Aerosol particle deposition rates on surfaces inside a clean room are predicted by a model developed to account for particle convection, diffusion and sedimentation. External forces acting on the particle also influence the rate of deposition. Both electrical charge build up on product surfaces and temperature gradients in the air near the product surface are known to effect the rate of deposition. A description of an electrostatic and thermophoretic force on the particle is thus included in the model. The equations governing the particle deposition process and the approach used in obtaining a solution to these equations are both described. A finite element numerical solution is detailed, followed by a description of the electrostatic force models. Finally, predictions of the model are presented with a comparison to data experimentally obtained by other researchers.

No description available.

Call number: LD2668 .T4 1986 L43 / Master of Science / Civil Engineering

In an attempt to achieve more efficient designs, the technological frontier is pushed further and further. Every year science probes for a better understanding of natural phenomena, discovering new and improved methods to perform the same task more efficiently and with better results. One of the new technologies is the nonlinear analysis of structural systems using inelastic post-buckling member performance. Inelastic post-buckling member performance is defined as the constitutive relationship between axial load and displacement after the ultimate member capacity has been exceeded. A nonlinear analysis is able to predict the failure behavior of a structural system under ultimate loads more accurately than the traditionally used linear elastic analysis. Consequently, designs can be improved and become more efficient, which reduces the realization cost of a project. An improved nonlinear analysis solution algorithm has been developed, that allows the analyst to perform a nonlinear analysis using post-buckling member performances faster than previously possible. Furthermore, the original post-buckling member performance database was expanded using results obtained from physical member compression tests. Based on the experimental results, new post-buckling member performance model curves were developed to be used together with the improved nonlinear solution algorithm. In addition, a program was developed that allows the analyst to perform a valid nonlinear analysis using a finite element program (LIMIT). The program combines a numerical pre-processor, and input and output data evaluation modules based on human expertise together with the LIMIT analysis package. Extensive on-line help facilities together with graphical pre- and post-processors were also integrated into the program. The resulting analysis package essentially combines all of the necessary components required to perform a nonlinear analysis using post-buckling member performances into one complete analysis package.

Brushless Doubly Fed Machines (BDFM) have potential advantages in variable speed generation and adjustable speed drive applications. The most significant of these advantages is a reduction in the power electronic converter rating, and therefore a reduction in overall system cost. Presently, efforts are being directed at optimizing the design of the BDFM and investigating areas of commercial feasibility. One possible aid in the investigation of design alternatives is finite element analysis. Finite element analysis is a numerical method for determining the field distribution in a dimensional model. Finite element techniques have been successfully used for some time in the design of induction, reluctance and permanent magnet machines. However, the characteristics of the BDFM require adjustment of the finite element design procedure used for conventional singly-fed induction machines. In this thesis, a three-dimensional finite element design procedure for modeling the BDFM has been developed. This design procedure avoids the difficulties previously associated with finite element modeling of the BDFM. The three-dimensional finite element design procedure developed in this thesis was used to model the 6/2 pole 5 horsepower BDFM laboratory machine. From the simulation results, the induced currents in the BDFM rotor bars were calculated. In the course of investigating three-dimensional finite element analysis for the BDFM, two different commercially available finite element analysis software packages were examined and tested. The first was Maxwell 3D Field Simulator produced by Ansoft Corporation, and the second was MSC/EMAS (Electromagnetic Analysis System) and MSC/XL by MacNeal-Schwendler Corporation. These two software packages are compared and their advantages and disadvantages/limitations are discussed. A tutorial for setting up and solving a three-dimensional BDFM model using MSC/XL and MSC/EMAS is presented. This goal of this tutorial is to guide a new user of MSC/XL and MSC/EMAS through the creation, setup, simulation, and analysis of a BDFM model. This tutorial contains condensed information included in the MSC/XL and MSC/EMAS program documentation provided by MacNeal-Schwendler. In addition, modeling techniques particular to the BDFM, which are not included in the program documentation, are described. This tutorial is applicable only to those individuals interested in learning how to use MSC/XL and MSC/EMAS in order to simulate a BDFM model. / Graduation date: 1995

No description available.

No description available.

Experimental characterisation of rubber in uniaxial, equi-biaxial and planar tension under cyclic quasi-static loading shows strain-induced stress softening, hysteresis and unrecoverable strain. The objective of this work is to study the applications and limitations involved in predicting the behaviour of rubber with hyperelastic models. To assume a preconditioned perfectly elastic material, the data obtained from experiments must first be simplified. The data is then fitted to popular hyperelastic models in the finite element analysis (FEA) software ANSYS(TM). A single hyperelastic model (with given coefficients) is shown to only provide a good fit to a single characterisation test and level of preconditioning at the time. A two-iteration preconditioning method is developed using different hyperelastic models for a given material to approximate the softening effect of cyclic loading in a static FEA simulation. A biaxiality test is developed, providing information on the dominant mode of simple strain in the elements of a FE model. FEA simulations and experimental tests of a cantilevered rubber plate subjected to a bending load at its free end as well as a rubber guide lug subjected to a transverse deflection are presented and discussed. It is shown that using a single hyperelastic model is insufficient to predict the behaviour of these experiments in FEA simulations. The preconditioning iteration, when applied to these simulations, shows very good agreement with the experiments, both qualitatively and quantitatively. The biaxiality test provides insight on which characterisation test is the most appropriate for curve fitting hyperelastic models for a given analysis.

No description available.

No description available.