MMS5A Design and Failure of Materials Systems under Multiaxial Loads

1

• MMS5A Design and Failure of Materials Systems
  under Multi-axial Loads

2
Driveshaft Failure

• We designed these CFRP driveshafts using the
  maximum strain criterion but the fail at two
  thirds of the expected load..

3
Tank Failure - Dec 2001

• We built these GRP structures according to the
  ruleswe dont know why they have failed...

4
Failure vs Design

• We have looked at the world-wide failure theory
  exercise we dont know what it is telling us.
  It does not seem to relate to the questions we
  are interested in..

5
MMS 5 Project

• Team
• ESR Technology
• NPL
• NetComposites
• Objectives
• Compilation of failure theory data
• Augmentation of materials database
• Input to standards
• Interactive knowledge base

Task 1
Failure Theory Data
Task 3
Task 2
Development of Lifetime
Extension of Database
Methodologies
Task 4
Task 5
Development of IKB
Dissemination
6
Interactive Knowledge Base

• More than a presentation of data.
• More than a comparison of competing failure
  criteria
• Providing knowledge to answer questions like.

Main Menu
Browse Criteria
Recommend a Criteria
Interactive Knowledge Base
Browse all Criteria
Design of Pressurised Parts


Maximum stress
Method
Application
Theory
Data
Applicability
Background


Maximum strain
Laminate Construction
-

Hart Smith
-

Eckold
The optimum angle for cylindrical parts under


Distortional Energy
internal pressure with closed ends is considered
to be
-

Tsai-Azzi

55
o
.
-

Hoffman
-

Puck
If there is no liner present failure is deemed to
be
-

Puppo-Evensen
the onset of micro-cracking as this will initiate
-
leakage and allow the contents to gain access to

Swanson
the reinforcement, thereby accelerating
degradation.


Tensor
With a liner fibre failure can be taken as the
governing
-

Tsai-Wu
factor for design.
-

Haung
-

Goldenblat
Design codes employ design methods based on the


Micromechanics
maximum strain criteria with limiting values set
at the
-

Shear lag
range between 0.2 and 0.3.
-

McCartney


Damage Mechanics
In prediction for the onset of micro-cracks the
-

Tareja
distortional failure criteria are conservative in
the
Examples of failure predictions for

55
o

laminates


Joshi
What information can I use to design my
composite beam.?
(NB. See for an example www.aeat.co.uk/ndt/IKB)
7
(No Transcript)
8
(No Transcript)
9
Compiling of Failure Theory Data

• gt50 failure theories, lots of data, numerous test
  methods, conflicting interpretations, disparate
  correlations..
• What theory can be used ..
• With what load scenarios?
• With what materials?
• With what confidence?
• And validated with what test method?

10
Compiling of Failure Theory Data

• 20 main failure criteria catalogued
• Browse methods for failure criteria formulated
• Catalogued failure criteria for IKB and
  standalone database
• Assessed results of World Wide Failure Exercise
  in terms of conservatism for design
• Catalogued design criteria/methods for IKB

11
Failure Criteria

• Chang failure criterion
• Fracture mechanics based failure criteria
• Grant-Saunders failure criterion
• Hashin failure criterion
• Maximum strain failure criterion
• Maximum stress failure criterion
• MDE failure criterion
• Multicontinuum failure criterion
• Puck failure criterion
• Puck modified failure criterion
• Puppo-Evenson failure criterion

• Rotem failure criterion
• Stress Based Grant-Saunders failure criterion
• Sun failure criterion
• Ten percent rule failure criterion
• Truncated maximum strain failure criterion
• Tsai-Hill failure criterion
• Tsai-Wu failure criterion
• Yamada and Sun failure criterion
• Zinoviev failure criterion

12
(No Transcript)
13
Failure Criteria Database

• Information available on the web or in a
  standalone MS Access database
• Contains all the information on the failure
  criteria reviewed
• Includes advantages and disadvantages
• MS Access database available for purchase for
  incorporation into other software packages.

14
Failure Criteria into Design Criteria

• Failure criteria attempt to predict when and how
  a composite will fail
• Design criteria need to conservatively calculate
• how strong a composite will be
• how long a composite will last
• Design criteria incorporate probabilistic and/or
  pragmatic factors
• Design codes incorporate industry knowledge of
  failure modes/environment/loading

15
(No Transcript)
16
(No Transcript)
17
What do we do where design codes dont exist?

• Recommend a criterion attempts to help

18
Shafts Automotive front wheel drive car
Functional

• Torsional Fatigue
• Ultimate torsion
• Torsional and out of plane impact
• Environmental
• Heat
• Moisture
• Corrosion
• Sufficient stiffness for driveline stability
• Torsional creep?

19
Shafts Automotive front wheel drive car CFRP q

• First ply failure
• End of useful life
• Ultimate failure
• Safety related
• Onset of delamination
• Load transfer
• Impact
• Environmental
• Stiffness?

