SCI Publications

We present a novel simulation algorithm based on tempering a fraction of relaxation-limiting interactions to accelerate the process of obtaining uncorrelated equilibrium configurations of self-associating polymer solutions. This approach consists of tempering (turning off) the attractive interactions for a fraction of self-associating groups determined by a biasing field h. A number of independent configurations (replicas) with overlapping Hamiltonian distributions in the expanded (NVTh) ensemble with constant NVT but different biasing fields, forming a chain of Hamiltonians, were simulated in parallel with occasional attempts to exchange the replicas associated with adjacent fields. Each field had an associated distribution of tempered interactions, average fraction of tempered interactions, and structuraldecorrelation time. Tempering parameters (number of replicas, fields, and exchange frequencies) were chosen to obtain the highest efficiency in sampling equilibrium configurations of a self-association polymer solution based on short serial simulation runs and a statistical model. Depending on the strength of the relaxation-limiting interactions, system size, and thermodynamic conditions, the algorithm can be orders of magnitude more efficient than conventional canonical simulation and is superior to conventional temperature parallel tempering.

Taylor impacts tests were originally devised to determine the dynamic yield strength of materials at moderate strain rates. More recently, such tests have been used extensively to validate numerical codes for the simulation of plastic deformation. In this work, we use the material point method to simulate a number of Taylor impact tests. The goal is to par- tially validate some plasticity models used by the UINTAH multi-physics code. In addition, we would like to determine the plasticity model that is most appropriate for fire-structure interaction problems that are being simulated using UINTAH. We compare the Johnson- Cook, Steinberg-Cochran-Guinan-Lund, Zerilli-Armstrong, Mechanical Threshold Stress, and the Preston-Tonks-Wallace plasticity models. We evaluate these models for OFHC cop- per, 6061-T6 aluminum alloy, and 4340 steel alloy at various temperatures and strain rates. A number of validation metrics are presented for quantitative comparisons of numerical simulations and experimental data. It is observed that the accuracy of all the models drops when the initial conditions involve high temperatures and high impact velocities.

Foamed materials are increasingly finding application in engineering systems on account of their unique properties. The basic mechanics which gives rise to these properties is well established, they are the result of collapsing the foam microstructure. Despite a basic understanding, the relationship between the details of foam microstructure and foam bulk response is generally unknown. With continued advances in computational power, many researchers have turned to numerical simulation to gain insight into the relationship between foam microstructure and bulk properties. However, numerical simulation of foam microscale deformation is a very challenging computational task and, to date, simulations over the full range of bulk deformations in which these materials operate have not been reported. Here a particle technique is demonstrated to be well-suited for this computational challenge, permitting simulation of the compression of foam microstructures to full densification. Computations on idealized foam microstructures are in agreement with engineering guidelines and various experimental results. Dependencies on degree of microstructure regularity and material properties are demonstrated. A surprising amount of porosity is found in fully-densified foams. The presence of residual porosity can strongly influence dynamic material response and hence needs to be accounted for in bulk (average) constitutive models of these materials.

A methodology for simple multiscale modeling of mechanical properties of polymer nanocomposites has been developed. This methodology consists of three steps: (1) obtaining from molecular dynamics simulations the viscoelastic properties of the bulklike polymer and approximating the position-dependent shear modulus of the interfacial polymer on the basis of the polymer-bead mean-square displacements as a function of the distance from the nanoparticle surface, (2) using bulk- and interfacial-polymer properties obtained from molecular dynamics simulations and performing stress–relaxation simulations of the nanocomposites with material-point-method simulations to extract the nanocomposite viscoelastic properties, and (3) performing direct validation of the average composite viscoelastic properties obtained from material-point-method simulations with those obtained from the molecular dynamics simulations of the nanocomposites.