Mechanical and Optimization Analyses for Novel Wound Composite Axial Impeller

A mechanical and optimal analyses procedure is developed to assess the stresses and deformations of Novel Wound Composite Axial-Impeller under loading conditions particular to centrifuge. This procedure is based on an analytical method and Finite Element Analysis (FEA, commercial software ANSYS) results. A low-cost, light-weight, high-performance, composite turbomachinery impeller from differently designed patterns will be evaluated. Such impellers can economically enable refrigeration plants using water as a refrigerant (R718). To create different complex patterns of impellers, MATLAB is used for creating the geometry of impellers, and CAD software UG is used to build three-dimensional impeller models. Available loading conditions are: radial body force due to high speed rotation about the cylindrical axis and fluid forces on each blade. Two-dimensional plane stress and three-dimensional stress finite element analysis are carried out using ANSYS to validate these analytical mechanical equations. The von Mises stress is investigated, and maximum stress and Tsai-Wu failure criteria are applied for composite material failure, and they generally show good agreement.

A Multiscale NDT System for Damage Detection in Thermal Barrier Coatings

Progress in aero-engines and land-based gas turbines is continuously linked with a rise of the operating temperature. TBCs are multilayered structures which function together to effectively lower the temperature of its load-bearing superalloy substrate while simultaneously providing oxidation protection against high temperature combustion environments during operation. They typically comprise of a ceramic top coat for thermal insulation and a metallic bond coat that provides oxidation/corrosion resistance and enhances the adhesion of the YSZ to the superalloy substrate. Due to high-temperature oxidation of the bond coat, a thermally grown oxide (TGO) scale of continuous Al2O3 is formed between the ceramic top coat and the bond coat. The formation and growth of the TGO increases the mismatch of thermal expansion coefficients among the multilayered TBC and induce high thermal stresses leading to spallation of the YSZ coat from the underlying metal. Hence, nondestructive diagnostic tools that could reliably probe the subsurface damage state of TBCs are essential to take full advantage of these systems. In this contribution, a new concept of multiscale NDT system is presented. The instrument uses a combination of imaging-based methods with photoluminescence piezospectroscopy, a laser-based method. Imaging-based methods like mid-infrared reflectance, laser optical backscatter and infrared tomography were used to predict the overall lifetime of the coated component. When TBCs approach the end of life, micro-crack nucleation and propagation at the top coat/bond coat interface increases the amount of reflected light. This rise in reflectance was correlated with the lifetime of the component using a neural network that merges the mean and standard deviation value of the gray level. Photoluminescence piezospectroscopy was subsequently used to give information about the structural integrity of the hot spots identified in the image analysis. This laser-based technique measures in-situ the residual stress in the TGO at room temperature. Damage leads to a relaxation of the local stress which is in turn reflected in the luminescence spectrum shape. However, presently there is no agreement on the best spectral parameters that should be used as a measure of the damage accumulation in the coatings. Therefore, the evolution of luminescence spectrum from as-manufactured to critically damaged TBCs was determined using the finite element method. This approach helped to identify the most suitable spectral parameters for damage detection, improving the reliability of photoluminescence piezospectroscopy as a failure assessment tool for TBCs.

Towards Truly Multiscale Simulation of Fatigue Processes in Metallic Structures

The fatigue resistance of metallic structures is inherently random due to environmental and boundary conditions, and microstructural geometry, including discontinuities, and material properties. A new methodology for fatigue life prediction is under development to account for these sources of randomness. One essential aspect of the methodology is the ability to perform truly multiscale simulations: simulations that directly link the boundary conditions on the structural length scale to the damage mechanisms of the microstructural length scale. This presentation compares and contrasts two multiscale methods suitable for fatigue life prediction. The first is a brute force method employing the widely-used multipoint constraint technique which couples a finite element model of the microstructure within the finite element model of the structural component. The second is a more subtle, modified multi-grid method which alternates analyses between the two finite element models while representing the evolving microstructural damage. Examples and comparisons are made for several geometries and preliminary validation is achieved with comparison to experimental tests conducted by the Northrop Grumman Corporation on a wing-panel structural geometry.

