2015 Simulia Uk David Winfield

FREUDENBERG OIL & GAS TECHNOLOGIES

HAVING A BLAST! SIMULIA UK RUM 2015 Park Royal Hotel, Cheshire 3rd November 2015 Dr. David Winfield & John Stobbart

OIL & GAS TECHNOLOGIES

Contents

 Introduction  The Blast Chamber

 FEA Simulation Methodology  Material Model Development  Simulation Limitations & Validation

 FEA Simulation Results  Conclusions & Future Work

OIL & GAS TECHNOLOGIES Page 2

11/9/2015

Introduction

OIL & GAS TECHNOLOGIES 3

11/9/2015

Introduction Project Scope  FO> have moved into a larger upgraded facility.

 FO> to bring all component qualification programs inhouse for the first time (subsea, topside, onshore, offshore).

 Provision of a bespoke testing program for individual, high profile clients.

 Blast chamber and auxiliary support equipment (pumps, monitors, piping etc…); £350,000.

 Design a blast chamber to contain 0.2 MJ of energy.

 Initial developmental ‘Stage 1’ FE simulation of the blast chamber in conjunction with a physical test program. OIL & GAS TECHNOLOGIES Page 4

11/9/2015

The Blast Chamber


11/9/2015

The Blast Chamber Construction & Key Features Steel Girder Polyurethane

Pine

Dampers

(95x250 mm)

Disc Springs

6 mm Blast

Plate

L-Frame Vertical Steel

Concrete Base 20 mm

Column & Base


11/9/2015

Oak

Blast

(120x250 mm)

Plate

The Blast Chamber ….because size is everything!

 ~450 m3

x4 OIL & GAS TECHNOLOGIES Page 7

11/9/2015

x 643,000 x 11,250,000

FEA Simulation Methodology


11/9/2015

FEA Simulation Methodology Engineering Hurdles

 Difficult to design a blast chamber due to the lack of available performance data for materials subjected to ballistic tests.

 Assumptions made in hand calculations are extremely specific and do not consider elasto-plastic material properties.

 Interaction of components under flexure is difficult to predict.

 Blast chamber design must be conservative. High safety factors are required to account for unexpected events.

 0.2 MJ; family car travelling at ~40 mph.


11/9/2015

FEA Simulation Methodology ….How to get from A to B


11/9/2015

FEA Simulation Methodology Sub-modelling Architecture

Series of simple sub-models benchmarking factory tests to develop material properties. Series of less complex sub-models of interacting parts with damage and erosion. Single panel impact on nominal geometry to check speed of the solution.

Creation of 3 panel simulation with exact geometries and component positioning. Multiple simulation runs considering projectile position, mass and velocity.


11/9/2015

FEA Simulation Methodology Mesh Discretization

 1.02x106 Nodes  795x103 Elements  91 % C3D8R  9 % S4R


11/9/2015

FEA Simulation Methodology Boundary Conditions

 Friction co-efficient 0.15 on all contacting surfaces.  Ambient temperature (20°C).  Eroding surface contact specified.  Projectile fired 1 m above ground level at panel center.

6in OD

8.5 kg @ 217 ms-1


11/9/2015

26in OD

0.2 MJ

1,778 kg @ 15 ms-1

Material Model Development


11/9/2015

3 Point Bend Test Rig Physical Testing vs. FEA; Pine & Oak

 Test; 57 mm, FEA; 61.6 mm, + 8.1 % OIL & GAS TECHNOLOGIES Page 15

11/9/2015

 Test; 173 mm, FEA; 194 mm, + 12.6 %

Ballistic Testing Steel Blast Plates [7]

 3D solid plate

 3D shell plate

 19.3 mm thick

 26.7 mm thick

 19.8 kg projectile

 19.8 kg projectile

 80 ms-1

 102 ms-1

[7] Health & Safety Executive, Pressure Test Safety; Contract Research Report 168/1998, Department of Chemical Engineering & Chemical Technology and BJS Research.


