The Multidisciplinary-design Adaptation and Sensitivity Toolkit (MAST) is a sensitivity-enabled, multiphysics FEA tool developed to support computational design and engineering analysis.
In addition to analysis of complex, nonlinear physics on large built-up models, MAST supports efficient analytical gradient/sensitivity calculation using direct and adjoint methods. As a result, it is well suited for gradient-based multidisciplinary design optimization processes.
MAST is developed at the Computational Dynamics and Design Laboratory at Mississippi State University in collaboration with the Air Force Research Laboratory (AFRL) Multidisciplinary Science and Technology Center (MSTC).
The MAST website is https://mastmultiphysics.github.io and includes examples/tutorials, API documentation, and a growing theory/users guide.
An abbreviated list of capabilities is given below.
- 1D, 2D, and 3D analysis in multiple disciplines
- Heat Transfer
- 1D/2D/3D nonlinear heat conduction
- Radiation and convection boundary conditions
- Temperature-dependent material properties
- Structures (Elasticity)
- Beam/plate structural and 2D/3D continuum elements
- Loading: concentrated force, surface pressure, & thermoelastic
- Single-point boundary conditions and tie/connector constraints
- Nonlinear static and transient analysis
- Modal vibration analysis (including about a nonlinear equilibrium)
- Bifurcation buckling
- Advanced continuation and load-stepping in nonlinear analysis
- Nastran Bulk Data mesh input using pyNastran
- other mesh formats also available including Exodus-II and manual definition
- Fluids
- SU/PG discretization of compressible Euler equations
- Small-disturbance linearized time-domain and frequency-domain solvers for Euler equations
- SU/PG discretization of compressible Navier-Stokes equations (experimental)
- Fluid-structure Interaction (FSI)
- Small-disturbance flutter solution through coupling of structural and fluid discretizations
- Time-accurate fluid-structure interaction (experimental)
- Aeroelasticity
- U-g flutter solver with mode tracking
- Time-domain flutter solver for piston-theory aerodynamics
- Analytical sensitivity analysis for nearly all analysis capabilities
- Interfaces to multiple optimizers (GCMMA, DOT, NPSOL)
- Topology Optimization (TO)
- Level-set based approaches
- Density-based approaches
Detailed installation/build instructions for MAST and its dependencies are located on the website or can be located in the source repository in doc\install\*.dox.
To keep the size of the main MAST repository smaller, a git submodule is used to store large media/assets such as images and animations used for the documentation in a separate repo (doc/assets). To build the documentation locally, you must update the submodule. To do this, simply run the following commands from inside the root level of this main repository:
git submodule init
git submodule update
Copyright (C) 2013-2020 Manav Bhatia and MAST authors
Contributions to MAST by AFRL have been cleared for public release by case numbers: 88ABW-2016-5689, 88ABW-2017-5258, 88ABW-2020-0297.