The conventional design of steel beams considers that any deformation of a member due to applied load must be a combination of the four rigid body modes (axial deformation, major axis bending, minor axis bending and twisting) i. e. the member retains its cross sectional shape without distortion. In a hot rolled member the warping stresses which arise due to violation of the assumption that plane sections remain plane can often be neglected. In thin walled sections, however, these warping stresses are typically of the same order of magnitude as the primary bending stresses induced in the member by the applied loading and therefore cannot be neglected. In addition, if plane sections do not remain plane, the cross section distorts when a load is applied. The first part of this Thesis presents a method of analysis for any open unbranched thin walled section which considers both rigid body movement and cross section distortion (including local buckling). The method is such that the four rigid body modes are automatically identified and separated from the remaining cross section distortion modes. The second part of this Thesis develops a finite difference method of analysis, in conjunction with the theory of part I, to consider the behaviour of a member subject to any arbitrary loading condition and end restraint. Both first order linear problems and second order elastic critical buckling problems are solved, including the interaction of local buckling, overall buckling and cross section distortion.

M.Ing. (Civil Engineering) / The results of a study presented on the behaviour of cold-formed stainless steel beams subjected to shear taking into account elastic shear buckling, inelastic shear buckling and shear yielding, as well as an investigation into the interaction relationship between shear and bending are presented. From this investigation the results obtained show good relation to the theory. The local shear buckling stress was experimentally determined and it was found that for unreinforced beam webs the shear buckling coefficient is that of an infinitely long plate, namely k=5,34. A good agreement between the experimental ultimate shear strength and the predicted ultimate shear strength was found. Stainless steel beams comprising of lipped channels were manufactured and tested to failure. The types of stainless steels used in this investigation were Types 304, 316, 430 and Type 3CR12 corrosion resiting steel, a modified Type 409 stainless steel. The stress-strain relationship for stainless steels differs from that of carbon steel in that stainless steel is a gradual yielding material. It was concluded in this investigation that Gerard's plasticity reduction factor, Gs/G o, should be used as a plasticity reduction factor in calculations concerning shear. It was found that present design criteria are adequate.

The sensitisation of a 12% chromium ferritic stainless steel, conforming to EN 1.4003 and available commercially in South Africa under the trade name of 3CR12, was investigated during the course of this project. 3CR12 was designed to pass through the (<font face="symbol">a</font>+<font face="symbol">g</font>) phase field on cooling, with the austenite transforming to martensite on subsequent cooling to room temperature. The aim of this investigation was to verify that 3CR12 can sensitise during continuous cooling after welding, provided that low heat input levels are used. Two grades of 3CR12 with slightly different chemistries, designated 41220 (A) and 41311 (B), were evaluated. Grade 41220 has a higher austenite potential than grade 41311. 3CR12 plate was joined autogenously to AISI 316L by means of a series of square butt welds. Gas tungsten arc welding with argon shielding was used, and the heat input was varied from approximately 30 J/mm to 450 J/mm, in conjunction with welding speeds ranging from 2.36 mm/s to 33.3 mm/s. Rosenthal’s heat flow equations were used to calculate the cooling rate from 1500ºC to 800ºC for each experimental weld, and to illustrate the influence of the effective heat input and welding speed on the martensite content of the high temperature heat-affected zone. An increase in welding speed reduces the heat input and increases the cooling rate after welding. At lower heat input levels (less than approximately 100 J/mm), austenite nucleation was found to be suppressed by the rapid cooling rates, and a continuous network of ferrite-ferrite grain boundaries formed in the high temperature heat-affected zone. Higher heat inputs resulted in slower cooling with more martensite in the high temperature heat-affected zone after cooling. At heat input levels above approximately 250 J/mm, enough martensite formed during cooling to eliminate a continuous network of ferrite-ferrite grain boundaries in the high temperature heat-affected zone. Sensitisation was evaluated using an electrolytic oxalic acid etch (ASTM 763-99, Practice W), and a potentiostatic etch in 0.5M H2SO4. During the potentiostatic etch test, the potential was maintained at 0 VSCE to reveal the presence of any chromium depleted zones. Both grades of 3CR12 were found to be sensitised when a continuous network of ferrite-ferrite grain boundaries was present in the high temperature heat-affected zone (i.e. after welding at low heat input levels). When the heat input during welding was high enough to ensure the presence of martensite on the majority of the heat-affected zone grain boundaries, thereby effectively eliminating continuous ferrite-ferrite grain boundary networks, the welds were not in the sensitised condition. The austenite that forms during cooling acts as a carbon sink, absorbing any excess carbon. This prevents supersaturation of the ferrite and subsequent carbide precipitation that can lead to chromium depletion and sensitisation. Due to its higher austenite potential, grade 41311 can be welded at lower heat input levels and with faster cooling rates than grade 41220 without inducing continuous carbide precipitation and sensitisation. In order to prevent sensitisation, a fusion-line cooling rate of 80ºC/s should not be exceeded in 3 mm 3CR12 plate during welding. / Dissertation (MSc(Applied Science))--University of Pretoria, 2007. / Materials Science and Metallurgical Engineering / unrestricted

