Using CAE & Optimisation Tools to Design Lighter and Safer ...NAFEMS World Congress NAFEMS UK...


Seats Background

• M100 seat is a pedestal and wall mounted design, developed for Wabtec. Tested in

2013 at MIRA and passed all requirements of GM/RT2100 Issue 5.

• H200 is a development of M100 for HITACHI Rail Europe. Developed in 8 months,

passing the injury and integrity tests in 2014.

• M101 uses a combination of both previous designs and learnings from simulation and

tests.

• Each seat passed injury and integrity requirements of GM/RT2100 Issue 5 first time.

M100

H200

M101


GM/RT 2100 Issue 5

• New requirements for rail vehicle structures

• Document mandates requirements for the design and integrity of rail vehicle

structures for both primary and secondary structures, including interior

crashworthiness.

• Conflicting requirements between seat strength and passenger injury.

• At time of development, no other seats had successfully met requirements.


Previous test knowledge

Material Data

Concept CAD & CAE scoping

studies

Master CAE

model

Design Design Design

Simulation Simulation Simulation

Test

Methodology


Modal Analysis / Durability

Rearward & Forward Integrity

Forward Injury

Abuse Load Cases

Fatigue Load Cases

Methodology


1.5kN Forward1.5kN Forward

Injury

50%ile Rearwards

Integrity

95%ile Rearwards

100mm Lateral

2kN down

1.5kN Rear 1.5kN Rear

2kN down

1.5kN Grab Handle

GM/RT 2100 Seat Load Cases

Integrity

95%ile Forwards

Injury

50%ile Forwards

• All projects use finite design and

simulation resources.

• GRM’s approach has been to prioritise

load-case development

– Red – Checked & developed daily

– Amber – Checked weekly

– Green – Designed in from start


Methodology

Manufacturability

Lightweighting

Cost

Strength & stiffness

Concept Design

Optimisation Tools


Armrest Cover Ribs Optimisation

Abuse load cases to represent

150kg individual standing on trim

Initial Aluminium

armrest cover

Creation of candidate material

using RDM tool in Genesis

Topology

Optimisation Results

Definition of

manufacturing

constraints

Topology Optimisation

for multiple load cases

in Genesis

Optimisation keeps

solid element at the

locations where

reinforcements are

required

Initial Aluminium

armrest cover


Armrest Cover Ribs Optimisation


Reinforcement Derivation Method

• An existing FE structure is

our starting point

• The RDM® region is

automatically created in

Design Studio over the top at

user-defined limits

• Genesis carries out an

optimisation for multiple load-

cases, removing superfluous

material and only leaving the

most important parts


Cantilever Pedestal Design

Package Space

DefinitionTopology

optimisation Design

Interpretation

Folded Sheet

Stiffness &

Modal

Final DesignSizing & Topology

optimisation refinement

Sheet thick. (mm)


• Forward and Rearward Integrity tested with two 95th percentile dummies on various

pitch and seat configurations (single/double/triple, 1st & standard class) to determine

the worst case scenario to be tested.

• Axial and shear forces measured at wall fixing points for bolt design.

• Plastic strain in pedestal and wall brackets recorded to ensure that the seat will remain

attached to the carriage. Seat-back longitudinal deflection recorded.

Crashworthiness & Seat Integrity

Loads (kN)

A B C D E F

Axial 32 14 -16 -13 33 9

Plastic Strain


• Injury values extracted from simulation of forward impact. Design changes were

quickly implemented and the subsequent change on injuries determined.

• Use simulation to determine the injury criteria closest to the legal standards defined in

GM/RT2100 Issue 5.

CriteriaGM/RT2100

LimitSimulation

Femur Load Left (kN)5.7

4.8

Femur Load Right (kN) 4.5

Knee Shear Left (mm)16.0

9.7

Knee Shear Right (mm) 10.9

Tibia Load Left (kN)8.0

1.1

Tibia Load Right (kN) 0.98

Tibia Index Left1.3

0.62

Tibia Index Right 0.75

Neck Tension (kN) 4.17 0.57

Neck Compression (kN) 4.0 1.10

Neck Flexion (Nm) 310 95

Neck Extension (Nm) 135 26

Neck Injury Criteria 1.0 0.61

HIC 500 96

3ms Clip 80 46

Injury Prediction

Peak force extracted

Cross-section View


• Adhesive testing was carried out by 3M on various

adhesive materials.

