Stevens, Karl K.

The equivalent end deflections and rotations for beams with integral, but dissimilar, elastic supports were determined. Finite element analysis was used to generate the midsurface deflection of the beam. Numerical results were then fit to the analytical solution for the deflection of a beam, yielding the equivalent end slope resulting from deformations in the support. The lateral deflection at the support was available directly from the finite element calculation. The approach used for modeling of the supports is discussed. It was found that the slope and deflection at the support increase as the relative stiffness of the support decreases, as would be expected. Results are presented for both cantilever and beams with fixed ends, are valid for slender beams with small deflection.

Dynamic modeling of Printed Circuit Boards (PCBs) with mounted components, was
investigated via the example of a PC network card by removing components in different stages and
examining the resulting effects on the modal properties. Modal test results were compared with those
from an ANSYS finite element analysis.
Questions considered were: a) Do added components have a significant effect on the modal
properties of a PCB and what are the effects ? b) How much variation is there in natural frequencies
from board to board and test to test for a single board? c) Can a board with attached components
reasonably be modeled as a uniform elastic plate with an "equivalent" density and modulus of
elasticity?
Results obtained indicate that added components do have significant effects on the board
modal properties, less so for the lower modes than for the higher modes. There was only slight
variation in the natural frequencies from board to board and from test to test for a single board. For
the first two modes of vibration, it was found that the board considered could be modeled as a
uniform elastic plate with "equivalent" properties, provided an appropriate value of equivalent elastic
modulus was used. General findings, applicable to any PCB design, are presented.

A study of the stress distribution in and fracture behavior of the hermetic glass seal in a typical Integrated Circuit package is presented herein. Finite Element Analysis and Fracture Mechanics approaches were found effective for this investigation. A prescribed load or displacement applied at the tip of the lead protruding from the package causes high stresses at the lead-glass interface, which can lead to cracking and fracture of the seal. An approach for finding the value of the allowable load or displacement applicable at the lead tip is discussed. A correlation with a standard crack shape is presented for the 3-D model of the package. An extension of the problem revealing the effects of crack propagation on the stress intensity factor for the glass material is presented in later chapters. The J-integral method from Fracture Mechanics is found to be extremely useful for this investigation. A decline in the stress intensity factor with crack growth was observed from this study.

Proper application of matrix perturbation theory to the structural modification of vibrating systems yields accurate results without resolving the eigenvalue problem. For this reason it seems that the method may require less computer time and cost in performing iterative type design work. This thesis investigates the application of first-order perturbation techniques to a torsional shaft system subject to viscoelastic damping treatment modifications. The modifications are of general form in that the coating length and thickness are not restricted. Large modifications are built up by a series of small first-order perturbations. In addition to developing a better understanding of the usefulness of this application of perturbation theory, the goal of this paper is also to understand the nonproportional damping effects of partial viscoelastic coatings.

Modal analysis testing techniques were used to determine
the modal parameters and mode shapes for flexural vibration of a
Fiberizer Discmill disc to help define a safe operating speed range.
The modal parameters and mode shapes were determined for the
first six diametral modes, the first concentric mode, and the first
rigid body shaft mode. The disc consisted of a circular aluminum
plate with eight equally and symmetrically spaced radial cutters on
one face of the disc and eight equally spaced peripheral cutters
located on the rim of the disc midway between the angular locations
of the radial cutters. All cutters were of tool steel. The
safe area of operation is determined to be between 3000 rpm and
3600 rpm. The experimental setup and procedure Is described in
detail with particular emphasis on validation of assumptions
required for the analytical approaches used in calculating the modal
data.

The problem of identification. of distributed static
loadings of a one-dimensional beam is considered. All
response experimental data were derived from computer
simulations. True response data yield highly accurate
identification of the resultant forces; the
identification is, however, very sensitive to errors in
the response data. The results of the experimental data
are accurate within a small range of error.

This thesis develops an approach for predicting the
loss factors and natural frequencies of circular plates with
partial and complete viscoelastic damping surface
treatments. Plates with fixed and simply supported edges
were considered. The analysis is based on an energy
approach to the free vibration of plates. Results predicted
were compared with those obtained experimentally by Lopes.
Both the experimental and analytical studies clearly
indicate that substantial amounts of material and added
weight can be saved by using partial damping surface
treatments.

The finite element method is widely used in many areas of
engineering and applied science, including structures, fluids, heat
transfer, and electromagnetism. Several commercial programs exist
which have extensive capabilites in these areas, but they usually
require special training and mainframe computers to use. This creates
a series of problems in an engineering organization because of the
inefficiencies of timeshared computer systems and remote processing.
These problems result in poor productivity of the analyst.
This work demonstrates the feasibility of using a desktop
microcomputer to perform finite element structural analysis. A program
is developed which is capable of solving two-dimensional linear static
problems using truss, beam and plane stress elements. The program
includes extensive pre- and post-processing capabilities and graphic
displays. It allows problems with up to 500 degrees-of-freedom to be
solved within a few hours and requires only about one-half hour of
instruction to use. The program theory, a user's manual, and
verification problems are presented.

Application of modal testing techniques in the
determination of the modal parameters and mode shapes of an
edge-fixed circular plate with a free-layer viscoelastic
damping treatment extending over a portion of the plate
surface is described.
The mode shapes and modal parameters were determined
for the first five flexural modes of the plate, which had a
free-layer viscoelastic damping treatment on one side in the
form of a central circular patch of varying size.
The experimental setup and procedure is described in
detail, with particular emphasis upon problems of modal
overlap resulting from the symmetry of the test structure.
Results comparing the predicted and measured natural
frequencies and modal loss factors for varying degrees of
damping treatment are presented for select cases.

An experimental investigation to determine the effectiveness
of partial constrained layer damping treatments for a
clamped rectangular plate is described. The impulse testing
technique was used with a Hewlett Packard '5423A structural
Dynamics Analyzer' to determine modal parameters of
the first five flexural modes. The results obtained are
compared with theoretical results and they are in agreement.
The results indicate that partial constrained layer damping
treatments, if properly used, can be more effective than
complete treatments.