20
Shafts Automotive front wheel drive car CFRPq
Quasi-static In plane shear

• Tsais failure criterion gives a reasonable fit
  to experimental data for in-plane shear.
• An appropriate design factor needs to be applied,
  selected on the basis of weight/reliability/confid
  ence. Values between 1.5 and 3 are appropriate.
• Environmental conditions (temperature/humidity/cor
  rosive fluids) will affect strengths requiring
  larger factors.
• Theoretical composite strengths should ideally be
  used as initial sizing indicators only.
• It is recommended that strengths obtained from
  components manufactured used production
  techniques and materials are used for final
  design.

21
Shafts Automotive front wheel drive car q
P
22
(No Transcript)
23
Extending the Dataset

• Materials of wide relevance
• Making maximum use of existing data, focusing on
  the gaps
• Instrumentation to track materials performance
• Measured data and failure predictions must relate
  to the same material behaviour
• Review on multi-axial test methods
• Input to standards
• Arguably the most effective means of dissemination

24
Extending the Dataset

• Modelling to extend the validity
• Understanding allows application with confidence

25
Lifetime Methodologies

• Case studies
• To confirm and demonstrate relevance
• Lifetime performance
• How can test data be applied to my application?
• How long will it last?
• What factors of safety are used?

Design Envelopes for PSX (Pipe)
Axial Stress (psi)
5000
4500
4000
3500
Sustained (
excl thermal)
3000
Sustained (
incl thermal)
2500
Occasional
2000
1500
1000
500
0
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Hoop Stress (psi)
26
Lifetime methodologies

• How to define the design allowables at the
  required design lifetime
• An issue for composites is that the design
  allowables can be a function of time (e.g. the
  longer the load is continuously applied, the
  lower the strength)
• Design allowables are values of stress or strain
  against which the calculated stress and strains
  for each design load case are compared. (Usually
  with a partial or safety factor included!)
• Need to calculate stress, strain within the
  composite due to applied loads

27
Component Case Studies

• Case Studies
• First case study composite driveshaft
• Second case study composite repair
• Third case study Designing thick ply composites
  by Marine Composites

28
Generic Case studies

• Composite components and standards
• Pipes and pipework ISO 14692 (BS 7519)
• Vessels PrEN 13121 (BS 4994), ASME X
• Composite repair (ASME PCC-2)
• Bridge strengthening (TR-55, Ciria C585)
• Advanced certification methodology
  (DOT/FAA/AR-96/111)
• Structural Materials Handbook ESA PSS-03-203

29
GRP pipes failure mode
What are we designing against?
The failure mechanism is weepage.
Weepage results in a loss of containment of the
pipe contents through an interconnected series of
through thickness matrix cracks an interlaminar
delaminations.
30
GRP pipes short term failure
Can we use short term data?
Short term internal pressure test non-linear
structural response of the pipe makes short term
tests difficult to interpret for long term design
use.
31
GRP pipes short term failure envelope

• Not efficiently designed
• Joint performance important
• Reduced performance with extra axial loads
• Can predict envelope with reasonable accuracy in
  first quadrant
• What is required is long-term design envelope
• Assume it can be contracted from short term
  envelope
• Verified by a few long-term tests

Failure criterion
32
GRP pipes approximate short term failure
envelope
Short term data used for design allowables

• Approximate short term envelope based on 2
  measured data points
• Knowledge of performance used to construct
  simplified short term failure envelope

33
GRP pipes short to long term failure
Long term data used for design allowables

• Regression curve for GRP pipes
• Based on internal pressure tests up to 10,000
  hours
• Design pressure based on extrapolation to
  required lifetime then application of safety
  factor
• Design pressure is less than onset of damage
  point in short-term test

34
GRP pipes summary of design methodology
Long term design allowables

• Derive short-term envelope from tests
• Derive short-term to long-term de-rating from
  regression curve
• Apply safety factor, 1.33, (uniformly) to
  determine long-term design envelope

35
Conclusions from GRP pipe case study

• To summarise the generic methodology that can be
  derived from the practical example of GRP pipes
  can be described as follows
•  
• Define failure mechanism (applicable to both long
  term and short term)
• Measure or define relevant define short term
  failure criterion or envelope
• Define or measure long term de-rating factor
• Infer long term failure criterion or envelope
• Define design factors to produce design envelope

36
Generic Case studies

• Composite components and standards
• Pipes and pipework ISO 14692 (BS 7519)
• Vessels PrEN 13121 (BS 4994), ASME X
• Composite repair (ASME PCC-2)
• Bridge strengthening (TR-55, Ciria C585)
• Advanced certification methodology
  (DOT/FAA/AR-96/111)
• Structural Materials Handbook ESA PSS-03-203

37
(No Transcript)