This work presents a framework to predict near-optimal anisotropic spacing functions suitable to perform simulations with unseen operating conditions or geometric configurations. The strategy consists of utilising the vast amount of high-fidelity data available in industry to compute a target anisotropic spacing and train an artificial neural network to predict the spacing for unseen scenarios. The trained neural network outputs the metric tensor at the nodes of a coarse background mesh that is then used to generate meshes for unseen cases. Examples are used to demonstrate the effect of the network hyperparameters and the training dataset on the accuracy of the predictions. The potential of the method is demonstrated for examples involving up to 11 geometric parameters on CFD simulations involving a full aircraft configuration.

This work presents the first method for generating triangular surface meshes in three dimensions for the NURBS-enhanced finite element method. The generated meshes may contain triangular elements that span across multiple NURBS surfaces, whilst maintaining the exact representation of the CAD geometry. This strategy completely eliminates the need for de-featuring complex watertight CAD models and, at the same time, eliminates any uncertainty associated with the simplification of CAD models. In addition, the ability to create elements that span across multiple surfaces ensures that the generated meshes are highly compliant with the requirements of the user-specified spacing function, even if the CAD model contains very small geometric features. To demonstrate the capability, the proposed strategy is applied to a variety of CAD geometries, taken from areas such as solid/structural mechanics, fluid dynamics and wave propagation.