Optimization and Robustness Analysis of a Stent€¦ · Robustness Analysis Isight Artery dimensions are varied DoE generates approximation Monte Carlo simulation uses approximation

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

1

Optimization and Robustness

Analysis of a Stent

Axel Reichert

Dassault Systemes Deutschland GmbH

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

2

1 CAD Modeling

2 Finite Element Analysis

3 Fatigue Analysis

4 Parameter Optimization

5 Shape Optimization

6 Robustness Analysis

Overview

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

3

CAD Modeling SolidWorks

2D sketch 2D geometry

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

4

Finite Element Analysis Abaqus

2D mesh

3D mesh

Wrapped 3D mesh

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

5

Finite Element Analysis

Artery (anisotropic-hyperelastic)

Stent (elastic-plastic)

Expansion/crimping cylinders

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

6

Finite Element Analysis Crimping and insertion Blood pressure loading

Preloading stress Cyclic stress

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

7

Fatigue Analysis fe-safe

Surgical stainless steel with

ultimate tensile strength

Brown-Miller (critical plane method)

with Gerber (fatigue reserve factor)

Infinite life estimate: 1 billion cycles

(around 20 years)

Fatigue envelope:

Worst FRF: 1.223 Mean stress (MPa)

Str

ain (1e−

6)

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

8

Parameter Optimization Isight

Design sweet spot

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

9

Parameter Optimization Shape Changes

Baseline Parameter optimized Comparison

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

10

Parameter Optimization Mises Stress

Parameter optimized Baseline

5%

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

11

Parameter Optimization Fatigue Reserve Factor

Parameter optimized Baseline

8%

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

12

Shape Optimization Tosca Structure.shape

Method

Controller approach

Objective

Minimize maximum Mises stress

over all load cases

Geometrical restrictions

Preserve manufacturability

Cyclic symmetry

Design nodes

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

13

Shape Optimization Shape Changes

Parameter optimized Shape optimized

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

14

Shape Optimization Mises Stress

Shape optimized Parameter optimized

14%

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

15

Shape Optimization Fatigue Reserve Factor

Shape optimized Parameter optimized

14%

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

16

Robustness Analysis Isight

Artery dimensions are varied

DoE generates approximation

Monte Carlo simulation uses

approximation

Radius Mean

(mm)

Standard deviation

(mm)

Outer 1.62 0.028

Inner 1.13 0.018

Inner radius

Outer

radius

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

17

Robustness Analysis Generating the Approximation

Response surface

DoE

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

18

Robustness Analysis Using the Approximation

Monte Carlo simulation Fatigue reserve factor

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

19

Summary Process Integration with the Power of the Portfolio

CAD Modeling

Finite Elements

Fatigue

Parametric

Non-Parametric

Design Optimization Drives and Analysis

3DS

.CO

M/S

IMU

LIA

© D

assa

ult S

ystè

mes

20