In classical description, the mechanical state of a material point depends on the variables defined at this point solely. It can integrate and catch some aspects of the material’s microstructure by conventional homogenization method. The application of the conventional continuum assumption results in a simplified description of the system which makes the large scale simulation of the material more efficient but at the expense of a loss of information at small length scales. Localization is a phenomena where a large degree of deformation occurs in highly concentrated regions. The conventional continuum theory with strain softening can not give the convergent solution as the size of the localization zone is completely determined by the mesh discretization. The multiresolution continuum theory (MRCT) is a higher order continuum theory where additional kinematic variables supplementing the conventional macroscopic displacement field are added to account for deformations at several distinct length scales. The direct inclusion of the length scale parameters in the material’s constitutive equations remedies the convergence problem. In crystalline materials the initiation of plastic flow and subsequent permanent plastic deformation is attributed to the presence and movement of dislocations and also the interactions between the dislocation themselves and different kinds of obstacles, inclusions, second phase particles and grain boundaries etc. Some of these defects can alsolead to damage initiation in the materials. For example, the stresses developed at the dislocation pile-ups contribute to the initiation of the microvoids and microcracks. A dislocation density based damage model has been developed and combined with a physically based flow stress model. They are calibrated and validated for 316L stainless steel at different temperatures and strain rates. These models have been implemented into the macroscopic material description of the MRCT element. / <p>Godkänd; 2016; 20160215 (haoqin); Nedanstående person kommer att disputera för avläggande av teknologie doktorsexamen. Namn: Hao Qin Ämne: Materialmekanik/Material Mechanics Avhandling: Multiresolution Continuum Theory and Dislocation Density Based Constitutive Relations Opponent: Professor emeritus Kenneth Runesson, Avd för material- och beräkningsmekanik, Institutionen för tillämpad mekanik, Chalmers tekniska högskola, Göteborg. Ordförande: Professor Lars-Erik Lindgren, Avd för material- och solidmekanik, Institutionen för teknikvetenskap och matematik, Luleå tekniska universitet, Luleå. Tid: Måndag 25 april, 2016 kl 09.30 Plats: E246, Luleå tekniska universitet</p>

Ductile fracture presents challenges with respect to material modelling andnumerical simulations of localization. The strain and damage localization maybe unwanted as it indicates a failure in the process or, as in the case ofmachining and cutting, a wanted phenomenon to be controlled. The latterrequires a higher accuracy regarding the modelling of the underlying coupledplastic and fracturing/damage behaviour of the material, metal in the currentcontext as well as the stability and robustness of the simulation procedure.This aim of this work is to develop, evaluate and implement formulations thatcan efficiently and reliably handle localization problems in machiningsimulations. The focus is on non-local models. The non-local models extendthe standard continuum mechanics theory by using non-local continuumtheory in order to achieve mesh independent results when simulating fractureor shear localization.The non-local damage model is implemented and various formulations areevaluated in a Matlab™ based finite element code. The chosen algorithm wasthen implemented in commercial software. The implementations remedy themesh sensitivity problem and gives convergent solution for metal cuttingsimulations with reasonable cost. The length scale associated with the nonlocalmodels are in the current context considered as a numericalregularization parameter. The model has been applied in machiningsimulations and compared with measurements from industry.Keywords: Finite element simulation; Non-local damage; Plasticity; Machining / Godkänd; 2016; 20160222 (oluabi); Nedanstående person kommer att disputera för avläggande av teknologie doktorsexamen. Namn: Olufunminiyi Abiri Ämne: Materialmekanik/Material Mechanics Avhandling: Non-Local Models in Manufacturing Simulations Opponent: Professor Miguel Vaz, Dept Mechanical Engineering, State University of Santa Catarina, Joinville, Brazil. Ordförande: Professor Lars-Erik Lindgren, Avd för material- och solidmekanik, Institutionen för teknikvetenskap och matematik, Luleå tekniska universitet, Luleå. Tid: Fredag 22 april, 2016 kl 09.30 Plats: E231, Luleå tekniska universitet

