ITER Design Criteria Jacket Stress Evaluation

1
ITER Design Criteria Jacket Stress Evaluation
Jacket Material Meeting February 14 2006

• Peter Titus
• MIT Plasma Science and Fusion Center, 185 Albany
  St. Cambridge Ma 02139
• http//www2.psfc.mit.edu/people/titus/

2
First, We have Design Guidance
It is evolving, and needs work, but it
exists. Also There is a New July 10 2005 Draft
Criteria that Substantially Re-Works the Present
Set of Criteria
3
Status of the Criteria
Current July 1 2005 criteria needs updating-
witness this entry in Table A2
Another is in the works
Magnet Structural Design Criteria  Part I Main
Structural Components and Welds  10 July 2005
Version 3 DRAFT 3 N.Mitchell  
This is only structural components for the TF and
excludes jackets
4
Material Qualification

• 4.2 Metallic Material Qualification
• The material properties must be measured over an
  adequate range of compositions to demonstrate
  that the minimum of the measured samples is
  representative of the minimum that will be
  obtained in a large production run. Therefore,
  generic ITER metallic materials must be qualified
  following the procedures defined in ASME BPVC
  2001 section II (Materials), Part A-1 (Material
  Specification Ferrous), Appendix 2, Mandatory
  Guideline on the Approval of New Materials under
  the ASME BPV Code. It is not required to submit
  the material to ASME for code approval but the
  equivalent documentation must be approved by
  ITER. The data must cover the operating
  temperature (4K).
• 4.3 Metallic Material Characterisation and
  Properties
• The material properties must be measured on the
  final product form after processing fully
  representative of that to be used for the
  components. The ASTM Standards that are
  applicable are
• A 370 Test Methods and Definitions for Mechanical
  Testing of Steel Products
• E 8M Standard Test Methods for Tension Testing of
  Metallic Materials Metric
• E 23 Standard Test Methods for Notched Bar Impact
  Testing of Metallic Materials
• E 813 Standard Test Method for JIc, A Measure of
  Fracture Toughness
• E 1450 Standard Test Method for Tension Testing
  of Structural Alloys in Liquid helium
• E 1820 Standard Test Method for Measurement of
  Fracture Toughness
• Equivalent test procedures from internationally
  accepted standards are acceptable.

5
The Qualification Process
Both Static Stress Evaluation and Fatigue
Evaluation are Required
6
Stress Components

• MC 2.5.1 Primary Stress (P)
• These are stresses that could (if sufficiently
  high) contribute to plastic collapse, as distinct
• from secondary stresses, which do not. They can
  also contribute to failure by fracture,
• fatigue, creep or stress corrosion cracking. They
  include all stresses arising from internal
• pressure and external loads. The primary stresses
  are divided into membrane, Pm, and
• bending, Pb components as follows.
• a) Membrane stress, Pm is the mean stress through
  the section thickness that is necessary to
• ensure the equilibrium of the component or
  structure.
• b) Bending stress, Pb is the component of stress
  due to imposed loading that varies linearly
• across the section thickness.
• The bending stresses are in equilibrium with the
  local bending moment applied to the
• component. For the purpose of this document, Pb
  is regarded as a stress superimposed upon
• Pm.
• MC 3.2.1 Limiting Stress Values (Sm)
• For structural material (including the jacket)
  and welds, the limiting stress for plastic
  collapse
• (equivalent to the 'limiting stress intensity'
  value in 1), Sm, at design temperature is
  defined

7
Static Stress Criteria

• MC 3.2.2 Plastic Collapse Allowable Limits
• The Tresca yield criterion has been adopted for
  combining stresses to compare with the
• plastic collapse criterion. The Tresca stress is
  calculated from the direct and shear stress
• components which must include all relevant
  contributions for the limit being considered.
• Based on elastic stress analyses, the following
  stress limits shall be met
• Primary membrane stress shall not exceed 1.0
  KSm
• Primary membrane plus bending stresses shall
  not exceed 1.3 KSm
• Primary plus secondary stress (not including
  residual stress or peak stress) shall not
• exceed 1.5 KSm . If the peak stress cannot be
  clearly differentiated from the secondary
• stress then both secondary and peak stress must
  be included.
• Residual and peak stresses are self limiting by
  local plasticity and no maximum value is
• specified.
• The multiplier K is dependent on the type of
  service conditions listed in MC 2.4 and Sm is the
  limiting stress value defined in MC 3.2.1. Thus
  in normal operation with Sm determined
• by the yield stress, the primary plus secondary
  stress does not exceed the yield value. This
• can be compared with the ASME code which would
  allow primary plus secondary stresses up
• to the ultimate value (1. section III,
  NB-3222.2) and the Russian nuclear code which
  would
• allow primary and secondary stresses up to twice
  the yield value 2.
• Residual stresses and peak stresses are
  considered for fatigue and fracture but not for
  static

