamics the required final mesh size on these surfaces are roughly the same size or even smaller than the gap. We can't mesh it coarser in order to walk over the gaps! This poses few challenges in pre-processing.

1. If you do refine the gap and mesh it properly, you will end up with very high cell-count, specially if you try to grow prism layers.
2. If you do grow prism layers in the gap, you can't grow many layers in the gap. Many stair-steps away from such gaps. There will be too many places where you have to stair-step. This will introduce robustness issues in the solver.

Standard Practice

Standard practice has been to manually patch the gaps before meshing. However, closing all the gaps on the exterior surfaces of a car is one of the biggest challenges of the pre-processing. It can easily take a full day or even more. 

New Automated

There are advance wrapper operations in Fluent Meshing to automatically detect and patch most of the gaps in a car body and these operations are already scripted.

The remaining gaps you can manually close, or let the automated meshing process deal with it. 

The automatic meshing process that we have introduced in these posts is able to mesh the gaps (without patching them) with continuous prism layers, as seen below:

But if you are concerned about reducing total cell count, then you have the option of removing most of those gaps automatically.