Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 18

CHULALONGKORN UNIVERSITY

Department of Civil Engineering

AN EFFICIENT METHOD FOR INTEGRATION


ABAQUS AND MATLAB FOR SOLVING
ENGINEERING PROBLEMS

Presenter: Dr. Vu Quang Viet


Faculty of Civil Engineering, Vietnam Maritime University

Bangkok, 25 January 2023


CHULALONGKORN UNIVERSITY
Department of Civil Engineering

1 Introduction of Abaqus2Matlab

2 Specific problems can be solved by Abaqus2Matlab

3 Typical publications using Abaqus2Matlab

4 Future work using Abaqus2Matlab

Bangkok, 25 January 2023


INTRODUCTION OF ABAQUS2MATLAB

 Abaqus2Matlab is a computational tool that allows for automatic linking


between ABAQUS and MATLAB, in terms of transferring data and creating
the necessary files for solving engineering problems.

MATLAB Abaqus2Matlab ABAQUS


- Automatically create *inp file Perform the analysis
- Build the coding for solving through Abaqus input file
specific problems; (*inp)
- Store and show the results
obtained from Abaqus, etc.

Solving specific engineering problems

Bangkok, 25 January 2023


INTRODUCTION OF ABAQUS2MATLAB

Abaqus Graphical User Interface

Abaqus2Matlab

Abaqus Input file (*inp) New Abaqus Input file (*inp)


Bangkok, 25 January 2023
INTRODUCTION OF ABAQUS2MATLAB

 3 ways to create an Abaqus model

Python script Abaqus input file

Abaqus GUI

Bangkok, 25 January 2023


INTRODUCTION OF ABAQUS2MATLAB

Detailed engineering problems can be solved by Abaqus2Matlab:

 Optimization problems (e.g. size, shape, and topology optimization)

 Fragility, Reliability, Uncertainty analysis

 Inverse analysis

 Hybrid Experimental & FE analysis

 Convert Abaqus model to Artificial Neural Network or other AI objects

 Generate orphan meshes for Abaqus (e.g. 2D or 3D surfaces)

 Adaptive meshing and remeshing

 Consecutive Abaqus analyses

Bangkok, 25 January 2023


A LARGE VARIETY OF PROBLEMS CAN BE SOLVED BY ABAQUS2MATLAB

 Optimization (no limitations in the problem formulation or size)


Abaqus2Matlab can solve nearly all optimization problems

Function input Code inside function or subfunction

Coupling Abaqus & Matlab in series


inside functions:

 Objective function

 Constraint function

 Other customized function(s) or


subfunction(s)
Function output

Bangkok, 25 January 2023


A LARGE VARIETY OF PROBLEMS CAN BE SOLVED BY ABAQUS2MATLAB

 Fragility analysis, Reliability analysis, Uncertainty analysis

Code architecture: Coupling Abaqus and Matlab in series inside a loop

 Generate n random cases


 Scan one by one through for loop
for i=1:n  Generate Abaqus model
 Solve
 Extract results to Matlab
 Store results
 Calculate probability, statistic
measures, fragility curve, etc.

Bangkok, 25 January 2023


A LARGE VARIETY OF PROBLEMS CAN BE SOLVED BY ABAQUS2MATLAB

 Inverse analysis: minimize Err (modelling error)


START
 Make initial guess

 For initial guess or if Err > tol

 Generate Abaqus model

 Solve

 Extract results to Matlab


END  Calculate Err

 If Err < tol

Err<tol  Save or postprocess calibrated


if true if false
model

Bangkok, 25 January 2023


A LARGE VARIETY OF PROBLEMS CAN BE SOLVED BY ABAQUS2MATLAB

 Hybrid Experimental & FE analysis


 Merge experimental and numerical data to develop reliable FE models

Experiments FE analyses

Assess data
Data analysis
sensitivity

Compare with
experiment

Bangkok, 25 January 2023


A LARGE VARIETY OF PROBLEMS CAN BE SOLVED BY ABAQUS2MATLAB

 Convert Abaqus model to Artificial Neural Network or other AI objects

FE analysis  Example for a


single neuron:
 Generate and solve
Abaqus model
 Calculate output1
using Abaqus
results
 Calculate output2
if true using ANN
Err<tol
 Err = output1 –
output2
if false  If Err < tol :
ANN can be used
for FE analyses
without Abaqus
(Abaqus is replaced
Adjust weights by ANN)

Bangkok, 25 January 2023


A LARGE VARIETY OF PROBLEMS CAN BE SOLVED BY ABAQUS2MATLAB

 Generate orphan meshes for Abaqus (e.g. 2D or 3D surfaces)

 Develop geometry in Matlab

 Convert in Abaqus input file using


Abaqus2Matlab

 Import the input file in Abaqus

 Continue the FEA in Abaqus

Lim Young bin (2023). Generating Rough surface for Abaqus using MATLAB,
MATLAB Central File Exchange. Retrieved January 24, 2023.

Bangkok, 25 January 2023


A LARGE VARIETY OF PROBLEMS CAN BE SOLVED BY ABAQUS2MATLAB

 Adaptive meshing and remeshing

 Modify an Abaqus model based on adaptive meshing or remeshing

Adaptive (re)mesh

No
Mesh constraints
START satisfied?

Yes

END [Abaqus Documentation]

Bangkok, 25 January 2023


A LARGE VARIETY OF PROBLEMS CAN BE SOLVED BY ABAQUS2MATLAB

 Consecutive Abaqus analyses

 Generate consecutive Abaqus models in a way similar to successive


steps, restart and import capabilities (without the limitations of the
latter!)

Step 1

Step 2

… Step n

Bangkok, 25 January 2023


TYPICAL PUBLICATIONS USING ABAQUS2MATLAB

1. Optimization problems

Quang-Viet Vu, George Papazafeiropoulos, Carlos Graciano, Seung-Eock


Kim. Optimum linear buckling analysis of longitudinally multi-stiffened steel
plates subjected to combined bending and shear. Thin-Walled Structures
2019. 136, 235-245.

Seung-Eock Kim, George Papazafeiropoulos, Carlos Graciano, Viet-Hung


Truong, Quang Thang Do, Zhengyi Kong, Quang-Viet Vu*. Optimal design of
longitudinal stiffeners of unsymmetric plate girders subjected to pure bending.
Ocean Engineering 2021. 221, 108374

V T Pham, Q V Vu*, G Papazafeiropoulos, V T Ngo. Efficiency of


Abaqus2Matlab toolbox for structural optimization problems. IOP Conf.
Series: Materials Science and Engineering 869 (2020) 022025.

Bangkok, 25 January 2023


FUTURE WORK USING ABAQUS2MATLAB

1. Optimization problems (Size, shape, and topology optimizations):

- Finding optimum results (size, shape, and topology) of structures

considering its nonlinear inelastic behavior.

2. Uncertainty, Fragility, and Reliability analyses of structures

- Perform uncertainty, fragility, and reliability analyses of any structures

subjected to uncertain loading conditions (wind, earthquake, wave, rain,

snow, etc.), uncertain contact and boundary conditions considering its

nonlinear inelastic behavior.

Bangkok, 25 January 2023


FUTURE WORK USING ABAQUS2MATLAB

3. Inverse analysis

- Establish constitutive relationship models for materials

4. Generation of numerical big data

- Generate numerical big data from numerical finite element models for

application of AI, Machine Learning in engineering problems.

Bangkok, 25 January 2023


CHULALONGKORN UNIVERSITY
Department of Civil Engineering

THANK YOU
FOR YOUR ATTENTION!

Bangkok, 25 January 2023

You might also like