{ "brief_id": "SPR-2026-B151", "generated_at": "2026-04-12", "domains": [ "Aerospace Engineering", "Materials Processing" ], "original_hypothesis": "The high-fidelity, multi-physics simulation and control frameworks developed for liquid rocket engine operation can be adapted to model and control the dynamic, non-equilibrium solidification and microstructure evolution in advanced twin-roll casting processes.", "grounding": { "novelty_assessment": { "score": 0.8, "closest_existing_work": [ { "paper_id": "31bbd39c31dd5cfe1c79614a3bb2abcad640ece2", "title": "Correlation of digital twin and roll surface sensor results for AZ31 alloy TRC process", "doi": "10.1007/s10853-024-09733-y", "year": 2024, "similarity": "related", "key_difference": "This paper applies a digital twin to TRC for sensor correlation and process understanding, but does not propose or demonstrate the adaptation of high-fidelity, multi-physics control frameworks from aerospace (like LREs) for active, ML-driven microstructure control. It focuses on static correlation rather than dynamic, predictive control.", "authors": [ "S. Kwiecień", "M. Weiner", "O. Lypchanskyi", "L. Hamm", "M. Schmidtchen", "M. Ullmann", "Welf-Guntram Drossel", "U. Prahl" ] }, { "paper_id": "1eb0b7c60862291b37fb2d7bd6c31ab674dc9f92", "title": "Digital-Twin-Based Coordinated Optimal Control for Steel Continuous Casting Process", "doi": "10.3390/met13040816", "year": 2023, "similarity": "analogous", "key_difference": "This paper presents a digital-twin-based optimal control system for continuous casting, which is a related materials processing domain. However, the hypothesis specifically calls for the transfer of methodologies developed for the extreme, transient multi-physics environments of liquid rocket engines (a distinct aerospace domain) to the rapid solidification context of TRC. The control paradigms and physical fidelity required for LREs are more specialized.", "authors": [ "Jian Yang", "Zhenping Ji", "Wenhong Liu", "Zhiyong Xie" ] } ], "verdict": "novel" }, "evidence_base": [ { "paper_id": "31bbd39c31dd5cfe1c79614a3bb2abcad640ece2", "title": "Correlation of digital twin and roll surface sensor results for AZ31 alloy TRC process", "doi": "10.1007/s10853-024-09733-y", "year": 2024, "citation_count": 1, "support_type": "direct", "relevance": "Directly applies a digital twin framework to a Twin-Roll Casting (TRC) process, addressing the challenge of unknown parameters in the roll gap. It demonstrates the foundational step of integrating sensors and models for TRC.", "key_finding": "Demonstrates the correlation between digital twin predictions and roll surface sensor measurements for the AZ31 alloy TRC process, validating the use of a digital twin to infer internal process states.", "authors": [ "S. Kwiecień", "M. Weiner", "O. Lypchanskyi", "L. Hamm", "M. Schmidtchen", "M. Ullmann", "Welf-Guntram Drossel", "U. Prahl" ] }, { "paper_id": "1eb0b7c60862291b37fb2d7bd6c31ab674dc9f92", "title": "Digital-Twin-Based Coordinated Optimal Control for Steel Continuous Casting Process", "doi": "10.3390/met13040816", "year": 2023, "citation_count": 11, "support_type": "analogous", "relevance": "Provides a strong analogous example in materials processing (continuous casting) where a digital twin is used for coordinated, model-based optimal control of a solidification process, involving real-time heat transfer models and online calibration.", "key_finding": "Presents a digital-twin-model-based optimal control system for the steel continuous casting process, designed for coordinated optimization and dynamic control of secondary cooling and final electromagnetic stirring.", "authors": [ "Jian Yang", "Zhenping Ji", "Wenhong Liu", "Zhiyong Xie" ] }, { "paper_id": "4047d5244ae957c92b6189f0a1b3cce48baae158", "title": "Towards a Digital Twin Framework in Additive Manufacturing: Machine Learning and Bayesian Optimization for Time Series Process Optimization", "doi": "10.48550/arXiv.2402.17718", "year": 2024, "citation_count": 84, "support_type": "indirect", "relevance": "Demonstrates the integration of machine learning (LSTM), Bayesian optimization, and physics-based modeling within a digital twin framework for managing heat accumulation—a key multi-physics challenge—in an additive manufacturing process. This supports the methodological transfer of ML-based control to complex thermal processes.", "key_finding": "Develops a digital twin framework incorporating a surrogate model using LSTM-based machine learning with Bayesian Inference to predict temperatures in laser-directed-energy deposition, promoting integration of real-time monitoring and physics-based modeling for control.", "authors": [ "V. Karkaria", "Anthony Goeckner", "Rujing Zha", "Jie Chen", "Jianjing Zhang", "Qi Zhu", "Jian Cao", "R. X. Gao", "Wei Chen" ] }, { "paper_id": "c760f2c7cc6c0b7f33f4a84edd628acce07f2939", "title": "Probabilistic Digital Twin for Additive Manufacturing Process Design and Control", "doi": "10.1115/1.4054521", "year": 2022, "citation_count": 25, "support_type": "indirect", "relevance": "Proposes a detailed methodology for constructing a digital twin