Multiscale Modeling and Simulation

Materials and Process Simulation Center (MSC)

California Institute of Technology

Research
Members
Archive
Opportunities
Related links
Search
Contact Information

 

 

 

 

 

 

 

 

Multiscale Multiparadigm Simulation Strategy (AJB, Caltech)

Figure above.  Our Multiscale, Multiparadigm Simulation Strategy

Motivation and General Overview

Understanding natural phenomena from science or optimizing processes from engineering requires, by today's standard, synchronized contributions from theory, experiment and computation.  In an important number of cases, computer simulations -based on fundamental theory- supplement experiment, but in many others, they are the enabling tool for the study and comprehension of complex systems and phenomena that would otherwise be too expensive or dangerous, or even impossible, to study by direct experimentation.

 
In principle, all properties of all materials and phenomena are describable by quantum mechanics (QM) [Schrodinger's equation], unfortunately direct use of QM is impractical for solving applications that involve a large number of atoms (> ~1K).  Classical and quantum-based, adiabatic and non-adiabatic, approximations to Schrodinger's equation lead to simplified equations of motion [molecular mechanics/dynamics - MM/MD] that are applicable to much larger systems while still retaining the atomistic and electronic degrees of resolution (~millions of atoms and electrons).  Further classical approximations on these systems include coarse-grain models based on rigid-bodies of constrained particles, suitable, for example, in sampling the dihedral conformational space of large-scale molecular systems, Monte Carlo techniques, averaging and homogenization techniques to explore larger and longer length- and time-scales, close to or within the mesoscale regime.  At the higher-end of the length- and time-scales, phenomenological-based continuum-level methods [including, Finite Elements] are the norm, yet these are incapable of capturing fundamental nanoscale properties and phenomena that are critical to understanding, elucidating and optimizing the behavior of matter at the macroscale.

Our research involves developing first principles-based theory, methods and efficient multiparadigm computational algorithms and tools capable of seamlessly bridging length and time scales to enable de novo design, characterization and prediction of material properties and processes and their application into solving currently "impossible" problems.  We also leverage and extend on other legacy achievements at the MSC, including, eFF electron force fields, ReaxFF reactive force fields, Dreiding and Universal (UFF) non-reactive force fields, CoMoDyn constrained Molecular Dynamics, the Computational Materials Design Facility (CMDF), among many others in computational chemistry and physics.

Recent Milestones

bullet04/2013: We determine atomistic mechanisms that interfere with low-T diamond CVD deposition and propose alternative halogenated precursors to avoid these
bullet12/2011: We determine the atomistic mechanisms involved that lead to electronic emission during high-strain rate brittle fracture of Silicon using eFF
bullet11/2011: We explain the effect of excited electrons on the conductivity of high-pressure polyethylene using our non-adiabatic dynamics methods
bullet07/2011: We postulate the molecular origin of fracture nucleation in Portland cement, from simulations using our new Ca/Al/H/O/S ReaxFF reactive force fields
bullet10/2010: Our reactive dynamics simulations reveal possible composition of Enceladus' south pole plume, consistent with Cassini's INMS data
bullet03/2010: Experimentally validated dynamic shock Hugoniots for Li metal, using explicit electron molecular dynamics
bullet07/2009: Performed first large-scale (millions of nuclei and electrons), long-term (10's ps), non-adiabatic excited electron dynamics simulation of hypervelocity collisions
bullet02/2009: Elucidated key mechanism involved in enhanced mechanical response of DN hydrogels for tissue scaffolds
bullet more ...
 

What's  New

bullet11/2013: JPL/Caltech ICEE Technology Maturation funded for MARINE (Mass Analyzer for Real-time Investigation of Neutrals at Europa)
bullet09/2013: DARPA funds 3-year LoCo program on low-temperature thin film deposition
bullet07/2012: Intel (Santa Clara, CA) funds 2-year effort in semiconductors [confidential]
bullet01/2012: DARPA (US Defense Advanced Research Projects Agency, Arlington, VA) funds seedling effort to study low-temperature (<100C) thin film growth
bullet04/2011: Dow Chemical (Spring House, PA) funds effort for in colloids [confidential]
bullet10/2010: NSF-MRI award (#1040558) funds large-scale GPGPU cluster (32,256 Fermi cores) for soft matter simulations.  6 Caltech groups involved, including ours.
bullet08/2010: Samsung (South Korea) funds modeling effort in graphene-based nanodevices [confidential].
bullet06/2010: (DOT) Federal Highway Administration funds collaborative effort to optimize the nano-mechanical response of stone-based cement mixtures
bullet03/2010: JPL-NASA collaborative effort on Cassini-Huygens mass spectrometer impact ionization INMS data analysis
bullet more ...
 

Home | Research | Members | Archive | Opportunities | Related links | Search | Contact Information

 (C) Materials and Process Simulation Center, Caltech, 2007.
Contact: Andres Jaramillo-Botero [ajaramil at wag.caltech.edu].
Last updated: 04/04/12.