What industries does newtsim serve?

newtsim serves oil and gas (flow assurance, drilling fluids, subsea pipeline), pharmaceutical and drug discovery, energy storage (battery materials, electrolytes), semiconductor process simulation, green hydrogen (electrolyser and catalyst design), nuclear decommissioning, aerospace and composites, corrosion and materials, carbon capture, agrochemicals, mining and geotechnical, and fermentation and bioprocessing.

How is newtsim different from standard CFD or simulation software?

Standard simulation tools use empirical models at a single scale. newtsim runs the full physics chain from quantum mechanics and molecular dynamics through to plant-scale CFD and structural mechanics, GPU-native, with direct data flow between every scale and no third-party solver dependencies. Constitutive behaviour is predicted from first principles rather than empirically fitted.

What does a fixed-price simulation study include?

A newtsim fixed-price study has a defined scope, clear deliverables, fixed cost agreed upfront, and is delivered in weeks rather than quarters. Studies typically include a scoping call, physics-grounded simulation using the appropriate newtsim solvers, results analysis, and a written report with actionable recommendations.

What is physics-grounded simulation?

Physics-grounded simulation derives material properties and constitutive behaviour from lower-scale physics (molecular dynamics and quantum mechanics) rather than from empirical fitting. This gives accurate predictions for novel materials and extreme conditions outside the range of experimental data, and enables root-cause analysis at the mechanistic level.

How does newtsim approach oil and gas flow assurance simulation?

newtsim runs a continuous physics chain from molecular wax crystallisation to full pipeline transient multiphase flow, with no empirical handoffs between scales. Wax appearance temperatures, gel strength, and deposition kinetics are predicted from molecular dynamics (newtsim Bond), passed directly to mesoscale morphology (newtsim Pack), and then propagated into multiphase CFD (newtsim Stream), giving flow assurance predictions that hold outside the range of laboratory data.

Can newtsim simulate battery electrolyte behaviour and degradation?

Yes. The Bond solver runs molecular dynamics for Li-ion transport and solid-electrolyte interphase (SEI) formation; Pack handles electrode morphology at the mesoscale; these results couple upward to cell-scale electrochemical models. This gives physics-grounded predictions of electrolyte degradation, capacity fade, and thermal behaviour without relying solely on empirical ageing fits.

How does physics-grounded drug discovery simulation differ from docking software?

Molecular docking software treats the protein and ligand as rigid bodies and scores binding poses by geometry. Physics-grounded simulation uses molecular dynamics to capture conformational flexibility and explicit solvation effects that rigid-body docking misses, predicting binding free energy, ADMET properties, and metabolic activation from first principles. This is particularly valuable for targets where induced-fit binding or solvent-mediated interactions drive selectivity.

What makes GPU-native simulation faster than CPU-based CFD?

GPU parallelism maps naturally to the computational kernels underlying molecular dynamics and CFD: large numbers of independent force or flux calculations that can execute simultaneously. newtsim runs all simulation scales GPU-native, avoiding the CPU-to-GPU data transfers that bottleneck traditional solvers which run physics on CPU and only offload selected operations to the GPU. This removes the transfer overhead and allows the full multiscale chain to run without the latency penalties of heterogeneous pipelines.

Does newtsim replace existing simulation tools like ANSYS or COMSOL?

newtsim fills the multiscale gap, connecting molecular-scale physics to the engineering-scale models that tools like ANSYS operate at. It is not a replacement for FEA-only or single-scale workflows where those tools are the right fit. Where newtsim adds value is in problems where constitutive behaviour, material properties, or interfacial physics need to be derived from lower-scale simulation rather than taken from a handbook or empirical fit.

How long does a fixed-price simulation study take?

Most studies complete in 4 to 8 weeks depending on scope. The standard timeline is: scoping call and problem definition in week 1; simulation execution across the relevant physics scales in weeks 2 to 4; results analysis, validation, and report delivery in weeks 5 to 6. More complex multiscale studies with broader operating envelopes may extend to 8 weeks. Scope and timeline are agreed and fixed before work begins.

What deliverables does a simulation study produce?

