Search Results (1 - 4 of 4 Results)

Sort By  
Sort Dir
 
Results per page  

Valiveti, Dakshina M.INTEGRATED MULTISCALE CHARACTERIZATION AND MODELING OF DUCTILE FRACTURE IN HETEROGENEOUS ALUMINUM ALLOYS
Doctor of Philosophy, The Ohio State University, 2009, Mechanical Engineering

Ductile fracture in heterogeneous materials is strongly dependant on size, shape and distribution of heterogeneities in microstructure. Modeling mechanical response of such materials with explicit representation of microstructure morphology is computationally expensive and hence constrains the size of model. But analysis of a small microstructural model is not sufficient for prediction of fracture in a heterogeneous material. This demands the necessity of multi-scale analysis that can capture the material response over a large microstructural domain with explicit micro-mechanical analysis only in regions of dominant fracture.

Addressing various pre-requisites to multi-scale modeling, a three step Morphology based Domain Partitioning (MDP) is introduced. The first step in MDP is generating high resolution microstructure images of the entire computational domain. Subsequently, morphology based metrics are identified to relate microstructural features to mechanical response of material. In the third step, entire computational domain is partitioned to identify statistically homogeneous and inhomogeneous regions. This delineation forms the basis for macroscopic and microscopic length scales respectively, for a coupled multi-scale analysis.

The multiscale finite element model for ductile fracture developed in this work performs coupled analysis of macroscopic (level-0), and microscopic (level-2) length scales along with an intermediate swing region (level-1). The constitutive relations for macroscopic length scale are obtained by homogenization of microscopic representative volume element. Microscopic regions are analyzed with explicit representation of microstructure, using locally enhanced Voronoi Cell Finite Element Model (VCFEM) for ductile fracture. During the analysis, a macroscopic level-0 element adaptively evolves into level-1 and level-2, based on accumulation of damage. Macroscopic discretization error is minimized with h-adaptivity and modeling error is minimized by successive level change. The macroscopic and microscopic length scales are connected with an interface layer. Coupling is performed across the interface layer with Lagrange multipliers using hybrid variational formulation.

The multiscale model developed in this work captures detailed microscopic crack initiation and propagation in a large domain with minimal computational expense. The effectiveness of this multiscale characterization-modeling framework is demonstrated by studying structure-property relation and simulating ductile fracture in cast Aluminum alloy A319.

Committee:

Somnath Ghosh, Professor (Advisor); June K Lee, Professor (Committee Member); Stephen Bechtel, Professor (Committee Member); Carol Smidts, Professor (Committee Member)

Subjects:

Mechanical Engineering

Keywords:

Multiscale modeling; Finite Element Analysis; Domain Partitioning; Ductile fracture; Heterogeneous material; Aluminum alloys; microstructure modeling

Polasik, Alison KThe Role of Microstructure on High Cycle Fatigue Lifetime Variability in Ti-6Al-4V
Doctor of Philosophy, The Ohio State University, 2014, Materials Science and Engineering
The microstructural sources of fatigue lifetime variability were investigated for four different microstructural variations of Ti-6Al-4V. Specimens were tested at lower stresses to investigate the behavior in the HCF (high cycle fatigue) regime, which is characterized by lifetimes near or in excess of 10^6 cycles. Fractography and replication analyses confirmed that the lifetime was dominated by crack nucleation, and thus variations in the lifetime between individual test specimens are primarily attributed to variability in the time to nucleate a dominant crack. Stereology was used to quantify key microstructural features for each tested specimen. These values were used as inputs for a series of microstructurally-based fuzzy logic neural network models. Using these models, virtual experiments were conducted to investigate the individual effect of each microstructural feature on the lifetime, an investigation that is impossible to conduct empirically because of the complex microstructure in these alloy systems. These virtual experiments demonstrated that colony size and alath thickness have the greatest effect on HCF lifetime of ß-processed Ti-6Al-4V alloys, and that colony size is more important that a lath thickness. For the a/ß – processed microstructures, the volume fraction of primary a and the a lath thickness were shown to affect the lifetime, while the ap grain size was not. Defect analyses of failed specimens indicated that damage accumulation is often localized during cyclic loading, with dislocation densities varying from one a lath to another. For all specimens, a-type dislocations are seen and c+a - type dislocations were observed only in regions of localized plastic strain. Investigation of site-specific TEM foils extracted from the crack nucleation region of a/ß – processed specimens provided information about the nature and behavior of dislocations during the crack nucleation event. A comparison of short- and long- life specimens provides information about differences in the evolution of the dislocation structure prior to crack nucleation. The potential of this combinatorial approach for future fatigue lifetime investigations is discussed. In particular, the project demonstrates that such an approach could be useful in developing a quantitative understanding of the role variations in microstructural features have on variations in HCF lifetime.

