JEC COMPOSITES MAGAZINE - Issue #134 - May/June 2020 - 32

FEATURE
modelling

Unified modelling and simulation of
composite materials and structures

Wenbin Yu,
CTO, AnalySwift; Professor
Xin Liu,
Graduate Research Assistant
Orzuri Rique Garaizar,
Graduate Research Assistant,
Purdue University

C

ommercially
launched
by
AnalySwift in 2015, SwiftComp
has been used for the modelling
and simulation of a wide variety of
composites, ranging from deployable
space structures in satellites to printed
circuit boards in mobile phones. Due to
its general-purpose nature, the applications extend much further.
SwiftComp offers users an unprecedented combination of efficiency and
accuracy for multiscale, multiphysics
modelling. Most notably, by saving
orders of magnitude in computing time
without a loss of accuracy, engineers are
able to consider more design options
and arrive at the best solution more
quickly. With the release of a new version, engineers modelling composites
have a powerful addition in store for
their composites simulation toolbox.
The source of the technology's unique
advantages can be traced back to the
recently-developed
mechanics
of
structure genome (MSG), a disruptive, multiscale modelling theory for

32

SwiftComp is a novel simulation code for unified modelling
of composite materials and structures. The recently released
SwiftComp 2.0 added many new capabilities, including the prediction of viscoelastic properties, thermoviscoelastic properties
and composites failure, as well as AI-assisted composites
simulation.

efficient high-fidelity modelling and
simulation of composite materials and
structures developed by the Purdue
University Composites Manufacturing
and Simulation Center.
MSG simplifies the design and analysis
of composite materials and structures
by minimizing the loss of information
during multiscale homogenization,
which is essential for composites modelling. MSG unifies micromechanics
and structural mechanics to compute
equivalent structural properties in
terms of microstructural details, so it is
suitable for the modelling and simulation of all types of composite structures
including composite beams, panels,
shells, and 3D solids.
The Purdue team implemented this
theory into a companion code called
SwiftComp. More recently, SwiftComp
2.0 was released, which implements
many new cutting-edge capabilities
developed over the last three years
at Purdue including a multiscale
Timoshenko model, efficient and
accurate multiscale prediction of
failure indexes and envelopes, MSGbased thermo-viscoelastic multiscale
modelling, and AI-assisted multiscale
modelling.

jec composites magazine / N°134 May - June 2020

Viscoelastic analysis
Due to the inherent viscoelastic
behaviour of polymers, the long-term
performance of composites is a vital aspect in the design and analysis of many
composite structures.
For example, the viscoelastic behaviour
of thin-ply textile composites used
in deployable structures for space
applications may cause a reduction
in the strain energy stored in the
composites over time, hindering the
self-deployment of the structure. From
the modelling perspective, accurately
predicting the viscoelastic behaviour
of thin-ply textile composites involves
several challenges.
First, the viscoelastic properties of
yarns are not isotropic.
As most software tools cannot perform
viscoelastic analysis with a non-isotropic material, user-defined functions such
as UMAT subroutines in Abaqus CAE
must be developed, which requires
some coding skills.
Second, there are only a few layers in
thin-ply structures and the commonly-used periodic boundary conditions
(PBCs) in the thickness direction of the
structure are not valid. In other words,
the prediction based on the widely-used
representative volume element analysis
JEC COMPOSITES MAGAZINE - Issue #134 - May/June 2020

Table of Contents for the Digital Edition of JEC COMPOSITES MAGAZINE - Issue #134 - May/June 2020