A Smart Device for Harnessing Energy From Aerodynamic Flow Fields

In this work, inspired by music playing harmonicas, we conduct a conceptual investigation of a coupled aero-electromechanical system for wind energy harvesting. The system consists of a piezoelectric cantilever unimorph structure embedded within an air chamber to mimic the vibration of the reeds in a harmonica when subjected to air flow. In principle, when wind blows into the air chamber, the air pressure in the chamber increases and bends the cantilever beam opening an air path between the chamber and the environment. When the volumetric flow rate of air past the cantilever is large enough, the energy pumped into the structure via the nonlinear pressure forces offset the intrinsic damping in the system setting the beam into self-sustained limit-cycle oscillations. These oscillations induce a periodic strain in the piezoelectric layer which produces a voltage difference that can be channeled into an electric load. Unlike traditional vibratory energy harvesters where the excitation frequency needs to match the resonant frequency of the device for efficient energy extraction, the nonlinearly coupled aero-elasto dynamics of this device guarantees autonomous vibration of the cantilever beam near its natural frequency as long as the volumetric flow rate is larger than a certain threshold. Experimental results are presented to demonstrate the ability of this device to harvest wind energy under normal wind conditions.

Length Scale Effects in Materials Deformation of Nano and Microscale Au Structures

In this study, intrinsic size effect — strong size dependence of mechanical properties — in materials deformation was investigated by performing atomistic simulation of compression on Au (114) pyramids. Sample boundary effect — inaccurate measurement of mechanical properties when sample size is comparable to the indent size — in nanoindentation was also investigated by performing experiments and atomistic simulations of nanoindentation into nano- and micro-scale Au pillars and bulk Au (001) surfaces. For intrinsic size effect, dislocation nucleation and motions that contribute to size effect were analyzed for studying the materials deformation mechanisms. For sample boundary effect, in both experiments and atomistic simulation, the elastic modulus decreases with increasing indent size over sample size ratio. Significantly different dislocation motions contribute to the lower value of the elastic modulus measured in the pillar indentation. The presence of the free surface would allow the dislocations to annihilate, causing a higher elastic recovery during the unloading of pillar indentation.

A Novel Way to Characterize the Non-Specific Surface Adhesion of Cancer Cells and Understand Cancer Metastasis

Cancer deaths are mostly caused by the metastasis of the malignant cells, not by the primary tumor itself. During metastasis, cancer cells detach from the primary tumor, spread to different tissues via blood circulation or lymph system, and reattach to invade new tissues and organs. In this project, we hypothesize that cancer cells manage their invasion by changing their surface adhesivity. To study the cell surface adhesivity, a novel and versatile microelectromechanical systems (MEMS) force sensor is developed to quantify the strength of adhesion between living cancer cells and a probe. The Silicon sensors consist of a probe and 2 flexible cantilever beams, while the probe is used to contact the cancer cell and the flexible beams are used to measure the cell force response in the range from nN to uN. The spring constant of the sensor is 14 nN/ μm. Our results demonstrate that the aggressive HCT-8 cells (from human colon adenocarcinoma) show high nonspecific adhesivity when they aggregate into cell islands, and low surface non-specific adhesivity after they disassociate from the cell islands. The surface adhesivity of less aggressive Caco-2 cells (from human colon carcinoma) and normal MA104 cell (from monkey kidney) are found to be lower than that of before-disassociation HCT-8 cells. Furthermore, the adhesion force response of cancer cells is found to show 2-slope force behavior, which is different from previous results of focal-adhesion detachment experiments. The 2-stage force bearing model is proposed to interpret the underlying mechanism.

Truss Disposition Optimization Using the Linear Extended Interior Penalty Function

This paper describes the development of truss optimization by finding its optimum topology (member disposition). The objective of this work is to obtain the truss with minimum weight that able to meet its design requirement (design constraints). Optimization is a mathematical technique used to find an optimum design, a design that minimize a certain objective function but still meet its design constraints. In this research, the objective function was the weight of the structure while the constraints were strength and displacement. Optimization was done by removing members which do not contribute to the load carrying capacity. Through conventional member size optimization non contributing members were identified and deleted. Linear Extended Interior Penalty Function was utilized to accomplish the mission. The truss structural responses were found by Finite Element Analysis coupled with an approximation procedure. Examples of a three-bar and six bar truss structures were demonstrated. Results showed that the truss topology could be optimized by members deletion strategy using the optimization procedure reported in this paper.

