You are on page 1of 28

2011 Altair Optimization Contest

Seonho Park, Seunghoon Cho Graduate School of Mechanical Engineering, Hanyang University

Advisor : Dong-Hoon Choi Department of Mechanical Engineering, Hanyang University

Outline
1. Automation topic

2. Optimization topic B - Optimization of Transmission Center Mounting Bracket

3. Free subject -Topology optimization of hood reinforcement considering pedestrian protection

Applied Mechanics and Optimal Design Lab.

Optimization Topic B
- Optimization of Transmission Center Mounting Bracket

Introduction - Background
Objective
Design of transmission center mounting bracket to use the steel panel

Manufacturing
Press

Requirements
Rupture resistance Buckling resistance Fatigue Natural frequency Constraints (1a) Constraints (1b) Constraints (2) Constraints (3)

Applied Mechanics and Optimal Design Lab.

Design Problem Formulation


Minimize the sum of two panels weight subject to (1a) reaction force at B > Rupture reaction force function along with deflection at A (1b) reaction force at B > Buckling reaction force function along with deflection at A (2) fatigue life >1E6 (3) free-free normal mode 1st Hz > 300Hz Constraint (1a) R-(272.22*abs(dz)+14301) > 0

Resistance of rupture failure


Point A

Constraint (1b) R-(-26.651*dx2+1151.3*dx-2375.4) > 0 Resistance of buckling failure


R : reaction force magnitude at B dx : x axis displacement at A dz : z axis displacement at A Point B

Applied Mechanics and Optimal Design Lab.

Meshing
PIPE Upper & Lower Panel

Mixed mesh Quad 4 (size=4)

Quad4 + Tria3 (size : 2) Weld region : Quad4 (size : 4) Option : Align, Size

Refinement 1

Refinement 2

Autocleanup : to resolve inadequate mesh


6

Suppress line : to resolve inadequate mesh


Applied Mechanics and Optimal Design Lab.

Analysis Setting
Static Linear Analysis for Rupture
z axis load (Fz=-19600N) Reaction force =1-D Spring static force (DOF 3)

Static Linear Analysis for Buckling


x axis load (Fx=19600N) Reaction force =1-D Spring static force (DOF 1)

Fatigue Analysis
z axis harmonic load 1G3.5G TABLEFAT card : 0, -1, 0, 1, 0 FATLOAD card : LDM :1, SCALE :3.5, OFFSET :1 FATPARM card : TYPE : EN, STRESS COMBINE : SGVON, STRESS CORRECT : SWT, STRESSU : MPa, PLASTI : NEUBER, RAINFLOW : LOAD, CERTNTY [SURVCERT] : 0.5

Normal Mode Analysis


EIGRL card ND :7 Load step : normal mode No SPC (for free-free mode analysis) 7th mode observed

Applied Mechanics and Optimal Design Lab.

Initial Design Analysis (1/2)


Static Linear Analysis for Rupture
z axis load (Fz=-19600N)

Static Linear Analysis for Buckling


x axis load (Fx=19600N)

Constraint (1a) -2209.9N < 0 Infeasible

Constraint (1b) 6234.1N > 0 Feasible

Fatigue Analysis
z axis harmonic load 1G3.5G E-N analysis

Normal Mode Analysis


No SPC (free-free mode) First mode after six rigid mode

No damage Infinite life Overdesigned

531.9 Hz Feasible
Applied Mechanics and Optimal Design Lab.

Initial Design Analysis (2/2)


Summary

Inconsistency

Mass [T] Constraint (1a) [N] Constraint (1b) [N] Constraint (2) [Cycles] Constraint (3) [Hz]

6.6309E-3 -2209.9 6234.1 531.9 Infeasible Feasible Feasible Feasible

1D Spring element attached at point B

Reaction force = spring element force Distorting free-free mode frequency Cannot conduct simultaneous optimization

Not properly designed

Applied Mechanics and Optimal Design Lab.

Design Strategy
Sequential Optimization
Size Opt. Objective Initial Design Topology Opt. Topography Opt. Weight, Const. (1a), (1b) Weight, Const. (2), (3) Const. (1a), (1b) Size Optimization Topology Optimization

Topography Optimization

10

CAD Implementation
Applied Mechanics and Optimal Design Lab.

Size Optimization
Problem Formulation
Find thickness of two panel, to Minimize volume, subject to constraint (1a) for rupture > 0 constraint (1b) for buckling > 0 2.3mm* thickness 6mm.