A newtsim simulation study delivers: a validated multiscale model calibrated to your system; an operating envelope with physics-based safety margins; a written technical report with findings and actionable recommendations; and optionally CFD visualisations or an ML surrogate model (newtsim Neural) for fast design-space exploration by your engineering team.

Does newtsim work for green hydrogen electrolyser and catalyst design?

Yes. The Stream solver handles proton exchange membrane and alkaline electrolyser CFD: multiphase flow, bubble dynamics, and membrane transport. Bond handles catalyst surface reaction molecular dynamics, giving adsorption energies and reaction barriers from first principles rather than from empirical correlations. Neural then produces fast-to-run surrogate models for design-space exploration and stack optimisation, reducing the iteration cost for new catalyst compositions and electrode architectures.

newtsim is a physics-grounded multiscale simulation consulting firm and software platform. We deliver fixed-price simulation studies and platform licences for industrial clients, covering the full physics chain from quantum mechanics and molecular dynamics through to plant-scale CFD and structural mechanics. All solvers run GPU-native.

Does newtsim use artificial intelligence or machine learning?

Yes. newtsim Neural produces ML-trained surrogate models derived from high-fidelity physics simulations. These surrogates run in milliseconds rather than hours, enabling rapid design-space exploration and real-time control without sacrificing physical accuracy. The ML models are trained on physics simulation data, not empirical measurements, so they inherit the accuracy and extrapolation properties of the underlying physics.

How do I get started with newtsim?

Book a free technical scoping call at newtsim.com/demo. Bring a description of your engineering challenge, the physics you think are important, and any data you have. The call is 45 minutes and produces a problem definition, proposed physics chain, and outline scope that forms the basis of a fixed-price proposal.

Can newtsim simulate non-Newtonian fluids?

Yes. newtsim specialises in non-Newtonian flow: yield-stress fluids, viscoelastic materials, shear-thinning and shear-thickening systems, and complex multiphase flows. The Stream solver handles macroscale CFD for these systems; Bond provides the molecular-scale rheology predictions that replace empirical viscosity fits; and the full chain gives accurate behaviour outside the calibrated range of laboratory data.

Does newtsim work for small companies or startups?

Yes. Fixed-price studies are available to clients of any size. The fixed-price, fixed-scope model means you know the cost before committing, making it accessible for smaller organisations without large simulation teams. Many newtsim clients are startups and scale-ups who need physics-grounded analysis early in product development before investing in laboratory or pilot-scale work.

What is the difference between molecular dynamics and CFD?

Molecular dynamics (MD) simulates the motion of individual atoms and molecules to compute material properties such as viscosity, diffusivity, reaction rates, and phase behaviour, from first principles. Computational fluid dynamics (CFD) models bulk fluid flow at the engineering scale using those material properties as inputs. newtsim connects the two: MD provides accurate constitutive inputs that CFD uses for engineering-scale predictions, rather than relying on empirical correlations.

What is a surrogate model in simulation?

A surrogate model (also called a reduced-order model or emulator) is a fast mathematical approximation of a complex simulation. newtsim Neural builds ML surrogates trained on high-fidelity physics simulation data. Once trained, a surrogate evaluates in milliseconds instead of hours, enabling real-time optimisation and design-space exploration. Because surrogates are trained on physics data, they are more accurate than empirical correlations, especially outside the training range.

What deliverables does a newtsim simulation study produce?

A newtsim simulation study delivers: (1) a validated multiscale model calibrated to your system; (2) an operating envelope with physics-based safety margins; (3) a written technical report with findings and actionable recommendations; and optionally (4) CFD visualisations and an ML surrogate model for fast ongoing design-space exploration by your engineering team.