This work has concerned the preparation and properties of all-oxide composites. The most common examples of such materials are composites of oxide particles in an oxide matrix, continuous oxide fibres in an oxide matrix and layered composites consisting of alternate layers of different oxides. In the case of continuous fibre composites, the matrix and fibres can be of the same oxide since the mechanical properties of the composite are to a large extent influenced by the interface between the two. All-oxide composites are of interest as possible high temperature materials since they are inherently oxidation resistant. The emphasis of the thesis is on continuous fibre composites. The thesis, consisting of an extensive introductory review and six appended papers (listed in appendix 2), covers a number of aspects of these composites including the properties of candidate oxide constituents, methods of composite preparation, microstructure and mechanical properties. The first paper describes attempts made to prepare alumina composites reinforced with monocrystalline (sapphire) fibres using hot isostatic pressing. The second paper reports on the thermal expansion behaviour of candidate oxides and the consequences of differences in thermal expansion of composite constituents with respect to residual thermal stresses in the composite. Paper VI reports measurements made of the stress-strain and fracture behaviour of experimental sapphire fibre/alumina matrix composites. The materials studied included both unidirectional and 0/90° cross-ply composites, all with a thin layer of zirconia at the fibre/matrix interface designed to adjust the interfacial properties to provide optimum composite stress-strain behaviour. The results could be linked to measurements made of interfacial properties. Papers III, IV and V concern the stress strain behaviour of a commercial all-oxide composite consisting of fine-diameter, polycrystalline oxide fibres in a porous, aluminosilicate matrix. In these composites the fibres were woven in a 0/90° geometry and tested in both the 0/90° and ±45° orientation. The emphasis of the study was on the notch-sensitivity of test specimens containing a central circular hole. Aspects studied included failure mechanisms, the effects of notch size and the effects of high temperature thermal exposure on microstructure and strength degradation. The observed behaviour could be described successfully in terms of a simple model based on fracture mechanics. / Godkänd; 2001; 20060917 (cira)