Results
Volume history plot

Panel Thickness
Mass [T]
Lower PNL [mm] Upper PNL [mm]
2.300 mm 4.000 mm 4.122 mm 4.000 mm

Initial 6.6309E-3 -2209.9 6234.1 infinite 531.9 Constraint (1a) [N] Constraint (1b) [N] Constraint (2) [Cycles]
: Initial
: Optimum

Size Opt. 5.2644E-3 0.0 5914.2 1.8671E7 441.9

Constraint (3) [Hz]

* , 2011, ,

11

# of iterations : 22 Rupture const. : satisfy the constraint Frequency & fatigue : adequate designed
Applied Mechanics and Optimal Design Lab.

Topology Optimization
Design Region Reconfiguration

Considering the weld region Considering the manufacturing process

Problem Formulation
Minimize compliance subject to volume fraction 0.9 constraint (2) for fatigue >1E6 constraint (3) for frequency > 300Hz Volume fraction is determined by trial and error Compliance minimization is beneficial to constraint (1a), (1b)

12

Applied Mechanics and Optimal Design Lab.

Topology Implementation & Result


Topology Optimized Model
Option One plane symmetric Mindim : 15mm

Topology Optimization Results


Compliance history plot

Size Opt. <Upper PNL> Compl.[Nmm] Mass [T] Constraint (2) [Cycles] Constraint (3) [Hz] <Lower PNL> 1.3194E4 5.2644E-3 1.8671E7 441.9

Topology Var.[%] Opt. 1.2480E4 5.0375E-3 2.8021E7 452.3 -5.4 -4.3 30.4 2.4

# of iterations : 7 Mass : lighter Frequency & Fatigue : enhanced


13
Applied Mechanics and Optimal Design Lab.

CAD Implementation
CAD Model
<Upper PNL> Mass [T] <Lower PNL> Constraint (1a) [N] Constraint (1b) [N] Constraint (2) [Cycles] Constraint (3) [Hz]

CAD Model Analysis Results


Topology Opt. 5.0375E-3 N/A N/A 2.8021E7 Topology Opt. CAD Model 5.1421E-3 -786.3 5737.4 1.7467E7

<Upper PNL>

452.3

447.1

<Lower PNL>

Constraint (1a) : infeasible Topography optimization to satisfy constraint (1a) should be continued.
14

Threshold density : 0.25


Applied Mechanics and Optimal Design Lab.

Topography Optimization
Design Region Reconfiguration

Considering the weld region Considering the manufacturing process

Problem Formulation
Minimize compliance subject to constraint (1a) > 0 constraint (1b) > 0. Reconsider constraint (1a), (1b) Compliance minimization is beneficial to constraint (1a), (1b)

15

Applied Mechanics and Optimal Design Lab.

Topography Implementation & Results


Topography Optimized Model
Option Linear Height : 5mm Angle : 60 Width : 8mm Draw direction : Inner (to avoid interference with other components)

Topography Optimization Results


Compliance history plot

Topology Opt. Topography Var.[%] CAD Model Opt. Compl.[Nmm] <Upper PNL> 1.2374E4 1.1645E4 -5.9

Mass [T]
Constraint (1a) [N] Constraint (1b) [N]

5.1421E-3
-786.3 5737.4

5.2992E-3
1.6 6048.2

3.1

5.4

<Lower PNL>

# of iterations : 100 Mass : lighter Frequency & Fatigue : enhanced


16
Applied Mechanics and Optimal Design Lab.

Conclusion

Initial Design

Results
Initial Design

Size Optimization

Topology Optimization CAD Implementation

Size Opt.

Topology Topograph Topology Opt. CAD y Opt. Model Opt.

Var. [%]

Optimum Layout

Mass 6.6309E-3 5.2644E-3 5.0375E-3 5.1421E-3 5.2992E-3 -20.1 [T] Constraint -2209.9 0.0 N/A -786.3 1.6 (1a) [N] Constraint <Upper PNL> 6234.1 5914.2 5737.4 6048.2 N/A (1b) [N] Constraint (2) 1.8671E7 2.8021E7 1.7467E7 6.5339E7 [Cycles] Constraint (3) 531.9 441.9 452.3 447.1 435.0 [Hz] <Lower PNL> Applied Mechanics and Optimal Design Lab. 17

Free subject
-Topology optimization of hood inner panel considering pedestrian protection

Introduction - Background
Background*
In Europe, about 41000 fatalities occur in traffic participation. 6100 (15%) of these are pedestrians. For pedestrians, the most frequent injuries occur in the head, 62% of all fatalities are caused by head injuries

Requirement of manufacturing hood inner panel**


Vehicle durability Hood stiffness for bending and torsion Engine noise absorbing characteristic Manufacturing requirements for drawing For pedestrian protection, to design the hood inner panel with a more uniform stiffness distribution Design more stiff hood inner panel to protect pedestrians
19

<Traditional Hood Inner Panels>***


*OECD International Road Traffic and Accident Database (IRTAD), 2002
**Kerkeling et al., Structural hood and hinge concepts for pedestrian protection, GM Europe, 2005 ***www.zr1specialist.com
Applied Mechanics and Optimal Design Lab.