The microstructure of high-strength weld deposits has been investigated using thermodynamic analysis and phase transformation theory, backed by experimental confirmation. The microstructures of both the fusion and reheated zones of multirun alloy-steel weld deposits have been studied. The transformation mechanism of acicular ferrite has been established, and a theory for reaustenitisation in steel weld deposits has also been proposed. In addition to the studies on weld metals, some model alloys were fabricated and tested in order to confirm some of the predictions made by the new theory for reaustenitisation. All the weld metals studied in the investigation possess good hardenability. The primary microstructures of the fusion zone of these deposits consist mainly of acicular ferrite with very little allotriomorphic ferrite.

No description available.

No description available.

Two-phase composites have been prepared by infiltrating sintered iron compacts with liquid copper. The effects have been studied of iron particle size, matrix mean free path, and the volume fraction and micro-hardness of the iron-rich constituent, on the tensile properties of composites. It has been found that the strength of the composites is related to the amount of solution hardening of the iron component during infiltration. The results of tensile tests have suggested that the hardness of the iron-rich constituent is the dominant factor controlling yield strength, ultimate tensile strength and elongation. However, the ultimate strength has been found to depend also on the volume fraction of the hard constituent, and elongation has also been found to be a function of the interface area. / Applied Science, Faculty of / Materials Engineering, Department of / Graduate

This paper is a report of the investigations carried out to determine the effect of microstructure on the rate and nature of the formation of austenite in some commercial steels. Previous work is reviewed and compared with the results of current experiments. Experimental procedure is briefly discussed. Steels used for these experiments were similar to the following S.A.E. specifications; 1020, 1045, 1080, 4140, and 52100. Studies of the 1080 type form the greater part of the work. Results are presented in graphs of hardness against time-at-austenitizatlon-temperature and in a series of photographs showing various stages of transformation for different prior structures. Conclusions drawn from results are discussed. The processes of nucleation and growth of austenite in pearlite, spheroidite, bainite, martensite, and sorbite are examined. The effects of lamellar spacing, size and distribution of spheroids, and proeutectoid constituents are noted, together with general considerations such as stability of micro-structures. The effect of chromium is briefly discussed. Possibilities for future work on austenitizatlon are presented, with emphasis on induction heating and flame hardening. / Applied Science, Faculty of / Materials Engineering, Department of / Graduate

This study is carried out to find the effect of nickel on the beginning of isothermal transformation of austenite in an iron-carbon-molybdenum alloy containing 0.55 carbon and 0.35 molybdenum. An introduction describes the iron-carbon equilibrium system, the products of slow cooling of austenite, the relation between slow cooling and isothermal transformation at temperatures below equilibrium, and a full discussion of terminology used. A literature review discusses theories of transformation and previous work on the effect of nickel on austenite transformation. Development of experimental technique in isothermal transformation and melting of pure alloys is discussed. The isothermal transformation diagrams are shown for beginning of transformation of austenite of base composition 0.55 carbon and 0.35 molybdenum, relatively free from impurities (silicon, manganese, etc.), showing the effect of nickel on the beginning of transformation. Nickel additions used are 0, 2.13, 3.69, and 5.31 percent. Isothermal transformation is shown by photomicrographs which are discussed fully. The effect of increasing nickel on the isothermal transformation of an alloy containing 0.55 carbon and 0.35 molybdenum is as follows: (1) the pearlite reaction is delayed appreciably: (2) ferrite formed at intermediate temperatures (880 to 1000 deg. F.) becomes more prominently acicular, the acicular ferrite reaction taking the place of the upper bainite reaction of low nickel alloys: (3) the acicular ferrite reaction is followed first by rejection of carbide particles, and later by agglomeration and growth of the carbide phase: (4) the acicular ferrite and feathery bainite reactions as represented on the isothermal transformation diagram become separated by the appearance of a bay in the isothermal transformation curve. / Applied Science, Faculty of / Materials Engineering, Department of / Graduate

