Multi-Physics Numerical Modeling and Experimental Characterization of Materials Vincent Y. Blouin Assistant Professor Materials Science and Engineering Clemson University MILMI Bordeaux June 15, 2010

Background Diplome d’ingenieur (Hydrodynamic naval) PhD (Marine Engineering) Post-doc (Mech. Eng.) Research Assistant Professor Assistant Professor 2/36

Research Activities Multi-Physics Numerical Modeling Mechanical FluidsThermal Calibration Validation Characterization of Material Properties 3/36

1/25 Hunley Submarine

5/25 Corrosion-Erosion

Thermal Behavior of Buildings CFD Steady state fluid flow analysis Input: Wall temperatures Output: Heat fluxes FEA Transient thermal analysis Input: Heat fluxes Output: Wall temperatures Coupling between CFD and FEA. 6/36

Cooling of Precision Glass Molding Glass lens Cooling channels (N 2 flow) 7/36

3D heat transfer model of assembly 8/36

Coupled 3D fluid flow / thermal analysis Fluid flow analysis (CFD) Required parameters:  Material properties Thermal analysis (FEA) Required parameters:  Material properties  Surface conductance values N 2 flow rates Boundary conditions Surface heat fluxes Surface temperatures 9/36

Temperature profile Gradient of 10 o C through the lens This validates the axisymmetric assumption 10/36

Heat fluxes Can visualize heat fluxes Heat is drawn to the center of the assembly (inlet of cooling channels) 1/25

Modeling Precision Glass Molding Precision glass molding requires full understanding of – Thermal properties – Interaction properties Friction coefficient Heat exchange – Viscosity as function of temperature – Structural relaxation – Stress relaxation 12/36

Creep test 13/25 Constant force Glass sample

Experimental Characterization of Stress Relaxation Properties Three issues Separate shear and hydrostatic behavior Manufacture samples Numerical treatment to extract stress relaxation properties 14/36

Shear constitutive law Hydrostatic constitutive law Relaxation functions: Separate stress into shear and hydrostatic parts A considerable experimental effort is required to define visco-elastic behavior of glass (Prony Series)

Stress Relaxation Basics Instantaneous Elastic Strain Instantaneous Elastic Strain Delayed Elastic Strain Delayed Elastic Strain Strain due to viscous flow CREEP RECOVERY Strain produced by viscous flow. The viscosity is related to the slope. 16/36

Shear and Hydrostatic Deformations Shear or deviatoricHydrostatic or dilatation Shape changeVolume change Comparatively easy to conduct experiments involving pure shear Experiments involving pure hydrostatic component is complicated 17/36

Shear and Uni-axial Tests Shear test Shear deformation only (pure shear) Uni-axial test Shear deformation Hydrostatic deformation +

Literature Rekhson S.M. (1980) Extension of theory of linearity to complex glasses and temperature dependent viscosity values of Pyrex ® glass Scherer. G (1986) Fundamentals of viscoelasticity Gy R. et al. (1994) Retardation to Relaxation conversions Gy R. et al. (1996) Concept of viscoelastic moments and constants Duffrène et al. (1997) Overview of creep testing, methodology and experimental requirements Pascual M.J. et al. (2001) Temperature dependent viscosity data for Pyrex ® glass Spinner S. (2006) Temperature dependent mechanical properties of Pyrex ® glass 19/36

Overview of Characterization Process

Creep tests on Helical Spring sample (At different loads and temperatures) Time (sec) Displacement(mm) 21/36

Curve Fitting

Stress Relaxation Module Data processing based on Mathematical formulations Create dog-bone specimen Conduct creep relaxation test at various temperatures Strain vs time data Create spring specimen Conduct creep relaxation test at diff. temperatures Displacement vs time data 563 o C588 o C w 1j τ 1j w 1j τ 1j 3.06E E

Alternative Geometries Helical spring (pure shear) Shaft under torsion (pure shear) Tension/compression uniaxial test (shear + hydrostatic) 3-point bending (shear + hydrostatic) 24/36

25 Equipment Creep frameParallel-plates viscometer (PPV)

Manufacturing Samples PyrexBK7LBAL35

Glass Manufacturing 27/36

Manufacturing Samples of Low T g Glass Manufacturing process Short thick rod Create ballStretch into long rod Wrap long rod around metal rod to create spring Heat treatment and large deformations may alter optical and thermo-mechanical properties BK7 was successful with some defects L-BAL35 was not successful BK7

Glass Manufacturing 29/36

Glass Manufacturing 30/36

Glass Manufacturing

Optical Glasses (Low T g ) Sensitive to thermal shocks Impossible to fix once broken Optical glasses usually come as short rods (<20cm x 1cm) Prone to formation of bubbles when melted and extended Sand-blasted finish is harder to work with than smooth finish Current and future work: Use simpler geometries (not as accurate, requires more numerical treatment) Develop setup to manufacture samples at controlled temperature 35/36

Current and Future Work Use simpler geometries (not as accurate, requires more numerical treatment) Develop setup to manufacture samples at controlled temperature Experimental validation of numerical simulation Development of automatic numerical treatment of experimental data for extracting properties Use PPV for better temperature control (small samples) 36/36