Three parallel algorithms for classical molecular dynamics are presented. The first assigns each processor a fixed subset of atoms; the second assigns each a fixed subset of inter-atomic forces to compute; the third assigns each a fixed spatial region. The algorithms are suitable for molecular dynamics models which can be difficult to parallelize efficiently—those with short-range forces where the neighbors of each atom change rapidly. They can be implemented on any distributed-memory parallel machine which allows for message-passing of data between independently executing processors. The algorithms are tested on a standard Lennard-Jones benchmark problem for system sizes ranging from 500 to 100,000,000 atoms on several parallel supercomputers--the nCUBE 2, Intel iPSC/860 and Paragon, and Cray T3D. Comparing the results to the fastest reported vectorized Cray Y-MP and C90 algorithm shows that the current generation of parallel machines is competitive with conventional vector supercomputers even for small problems. For large problems, the spatial algorithm achieves parallel efficiencies of 90% and a 1840-node Intel Paragon performs up to 165 faster than a single Cray C9O processor. Trade-offs between the three algorithms and guidelines for adapting them to more complex molecular dynamics simulations are also discussed.

Fast parallel algorithms for short-range molecular dynamics

Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Data Mining Practical Machine Learning Tools and Techniques

Additive manufacturing of metallic components – Process, structure and properties

Mechanical properties of nanocrystalline materials

A Taylor-type polycrystalline model, together with a new fully-implicit time-integration scheme has been developed and implemented in a finite element program to simulate the evolution of crystallographic texture during bulk deformation processing of face centered cubic metals deforming by crystallographic slip. The constitutive equations include a new equation for the evolution of slip system deformation resistance which leads to macroscopic strain hardening behavior that is in good accord with experiments performed on OFHC copper. The good predictive capabilities of the constitutive equations and the time-integration procedure for simulating the stress-strain behavior and the evolution of texture under both homogeneous and non-homogeneous deformation conditions are demonstrated by comparing numerical simulations against experimental measurements in simple shear and a simple plane-strain forging experiment on copper.

Crystallographic texture evolution in bulk deformation processing of FCC metals

Deformation twinning in AZ31: Influence on strain hardening and texture evolution

A new constitutive framework, together with an efficient time-integration scheme, is presented for incorporating the crystallography of deformation twinning in polycrystal plasticity models Previous approaches to this problem have required generation of new crystal orientations to reflect the orientations in the twinned regions or implementation of “volume fraction transfer” schemes, both of which require an update of the crystal orientations at the end of each time step in the simulation of the deformation process In the present formulation, all calculations are performed in a relaxed configuration in which the lattice orientation of the twinned and the untwinned regions are pre-defined based on the initial lattice orientation of the crystal The validity of the proposed constitutive framework and the time-integration procedures has been demonstrated through comparisons of predicted rolling textures in low stacking fault energy fcc metals and in hcp metals with the corresponding predictions from the earlier approaches as well as through qualitative comparisons with the measurements reported previously

Incorporation of deformation twinning in crystal plasticity models

This article investigates the microstructural variables influencing the stress required to produce deformation twins in polycrystalline fcc metals. Classical studies on fcc single crystals have concluded that the deformation-twinning stress has a parabolic dependence on the stacking-fault energy (SFE) of the metal. In this article, new data are presented, indicating that the SFE has only an indirect effect on the twinning stress. The results show that the dislocation density and the homogeneous slip length are the most relevant microstructural variables that directly influence the twinning stress in the polycrystal. A new criterion for the initiation of deformation twinning in polycrystalline fcc metals at low homologous temperatures has been proposed as (σ
 tw
 −σ
 0)/G=C(d/b)A, where σ
 tw
 is the deformation twinning stress, σ
 0 is the initial yield strength, G is the shear modulus, d is the average homogeneous slip length, b is the magnitude of the Burger’s vector, and C and A are constants determined to have values of 0.0004 and −0.89, respectively. The role of the SFE was observed to be critical in building the necessary dislocation density while maintaining relatively large homogeneous slip lengths.

Surya R. Kalidindi

Papers

Crystallographic texture evolution in bulk deformation processing of FCC metals

Deformation twinning in AZ31: Influence on strain hardening and texture evolution

Incorporation of deformation twinning in crystal plasticity models

Influence of grain size and stacking-fault energy on deformation twinning in fcc metals

Strain hardening of titanium: role of deformation twinning