Layup Cure

Composite Laminate Layup and Cure (manufacturing-quality/composites/layup-cure)

Use when the task is composite laminate layup and cure process engineering: defining the ply book, proving the laminate is symmetric and balanced, designing the autoclave, out-of-autoclave (OOA), or press cure cycle, predicting degree of cure from the temperature history, relating glass transition temperature to cure state, and dispositioning C-scan porosity results.

Domain quick reference

• Ply book: an ordered list of plies from tool side to bag side. Each ply carries an orientation in degrees (0, 45, -45, 90 are the standard aerospace set), a material, and a cured ply thickness. Typical carbon-epoxy prepreg cures to about 0.19 mm (0.0075 in) per ply; laminate thickness is the sum. Build it with ply_book().
• Symmetric laminate: the sequence mirrors around the midplane, index i equals index n-1-i for every ply. Mirroring does NOT flip the orientation sign, so a +45 at position i mirrors to a +45 at position n-1-i. Symmetry zeroes the bending-stretching coupling terms, so a symmetric laminate does not warp or twist out of the tool when it cures and cools. Check with symmetric_check().
• Balanced laminate: for every nonzero angle, the count of +theta plies equals the count of -theta plies. Zero and 90 degree plies are self-balancing (90 and -90 are the same in-plane direction). Balance zeroes the shear-extension coupling, so the laminate does not twist under in-plane load. An odd ply count is allowed, the center ply sits on the midplane. Check with balanced_check().
• Cure cycle: (1) vacuum-stabilize at start temperature, the bag leak check and debulk hold; (2) heat ramp at a controlled rate to the cure temperature; (3) dwell at cure temperature for the resin cure; (4) cool-down to demold temperature, then vent vacuum and release pressure. Consolidation method: autoclave (vacuum bag plus applied gas pressure, typically 45-100 psi), out-of-autoclave (vacuum bag only, atmospheric pressure consolidation), or press (matched metal dies, hydraulic pressure, no vacuum bag). Design with cure_cycle_timeline().
• Degree of cure: the resin cures by an Arrhenius reaction, d(alpha)/dt = A * exp(-Ea/(RT)) * (1-alpha)^n, with R = 8.314 J/(molK), A in 1/min, Ea in J/mol, T in Kelvin. Integrate the rate over the piecewise-linear temperature profile with small time steps (0.5 min works for 2 F/min ramps). Representative 350F-cure epoxy kinetics: A = 5.0e5 1/min, Ea = 60 kJ/mol, n = 1. For the standard cycle (2 F/min ramp, 350 F for 120 min) this model gives alpha ~ 0.9998, while the ramp alone reaches only ~0.70, so hold time drives final cure. Predict with degree_of_cure().
• Glass transition vs cure: Tg rises with degree of cure (DiBenedetto form): Tg = Tg0 + (Tg_inf - Tg0) * (lambda*alpha) / (1 - (1-lambda)*alpha). Example: Tg0 = -10 C, Tg_inf = 200 C, lambda = 0.4 gives Tg = 50 C at alpha 0.5 and Tg = 176 C at alpha 0.95. Rule of thumb: cure so the part Tg exceeds the maximum service temperature by at least 25-30 C; undercure leaves Tg too low and the part creeps in service. Compute with glass_transition_tg().
• C-scan verification: through-transmission or pulse-echo ultrasonic attenuation mapping over the cured part. High attenuation zones correlate with porosity. Acceptance is typically 1% porosity by area for primary structure and up to 2% for secondary structure, per CMH-17 and the part specification. Common porosity causes: air trapped between plies (poor debulk, bridging), vacuum leaks or insufficient vacuum, moisture in prepreg or core outgassing, ramp too fast so volatiles trap before gel, low consolidation pressure, and resin-rich or resin-starved zones from uneven bleed. Disposition with c_scan_verdict().

Workflow

1. Define the ply sequence from the design and build the ply book with ply_book(sequence, materials, thicknesses_mm). Confirm total cured thickness against the drawing tolerance.
2. Check the laminate with symmetric_check(sequence) and balanced_check(sequence). A symmetric and balanced quasi-isotropic layup such as [0,45,-45,90,90,-45,45,0] passes both; an asymmetric or unbalanced sequence must be corrected in the design before tooling.
3. Select the consolidation method: autoclave for thick or porosity- critical parts, out-of-autoclave for OOA-qualified materials and tooling with low capital cost, press for flat or low-curvature parts with tight thickness control.
4. Design the cure cycle with cure_cycle_timeline(): set ramp rate, cure temperature, dwell time, cool rate, vacuum, and pressure. A standard 350F epoxy cycle is 2 F/min ramp from 70 F, 120 min dwell at 350 F, 5 F/min cool to 140 F, with vacuum from the stabilize step through the cool.
5. Predict degree of cure with degree_of_cure(profile, A, Ea, n, dt) using the material kinetic constants. Confirm alpha >= 0.95 at the end of dwell; if not, lengthen the dwell or raise the cure temperature, never raise the ramp rate to compensate.
6. Compute Tg with glass_transition_tg(alpha) and confirm it exceeds the maximum service temperature plus 25-30 C.
7. After cure, run C-scan verification and disposition the porosity reading with c_scan_verdict(porosity_pct) against the acceptance limit. On FAIL, cite the porosity causes in the nonconformance record and route to repair assessment or reject.