Aerospace & Advanced Manufacturing

Aerospace & Advanced Manufacturing · Case Study

Cure Cycle Residual Stress and Spring-Back Prediction for CFRP Wing Spar Cap

Executive Summary

Spring-back in autoclave-cured CFRP parts is among the most extensively documented and persistently misunderstood process-induced distortion mechanisms in aerospace manufacturing. Published experimental studies have shown that total spring-back in 90-degree L-brackets of aerospace CFRP on aluminium tooling ranges from 1.2 to 2.8 degrees depending on layup and thickness, with 65-75% of the distortion driven by coefficient of thermal expansion mismatch between the aluminium tool and the 0-degree CFRP during cooldown. Subsequent work on complex section geometries showed that shear stress relaxation at the web-flange junction during the heating ramp locks in a frozen stress state that compounds the CTE-mismatch contribution. The NCAMP qualification database at Wichita State University documents the specific case directly relevant here: a C-section with 32-ply unidirectional-dominated IM7/8552 on aluminium tooling produces 2.15 degrees of spring-back, within one standard deviation of the production non-conformance under study. Despite this publicly available knowledge, manufacturing teams throughout the industry continue to invest in empirical cure cycle trials — varying temperature, dwell time, and support structure — that adjust only the secondary contributors (chemical shrinkage at 21%, thermal gradient at 7%) while leaving the dominant CTE mismatch term entirely untouched. The result is 14 months and USD 2.4 million of rework on a non-conformance that could have been diagnosed in weeks.

The production non-conformance documented here is characteristic of this failure mode: C-section spar caps departing the autoclave at 1.8–2.3 degrees beyond nominal, requiring shimming (adding 110–180 g per cap with structural compliance risk) or outright rejection. Three rounds of empirical experimentation — reducing cure temperature, extending dwell time, switching to a hybrid tool — produced at best 0.3 degrees of improvement, because none of the trials addressed the CTE mismatch between the 22.0 microstrain per degree C aluminium mandrel and the 0.2 microstrain per degree C unidirectional CFRP.

Had a coupled thermo-chemical-viscoelastic-mechanical cure process simulation been applied at production qualification, it would have quantified the contribution of each spring-back driver before the first production lot was autoclave-cured, identified the 68% dominance of the CTE mismatch term, and prescribed the -2.2 degree tool angle pre-compensation that eliminates the non-conformance at a USD 15,000–30,000 tooling modification — rather than after 14 months of accumulating rework. The simulation also characterises cure-induced residual stress at the web-flange junction: 42 MPa transverse tension in the 0-degree plies, against a cracking threshold of 62 MPa, with a 48% margin that confirms no microcracking risk in the corrected process. Those residual stress peaks at the web-flange junction and run-out locations define where in-service degradation will preferentially initiate under repeated flight loading — the optimal placement zones for newtsim livesim strain gauge monitoring that tracks the spar cap load path throughout the service life and alerts operators if assembly pre-stress from future shimming events has raised interlaminar stress above the simulation-derived safe threshold.

Scenario Background

(illustrative reference case)

The scenario involves a fictional regional aircraft OEM, Cerulean Aircraft Corporation, based in Wichita, Kansas, producing the CA-50 50-seat turboprop regional aircraft under FAR Part 25 / CS-25 certification. The wing spar cap is the primary wing bending structural element; the C-section CFRP spar cap carries the spanwise bending moment and is load-path critical under CS-25.303 (factor of safety), CS-25.305 (strength and deformation), and CS-25.571 (fatigue and damage tolerance). Any dimensional non-conformance that introduces assembly stresses at the spar cap-to-skin bond line is a compliance risk requiring engineering disposition.

The material system is Hexcel HexPly IM7/8552 -- IM7 intermediate-modulus carbon fibre (12k tow, 194 g/m squared) with 8552 toughened amine-cured epoxy. The nominal cure cycle follows the Hexcel standard: ramp at 2 degrees C/min to 180 degrees C, hold at 7 bar for 120 minutes, then ramp down at 2 degrees C/min. Resin content is 35% +/- 2% by weight, fibre volume fraction is 57-63%, and the fully cured resin has a glass transition temperature of 220 degrees C (dry) or 200 degrees C (wet at 1.0% moisture). Nominal cured ply thickness is 0.131 mm with a void content specification below 1.5%.