This doctoral thesis covers investigation of the surface phenomena of Ti-6Al-4V and Ti-6Al-2Sn-4Zr-2Mo alloys related with oxidation at elevated temperature and corrosion duringchemical treatment in hydrofluoric-nitric acid (HF/HNO3) solutions. The explored phenomenaare related to manufacturing and service conditions of aero-engine components. Duringmanufacturing and operation, the alloys are running at elevated temperatures in oxygencontaining environment. Under these conditions there are formations of an oxide on the surfaceand an oxygen enriched layer below the oxide, commonly referred as alpha-case. The alpha-caseis a hard and brittle layer that is detrimental to the mechanical properties and must therefore beminimized or completely eliminated. A conventional method for elimination of alpha-case ischemical processing in HF/HNO3 solutions, known as chemical milling.Isothermal oxidation treatments in air at 500, 593 and 700 ºC for up to 500 hours were carriedout in this study. Both alloys developed rutile type of oxide structure and Ti-6Al-2Sn-4Zr-2Moexhibited stronger oxidation resistance than Ti-6Al-4V. Transition from parabolic to linearoxidation rate was observed at 700 ºC and ≥ 200 hours for both alloys. The difference in theoxidation kinetics of the two alloys is suggested to be related with the chemical composition ofthe alloys. The oxygen enriched layer, i.e. alpha-case layer, was characterised and its thicknesswas measured using conventional metallographic and microscopic techniques. Parabolicrelationship of the alpha-case layer growth rate with time was observed for both alloys. Theoxygen diffusion parameters and activation energies were estimated in the temperature range of500-700 ºC. Additionally, the oxidation at 700 ºC for 500 hours resulted in microstructuralchanges and element re-distribution. The bulk and alpha-case layer hardness at micro- and nanoscalewere measured using microhardness and nanoindentation techniques. The alpha-case layerhad higher hardness due to the solid solution strengthening effect of the diffused oxygen.The effect of chemical milling in 1:11 HF/HNO3 solution on the surface integrity, and theinfluence on low cycle fatigue (LCF) strength of cast Ti-6Al-2Sn-4Zr-2Mo alloy wasinvestigated. Short and long chemical processing times (5 and 60 minutes) and three imposedtotal strain ranges in fatigue tests were evaluated. Significant drop in fatigue life was observedfor the samples etched before LCF testing, as compared to the non-etched samples. The influencefrom etching was found to be most detrimental for fatigue samples tested at the lowest strainranges. The fatigue life reduction was correlated with the number of crack initiation sites.Multiple crack initiation sites were observed for the etched samples, whereas only one crackinitiation site was discerned in the non-etched samples. Inspection of the surface of the etchedsamples revealed selective and severely etched prior β grain boundaries and pit formation at thetriple joints of the prior β grain boundaries. These surface defects were considered as stressraisers promoting an earlier fatigue crack initiation.The influence of two different molar concentration ratios (1:3 and 1:11) of HF and HNO3acids on the corrosion behaviour of cast Ti-6Al-4V and Ti-6Al-2Sn-4Zr-2Mo were investigatedusing electrochemical and atomic force microscopy (AFM) techniques. The corrosion of the twoalloys was a function of the HF/HNO3 concentration ratio and also of the alloys’ chemicalcompositions. The AFM measurements revealed selective and faster dissolution rate of the α-phase than the β-phase in the Widmanstätten microstructure. It was considered that the reason forselective dissolution was the formation of local micro-galvanic cells between the α-phase and theβ-phase. Moreover, the Volta-potential was measured using scanning Kelvin probe forcemicroscopy (SKPFM) and the obtained maps revealed difference in the Volta-potential betweenvithe α-laths and the β-laths in both alloys. This observation strengthened the likelihood forformation and operation of micro-galvanic cells between the α-phase and the β-phase when thealloys were in contact with HF/HNO3 solution.Keywords: Titanium alloys, oxidation, oxide, alpha-case, chemical milling, corrosion.

Theoretical, physically-based models of fusion welding are developed, and calibrated using experimental data from practical welds. The following heat-affected zone phenomena are investigated: grain growth, precipitate dissolution and coarsening, martensite formation and hardness. A model is also developed to describe the effects of the welding arc and their dependence on welding conditions. All calculations are carried out using a microcomputer which readily allows the influence of a large number of material and welding variables to be taken into account. The results are presented as read-outs in the form of various types of Welding Process Diagrams, these providing information on weld geometry, H.A.Z. microstructure and hardness in a form understandable to the welding metallurgist and engineer alike. In addition, Implant Testing Diagrams are developed, based on similar modelling, which help to exactly locate the notch with respect to the grain growth zone and weld process. Diagrams can be constructed showing microstructural variation over a wide range of energy inputs corresponding to different welding processes, or for a particular process showing weld geometry and providing a physical picture of the weld. The programs are written such as to allow easy interaction between the operator and computer concerning choice of welding parameters, steel composition etc., and the storage of material data which can be readily called up by the operator. It is thus shown that the diagrams can be used to help optimize welding conditions, as well as supply information on the H.A.Z. microstructure and hardness. / Godkänd; 1984; 20070502 (ysko)

