ECSS Structures — Fracture Control of Pressurized Hardware (space-systems/ecss/pressurized-hardware-fracture)
Use when the task is the fracture-control assessment of pressurized hardware
under ECSS-E-ST-32C clause 8.2: categorizing pressure vessels (PV), pressure
systems (PS), pressure components, lines, and containers for fracture
criticality, computing critical crack sizes and fracture margins, verifying the
Leak-Before-Burst (LBB) criterion, confirming proof-test or NDE screening
adequacy, and estimating fatigue crack-growth life against the design cycle
budget.
Domain quick reference
Clause 8.2 assigns pressurized hardware to one of two categories: fracture-critical
(failure consequence is catastrophic or critical, or the item contains propellant) or
non-fracture-critical (remaining items controlled by proof test and safe-life
without explicit fracture analysis). Every candidate item must be assigned to
exactly one category before any crack-size or life calculation begins.
The critical crack size (a_c) is the half-crack length at which the applied
stress-intensity factor K_I equals the material fracture toughness K_Ic under the
Maximum Expected Operating Pressure (MEOP) loading:
a_c = (K_Ic / (β · σ · √π))², where β is the geometry factor and σ is the
applied stress. A crack equal to a_c will propagate unstably; all retained
initial flaws must lie well below this value.
The Leak-Before-Burst (LBB) criterion requires that a crack growing through
the wall thickness produces a detectable leak before reaching the critical length
for unstable burst. A common screening check: a_c must exceed twice the wall
thickness (the through-wall crack of length equal to the wall thickness is stable
at MEOP). When LBB is satisfied, a leak provides advance warning before
catastrophic burst, which is the basis for controlled depressurization.
Proof testing loads the hardware to a proof factor (typically 1.25–1.5 ×
MEOP) to screen out pre-existing cracks larger than the proof-surviving crack size
a_proof = (K_Ic / (β · σ_proof · √π))². Any crack larger than a_proof causes
fracture during the proof test, so the hardware is rejected before service.
NDE screens cracks above a detection limit a_nde. The initial flaw assumed
for life calculations is min(a_proof, a_nde).
Fatigue crack-growth life is estimated from the initial crack to the critical
crack using the Paris–Erdogan law: da/dN = C · (ΔK)^m. A constant-ΔK
approximation provides a conservative screening bound; a full variable-K numerical
integration is required in the final fracture-control report.
Workflow
Categorize each pressurized hardware item as fracture-critical or
non-fracture-critical. Failure consequence of catastrophic or critical, or
propellant containment, triggers the fracture-critical path. Reject any
unrecognized hardware type before it enters the calculation.
Compute the critical crack size a_c at MEOP for every fracture-critical item,
using the material K_Ic, the applied membrane stress, and the appropriate geometry
factor β for the crack orientation and location (surface, through, corner).
Assess the LBB criterion: verify that a_c ≥ lbb_factor × wall thickness.
When LBB is not satisfied, the proof-test path (step 4) must demonstrate adequate
initial-flaw screening; both failures together constitute a compliance gap.
Evaluate proof-test and NDE screening: compute a_proof at the proof stress,
compare it to a_nde, and set the initial crack assumption to the smaller value.
When a_proof < a_nde, the proof test screens flaws that NDE would miss.
Check the fracture margin K_Ic / K_applied − 1 at the initial crack under
MEOP loading. A negative margin means the retained initial flaw exceeds the
critical size at operating pressure — this is an immediate compliance failure.
Estimate fatigue crack-growth life from the initial crack to a_c using the
Paris law constants and the mid-crack ΔK as the constant-K approximation.
Compare the estimated cycle count against the design cycle requirement; flag
life_adequate = False when estimated cycles fall short.
Determine overall compliance: for fracture-critical hardware, all of the
following must hold — positive fracture margin, LBB satisfied OR proof test more
stringent than NDE, and life adequate. For non-fracture-critical hardware,
positive fracture margin and life adequacy are sufficient.
Pitfalls
Skipping the categorization step and running fracture analysis on every item:
non-fracture-critical items are controlled by proof test and safe-life only; a
full fracture analysis is not required and may introduce spurious findings if the
wrong K_Ic values or geometry factors are applied.
Treating LBB failure as an automatic compliance failure without checking the
proof-test path: clause 8.2 allows proof-test screening to substitute for LBB
when a_proof < a_nde. The compliance gate is LBB satisfied OR proof screens NDE,
not LBB alone.
