Multiscale Computational Modeling of CNT- Based Composite Materials Greg Odegard Richard and Elizabeth Henes Professor of Computational Mechanics Director, NASA STRI for Ultra-Strong Composites by Computational Design Michigan Technological University Workshop on Multiscale Modeling of Carbon Materials August 20-21, 2018
26
Embed
Multiscale Computational Modeling of CNT- Based Composite ...€¦ · Outline • Introduction: Institute for Ultra-Strong Composites by Computational Design (US-COMP) • Project
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Multiscale Computational Modeling of CNT-Based Composite Materials
Greg Odegard
Richard and Elizabeth Henes Professor of Computational Mechanics Director, NASA STRI for Ultra-Strong Composites by Computational Design
Michigan Technological University
Workshop on Multiscale Modeling of Carbon Materials August 20-21, 2018
Outline
• Introduction: Institute for Ultra-Strong Composites by Computational Design (US-COMP)
• Project example 1: MD Modeling of CNT/Epoxy Composites
• Project example 2: Multiscale modeling of PEEK
• Validation vs material exploration dilemma
2
3
Current carbon fiber composites lack strength/toughness (per unit mass) for
manned missions to deep space
acpsales.com
www.nasa.gov
MaterialsDevelopmentExperimental development and characterization of composites • High material and labor costs • Difficulty in testing under extreme conditions
(deformation, temperature, pressure) • Lack of methods to fully probe molecular-scale
behavior • Trial and error approach (Edisonian method)
4
“I have not failed. I've just found 10,000 ways that won't work.”
- Thomas A. Edison
ComputationalModeling
Computational modeling can • Provide efficient means to explore design space • Predict material behavior under a wide range of
conditions • Provide physical insight into observed behavior
5
Z1 – 1936 (computerhope.com)
IBM PC – 1981 (vintage-computer.com)
SUPERIOR HPC – 2013
MaterialsGenomeInitiative(MGI)
6
www.datanami.com
NASASTRIsolicitationrequirements• Next generation composite materials with
– Three-fold increase in tensile properties • Quasi-isotropic Specific Tensile Strength: 3 GPa/(g/
cm3) • Quasi-isotropic Specific Tensile Modulus: 150 GPa/(g/
• Institute for Ultra-Strong Composites by Computational Design
• First generation of NASA Space Technology Research Institutes (STRIs)
• Total funding: $15M over 5 years (starting summer 2017) • Partners
– 11 universities (Michigan Tech is lead, Prof. Odegard PI) – NASA (multiple centers) – Air Force Research Laboratory – 2 materials manufacturers (Nanocomp, Solvay) – 3 aerospace companies (Boeing, Lockheed Martin, Orbital
ATK)
Universityparticipants
• Michigan Tech, PI: Greg Odegard • Florida State University, PI: Richard Liang • MIT, PI: John Hart • University of Utah, PI: Mike Czabaj • Georgia Tech, PI: Satish Kumar • Johns Hopkins, PI: Jamie Guest • University of Minnesota, PI: Traian Dumitrica • University of Colorado, PI: Hendrik Heinz • Virginia Commonwealth University, PI: Ibrahim Guven • Florida A&M University, PI: Tarik Dickens • Penn State, PI: Adri van Duin
Sun et al, Carbon (2008) 46(2): pp. 320 Wang et al, Nanotechnology (2006) 17(6): pp. 1551 Wang et al, Polymer composites (2009) 30(8): pp. 1050 Gojny et al, Composites Science and Technology (2005) 65(15): pp. 2300
Bulk-levelComparison
Designmap
Project2–MultiscalemodelingofPEEK• Motivation
– PEEK polymers are used for internal structures in aircraft – PEEK is a multiscaled material – Improvement of PEEK composites requires a multiscale
modeling strategy • Objectives: Predict bulk mechanical properties of PEEK
using molecular-and micro-structure • Sponsor: NSF I/UCRC for Novel High Temperature/Voltage
Materials and Structures • Collaborators: Will Pisani, Evan Pineda (NASA GRC), Brett
Bednarcyk (NASA GRC), Greg Odegard
19
PEEKmicrostructure
Wang et al, RSC Advances, 2016
Molecularmodeling
Amorphous phase Crystal phase
• LAMMPS MD software • ReaxFF force field used • Multiple samples simulated for statistical evaluation
• Predicted results agree well with experiment • Apparent lack of strain rate effect • Relatively large amounts of crystalline
phase • May be obscured by variance
Validationvsmaterialexploration
• Most journals (e.g. Composites Science and Technology) require experimental validation of modeling based papers
24
How can we publish our material exploration research if the designed
materials cannot yet be fabricated for validation?
• One purpose of computational modeling is to efficiently explore new material designs with desired properties that have not been made (or cannot be made) in the laboratory
Acknowledgements
U.S. Air Force Office of Scientific Research Low Density Materials Program (Grant FA9550-13-1-0030)
SUPERIOR computing cluster Michigan Tech
National Aeronautics and Space Administration Aeronautical Sciences Program (Grant NNX11AI72A)
National Science Foundation I/UCRC program (Grant IIP-1362040)