Abstract:
A mid-to-long range roadmap for industrial simulation software must include many of the considerations outlined in the CFD Vision 2030 document.  Changing hardware platforms and cloud-based service delivery demand evolution of the CFD algorithms themselves in order to leverage the available computing power.  The increasing demand for tractable high-fidelity unsteady flow simulations also places demands on modeling and post-processing.  Mesh generation, as always, requires technological advances that accompany increases in geometric complexity and the requirements of the solver evolving algorithms.  Advances in adjoint methods increase the access to technologies such as multi-objective design, robust design and topology optimization. Finally, in the industrial context, many applications demand a multidisciplinary approach to engineering simulation involving high fidelity methods, as well as system modeling including the use of advanced Reduced Order Models. This talk will attempt to bridge from some of the key recommendations of the NASA study to the commercial software development context, keeping in mind the evolving and anticipated needs of of engineering simulation software customers.