Committee:

Hamish Fraser, PhD (Advisor); Michael Mills, PhD (Committee Member); Stephen Niezgoda, PhD (Committee Member)

Subjects:

Aerospace Materials; Engineering; Materials Science

Keywords:

Fatigue; Titanium; Fuzzy Logic Modeling; Ti-6-4; stereology; microstructure modeling

Riggs, Bryan EMULTI-SCALE COMPUTATIONAL MODELING OF NI-BASE SUPERALLOY BRAZED JOINTS FOR GAS TURBINE APPLICATIONS
Doctor of Philosophy, The Ohio State University, 2017, Welding Engineering
Brazed joints are commonly used in the manufacture and repair of aerospace components including high temperature gas turbine components made of Ni-base superalloys. For such critical applications, it is becoming increasingly important to account for the mechanical strength and reliability of the brazed joint. However, material properties of brazed joints are not readily available and methods for evaluating joint strength such as those listed in AWS C3.2 have inherent challenges compared with testing bulk materials. In addition, joint strength can be strongly influenced by the degree of interaction between the filler metal (FM) and the base metal (BM), the joint design, and presence of flaws or defects. As a result, there is interest in the development of a multi-scale computational model to predict the overall mechanical behavior and fitness-for-service of brazed joints. Therefore, the aim of this investigation was to generate data and methodology to support such a model for Ni-base superalloy brazed joints with conventional Ni-Cr-B based FMs. Based on a review of the technical literature a multi-scale modeling approach was proposed to predict the overall performance of brazed joints by relating mechanical properties to the brazed joint microstructure. This approach incorporates metallurgical characterization, thermodynamic/kinetic simulations, mechanical testing, fracture mechanics and finite element analysis (FEA) modeling to estimate joint properties based on the initial BM/FM composition and brazing process parameters. Experimental work was carried out in each of these areas to validate the multi-scale approach and develop improved techniques for quantifying brazed joint properties. Two Ni-base superalloys often used in gas turbine applications, Inconel 718 and CMSX-4, were selected for study and vacuum furnace brazed using two common FMs, BNi-2 and BNi-9. Metallurgical characterization of these brazed joints showed two primary microstructural regions; a soft, ductile a-Ni phase that formed at the joint interface and a hard, brittle multi-phase centerline eutectic. CrB and Ni3B type borides were identified in the eutectic region via electron probe micro-analysis, and a boron diffusion gradient was observed in the BM adjacent to the joint. The volume fraction of centerline eutectic was found to be highly dependent on the extent of the boron diffusion that occurred during brazing and therefore a function of the primary process parameters; hold time, temperature, FM/BM composition, and joint gap. Thermo-CalcTM and DICTRATM simulations were used to model the BM dissolution, isothermal solidification and phase transformations that occurred during brazing to predict the final joint microstructure based on these process parameters. Good agreement was found between experimental and simulated joint microstructures at various joint gaps demonstrating the application of these simulations for brazed joints. However, thermodynamic/kinetic databases available for brazing FMs were limited. A variety of mechanical testing was performed to determine the mechanical properties of CMSX-4/BNi-2 and IN718/BNi-2 brazed joints including small-scale tensile tests, standard-size butt joints and lap shear tests. Small-scale tensile testing provided a novel method for studying microstructure-property relationships in brazed joints and indicated that both joint strength and ductility decrease significantly with an increase in the volume fraction of centerline eutectic. In-situ observations during small-scale testing also showed strain localization and crack initiation occurs around the hard, eutectic phases in the joint microstructure during loading. The average tensile strength for standard-size IN718/BNi-2 butt joints containing a low volume fraction of centerline eutectic was found to be 152.8 ksi approximately 90% of the BM yield strength (~170 ksi). The average lap shear FM stress was found to decrease from 70 to 20 ksi for IN718/BNi-2 joints and from 50 to 15 ksi for CMSX-4/BNi-2 as the overlap was increased from 1T to 5T due to non-uniform stress/strain distribution across the joint. Digital image correlation techniques and FEA models of the lap shear brazed joints were developed to assess the strain distributions across the overlap. Results were used to validate the use of damage zone models for predicting the load carrying capacity of lap shear brazed joints and suggest that the damage zone is independent of the overlap length. To account for the presence of flaws and defects in fitness-for-service assessments of brazed joints determination of the average fracture toughness (KIC) is necessary. Currently no standard exists to measure the KIC for brazed joints, so three test methods were evaluated in this investigation on IN718/BNi-2 brazed joints. The compact tension and double cantilever beam test methods were found to give the most conservative KIC values of 16.42 and 14.42 ksivin respectively. Linear-elastic FEA models of the test specimens were used to validate the calculated KIC values. Similar to joint strength the fracture toughness appeared to be strongly influenced by the volume fraction of centerline eutectic phases. The data and methodology generated in this initial study provides validation for the proposed multi-scale computational model by demonstrating microstructure-property relationships in brazed joints and the ability to predict joint microstructure using simulation tools. Furthermore, an experimental framework and new techniques including small-scale tensile testing, digital image correlation and fracture mechanics were established to assist in future modeling efforts. Ultimately, successful development and implementation of multi-scale computational models for brazed joints will allow for the optimization of BM/FM compositions, brazing process optimization, improved reliability of brazed joints, and more efficient design and analysis of brazed components by accounting for the properties of the joint. In addition, the overall multi-scale modeling approach demonstrated in this investigation may also be applied for dissimilar joints in general, for example dissimilar metal welds used in oil and gas, petrochemical, nuclear and power generation industries.