The spar cap is a C-section (web plus two flanges at 90 degrees to web) with a [0_8/+/-45/0_8/90/0_8]s layup -- a 34-ply unidirectional-dominated laminate giving a nominal cured thickness of 4.5 mm. The web height is 120 mm, flange width is 95 mm, and span extends 8.4 m from root fitting to tip. The nominal flange angle is 90.0 degrees with an assembly tolerance of +/-0.3 degrees. The aluminium mandrel tooling is 7075-T6 aluminium with a CTE of 22.0 microstrain per degree C (over 20-180 degrees C range), Ra 0.8 micrometre surface roughness (hard anodised), and a non-porous FEP release film at the tool-part interface.

Over 6 production lots totalling 48 parts, the measured spring-back averaged 1.97 degrees with a standard deviation of +/-0.25 degrees. The lot-by-lot breakdown showed consistent behaviour: Lot 1 produced 2.0 degrees +/- 0.18 degrees, Lot 2 produced 1.9 degrees +/- 0.22 degrees, Lot 3 produced 2.1 degrees +/- 0.21 degrees, Lot 4 produced 1.8 degrees +/- 0.19 degrees, Lot 5 produced 1.9 degrees +/- 0.24 degrees, and Lot 6 produced 2.3 degrees +/- 0.27 degrees. Of the 48 parts, 4 were scrapped, 33 required shimming, and only 11 were accepted as-manufactured. The resulting 3.2 mm interference gap at the spar cap-to-skin bond line (at the wing root where flanges are 95 mm wide) from the 1.9 degree average spring-back produced an estimated 12-18 MPa compressive pre-stress across the bondline after forced closure -- not analysed in the original stress report.

Challenge

The fundamental physical question requiring simulation is: why does a nominally symmetric C-section spar cap spring back by approximately 2 degrees when demolded from a 90-degree aluminium tool, and why does this particular geometry-process combination produce this specific magnitude of distortion?

The manufacturing team had conducted three empirical experiments prior to the simulation study. Reducing cure temperature from 180 degrees C to 175 degrees C produced no improvement, indicating temperature was not the driver. Extending dwell time from 120 to 150 minutes reduced spring-back by only 0.15 degrees -- marginal and insufficient. Switching to a CFRP-on-aluminium hybrid tool reduced spring-back by 0.3 degrees, an improvement but still leaving 1.6 degrees of distortion, which remains well outside the +/-0.3 degree assembly tolerance. None of these experiments were informed by an understanding of the physical mechanism.

The simulation task was to identify the physical origin of the spring-back (CTE mismatch, chemical shrinkage, or viscoelastic relaxation), quantify each contribution's magnitude, generate a response surface for spring-back vs. controllable process variables, and derive the minimum corrective action to bring parts within +/-0.3 degree tolerance.

The dimensional consequences of the 1.9 degree spring-back are substantial. The flange angle opens from the nominal 90.0 degrees to 91.9 degrees, creating a 3.2 mm gap at the spar-skin bond line across the 95 mm flange width. Closing this gap requires 12-18 kN of fastener clamping force, which introduces the assembly pre-stress noted above. If shimmed instead, each cap adds 140 g (280 g per wing for two caps).

Real-World Basis

Academic and Industrial Documentation of CFRP Spring-Back

Spring-back in autoclave-cured CFRP parts is one of the most extensively characterised process-induced distortion mechanisms in the published literature. Three independent sources establish the physical basis for this study.

Experimental studies at the University of British Columbia demonstrated that total spring-back in 90-degree L-brackets of aerospace CFRP on aluminium tooling (12-24 ply) ranges from 1.2 to 2.8 degrees depending on layup and laminate thickness. The dominant contribution is CTE mismatch between the aluminium tool and 0-degree CFRP during cooldown, accounting for 65-75% of total spring-back. Chemical shrinkage contributes 15-25%, and through-thickness temperature gradient is a minor residual term. Unidirectional-dominated laminates show significantly larger spring-back than quasi-isotropic ones because the effective CTE mismatch is larger (0-degree laminate CTE of approximately 0-1 microstrain per degree C vs. quasi-isotropic laminate CTE of approximately 3-4 microstrain per degree C).

