At Sandia/California, we apply our multidisciplinary expertise to an broad array of engineering projects — from atomic-scale materials design to macroscale systems engineering and prototype fabrication. We adopt a science-based approach, tightly integrating computer simulation with experiments to rapidly find the optimal solutions to national-security challenges. Our multiphysics models solve complex problems, coupling chemical, electrical, thermal, and mechanical behavior to yield accurate, detailed results. As needed, we develop new algorithms and computer codes to facilitate our work. With our seasoned scientists and engineers, high-performance supercomputers such as Red Storm, and state-of-the-art laboratory capabilities, Sandia/California offers the convenience of “one-stop shopping” for answers to complex engineering problems.
Full-scale Explosive Destruction System (EDS) vessel without the munition (top left) and with a British 4.2-inch munition (top right). Simulation of the EDS involves calculating pressure histories along the vessel walls from the internal detonation (lower left) and then performing a structural analysis of the vessel based on those pressures to obtain field quantities, such as effective stresses (lower right) and equivalent plastic strains.
We provide modeling support for the Army’s Explosive Destruction System (EDS), which was developed and designed at Sandia. The mission of EDS is to destroy and decontaminate recovered, chemically configured munitions in a safe and environmentally sound manner. Finite-element analyses are performed to assess the structural integrity of the vessel and seal after being subjected to the impulsive and quasistatic loads that result from an internal detonation. Key topics of investigation include concerns about excessive plastic straining of the vessel walls, which could lead to plastic collapse; critical fatigue-induced crack growth during the ring-down associated with the vessel’s breathing modes; and preventing leakage by mitigating the failure and buckling of the metallic seal.
We have developed a unique microfluidic mixing device based on the principle of induced-charge electroosmosis. By enabling mixing to be turned on or off, this approach prevents sample dilution, a common problem with sample mixing on microfluidic platforms. High-performance computation of electric field, fluid flow, and mass transport in multispecies liquids enabled us to quickly create prototypes of a wide range of device designs and to identify the geometries that promised best performance.
Our advanced constitutive models allow us to track material microstructure and residual stresses through each step of the manufacturing process and to use this information to assess aging issues. By embedding optimization software into the analysis process, we can produce first-time-correct parts with optimal material properties in just a fraction of previous procurement times.
The performance or failure of materials used in weapons systems depends on the mechanics of events down to the atomic scale, yet these complex systems typically require the integrated simulation of multiple parts at once. Some of these inherently multiscale problems involve processes too complex to be replicated by simple homogenization; instead, concurrent simulation is required. From this class of problems, we choose to target the model problems where highly localized heat generation and deformation are present, such as crack propagation, friction, spot welding, and solidification events. The concurrent, dual-paradigm approach will have the distinct advantage of using atomistic simulation in areas without perfect knowledge of the continuum phenomenology and efficient continuum finite-element representation where the material is behaving according to well-established constitutive theory.
Many of the most difficult problems in computational mechanics involve the interaction of macroscale material behavior with phenomena occurring at the atomic scale. Examples of these problems include the fracture and failure of solids; surface interactions; and the design of nanostructured materials, such as thermoelectrics. A promising approach to solving these problems, still being researched, is to use multiscale simulation techniques where the continuum behavior of the material is modeled with finite elements and then coupled to molecular-dynamics simulations to model the atomic-scale phenomena. We are involved in several projects to develop and apply these atoms-to-continuum coupling methods.
We are using cohesive element methods in Sandia’s SIERRA mechanics applications to model interlaminar fracture in composite materials. We have finished verifying the code and have shown that the simulations provide mesh-independent results. We have also performed sensitivity and preliminary uncertainty quantification analyses. Future work includes using mode-I, mode-II, and mixed-mode experimental results to characterize the model parameters and conducting validation tests.