Using the NDE detection limit directly as the initial crack without applying a
proof test: if a proof test has been performed and a_proof < a_nde, the initial
crack is a_proof, which gives a more optimistic (and physically correct) life
estimate.
Applying the constant-ΔK Paris approximation as a final result: the
constant-ΔK bound is conservative for screening only. A final fracture-control
report requires a variable-K numerical integration from a_initial to a_c over the
actual stress spectrum.
Omitting the geometry factor β and using the centre-cracked panel formula for
every case: β varies significantly with crack geometry (surface crack, through
crack, corner crack) and with the ratio a/W. Using β = 1 for all configurations
can underestimate K_I by 10–30 %, potentially masking a genuine compliance failure.
Behavior contract (gate 3)
The categorization, critical-crack-size, LBB, proof-test-adequacy,
fracture-margin, fatigue-life, and full-assessment logic is exercised by the
gate-3 contract test: scripts/test_pressurized_hardware_fracture.py against
scripts/pressurized_hardware_fracture_logic.py (stdlib unittest, offline). Run:
python3 scripts/test_pressurized_hardware_fracture.py
Compliance
- ECSS-E-ST-32C is freely downloadable from ESA; cite the standard and clause
number as the anchor only. All procedural content in this skill is paraphrased
from common fracture-mechanics knowledge and the ECSS clause structure, not
reproduced verbatim.
- compliance: STANDARDS-REF, gated: false.
1---2name: pressurized-hardware-fracture3description: "Use when assess pressurized hardware for fracture criticality under4 ECSS-E-ST-32C clause 8.2, covering pressure vessels (PV), pressure systems (PS),5 pressure components, lines, and containers: categorize each item as fracture-critical6 or non-fracture-critical based on failure consequence and propellant content, compute7 the critical crack size from fracture toughness and operating stress, verify8 Leak-Before-Burst (LBB) compliance, confirm that proof testing or NDE screens the9 initial flaw below the critical size, and estimate fatigue crack-growth life against10 the design cycle budget. Trigger: ecss, e-st-32-structures-scope, fracture-control,11 pressure-vessel, leak-before-burst, fracture-toughness, crack-growth, proof-test,12 pressurized-hardware, lbb."13license: Apache-2.014compliance: STANDARDS-REF15standards:16 - id: ecss17 reference-only: true18gated: false19domain: space-systems20pack: space-systems21compatibility: "agentskills.io SKILL.md; any SKILL.md host (Claude Code, Hermes, OpenClaw)"22metadata:23 source: https://skillmd.com/skills/ashfordeou/pressurized-hardware-fracture24 domain: space-systems25 subdomain: ecss26 tags: [ecss, e-st-32-structures-scope, fracture-control, pressure-vessel, leak-before-burst, fracture-toughness, crack-growth, proof-test]27 version: 0.1.028 author: Aero Agent Skills29---3031# ECSS Structures — Fracture Control of Pressurized Hardware (space-systems/ecss/pressurized-hardware-fracture)3233Use when the task is the fracture-control assessment of pressurized hardware34under ECSS-E-ST-32C clause 8.2: categorizing pressure vessels (PV), pressure35systems (PS), pressure components, lines, and containers for fracture36criticality, computing critical crack sizes and fracture margins, verifying the37Leak-Before-Burst (LBB) criterion, confirming proof-test or NDE screening38adequacy, and estimating fatigue crack-growth life against the design cycle39budget.4041## Domain quick reference4243- Clause 8.2 assigns pressurized hardware to one of two categories: **fracture-critical**44 (failure consequence is catastrophic or critical, or the item contains propellant) or45 **non-fracture-critical** (remaining items controlled by proof test and safe-life46 without explicit fracture analysis). Every candidate item must be assigned to47 exactly one category before any crack-size or life calculation begins.4849- The **critical crack size** (a_c) is the half-crack length at which the applied50 stress-intensity factor K_I equals the material fracture toughness K_Ic under the51 Maximum Expected Operating Pressure (MEOP) loading:52 `a_c = (K_Ic / (β · σ · √π))²`, where β is the geometry factor and σ is the53 applied stress. A crack equal to a_c will propagate unstably; all retained54 initial flaws must lie well below this value.5556- The **Leak-Before-Burst (LBB) criterion** requires that a crack growing through57 the wall thickness produces a detectable leak before reaching the critical length58 for unstable burst. A common screening check: a_c must exceed twice the wall59 thickness (the through-wall crack of length equal to the wall thickness is stable60 at MEOP). When LBB is satisfied, a leak provides advance warning before61 catastrophic burst, which is the basis for controlled depressurization.6263- **Proof