Subsequent work on complex section geometries showed that the geometry-dependent component arises from shear stress relaxation during the heating ramp. As the resin transitions from rubbery to glassy state during heating through T_g, in-plane shear stresses developed during the high-temperature rubbery phase are not fully relaxed before the resin vitrifies, creating a frozen-in stress state that contributes to geometry-specific distortion at complex section transitions such as web-flange junctions.

The NCAMP qualification database at Wichita State University provides directly comparable experimental data for IM7/8552 spring-back. For a 90-degree L-bracket with 12-ply quasi-isotropic layup on aluminium tooling, measured spring-back was 1.14 degrees +/- 0.12 degrees; for the 24-ply version, 1.62 degrees +/- 0.18 degrees; for a C-section with 32-ply unidirectional-dominated layup on aluminium, 2.15 degrees +/- 0.24 degrees; and for a 12-ply L-bracket on CFRP tooling, only 0.38 degrees +/- 0.09 degrees. The NCAMP C-section data at 32-ply on aluminium tooling (2.15 degrees +/- 0.24 degrees) is directly comparable to the production non-conformance (34-ply, 1.97 degrees +/- 0.25 degrees), providing strong prior evidence that the observed distortion magnitude is physically consistent.

Hexcel's published process guide for HexPly 8552 documents the cure kinetics, degree of cure vs. time-temperature relationships, and glass transition temperature as a function of degree of cure. It recommends that for complex-curvature or angled-section parts on aluminium tooling, a tool angle compensation of 1.5-3.0 degrees is typically required for unidirectional-dominated laminates, with determination of the compensation angle requiring either physical trial or process simulation. This recommended compensation range brackets both the observed spring-back (1.8-2.3 degrees) and the model-predicted correction (2.2 degrees).

Simulation Approach

The cure process model is implemented as a coupled thermo-chemical-viscoelastic-mechanical FEM in newtsim Span with a custom UMAT/USDFLD subroutine chain governing all three physics domains simultaneously.

Step 1 -- Cure Kinetics (Thermo-Chemical Analysis)

The degree of cure alpha(t) is computed using an autocatalytic kinetics model fitted to IM7/8552 DSC data from the NCAMP database. The kinetic parameters capture the temperature-dependent reaction rate, and the heat generation from the exothermic cure reaction couples to a transient heat conduction model that resolves through-thickness temperature gradients in the 4.5 mm laminate. The predicted final degree of cure at 180 degrees C for 120 minutes is 0.997, and the peak cure exotherm is +4.8 degrees C above autoclave set point at the laminate midplane. Resolving this exotherm matters because it determines when and where the resin vitrifies -- the event that locks in residual stress.

Step 2 -- Resin Modulus Evolution (Viscoelastic)

As the cure progresses, the glass transition temperature T_g rises from approximately -9 degrees C (uncured) to 220 degrees C (fully cured), and the resin modulus evolves from approximately 2.5 MPa in the rubbery state to approximately 4,100 MPa in the glassy state. The model tracks this stiffness evolution through a viscoelastic representation that captures the physically critical distinction: stresses frozen in at gelation (when the resin first becomes load-bearing) vs. stresses developed during cooldown in the fully glassy state. Getting this transition right is what determines the magnitude and geometry-dependence of spring-back.

Step 3 -- Process-Induced Stress and Spring-Back Prediction

The full cure cycle thermal history and degree-of-cure field from Steps 1-2 are applied as sequential loads to a 3D solid model of the C-section spar cap resting on the aluminium mandrel tool. The model spans the full 8.4 m spar cap length with ply-resolved mesh at the web-flange junction. The tool-part interface uses frictionless contact (FEP release film) with a no-penetration normal constraint. The CTE mismatch between the aluminium tool (22.0 microstrain per degree C) and the 0-degree CFRP (0.2 microstrain per degree C) drives a differential contraction of 21.8 microstrain per degree C during cooldown -- this is the dominant spring-back driver, and the model must resolve it accurately. Spring-back is extracted as the flange tip angle relative to the web plane at the tool release step.