8
Qualification Basis Jackets, must pass
Monotonic/Static Stress and Fatigue Criteria
Deterministic LEFM, Probabilistic LEFM ,SN
Qualification, are allowed,

• Monotonic Stress Check

Allowables (All based on Sm)
Deterministic LEFM Allowables
Sm2/3yield -No Check on Ultimate
The multiplier K is dependent on the type of
service conditions listed in MC 2.4
9
Applied Jacket Max Principal Stress
Faulted
Normal Operating
Jacket Metal Tensile stress at the ID of the
Central Solenoid 400 MPa
Faulted Stress is 50 Higher 600 MPa
10
Required Yield
ID Hoop is Primary Membrane
Mid Build is Primary Membrane
Static Stress Criteria is not Limiting
11
Sm Based on Yield only and No Ductility
Requirement (As in Fire Criteria)

• All the discussions of self relieving secondary
  stresses and the logic behind stress categories,
  including relaxation of requirements for faulted
  conditions are meaningless.
• This would be a problem, except that Fatigue
  limits overwhelm static stress limits.

12
Fatigue Governs Jacket Qualification
Probabilistic Fracture Mechanics
S-N Evaluation Still allowed, 2 on stress and
20 on cycles but not intended for jacket
qualification
Addresses Statistical Uncertainty of Paris
Constants, Inspectabilty and flaw size
determination and can address systems
interactions.
Deterministic LEFM Qualification
Factors of Safety
ASME XI rules for Inservice Inspection Invoked by
proposed revision to the criteria First Use
only Preservice. IWA-3300 Flaw
Characterization IWB Acceptance Standards
(IWB-3410.2 essentially allows developing your
own flaw acceptance standards
13
Probabalistic Fracture MechanicsJacket
Reliability (Uncertainty) Analysis
This is allowed by the criteria and is
essentially Juns Monte Carlo Simulation
Jun Feng
PSFC/RR-03-7
14
Residual Stresses Add to Applied Stresses in a
Fatigue Evaluation

• Nb3Sn conductor jacket (steel)
• The same stress levels as the NbTi steel jacket
  apply if the welding and jacketing is
• performed after the Nb3Sn heat treatment. If the
  jacketing is performed before, the
• reduction of yield stress at the heat treatment
  temperature and creep are assumed to
• reduce the stress to about 25MPa.
• Nb3Sn conductor jacket (incoloy)
• The stresses due to application of the jacket to
  the superconducting cable and winding
• have been calculated 31. The effect of the heat
  treatment is complex as the incoloy
• undergoes precipitation hardening. Before
  hardening occurs, there is stress relaxation, but
• the hardened material has a high high-temperature
  yield stress and low creep. Residual
• stresses after hardening have been measured as
  part of the work in 31 although they are
• not reported, and a reasonable average value
  appears to be about 50MPa.

15
For Deterministic Fracture Mechanics, The
Criteria Provide Guidelines for Defect Sizing
A factor of 2.0 is applied to the Initial Defect
with a 95 Detection Probability
The Criteria Document Guidelines for Defect
Sizing must be Verified by RD
16
MC 3.4.2.2.2 Probabilistic Assessment Leak
Before Break

• ..These examples indicate the advantage of
  the probabilistic method over the fixed defect
  method, as the failure probability is weighted
  according to the amount of the 'highest risk
  element that is present, and can be assessed
  according to the type of risk. Generally, a
  probability of failure below 0.1 may be adequate
  for a component that fails by 'leak before break'
  i.e. the crack penetrates the wall thickness
  before reaching the critical size, and there is
  no possibility (due to the limited physical size
  of the component) of the crack continuing to grow
  to critical size. This probability is
  appropriate, for example, for the conductor
  jacket in a coil that is replaceable. For
  components where catastrophic failure is
  possible, the failure probability must be less
  than 0.0001. The safety factor of 2 on the number
  of cycles is chosen to be adequate for a range of
  components, regardless of amount of 'risky'
  elements. It can be expected to ensure failure
  probabilities less than 0.001 with a fixed defect
  type of assessment.