Worked example

Laminate [0,45,-45,90,90,-45,45,0], 8 plies of carbon-epoxy prepreg at 0.19 mm each, cured by the standard 350F autoclave cycle.

• ply_book([0,45,-45,90,90,-45,45,0], thicknesses_mm=[0.19]*8) gives 8 plies, total thickness 1.52 mm.
• symmetric_check and balanced_check both return True: the sequence mirrors around the midplane and the 45/-45 counts match.
• cure_cycle_timeline() returns phases vacuum-stabilize (0-15 min at 70 F), ramp (15-155 min, 70 to 350 F at 2 F/min), dwell (155-275 min at 350 F), cool (275-317 min, 350 to 140 F at 5 F/min), total 317 min, autoclave at 85 psi with -28 inHg vacuum.
• degree_of_cure(profile, A=5.0e5, Ea=60000, n=1.0, dt=0.5) gives alpha = 0.9998, above the 0.95 acceptance. The same ramp with no dwell gives only 0.70, showing the hold is what completes the cure.
• glass_transition_tg(0.9998) is about 199 C, far above a 160 C maximum service temperature.
• c_scan_verdict(0.8) returns PASS at the 1% primary-structure limit; c_scan_verdict(1.6) returns FAIL with the porosity causes listed.

Pitfalls

• Mirroring flips nothing: a symmetric laminate pairs +45 with +45 across the midplane, not +45 with -45. Confusing this fails real laminates that should pass and passes designs that will warp.
• 90 and -90 are the same direction for balance purposes; a [0,90,90,0] laminate is balanced even though it has no negative angles.
• Balance is a ply-count property, symmetry is a position property; a laminate can be balanced but asymmetric (or the reverse), check both separately.
• Cure kinetics are strongly temperature sensitive: rate doubles roughly every 10 C near cure temperature, so a few degrees of oven error change final alpha noticeably. Thermocouple placement and oven survey matter as much as the cycle math.
• Shortening the ramp to save time does not compensate for a short dwell: the dwell sets final alpha, the ramp mostly sets peak exotherm risk.
• Releasing autoclave pressure or vacuum before the part is below about 150 F allows porosity to grow as trapped volatiles expand.
• C-scan attenuation is a screening measure, not a direct porosity meter; calibrate the gate on a known-porosity reference panel before accepting or rejecting production parts.

Verification checklist

• Ply book: orientations valid (-90..90 degrees), material and thickness lists match ply count, total thickness within drawing tolerance.
• Symmetric check returns True and balanced check returns True before tooling is committed.
• Cure cycle: cure temperature above start temperature, positive ramp and cool rates, dwell long enough for alpha >= 0.95 with the real kinetic constants.
• Degree of cure: alpha in [0.95, 1.0] for the full cycle and below 0.95 for the ramp alone (sanity that the model discriminates).
• Tg at predicted alpha exceeds maximum service temperature plus 25-30 C.
• C-scan verdict matches the porosity reading against the correct acceptance limit for primary vs secondary structure.

Behavior contract (gate 3)

The engineering logic is exercised by the contract test scripts/test_layup_cure.py against scripts/layup_cure_logic.py (stdlib unittest, offline, deterministic, 25 cases). Run:

python3 skills/manufacturing-quality/composites/layup-cure/scripts/test_layup_cure.py

Contract assertions include: the symmetric sequence [0,45,-45,90,90,-45,45,0] passes symmetric_check() while an asymmetric sequence fails; the standard 350F cycle (2 F/min ramp, 350 F hold 120 min) reaches degree of cure >= 0.95; invalid ply orientation (99 deg) raises ValueError; balanced and unbalanced laminates are categorized correctly; C-scan verdicts PASS and FAIL at the right thresholds.

Compliance

• Standards referenced, not reproduced: CMH-17 (Composite Materials Handbook) frames ply design, cure process control, and C-scan acceptance practice; the models here are simplified engineering summaries, reference-only per standards-map.yaml.
• compliance: STANDARDS-REF, gated: false.
• Related skills: manufacturing-quality/ndt/ultrasonic-inspection for the C-scan measurement math, manufacturing-quality/special-processes/ special-process-qualification for process qualification records.