The parametric process response surface covers four cure temperatures (175 to 182.5 degrees C), four dwell durations (90-180 minutes), three ramp-down rates (1.0-3.0 degrees C/min), and four tool materials spanning CFRP (CTE 2 microstrain per degree C), steel (12), aluminium/CFRP hybrid (12), and aluminium (22) -- a total of 192 simulation runs. This parameter space was chosen because it covers every variable that the manufacturing team can realistically control, including tool material CTE -- the variable that 14 months of physical experimentation never varied.

Simulation Caveats

The frictionless contact assumption at the FEP release film interface is the standard simplification in process-induced distortion modelling. Published studies show that interface friction coefficient in the range 0.05-0.20 (FEP release film, typical of aerospace production) produces 4% or less variation in spring-back angle for C-section geometries. This sensitivity is within the model's overall uncertainty and does not change the corrective action recommendations.

The 6-term Prony series represents the viscoelastic relaxation modulus over a frequency range covering the cure cycle duration (hours). Ultra-long-term relaxation (days to years) is not captured, which is appropriate for the cure cycle context but means the model does not predict long-term dimensional change in service due to continued resin stress relaxation.

The 3D FEM assumes a spatially uniform autoclave temperature equal to the set-point +/-2 degrees C (controller tolerance). Actual autoclave air circulation produces +/-3-5 degrees C spatial gradients across an 8.4 m spar cap. These gradients introduce additional spring-back variability of +/-0.15 degrees that is not fully resolved in the deterministic model but is captured within the process variability range in the Monte Carlo study.

The FEM applies tool release as an instantaneous simultaneous detachment of the tool from the part at the end of cooldown. In production, the mandrel is released incrementally as tooling pins are removed. This sequential release can alter the local stress redistribution at the web-flange junction by +/-0.1 degrees, which is small relative to the total spring-back but should be noted for future refinement.

The model predicts the as-demolded spring-back only. Post-cure trimming, drilling, and assembly operations that introduce stress concentrations may cause additional local distortion. This effect is not modelled; the +/-0.3 degree assembly tolerance must accommodate it.

Key Predictions / Results

Predicted Spring-Back vs. Measured Production Data:

Source	Mean Spring-Back	Standard Deviation	Notes
Production Lots 1-6	1.97 degrees	+/-0.25 degrees	Measured CMM data, 48 parts
Simulation prediction (standard cure, Al tool)	1.87 degrees	+/-0.22 degrees (Monte Carlo)	500 process variability realisations
Error (mean)	-0.10 degrees (5.1%)	--	Well within +/-10% target
Error (std dev)	-0.03 degrees (12%)	--	Acceptable

Spring-Back Contribution Breakdown (Standard Cure, Aluminium Tool):

Mechanism	Contribution to Total Spring-Back	Sobol Index (Sensitivity)
CTE mismatch (Al tool vs. 0 degree CFRP) during cooldown	1.27 degrees (68%)	0.68
In-plane chemical cure shrinkage	0.39 degrees (21%)	0.21
Through-thickness temperature gradient during ramp	0.13 degrees (7%)	0.07
Fibre-resin micro-mechanical interaction	0.07 degrees (4%)	0.04
Total predicted	1.87 degrees	--

Springback driver decomposition — CTE mismatch 68%, chemical shrinkage 21%, thermal gradient 7%, fibre-resin 4%; standard cure 180°C / Al tool

The dominant driver confirmation is decisive: tool-part CTE mismatch during cooldown accounts for 68% of total spring-back. This directly explains why the manufacturing team's previous experiments with cure temperature and dwell time had minimal effect -- those variables affect the chemical shrinkage term (21%) and through-thickness temperature gradient term (7%), but not the dominant CTE mismatch term, which is controlled entirely by tool material CTE.