Galvanized sheet is produced commercially by hot dipping steel sheet in a galvanizing bath to produce a thin zinc rich protective coating on the steel. This investigation is concerned with the solidification of the galvanized layer with particular reference to the growth of large "spangles" in the galvanized layer. The manner in which the galvanized layer solidifies was found to be dependent on a number of factors including melt supercooling, nucleation rate, dendritic growth, bath composition, bath impurities, and cooling rate. The present results show that the supercooling necessary to activate nucleating sites in a typical galvanizing bath is less than 1°C. The supercooling which occurs in the galvanized layer as it solidifies is also generally less than 1°C, contrary to results reported in the literature. This suggests that grains nucleate in the bulk of the galvanized layer, and not necessarily at the air or iron surfaces of the melt as has been reported. No clear evidence was obtained which shows that spangles, with large dendrite spikes, are associated with large melt supercooling. Accordingly, spangle formation cannot be attributed to higher dendritic velocities resulting from large thermal supercooling. The surface topography of the galvanized layer is a property of major consideration in industrial applications of the galvanized product. Present measurements show that the surface topography is strongly dependent on the solidification structure of the galvanized layer. Large spangles solidify dendritically with large variations in the surface topography. The spangles have shiny and frosty sectors, which are relatively smooth and rough respectively, and inclined to the steel sheet surface. Large depressions are present at the boundaries between adjacent spangles, termed "pulldown", which can markedly reduce the effective thickness of the galvanized layer, and cannot be removed by subsequent treatment of the galvanized sheet. Increasing the concentration of lead in the bath increased the pulldown. The mechanism of pulldown formation is not clear. Volume shrinkage during solidification cannot account for the large depressions observed. The hot dipped galvanized samples prepared in this investigation used galvanized sheets as starting material. Observation of the distribution of bath alloy additions in the samples, using microprobe analysis and radioactive tracers, clearly showed that the original galvanized layer was replaced by the metal from the bath. Solute segregation in a galvanized layer containing spangles is directly associated with the dendritic growth of the spangles. Solute is depleted in the dendrite spikes and concentrated between the spikes. No solute concentration was observed at the grain boundaries between adjacent spangles. Some preferential surface segregation, associated with shiny and frosty sectors of a spangle, was observed. The growth of large spangles in the galvanized layer is directly related to the galvanizing bath composition. Large spangles are obtained with alloying additions which have very limited solid solubility in zinc and relatively low liquid surface tensions. The diameter of the spangles decrease as the surface tension of the alloy addition increases. Spangle growth is associated with dendritic growth. Dendritic growth occurs as a result of constitutional supercooling at the dendrite tip due to solute segregation during growth at the solid/liquid interface. It is proposed that the presence of a thin layer of highly concentrated solute at the dendrite tip changes the curvature of the tip. The change is related to the liquid surface tension of the solute. Solutes with lower values of surface tension decrease the tip curvature which results in an increase in dendrite velocity and larger spangles. The orientation of the spangles is shown to vary appreciably. The basal plane of a spangle is observed to be inclined to the surface of the steel sheet at angles between 8 and 80 degrees. This differs from reports which indicate that the basal plane is nearly parallel to the steel sheet. The growth of a spangle is primarily associated with dendritic growth. Small (0001) platelets of solid form at heterogeneous nucleating sites in the melt from which <1010> spikes grow. Each spangle forms from a single nucleating source which is randomly oriented with respect to the surface of the steel sheet. As the dendrite spikes grow they shortly encounter the melt/air interface or the melt/steel interface. The spikes will continue to grow along the surfaces at a high velocity in a direction defined by the initial <1010> direction of the growing spikes. As the primary spike grows, secondary and tertiary spikes form, generally inclined to the melt surface and in <10l0> directions when possible. The secondary branches of a primary stalk which grow along the melt/air interface form shiny spangle segments. The secondary branches on the opposite side of the primary spike grow along the melt/steel surface and form frosty sectors. The difference is not due to an orientation difference between the spangles as reported in the literature. / Applied Science, Faculty of / Materials Engineering, Department of / Graduate