Mechanical Properties of Microscale Graphitic Carbon Foam Ligaments and Nodes

Carbon foam is recognized as having the greatest potential to replacement for metal fins in thermal management systems such as heat exchangers, space radiators, and thermal protection systems [1–5]. Carbon foam refers to a broad class of materials that include reticulated glassy, carbon and graphitic foams that are generally open-cell or mostly open-cell. They can be tailored to have low or high thermal conductivity with a low coefficient of thermal expansion and density. These foams have high modulus but low compression and tensile strength. Among the carbon foams, the graphitic foam offers superior thermal management properties such as high thermal conductivity. Graphitic foams are made of a network of spheroidal shell segments. Each cell has thin, stretched ligaments in the walls that are joined at the nodes or junctions. The parallel arrangement of graphene planes in the ligaments confers highly anisotropic properties to the walls of the graphitic foams. The graphene planes tend to be oriented with the plane of the ligaments but become disrupted at the junctions (nodes) of the walls. Since conduction is highest along parallel graphene planes, the thermal conductivity is highest in the plane of the ligaments or struts, and much lower in the direction transverse to the plane of these ligaments. In a previous study [6] extensive mechanical and thermal property characterization of carbon foams from Kopper Inc. (L1) and POCO Graphite, Inc. (P1) were reported. These foams were graphitic ones that are expected to have high thermal conductivity. Figure 1 shows sections of light microscopy images of the three foams of four foams. The most important thing to notice is that the images were not at the same magnification. The large cells in the GrafTech foam have an average diameter of only ∼100 μm but have a bimodal distribution cells with many small closed-cells few micrometers in diameter. Changes in density in the GrafTech foam was accompanied by a change in the large cells’ diameter — larger diameter giving greater porosity and lower density without changing the smaller cells’ sizes that filled the solid phase between the larger bubbles. The POCO foam has a fairly uniform size cell distribution of a few hundred micrometers. The Koppers’ foams show larger cells yet with the left (“L” precursor) having a uniform size while the right-hand (“D” precursor) is a less uniform and lower porosity.

Finite Element Analysis of a Locked Bolt With an Initial Crack

Bolted connections are largely employed in various types of engineering structures to transfer loads from one member to another. In particular, the off-shore industry has made extensive use of these connections, predominantly at the sub-sea level. In spite of their advantages, bolted joints are critical regions and may become sources of structural weakness due to large stress concentrations. Under severe operating conditions, micro-cracks can develop in the bolt, creating regions of elevated stress which may significantly reduce the integrity of the connection and ultimately lead to failure. This paper presents the three-dimensional finite element analysis of a steel locked bolt assembly aimed to assess the effect of micro-cracks on the structural integrity of the assembly using the commercial finite element package ANSYS. Non-linear contact between the bolt and nut threads is considered, where frictional sliding between components is allowed. A bi-linear isotropic hardening model is used to account for non-linear material behavior. The assembly is loaded by applying a pre-load of fifty percent of the yield stress of the material, according to the API-6A Norm. Two geometric models are investigated: a healthy locked bolt assembly with no initial cracks; and a damaged model, where a circular crack is introduced at the root of the bolt threads. The effect of the crack size is studied by modeling the crack with three different radius sizes. The J-Integral fracture mechanics methodology was used to study the stress concentrations in the damaged model.

Fatigue Behavior of Magnesium AZ31 in Friction Stir Spot Welds

In this study, the fatigue behavior of AZ31 magnesium friction stir spot welded joints is experimentally investigated. The friction stir spot welds employed here are representative of preliminary welds made in developing the joining process for potential use in automobile manufacturing. Load control cyclic tests were conducted on single weld lap-shear coupons and were fatigued until failure to determine stress-life properties. The fractured coupons were examined under optical and scanning electron microscopes with the intent to determine fatigue crack characteristics. Fractography analysis suggests that long crack growth accounts for a majority of the fatigue life. To predict the fatigue life of the lap-joint coupons, a long crack growth modeling approach, based on a kinked crack stress intensity solution, was employed. The fatigue model predictions compared well to the experimental stress-life results.