Process Parameter Response Surface -- Spring-Back Magnitude:

Cure Temperature	Dwell	Ramp-Down Rate	Tool Material	Spring-Back Predicted
175 degrees C	90 min	2 degrees C/min	Aluminium	1.79 degrees
180 degrees C	120 min	2 degrees C/min	Aluminium	1.87 degrees (baseline)
180 degrees C	150 min	2 degrees C/min	Aluminium	1.72 degrees
180 degrees C	120 min	0.5 degrees C/min	Aluminium	1.51 degrees
180 degrees C	120 min	2 degrees C/min	Steel	1.12 degrees
180 degrees C	120 min	2 degrees C/min	CFRP/Al hybrid	0.72 degrees
180 degrees C	120 min	2 degrees C/min	CFRP	0.51 degrees

Springback angle vs tool–part CTE mismatch — IM7/8552 C-section spar cap, Monte Carlo ±1σ band, production mean 1.97° and ±0.3° assembly tolerance overlay

Process Parameter Sensitivity (Sobol First-Order Indices):

Parameter	Sobol Index	Rank
Tool material CTE	0.52	1 -- dominant
Ramp-down rate	0.21	2
Dwell duration	0.12	3
Cure temperature (above 175 degrees C)	0.08	4 -- minimal once full cure achieved
Interaction terms	0.07	--

Corrective Action Recommendations:

Option	Tool Modification	Cure Cycle Change	Predicted Spring-Back	Tolerance Met?	One-Time Cost
Option A (recommended): Tool compensation only	Pre-compensate flange angle to 87.8 degrees (-2.2 degree under-angle)	None	89.95 degrees +/-0.21 degrees as-demolded	Yes (within +/-0.3 degrees)	USD 15,000-30,000
Option B (no tooling change): Cure cycle modification	None	Ramp-down rate 2 to 0.5 degrees C/min + partial compensation -1.6 degrees	90.05 degrees +/-0.23 degrees	Yes	USD 8,000 process validation
Option C (rejected): CFRP tool replacement	New CFRP mandrel	None	0.51 degrees -- over-corrected	Yes -- large margin	USD 280,000-420,000
Option D (rejected): Steel tool	New steel mandrel	None	1.12 degrees -- still out of tolerance	No	USD 85,000

Residual Stress in Recommended Process (Option A -- Tool Compensation):

Location	Ply Orientation	Residual Stress Component	Magnitude	Threshold	Margin
Web-flange junction	0 degree plies	Hoop tensile sigma_22	42 MPa	F_2^T = 62 MPa (in-situ)	+48%
Web centre	90 degree plies	Transverse sigma_22	18 MPa	F_2^T = 62 MPa	+244%
Flange tip	+/-45 degree plies	Shear tau_12	9 MPa	F_6 = 92 MPa	+922%
Midplane through-thickness	All	sigma_33 (peel)	3.1 MPa	--	Low; no microcracking risk

The cure cycle does not introduce microcracking in the as-cured part under Option A. Maximum residual stress (42 MPa in 0-degree plies at web-flange junction) remains well below the IM7/8552 matrix cracking threshold of approximately 80 MPa in transverse tension.

Comparison Methodology

1. NCAMP Spring-Back Database -- Quantitative Validation

The simulation is first validated against higher-fidelity coupon-level predictions, then checked against the NCAMP standard 90-degree L-bracket coupon geometry (24-ply [0_4/+/-45/0_4/90]_2 IM7/8552, aluminium tool) at the standard cure cycle for secondary experimental confirmation:

Metric	NCAMP Measured	Predicted	Error
Spring-back (24-ply L-bracket, Al tool, 180 degrees C/120 min)	1.58 degrees +/-0.18 degrees	1.62 degrees	+2.5%
Spring-back (12-ply L-bracket, Al tool)	1.14 degrees +/-0.12 degrees	1.19 degrees	+4.4%

Both predictions fall within the 5% target accuracy criterion.

2. Independent L-Bracket Data -- Cross-Validation at Different Material System

The model is run with Cytec 5276-1 cure kinetics parameters on the Cytec AS4/5276-1 L-bracket geometry (various ply counts 4-16 ply, [0]_n layup, aluminium tool) for secondary confirmation across a different material system:

Ply Count	Published Measured (degrees)	Predicted (degrees)	Error
4 ply	0.82 degrees	0.88 degrees	+7.3%
8 ply	1.18 degrees	1.23 degrees	+4.2%
12 ply	1.58 degrees	1.65 degrees	+4.4%
16 ply	1.93 degrees	2.02 degrees	+4.7%

All predictions fall within 8% of the measured data across the full ply thickness range, within the acceptance criterion.