11/9/2015

Further Sub-Modelling Dampers & Disc Springs  Polyurethane 80shoreA  Standard ASTM D412-06a test methods.  Elastomer material models developed further by Freudenberg GmbH to ensure highly accurate dynamic response.

Δ Dynamic travel (50 mm)  4 (spring) beam elements connecting 2 solid bodies.  k = 250 N/mm.  ASME VIII Div.2 ANNEX 3-D.


11/9/2015

Simulation Limitations & Validation


11/9/2015

Simulation Limitations

 Random nature of grain direction and presence of knots in wooden sleepers means material models are based on average data obtained from 3 point bend test.

 Ballistic testing of steel blast plates benchmarked with limited available test data (BRL [7]).

 Material models for oak, pine and steel are purposefully kept as simplistic as possible for initial stages of simulation development.

 Mass scaling used in very small sections of specific part geometries to improve overall solution timescale (further investigation required to assess impact on final solution).


11/9/2015

FEA Validation Key Features

DATA

THEORY

FEA

(+/-) %

NOTES

Deflection of polyurethane damper

18 mm

9.08 mm

-49 %

Average of 18 dampers around the steel plate before disc spring is energized.

Average bolt force

23,125 N

22,347 N

-4 %

To resist column moment at maximum deflection

Deflection of column

3.31 mm

7.83 mm

+137 %

Recorded at height 1 m above ground before disc spring is energized.

Stress in main support column

519.7 MPa

373.2 MPa

-28 %

Theory limited to elastic properties only

 Less flexure in dampers than expected.  Greater flexure in steel plate(s).  Hand calculations consider elastic material performance only.


11/9/2015

FEA Simulation Results


11/9/2015

FEA Simulation Results 1,778 kg @ 15 ms-1 @ Panel


11/9/2015


 Dynamic response of dampers  Compression of disc springs  Flexure of blast plates  Free movement in pine/oak sleepers  Sway in complete panel(s)  Recovery of panel(s)


11/9/2015



11/9/2015

FEA Simulation Results 1,778 kg @ 15 ms-1 @ Main Column


11/9/2015

FEA Simulation Results 1,778 kg @ 15 ms-1 @ Main Column


11/9/2015

FEA Simulation Results 1,778 kg @ 25 ms-1 ……THE SHOW STOPPER!!!


11/9/2015

FEA Simulation Results 1,778 kg @ 25 ms-1 ……THE SHOW STOPPER!!!


11/9/2015

Conclusions & Further Work


11/9/2015

Conclusion  Detailed ‘Stage 1’ explicit simulation model generated to represent the impact of a projectile into the walls of a custom blast chamber design.

 Implementation of simplistic and detailed dynamic material models derived based on physical tests for oak, pine, steel and elastomers.

 Blast chamber integrity is maintained up to a total impact energy of 0.2 MJ at multiple impact locations.

 Blast chamber can contain a range of projectile geometries and velocities.

 Sacrificial failure of multiple components ensures containment integrity. OIL & GAS TECHNOLOGIES Page 30

11/9/2015

Further Work

 Based on flexure seen from ‘Stage 1’ simulation, additional design changes and reinforcement will be made.

 Factory testing to be scheduled on a mock-up single panel wall section in the guise of a drop test.

 Model developed and correlated further based on the dynamics of the drop test.


11/9/2015

Acknowledgements

I would like to thank the following individuals and companies for their invaluable support during the project:

Mr. John Stobbart (Technical Director)

Mr. Laurence Marks & Support Team

Abaqus UK Headquarters, CSE


11/9/2015

OIL & GAS TECHNOLOGIES

Dr. David Winfield

FEA & CFD Development Engineer, R&D Freudenberg Oil & Gas Technologies Ltd – Metal Sealing Solutions - Vector Products

[email protected]

+44(0) 7952 055 963 +44(0) 1639 822 555 ext. 723


11/9/2015

2015 Simulia Uk David Winfield

Recommend Documents