|
|
|
Composite Mechanics
|
| |
Multi-Scale Dynamic Failure Prediction of Marine Composite Structures - Under the sponsorship of ONR and Navy Labs, GEM has developed a hybrid discrete and continuum damage model to capture the synergistic interaction between a debonding crack and its surrounding matrix cracking zone. The model has been numerically implemented within the LS-DYNA environment via a user-defined material model. A validation study has been conducted via its application to a circular plate (a) and a composite hat stiffener subjected to shock and low velocity impact loading. The complex delamination growth patterns at multiple crack initiation sites have been accurately captured by the numerical simulation tool. The tool has been used to predict the damage progression of a sandwich structure subjected to a dynamic impact.
Figure 1: Dynamic Failure Prediction of Composite Plate and Hat Stiffener Subjected to Impact Loading
Figure 2: Progression of Delamination Failure Using Cohesive Interface Element
Figure 3: Exploration of Synergistic Interaction of Discrete and Continuum Damage in Sandwich Beam Subjected to Impact Loading
Woven Fabric Analyzer - Under the sponsorship of ONR and Navy Labs, GEM has developed a woven fabric analyzer (TMAT) to determine thermal-mechanical properties of 2D woven fabric composites based on a given constituent properties and woven fabric architecture. Input parameters for TMAT include effective fiber and matrix properties determined through uni-directional composite testing along with fiber volume fractions, weave geometry, tow geometry, tow path, and waviness ratio. TMAT uses this specified tow architecture to automatically generate a finite element model of a woven fabric unit cell along with the required thermal-mechanical load cases for computing the equivalent material properties from the corresponding constituent material properties (fiber and matrix). Multiple point constraints (MPCs) are employed in the FEM-based micromechanics model to ensure periodic mechanical/thermal boundary conditions.
Figure 4: Hierarchical FEM Based Micromechanics Models for Woven Fabric Composites
Ballistic Impact Analysis of Z-Pinned Kevlar Fabric Layers – Under the sponsorship of AEWC Center at University of Maine, GEM has developed a ballistic impact damage model to study the effects of density of z-pinning on the ballistic resistance of a multi-layered fabric patch. To capture the energy dissipation during the z-fiber pull out, a 1-D spring and dashpot in x-, y-, z- direction has been used in connecting each layer at its interface nodal points. A physically-based continuum damage model is used to characterize the multiple failure mechanisms such as fiber breakage, buckling, and frictional sliding between tows. The effect of using various z-pin densities on the final damage pattern of the fabric has been studied to explore an optimal z-pin enhancement.
Figure 5: Comparison of Interface Damage of Z-Pinned Fabric Subjected to High Velocity Impact
Figure 6: Comparison of Impact Failure Mechanisms of Z-Pinned Fabric with Different Z-Pin Density
Validation of Cohesive Zone Model for Delamination Failure Prediction – Under this ONR sponsored composite joint reliability program, GEM has validated the cohesive zone model using UMaine’s coupon data obtained from DCB/ENF/MMB tests. The validated cohesive zone model is then applied to determine the delamination failure of both primary and secondary bonded composite structures. The model prediction of a MMB specimen is demonstrated here, where the time of failure is predicted from an energy based fracture criterion using the predicted time histories of strain energy release rates.
Figure 7: Simulation of a Mixed Model Delamination Failure in a MMB Specimen Using Cohesive Zone Model
Thermal-Mechanical Fire Damage Prediction Tool – Under this 3-year ONR sponsored program, GEM has developed a thermal-mechanical fire damage prediction tool for a loaded composite sandwich structure subjected to a fire. The time dependent temperature field, char formation in the solid skin and core, and material loss due to thermal decomposition are predicted using the newly developed 1D fire model for an E-glass vinyl easter laminate skin and balsa core sandwich material. Given the temperature and mass loss distribution in the skin and core layer, the thermal mechanical layer properties are determined from CELLMAT. The laminator, VAPAS, is then applied to generate the equivalent thermal-mechanical properties of a laminated plate from its individual layer properties along with its stacking sequence. After determining the thermal mechanical properties at each Gaussian point, a finite element based solution (LS-DYNA3D) is applied at the current mechanical loading stage to determine the stress and strain response at the structural level. A hybrid continuum and discrete damage model is employed to capture the continuum damage induced stiffness degradation and the delamination failure at the skin and core interface. To apply the constituent based failure criteria during the continuum damage prediction, the structure-level stress/strain components are decomposed into their corresponding in each of its constituent. The predicted stiffness degradation and delamination zone are used for the failure progression associated with the next loading stage. The fire damage and final failure pattern of a sandwich beam at various load levels have been studied using the developed thermal-mechanical fire damage prediction tool.
Figure 8: Comparison of Failure Configuration of a Pre-Loaded Composite Sandwich Beam Subjected to a Fire
Composite Durability Assessment Tool – Under this ONR sponsored program, GEM is working closely with Virginia Tech (VT) to develop a residual strength based composite durability assessment tool. GEM is integrating VT’s residual material residence module with a commercial FEM solver via a user-defined material subroutine (umat). Given the matured methodology for the stiffness degradation and residual strength prediction under constant amplitude loading, GEM is using this as a deterministic module to build a probabilistic life prediction tool at coupon and component level. Three major components under this program are: 1) propagate the coupon (material) level residual stiffness and strength prediction models to a FEM-based component level; 2) integrate the component level residual strength based life prediction with a probabilistic analysis solver; 3) determination of statistical distribution of component life at a given constant load amplitude; and 4) extended the tool capability from the constant loading to variable amplitude fatigue loading.
Figure 9: Overview of a Residual Strength Based Composite Durability Assessment Tool
|
|
GEM Capabilities
Computational Mechanics
Composite Mechanics
Fatigue and Fracture
Probabilistic Mechanics and Reliability Engineering
Stochastic Mechanics
|
|