A dissertation submitted to the Faculty of Engineering, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Master of Science in Engineering. ,Department of Metallurgy and Materials Engineering University of the Witwatersrand Johannesburg, 1990. / An instrumented hot deformation simulator was desLgned, cons'.:ructedand commissioned at Iscor's Pilot Plant. A modified servohydraulic machine, using plane strain compression, simulated industrial plate rolling schedules. The simulation test included induction heating, multiple pass plane strain compression and either air cooling or quenching. The movement of the specimen between the different test stages was computer controlled. Accurate control of specimen temperature, strain atrain rate and interpass hold times was achieved by means of sophisticated data acquds i+Lon equipment. A comput er programme was written to simulate the hot. rolling of plate in t erms of roll sepa ratIng f.orces, ffiE.\tallurgical changes during deformation as well as the final microstructure and mechanical properties of SGa carbon manganese steel. Multiple linear regression was pe.rformed on the results obtained from mUlti-pass plane strain compression tests. It was found that themical composition, finish temperature and finish strain were the most important process parameters affecting yield strength and impact energy of air cooled place. A computer model was developed to simulate the temperature distribution in the deformation zone of a plane strain compressLon specimen at any point during or after a mUlti-pass test.

The Oregon Department of Transportation is responsible for many steel deck truss bridges containing connection details that are fatigue prone. A typical bridge, the Winchester Bridge in Roseburg, Oregon, was analyzed to assess the loading conditions, stress levels, and fatigue life of the connection details. The analysis included linear-elastic beam analysis, 2D and 3D finite element modeling, and fatigue modeling. A field identification methodology was developed to expand the analysis to other steel deck truss bridges. Five retrofit strategies were investigated to determine their effectiveness in reducing the stress ranges developed in the connection details. / Graduation date: 1998

No description available.

The post-buckled performance of cross-braced single angles was experimentally determined. The results of this study will be used by the Bonneville Power Administration for the analysis of member performance in transmission towers.

No description available.

Industry is constantly searching for improved materials for use in highly demanding applications. The materials chosen must withstand a wide range of temperatures and extended exposure in aggressive environments, including hydrogen gas. Because of the risk of catastrophe if brittle failure occurs, careful material selection is imperative. Austenitic stainless steels may be a likely choice for hydrogen service because their behavior in high pressure hydrogen ranges from no apparent damage to relevant, but generally small ductility loss (13). Because of this Variation in behavior, a single category cannot be established to encompass all austenitic steels and studies must be performed on each type of steel to determine its behavior under specific circumstances. Two steels being currently under consideration for use in hydrogen are Armco 21-6-9 and Tenelon, both are fully austenitic stainless steels which may be used over a wide range of temperatures, including service at liquid nitrogen temperature. / Master of Science / incomplete_metadata

M. S.

To provide a basis for the development of interaction curves for tubular beam-columns of annular cross section, a general purpose beam-column computer program is developed, and used to determine ultimate load capacities. The paper presents the analytical model and the computer method. The analytical results are compared with published test data as well as experimental data obtained as part of this project.

As typically designed, the beam-column-brace connections of buckling-restrained braced steel frames have undesirable failure modes that compromise the integrity and performance of the frames and are costly to repair. To decrease the time and resources needed to repair the frames following an earthquake, a new connection design was developed that attempts to confine yielding to replaceable frame components. The design incorporates a gap in the beam beyond the edge of the beam-gusset weld that acts as a hinge and reduces moment forces transferred to the connection; it is bridged by splice plates that are bolted to the beam top flanges. The splice plates and buckling-restrained braces are the only frame components that are expected to yield. To investigate the performance of the proposed connection design, a prototype bay was designed and two test specimens were fabricated and tested. Each specimen represented a corner of the prototype braced bay and consisted of a beam, column, gusset plate, brace core extension assembly, splice plates, and lateral bracing angles. Both standard design procedures and newly developed criteria were used to design the connection. In preparation for testing, a method was developed for estimating the hysteretic response of a buckling-restrained brace. By using this method to program an actuator, the specimens could be tested without using actual braces, resulting in a significant reduction in testing cost. Testing was conducted using two 600 kip actuators; the first followed a static loading protocol with a maximum design drift of 6.5%, and the second replicated the prototype BRB's response. The tests yielded promising results: both specimens withstood the maximum displacements and avoided yielding in the beams, columns, and gusset plates; yielding did occur in the splice plates and BRB core extension assembly, as anticipated. Possible limitations in the design may arise under the presence of increased shear loads, concrete floor slabs, or out-of-plane loading. Additional testing is recommended.