• A model was built to replicate the adhesive test procedure

with tiebreak contact between the aluminium sample and

the adhesive, which allows the failure force to be specified.

• The failure force value was set-up to achieve the same

failure characteristic seen in the adhesive test.

• Calibrated adhesive was then used to build accurate FE

model with material failure.

Adhesive Testing & Modelling

Sample

1

Sample

2

Sample

3

Sample

4Sample

5


Seat Back Design

• Design sensitivity to seat-back shape / section profiles.

• Aim is to use the seat-back shape to absorb the energy of a crash to reduce

the leg injuries. Achieved through the unfolding behaviour of the section profile.

• Different shapes were tested to meet visual aspect requirements as well as

injury targets.

CriteriaGM/RT2100 Limit

CAE limit

(80%)Design 2 Design 8 Design 11

Femur Load Left (kN)

5.7 4.565.7 6.5 5.3

Femur Load Right (kN)

5.3 6.2 5.0

Knee Shear Left (mm)

16.0 12.813.2 13.0 11.9

Knee Shear Right (mm)

16.8 16.4 14.1

Design 2 Design 8 Design 11


Seat Back Model

• Variation of thickness due to stamping was mapped on the seat back, based

on material thickness of 1.6mm.

• Thickness results used in the injury and integrity impact simulations to improve

the performance prediction.

Thickness (mm)

1.6mm based seat-back

CriteriaGM/RT2100

LimitCAE limit

(80%)

No Thickness mapping

With thickness mapping

Test Results

Femur Load Left (kN)5.7 4.56

5.3 4.8 4.6

Femur Load Right (kN) 5.0 4.5 4.2

Knee Shear Left (mm)16.0 12.8

12.1 9.7 8.2

Knee Shear Right (mm) 13.7 10.9 10.9


Optimisation & Accurate

Predictive Modelling• CAE led design to meet all targets 1st time

• Optimisation tools to design lightweight, stiff and strong train seats

- RDM based on Topology Optimisation for internal reinforcement

- Topology Optimisation for cantilever pedestal design

- Combined Sizing & Topology for additional weight saving

• Test prediction highly depending on FE model quality

- Accurate adhesive model with failure from physical test correlation

- Stamping effect accounted for real seat-back thickness

- Dummy injury knowledge from previous automotive projects


Forward Injury Test

CriteriaGM/RT2100

LimitSimulation Test

Femur Load Left (kN)5.7

4.8 4.6

Femur Load Right (kN) 4.5 4.2

Knee Shear Left (mm)16.0

9.7 8.2

Knee Shear Right (mm) 10.9 10.9

Tibia Load Left (kN)8.0

1.1 0.93

Tibia Load Right (kN) 0.98 0.72

Tibia Index Left1.3

0.62 0.72

Tibia Index Right 0.75 0.99

Neck Tension (kN) 4.17 0.57 1.19

Neck Compression (kN) 4.0 1.10 0.18

Neck Flexion (Nm) 310 95 127

Neck Extension (Nm) 135 26 26

Neck Injury Criteria 1.0 0.61 0.42

HIC 500 96 150

3ms Clip 80 46 62

• Forward injury test shows good correlation with simulation. Differences are mainly due

to dummy positioning.


Rearwards Integrity Test• Rearwards integrity result is a pass as seat remains attached to rail structure and seat

is also well within the survival space envelope as predicted.


Conclusions

• CAE analysis and optimisation tools made it possible to design three

generations of lightweight train seats, each passing both integrity

and injury tests first time.

• FE analysis saved time and money allowing the team to deliver a

design 35% lighter than standard train seats.

Date post:	14-Mar-2020
Category:	Documents
Upload:	others
View:	9 times
Download:	0 times

Using CAE & Optimisation Tools to Design Lighter and Safer ...NAFEMS World Congress NAFEMS UK...

Documents