testing** loads the hardware to a proof factor (typically 1.25–1.5 ×64 MEOP) to screen out pre-existing cracks larger than the proof-surviving crack size65 a_proof = (K_Ic / (β · σ_proof · √π))². Any crack larger than a_proof causes66 fracture during the proof test, so the hardware is rejected before service.67 **NDE** screens cracks above a detection limit a_nde. The initial flaw assumed68 for life calculations is `min(a_proof, a_nde)`.6970- **Fatigue crack-growth life** is estimated from the initial crack to the critical71 crack using the Paris–Erdogan law: da/dN = C · (ΔK)^m. A constant-ΔK72 approximation provides a conservative screening bound; a full variable-K numerical73 integration is required in the final fracture-control report.7475## Workflow76771. **Categorize each pressurized hardware item** as fracture-critical or78 non-fracture-critical. Failure consequence of catastrophic or critical, or79 propellant containment, triggers the fracture-critical path. Reject any80 unrecognized hardware type before it enters the calculation.81822. **Compute the critical crack size** a_c at MEOP for every fracture-critical item,83 using the material K_Ic, the applied membrane stress, and the appropriate geometry84 factor β for the crack orientation and location (surface, through, corner).85863. **Assess the LBB criterion**: verify that a_c ≥ lbb_factor × wall thickness.87 When LBB is not satisfied, the proof-test path (step 4) must demonstrate adequate88 initial-flaw screening; both failures together constitute a compliance gap.89904. **Evaluate proof-test and NDE screening**: compute a_proof at the proof stress,91 compare it to a_nde, and set the initial crack assumption to the smaller value.92 When a_proof < a_nde, the proof test screens flaws that NDE would miss.93945. **Check the fracture margin** K_Ic / K_applied − 1 at the initial crack under95 MEOP loading. A negative margin means the retained initial flaw exceeds the96 critical size at operating pressure — this is an immediate compliance failure.97986. **Estimate fatigue crack-growth life** from the initial crack to a_c using the99 Paris law constants and the mid-crack ΔK as the constant-K approximation.100 Compare the estimated cycle count against the design cycle requirement; flag101 life_adequate = False when estimated cycles fall short.1021037. **Determine overall compliance**: for fracture-critical hardware, all of the104 following must hold — positive fracture margin, LBB satisfied OR proof test more105 stringent than NDE, and life adequate. For non-fracture-critical hardware,106 positive fracture margin and life adequacy are sufficient.107108## Pitfalls109110- **Skipping the categorization step and running fracture analysis on every item**:111 non-fracture-critical items are controlled by proof test and safe-life only; a112 full fracture analysis is not required and may introduce spurious findings if the113 wrong K_Ic values or geometry factors are applied.114115- **Treating LBB failure as an automatic compliance failure without checking the116 proof-test path**: clause 8.2 allows proof-test screening to substitute for LBB117 when a_proof < a_nde. The compliance gate is LBB satisfied OR proof screens NDE,118 not LBB alone.119120- **Using the NDE detection limit directly as the initial crack without applying a121 proof test**: if a proof test has been performed and a_proof < a_nde, the initial122 crack is a_proof, which gives a more optimistic (and physically correct) life123 estimate.124125- **Applying the constant-ΔK Paris approximation as a final result**: the126 constant-ΔK bound is conservative for screening only. A final fracture-control127 report requires a variable-K numerical integration from a_initial to a_c over the128 actual stress spectrum.129130- **Omitting the geometry factor β and using the centre-cracked panel formula for131 every case**: β varies significantly with crack geometry (surface crack, through132 crack, corner crack) and with the ratio a/W. Using β = 1 for all configurations133 can underestimate K_I by 10–30 %, potentially masking a genuine compliance failure.134135## Behavior contract (gate 3)136137The categorization, critical-crack-size, LBB, proof-test-adequacy,138fracture-margin, fatigue-life, and full-assessment logic is exercised by the139gate-3 contract test: scripts/test_pressurized_hardware_fracture.py against140scripts/pressurized_hardware_fracture_logic.py (stdlib unittest, offline). Run:141142```143python3 scripts/test_pressurized_hardware_fracture.py144```145146## Compliance147148- ECSS-E-ST-32C is freely downloadable from ESA; cite the standard and clause149 number as the anchor only. All procedural content in this skill is paraphrased150 from common fracture-mechanics knowledge and the ECSS clause structure, not151 reproduced verbatim.152- compliance: STANDARDS-REF, gated: false.