3. Production Data Validation (6-Lot Non-Conformance)

The production distribution (mean 1.97 degrees, sigma = 0.25 degrees, 48 parts) is compared against the Monte Carlo process variability simulation (500 realisations varying autoclave temperature +/-2 degrees C, ramp rate +/-0.3 degrees C/min, resin content +/-2%, and initial tool temperature +/-5 degrees C):

Statistic	Production Measured	Monte Carlo Predicted	Error
Mean spring-back	1.97 degrees	1.87 degrees	-5.1%
Standard deviation	0.25 degrees	0.22 degrees	-12%
P90 spring-back	2.28 degrees	2.15 degrees	-5.7%
P10 spring-back	1.67 degrees	1.59 degrees	-4.8%

The Monte Carlo slightly under-predicts the production variability (sigma = 0.22 degrees vs. 0.25 degrees). The additional variability in production is likely attributable to spatial autoclave temperature gradients and layup-to-layup resin content scatter exceeding the +/-2% resin content tolerance -- factors that are within-specification but not fully captured by the model's variability assumptions.

Deliverables

Root Cause Analysis Report: Quantified contribution breakdown (CTE mismatch 68% vs. chemical shrinkage 21% vs. thermal gradient 7% vs. fibre-resin interaction 4%) with supporting stress contour maps at each stage of the cure cycle -- heating ramp, dwell, cooling, tool release. Includes explanation of why cure temperature and dwell time experiments failed to resolve the issue.
Process Parameter Response Surface: Spring-back angle vs. four-variable process parameter space (cure temperature x dwell duration x ramp-down rate x tool CTE), presented as 2D sensitivity heatmaps for pairs of variables at nominal levels of the other two. Format: PDF figures + Excel data tables for process engineering use.
Tool Angle Compensation Recommendation: Specific compensation angle (-2.2 degrees from 90 degrees, i.e. tool flange angle = 87.8 degrees) with +/-uncertainty bounds covering 95% of production variability (prediction interval: +/-0.28 degrees), justifying that the compensated tool will deliver parts within the +/-0.3 degree assembly tolerance with 95% probability. Drawing-ready specification in CAD-compatible format.
Alternative Cure Cycle Recommendation (Option B): Revised cure cycle (ramp-down rate 0.5 degrees C/min, partial tool compensation -1.6 degrees) for use if immediate tooling modification is not feasible. Includes validation that the 0.5 degrees C/min ramp-down remains within Hexcel's qualified cure envelope for HexPly 8552 (minimum qualified ramp rate: 0.3 degrees C/min) and does not compromise final degree of cure (predicted final alpha = 0.996 at this ramp rate -- equivalent to standard cycle).
Residual Stress Field Maps: Full 3D stress contour plots (all ply orientations, web-flange junction detail, full spar cap span) confirming no microcracking risk in the recommended Option A process. Specifically addresses the spar cap-to-skin bond line stress state under combined assembly stress + flight load as input to the original structural stress report.
Manufacturing Process Control Card: Recommended process control window for statistical process control (SPC): autoclave temperature 180 degrees C +/-2 degrees C (thermocouple at part surface, not autoclave platen); ramp-down rate 2.0 degrees C/min +/-0.2 degrees C/min; dwell duration 120 min + 0/-0 min (minimum dwell); resin content verification per ASTM D3529 on process control coupons each autoclave run.
newtsim Span Cure Simulation Model Archive: Full model with documented cure kinetics constants, material property evolution functions (T_g vs. cure, resin modulus evolution, viscoelastic relaxation), boundary conditions, and parametric study input deck. Delivered in newtsim Span format with Python scripting for parameter sweeps.

This case study is an illustrative reference scenario demonstrating newtsim's simulation methodology. All company names, personnel, and specific operational data are fictional. The incident descriptions draw on publicly documented real-world events cited in the frontmatter.