Deformation and stresses in butt-welding of plates were studied. The work includes numerical simulation and experimental verification. The simulations were performed by use of the finite element method. Temperature dependence of material properties and phase transformations were considered. A thermo-elastoplastic material model was used. Plane stress conditions were assumed. Automatic butt-welding of plates without backing needs close tolerances of joint geometry. The thermally induced deformations and stresses are of great importance for joint geometry during welding. Therefore the change of gap width in front of the moving arc has been of special interest in these studies. The residual stresses, which may affect inservice behaviour of welded plates, were also calculated and measured. The tack-welds were found to influence the change in gap width in front of the moving arc. A proper tack-welding procedure is important in order to avoid large changes in gap width during butt-welding. The tack-welds should be made as soon after each other as possible. The sequence in which the tack-welds are made also affect the change in gap width. The gap width increased during the last part of the butt-welding in the simulations performed in this work. This increase was larger for wide plates than for narrow plates. Residual stresses close to the weld were large. The effective stress reached the yield limit of the material in the weld line. / Godkänd; 1985; 20070424 (ysko)

Iron is one of the most important resources that can be found in the lithosphere; 90 % of all metal ores extracted are iron ore. Many steps are included in the extraction from iron ore to metallic iron, where the processes vary between different producers. Iron ore pellets, are a prepared burden material for ironmaking in the blast furnace. Such pellets are commonly sintered in a grate-kiln furnace system, where the kiln usually is insulated with mullite containing bricks. Different mechanisms wear these bricks and they need to be replaced regularly and this causes production stops. The slag present in the kiln consists of ~95 % hematite, alkali-, alkaline earth- and other oxides, mainly from pellets that have disintegrated and adheres in chunks on the bricks. This study is focusing on the interaction between refractories and slags that occurs in kilns during the sintering process in the iron ore pellet production. Results are shown from lab scale experiments, and from samples collected in industrial furnaces, commonly called rotary kilns. Slag/brick compatibility tests were performed in a laboratory furnace at various temperatures, holding times and atmospheres. Slag collected from a production kiln and three commercial bricks, in powder or solid form, were used. Deliberate additions of alkali species were included in order to evaluate their influence. XRD, DSC, TG and in-situ mass spectrometry confirm that addition of alkali dissolves the mullite in the bricks, and forms the phase nepheline (Na2O•Al2O3•2SiO2), which disintegrate to an amorphous phase at elevated temperature. QEMSCAN were used to view mineralogical mappings of different chemical phases by field image scans. It was found that when alkali penetrates the surface of the brick, besides formation of nepheline, phases as kalsilite (K2O•Al2O3•2SiO2), leucite (K2O•Al2O3•4SiO2) and potassium â-alumina (K2O•11Al2O3) are formed. Also seen is that potassium penetrates deeper, and in larger amounts than sodium in the lining material. Formations of alkali containing phases as the feldspathoid minerals kalsilite and nepheline are coupled to an expansion in the lining material, observed by dilatometry, causing structural spalling observed as cracks in some of the slag/brick compatibility tests. Grains of hematite with sizes between 50- 100 ìm stay on the original surface of the brick, while micrometer sized hematite migrates through the dissolved brick by capillary infiltration and diffusion, and nucleates in needle formations deeper in the lining material. We propose a wear mechanism of the bricks in an iron ore pellet producing kiln that involves these chemical reactions in combination with erosion by the continuously flowing slag. / Godkänd; 2008; 20081120 (ysko)

