Publications

Model-based computer simulations have revolutionized product development in the last 10 to 15 years. Technologies that have existed for many decades or even centuries have been improved with the aid of computer simulations. Everything from low-tech consumer goods such as detergents, lubricants and light bulb filaments to the most advanced high-tech products such as airplane wings, wireless communication technologies and pharmaceuticals is engineered with the aid of computer simulations today. In this paper, we present a framework for describing computational tools and their application within the context of product engineering. We examine a few cases of product development that integrate numerical computer simulations into the development stage. We will discuss how the simulations were integrated into the development process, what features made the simulations useful, the level of knowledge and experience that was necessary to run meaningful simulations and other details of the process. Based on this discussion, recommendations for the incorporation of simulations and computational tools into product development will be made.

Abstract not provided.

Abstract not provided.

formulation to satisfy velocity boundary conditions for the vorticity form of the incompressible, viscous fluid momentum equations is presented. The tangential and normal components of the velocity boundary condition are satisfied simultaneously by creating vorticity adjacent to boundaries. The newly created vorticity is determined using a kinematical formulation which is a generalization of Helmholtz` decomposition of a vector field. Though it has not been generally recognized, these formulations resolve the over-specification issue associated with creating voracity to satisfy velocity boundary conditions. The generalized decomposition has not been widely used, apparently due to a lack of a useful physical interpretation. An analysis is presented which shows that the generalized decomposition has a relatively simple physical interpretation which facilitates its numerical implementation. The implementation of the generalized decomposition is discussed in detail. As an example the flow in a two-dimensional lid-driven cavity is simulated. The solution technique is based on a Lagrangian transport algorithm in the hydrocode ALEGRA. ALEGRA`s Lagrangian transport algorithm has been modified to solve the vorticity transport equation and the generalized decomposition, thus providing a new, accurate method to simulate incompressible flows. This numerical implementation and the new boundary condition formulation allow vorticity-based formulations to be used in a wider range of engineering problems.

A description of the flow field in an overflow wafer rinse process is presented. This information is being used in an initiative whose principal objective is to reduce the usage of water in wafer rinsing. The velocity field is calculated using finite-element numerical techniques. A large portion of the water does not contribute to wafer rinsing.

The objective of the ``Hydrodynamics of Maneuvering Bodies`` LDRD project was to develop a Lagrangian, vorticity-based numerical simulation of the fluid dynamics associated with a maneuvering submarine. Three major tasks were completed. First, a vortex model to simulate the wake behind a maneuvering submarine was completed, assuming the flow to be inviscid and of constant density. Several simulations were performed for a dive maneuver, each requiring less than 20 cpu seconds on a workstation. The technical details of the model and the simulations are described in a separate document, but are reviewed herein. Second, a gridless method to simulate diffusion processes was developed that has significant advantages over previous Lagrangian diffusion models. In this model, viscous diffusion of vorticity is represented by moving vortices at a diffusion velocity, and expanding the vortices as specified by the kinematics for a compressible velocity field. This work has also been documented previously, and is only reviewed herein. The third major task completed was the development of a vortex model to describe inviscid internal wave phenomena, and is the focus of this document. Internal wave phenomena in the stratified ocean can affect an evolving wake, and thus must be considered for naval applications. The vortex model for internal wave phenomena includes a new formulation for the generation of vorticity due to fluid density variations, and a vortex adoption algorithm that allows solutions to be carried to much longer times than previous investigations. Since many practical problems require long-time solutions, this new adoption algorithm is a significant step toward making vortex methods applicable to practical problems. Several simulations are described and compared with previous results to validate and show the advantages of the new model. An overview of this project is also included.

A numerical method to simulate viscous diffusion of vorticity using vortex blobs (i.e., without a grid) is presented. The method consists of casting the effects of viscous diffusion into an effective ``diffusion velocity`` at which vortex blobs convect. The diffusion velocity was proposed previously by Ogami and Akamatsu, but they did not consider the effects of the divergence of the diffusion velocity. In fact, the diffusion velocity is highly non-solenoidal, which significantly affects the area over which a vortex blob diffuses. A formulation is presented that relates the area expansion to the diffusion velocity divergence. By taking into account the area expansion, more accurate simulations of diffusion are obtained, as demonstrated by a comparison of numerical and analytical diffusion solutions. Results from simulations show that vortex areas expand significantly in regions of large vorticity gradients. As a result of the area expansion, adjacent vortices remain overlapped, thereby maintaining smooth solution fields. The non-solenoidal diffusion velocity method is easily implemented in vortex blob algorithms, thus facilitating the development of vortex methods to simulate flows with finite Reynolds numbers.