17
New Criteria Revision (-But for TF Structures
Only)

• 7.3 Defect Acceptance for Leak Before Break
• Leak before break is the conventional description
  applied to a fatigue crack that is limited in
  size by the finite component dimensions before it
  can reach the critical size for unstable growth
  (i.e. fast fracture). The concept is described in
  1 section 9.5.2. Conventionally it is expected
  that the 'leak' is immediately detected and that
  operation can be stopped. When translated into
  the ITER magnet situation this expectation is
  doubtful. There is unlikely to be any fluid
  leakage when a crack extends to the through
  thickness of a plate, and the multiple available
  load paths within the magnet system make it
  unlikely that the condition can be detected by
  structural monitoring (of displacements for
  example). Under these conditions, crack growth
  will continue although with significantly
  modified shape factors and loading, which may be
  sufficient to restrict further growth. The limit
  on crack dimensions may prevent the crack size
  for fast fracture to occur from being reached.
  The assumption of leak before break is therefore
  not generally acceptable for the ITER magnet
  structures. Crack growth may be affected by
  particular geometries and an individual
  assessment is required for such cases.

18
The Next Criteria Revision

• Section XI (In-service Inspection) (and
  equivalently API 579 1) offer a satisfactory
  alternative. The design philosophy is therefore
  based on a demonstration of fitness for service
  (FFS) for a defined operating life, at the
  completion of manufacturing, by appropriate
  inspection of components.
• A defect based approach will be followed based
  around section XI (and API 579) where all
  components will be assumed to contain crack like
  defects up to the detection level of the
  pre-service inspection procedures. The
  manufacturing inspections will be considered as a
  'pre-service inspection' that provides
  certification for the required operating life,
  without further inspection. Operation beyond this
  certified mechanical life would require extensive
  disassembly and inspection.
• For manufacturers, this has some significant
  implications
• manufacturing inspections are at the same time
  in-service inspections, and that an explicit
  characterisation of the minimum detectable defect
  is required.

19
Historical, Factor of SafetyEvolves over a
long history of successes and failures.Examples
Boilers ASME III, VIII, European PED and Civil
Structures AISC ASC
Comments on Philosophy

• There are basically two ways to qualify a system

Reliability Analysis Start with Mission Need
(availability or Success Probability) and Develop
Criteria. Examples Aerospace, Military
Applications, Manufacturing Quality Control.
Or
Fracture mechanics calculations and Testing can
be based on either approach
20
Reliability Analysis
Cons For a first of a kind system it is
certain that there will be insufficient
data Conservative assumptions will almost
certainly lead to a prediction of failure. There
is some history relating to this objection. This
contributed to the Shuttle programs
implementation of reliability analysis after the
fact. Codified guidelines are lacking ASME XI?
Aerospace, military, or production QA standards
may not fit a fusion magnet

• Pros
• Provides Framework for establishing required
  conservatism
• Has a meaningful relationship with the safety
  analysis. Faulted conditions can be evaluated in
  the context of safety commitments, Common FMEA.
• Establishes priorities for
• Analysis
• RD
• Provides Framework and interfaces for Design
  Integration/Control
• Integrates with Systems Codes (Galambos at Oak
  Ridge?)

21
Conclusions

• ITER has a special purpose Criteria, and
  invokes other industry standards by reference. It
  is, however undergoing substantial alterations to
  include ASME XI and a Fitness for Service
  approach.
• ITER retains static stress criteria. But for
  jackets, fatigue based on fracture mechanics
  governs.
• Fracture based criteria demand a high level of
  knowledge of the material properties, and flaw
  size distributions.
• ITER Criteria basis remains Historical rather
  than using Reliability