Professional Documents
Culture Documents
Implementation State of The Art
Implementation State of The Art
Anechoic termination OI
Boundary Perfectly matched layer
Conditions
Multilayer
Optimization
Topology
Papers
legacy paper
TM
condensation
Matlab 2DMatlab 2D axi
(Fabien implemented this) - Does not yet work for curved structures X
Validated with AlphaCell X
Works with periodic boundaries. Some trouble remains when I remove
periodicity. This prevents extension to curved layers at the moment X
Several bugs remain to be fixed. Able to get a good result using Abaqus –
but that’s useless for us.
Several validations done using AlphaCell X X
X
X
Yet to implement this
Isotropic, homogenous layers only. Code does not handle anisotropic,
heterogeneous materials yet. X
The tricky coupling conditions are working (Major contribution from Arasan
and Fabien. I only provide initial implementation of Kirchoff plate and MITC
shell) NA NA
symmetric flat multilayers (implemented by Arasan and Fabien). Results
validated with 3-D FEM and AlphaCell NA NA
(implemented by Vivek) – Covariance Matrix Adaptation Evolution Strategy
(CMA-ES)
(implemented by Vivek, I provide only FE implementation) – getting optimal
foam configuration for specific excitation, volume constraint. Validated
against paper and analytical wedge results from FX.
NA NA
X X X X
X
X
X X X X
X NA NA
X NA NA
NA NA Python
80%
check here
Elastic materials (static and
dynamic)
Visco-plasticity
Poromechanics
Physics (consolidation)
Brittle fracture (phase field
method)
Fracture of porous media
Vibroacoustics
Multiscale method
Complexity
reduction
Model order reduction
Parallelization
Finite elements
Virtual elements
Elements
Isogeometric elements
(without multiscaling)
implemented by Savvas
implemented by Savvas – Non-linear problem. Iterative computation of MS basis
implemented by Savvas – Coupled problem
Lower order 2-D elements are working. Extension to 3-D and higher order are work in progress. I implemented a
version of 2nd order elements for Savvas’ fracture problems, but not tested on acoustic problems. 3-D polyhedral
mesher available online. Polygonal/complex mesh shapes implemented at all scales.
Very nice applications in contact mechanics, fracture, topology optimization etc where complicated interfaces,
shapes are found. In acoustics, only incremental advantage – since wavenumber rules still need to be followed.
(to be cleaned)
Works only for star-convex elements. Has expensive quadrature integration rule. Easier to implement than VEM
(without multiscaling) – Inserted my master thesis work – Multi-patch Isogemetric method (IGA) for
contact/interface problems into the code to have smooth boundaries, exact geometries, better continuity and
convergence properties etc. Tricky to couple with existing FEM code since IGA mesh not conforming with FE
mesh. (to be cleaned)
Matlab 2D Matlab 3D
X X
X X
X X
X X
X X
X
2-D MsVEM
with TO
Topology Optimization
(TO) Training meta
model with
CNN
IGA-BEM
to reduce computation time – Solid Isotropic Material with Penalisation (SIMP) and Level
Set Method (LSM). Basic structure ready. Had some trouble getting good result – for
elasostatics (bridge design)
To get instantaneous TO solns. Generating training data was too expensive. Didn’t have
enough time.
Fractional calculus for transient wave propagation and stochastic excitation in porous
media – can match result with paper for rigid porous media. Not able to make it work for
poro-elastic
Tried inserting my IGA into open-source Boundary Element Method code to get
IGA-BEM for nice shape optimization features. Total failure 😊 .