FE Modeling
Extracting Hood Components 2001 Ford Taurus Model*

1057113 elements For NCAP front crash test


*www.ncac.gwu.edu **, , , , , , 2005 *** ASM Material Data Sheet, asm.matweb.com

Width : about 1500mm Height : about 1000mm Thickness : 3mm** Shell model (10282 element) Material : Aluminum 6061-T6*** E : 6.9E4 Mpa Poisson ratio : 0.330 density : 2.7E-9 T/mm3

20

Applied Mechanics and Optimal Design Lab.

Pedestrian Protection (1/2)


Head Impact Legal Test* Head Injury Criteria**

HIC (HPC) : standardized maximum integral value of the head acceleration Refered to determine the load

Wrap Around Distance***

Hood Pedestrian protection


*Kerkeling et al., 2005, Structural hood and hinge concepts for pedestrian protection, GM Europe ** Crash Analysis Criteria Description Ver. 1.6.2, 2005 ***Euro NCAP pedestrian testing protocol, 2011

21

Applied Mechanics and Optimal Design Lab.

Pedestrian Protection (2/2)


Headform Test Zone

Load Definition
Geometrical symmetric Constraint : hinge point

Hinge point Hinge point

12 headform test zone Export to load case

Problem Formulation
Minimize weighted compliance subject to volume fraction < f

Load (6 load step) Adult Mag.: 2.5N/node Direction : 65 Child Mag.: 1N/node Direction :50
22

Hinge point

Applied Mechanics and Optimal Design Lab.

Topology Optimization Implementation


Problem Formulation for Pedestrian Protection
Minimize weighted compliance subject to volume fraction < f

Benchmark Problem Formulation*


Minimize compliance subject to maximum displacement < 2mm natural frequency >12Hz volume fraction < f

Options
Maximum number of iterations :1000 Obj. tolerance : 10E-5 Volume fraction f : 0.2, 0.3, 0.4 Mindim : 40, 60, 80

50N imposed for torsional stiffness test


100N imposed for bending stiffness test

Bending load

Torsional load
*, ,2010,

23

Applied Mechanics and Optimal Design Lab.

Benchmark Topology Optimization Results

f : 0.2 Mindim 40mm Compl. 7.0565E1

f : 0.2 Mindim 60mm Compl. 7.5256E1

f : 0.2 Mindim 80mm

Compl. 7.4665E10

f : 0.4 Mindim 40mm Compl. 4.4149E1

f : 0.4 Mindim 60mm Compl. 4.4346E1

f : 0.4 Mindim 80mm Compl. 4.9316E1

24

Applied Mechanics and Optimal Design Lab.

Current Topology Optimization Results

f : 0.2 Mindim 40mm Compl. 8.4609E5

f : 0.2 Mindim 60mm Compl. 1.1634E6

f : 0.2 Mindim 80mm Compl. 1.4148E6

f : 0.4 Mindim 40mm Compl. 1.2546E5

f : 0.4 Mindim 60mm Compl. 1.4484E5

f : 0.4 Mindim 80mm Compl. 1.6411E5

25

Applied Mechanics and Optimal Design Lab.

Result Comparison
Selected Results

<Benchmark Result>

<Current Optimum Result>

Selected results : Reasonable minimum compliance & Manufacturability Torsional stiffness :benchmark result is better Current Optimum : uniform stiffness distribution
26

Applied Mechanics and Optimal Design Lab.

Conclusion & Future Works


Conclusion
Hood inner panel design is conducted to protect pedestrian Load step and constraint was developed to represent the pedestrian impact into static analysis mindim option member size

Future Works
Validation check by impact test Material changes Hood hinge design is also considerable for pedestrian protection

manipulate mindim and volume fraction obtain proper result The result show the better stiff distribution than benchmark result

27

Applied Mechanics and Optimal Design Lab.

-THE END-

You might also like