This licentiate thesis adds a new piece to the puzzle that describes how the microstructural characteristics influence the hardness behavior of a multilayer coating. It contains a presentation of the manufacturing and the subsequent characterization of multilayer thin films. These multilayers consist of alternating layers of crystalline titanium nitride (TiN) and amorphous silicon nitride (Si3N4), deposited with a physical vapor deposition technique referred to as reactive magnetron sputtering. The microstructure of as-deposited films was examined with cross-sectional transmission electron microscopy (XTEM) and x-ray diffraction (XRD). XRD studies revealed a transition in preferred orientation for TiN, from a pure 002 orientation to a mixed 111/002 orientation as the TiN layer thickness increased from 4.5 nm to 9.8 nm. XTEM studies showed a microstructure consisting of equiaxed or elongated TiN grains, depending on layer thickness, limited in size by the amorphous interlayers. Selected area diffraction verified the observed transition in preferred orientation in TiN. For small silicon nitride layer thicknesses (~0.3 nm) an epitaxial stabilization of Si3N4 to the crystalline TiN lattice was observed through high resolution electron microscopy studies. Instead of amorphous interlayers a cubic silicon nitride rich phase (SiNx) was observed. This is to the present knowledge of the author the first time this phenomenon has been observed within this material system. In order to explain the observed behavior a model based on the involved energies were developed. Nanoindentation was performed to evaluate the mechanical behavior of the coatings as the layer thicknesses varied. All multilayers were harder than the monolithic TiN film, which had a hardness of 18 GPa compared to 32 GPa for the hardest multilayer. An interesting observation was that the hardest multilayer corresponds to the presence of cubic silicon nitride. Curvature measurements were performed and showed that the residual stresses within the multilayers were compressive and relatively constant, 1.3±0.7 GPa. In addition to the XTEM studies of as-deposited samples, XTEM studies of deformed multilayers were also conducted. The 300 mN load produced plastic deformation in the substrate under the indent. Cracks within the multilayer normally propagated along TiN/Si3N4 interfaces, which suggest that a lower energy is needed for cracking along an interface compared to intralayer cracking. The observed hardness increase can be ascribed to the multilayered structure of these films. By the interruption of TiN growth with intermittent Si3N4 layers the produced microstructure consisted of small TiN grains, separated in the growth direction by amorphous or crystalline interlayers. Small grains are known to contribute to hardening, but the interlayers also contribute, acting as dislocation obstacles either due to the amorphous tissue or to coherency stresses. / Godkänd; 2004; 20060917 (cira)

Godkänd; 2008; 20081121 (ysko)

Metastable stainless steels are promising engineering materials demonstrating good corrosion resistance and mechanical properties. Their mechanical properties are however significantly affected by the deformation induced martensitic transformation. Hence, in order to use these steels to their full potential it is vital to have profound knowledge on this martensitic phase transformation. The aim of this thesis was therefore to investigate the evolution of phase fractions, texture, microstrains and microstructure to improve the current understanding of the deformation induced martensitic transformation in AISI 301. To investigate the deformation behavior of AISI 301, in-situ high-energy x- ray diffraction during tensile loading has been performed on samples suffering different cold rolling reduction. Ex-situ transmission electron microscopy, electron back-scattered diffraction and optical microscopy were also used to characterize the microstructure at different deformation levels. The results show that parts of the austenite transforms to both ά- martensite and ε-martensite during deformation of AISI 301. The transformation behavior of ά-martensite is however completely different from the transformation behavior of ε-martensite. ε-martensite forms in a parabolic behavior, while the ά-martensite transformation can be divided in three characteristic stages. The third transformation stage of ά-martensite has previously not been reported and it is characterized by a series of rapid transformations, each of which is followed by a period of yielding without any transformation. Moreover, the lattice strain evolution in the austenite at high plastic strains was found to be oscillatory, which is correlated with the stepwise transformation of ά-martensite as well as changes in x-ray peak broadening. This behavior was also coupled with the evolution of microstructure, where a distinct banded structure consisting of slip bands and Ü-martensite was observed at low plastic strains. This banded structure was however broken at high plastic strains when the ά-martensite grew larger and formed a block- shaped morphology. These findings lead to the conclusion that the three stages of ά- martensite transformation is due to different stages of nucleation and growth. The ά-martensite will first form as small nucleus, mainly at dislocation pile-ups along slip bands. The nucleuses will grow moderately in size and the structure will become saturated with nucleuses. Hence, the only way more ά-martensite can form is by growth of the existing nucleuses. This growth is very localized and seen as bursts in the transformation curve. The oscillatory behavior observed for the lattice strains during martensite formation possibly originate when semicoherent boundaries between austenite and ά-martensite become incoherent as the ά-martensite grow large. / Godkänd; 2005; 20061213 (haneit)