Committee:

Boian Alexandrov, Ph.D. (Advisor); Avraham Benatar, Ph.D. (Advisor); Carolin Fink, Ph.D. (Committee Member)

Subjects:

Engineering; Materials Science

Keywords:

Brazing; Brazed Joints; Ni-base Superalloys; Microstructure; Modeling; Digital Image Correlation; Mechanical Properties

Makiewicz, Kurt TimothyDevelopment of Simultaneous Transformation Kinetics Microstructure Model with Application to Laser Metal Deposited Ti-6Al-4V and Alloy 718
Master of Science, The Ohio State University, 2013, Materials Science and Engineering
Laser based additive manufacturing has become an enabling joining process for making one-of-a-kind parts, as well as, repairing of aerospace components. Although, the process has been established for more than a decade, optimization of the process is still performed by trial and error experimentation. At the same time, deployment of integrated process-microstructure models has remained as a challenge due to some of the reasons listed below: (1) lack of good process models to consider the laser-material interactions; (2) inability to capture all the heat transfer boundary conditions; (3) thermo-physical-mechanical properties; and (4) robust material model. This work pertains to the development of robust material model for predicting microstructure evolution as a function of arbitrary thermal cycles (multiple heating and cooling cycles) that can be integrated into a process model. This study focuses on the development of a material model for Ti-6Al-4V and Alloy 718. These two alloys are heavily used in turbine engines and undergo complex phase transformations, making them suited to developing a material model for laser metal deposition (LMD). The model uses simultaneous transformation kinetics (STK) theory to predict the transformation of one parent phase into several products. The model uses calculated thermodynamic properties of the alloys for portions of the respective transformation characteristics. Being a phenomenological model there are several user defined calibration parameters to fit the predicted output to experimental data. These parameters modify the nucleation and growth kinetics of the individual transformations. Analyses of experimental LMD builds are used to calibrate the material model. A Ti-6Al-4V build made on a room temperature substrate showed primarily colony alpha morphology in the bottom half of the build with a transition to basketweave alpha in the top half. An increase in hardness corresponding to the microstructural transition was observed. This sample had an average of 340 HV hardness. Analysis of the calculated thermal profiles at the location of the morphology transition showed a transition from cooling below the beta transus to cooling above the beta transus. The Ti-6Al-4V STK model was calibrated using the experimental data from this sample. The substrate of a second build was heated above the Ti-6Al-4V beta transus. This build showed predominantly basketweave alpha without a microstructural transition. Large prior beta grains (>1mm) were observed growing epitaxially from the substrate. These large grains promoted the basketweave formation. Hardness testing showed an average of 344 HV. Samples built in this way were also fatigue tested in the as built condition. Results show that they match previous builds that had been stress relieved. A third build was performed at room temperature on a substrate with large prior beta grains. This build showed basketweave morphology like the second build even though the substrate was not thermally controlled. The hardness for this build averaged 396 HV which is ~50 HV higher than the previous two. This build shows that it may be possible to produce better mechanical properties by controlling the beta grain size rather than heating the substrate. Eighteen Alloy 718 builds were made using proprietary processing conditions. All of these builds were analyzed for nano-scale γ’ and γ’’ precipitates. Two of the builds were similar but had different laser powers. The low laser power build did not show nano-scale precipitates. The higher power build did show small amounts (<3%) of nano-scale precipitates and a corresponding increase in hardness at their locations. The higher power build was used to develop the STK model for Alloy 718. Sixteen of these builds were part of a design of experiments and are referred to as DOE samples. Eight of them have a single layer while the other eight have multiple layers. They were examined for nano-scale precipitates. The amounts of precipitates were correlated to hardness values and thermal profiles.

Committee:

Sudarsanam Babu (Advisor); Wolfgang Windl (Committee Member)

Subjects:

Aerospace Materials; Materials Science; Metallurgy

Keywords:

Simultaneous Transformation Kinetics; STK; Microstructure Modeling; Laser Additive Manufacturing; Laser Metal Deposition; aerospace repair; Ti-6Al-4V; Inconel 718; Alloy 718; Additive Manufacturing; LAM; LMD;