for a complex manufacturing process (LPBF) that incorporates model uncertainty and process variability, using surrogate models for faster computation. This aligns with the need for robust, predictive digital twins for process control as hypothesized.", "key_finding": "Proposes a methodology for an AM digital twin that incorporates model uncertainty and process variability, using a physics-based model replaced by a cheaper surrogate model for faster computation in uncertainty analysis and decision-making.", "authors": [ "Paromita Nath", "S. Mahadevan" ] }, { "paper_id": "72294ea552324b0c570e5ee5229886155e737eb3", "title": "A Digital Twin Framework With Bayesian Optimization and Deep Learning for Semiconductor Process Control", "doi": "10.1109/ACCESS.2025.3551332", "year": 2025, "citation_count": 6, "support_type": "analogous", "relevance": "Demonstrates a sophisticated digital twin framework integrating deep learning and Bayesian optimization for parameter control in a high-tech manufacturing process (semiconductors). This supports the core idea of using advanced ML for real-time optimization and control in complex processes.", "key_finding": "Introduces an intelligent optimization framework integrating Digital Twin technology, deep learning, and tailored Bayesian Optimization to enhance parameter control and yield in semiconductor manufacturing.", "authors": [ "Chin-Yi Lin", "T. Tseng", "Tsung-Han Tsai" ] } ], "counter_evidence": [ { "paper_id": "f4c240ea61f0eb37f9bd43ab172987f4d43d155d", "title": "A Digital Twin Approach for Soil Moisture Measurement with Physically Based Rendering Simulations and Machine Learning", "doi": "10.3390/electronics14020395", "finding": "The paper applies a digital twin to a slow, agricultural process (soil moisture monitoring) with very different physics and timescales compared to the rapid, extreme conditions of both LREs and TRC. It does not contradict the hypothesis but highlights that many digital twin applications are for simpler, slower processes, potentially questioning the necessity of LRE-level sophistication for TRC.", "severity": "minor", "authors": [ "Ismail Parewai", "Mario Köppen" ] } ], "gap_manifest_update": { "closed_gaps": [ "No gaps from the provided manifest are directly closed by the listed papers. The 'data_gaps' regarding dynamic, sensor-rich TRC data synchronized with microstructure remains open, as paper 31bbd39c... provides correlation but not the comprehensive dynamic dataset described." ], "new_gaps": [ "Methodological Transfer Gap: None of the provided papers explicitly demonstrate the transfer of control frameworks or ML methods specifically developed for liquid rocket engines (e.g., from DLR) to any materials processing domain. The hypothesis of a direct methodological transfer from aerospace propulsion remains unsupported by the evidence.", "Microstructure Control Gap: While several papers (1eb0b7c..., 4047d524..., c760f2c7...) discuss digital twins for process control, none explicitly demonstrate the active, in-situ control of final microstructure features (e.g., texture, grain size distribution) as the primary control objective. The link from process parameters to microstructure as a controlled variable is not evidenced." ], "data_available": [ "Paper 31bbd39c... (Kwiecień et al., 2024) provides a dataset involving digital twin predictions and roll surface sensor measurements for the AZ31 TRC process, which is relevant but static/correlative rather than the dynamic, multi-run dataset needed for ML model training." ] }, "search_queries": [ { "query": "digital twin microstructure control casting", "type": "novelty", "rationale": "Direct search for the core hypothesis of using a digital twin for active microstructure control in casting processes." }, { "query": "machine learning twin roll solidification control", "type": "novelty", "rationale": "Targets the specific application of ML for control in the twin-roll casting context mentioned in the hypothesis." }, { "query": "active in-situ microstructure engineering", "type": "novelty", "rationale": "Searches for the proposed paradigm shift from passive characterization to active control of microstructure during processing." }, { "query": "rocket engine control methods materials processing", "type": "novelty", "rationale": "Directly queries the cross-application of control frameworks from aerospace to materials science." }, { "query": "liquid rocket engine digital twin machine learning", "type": "evidence", "rationale": "Seeks evidence for the advanced ML and digital twin frameworks developed in LREs, as stated in the mechanisms." }, { "query": "combustion instability prediction control ML", "type": "evidence", "rationale": "Targets specific techniques (combustion instability control) from aerospace that are proposed for transfer." }, { "query": "twin roll casting microstructure modeling simulation", "type": "evidence", "rationale": "Finds foundational research on modeling the complex multi-physics (thermal, fluid, stress) in TRC that needs control." }, { "query": "rapid solidification process sensor data assimilation", "type": "evidence", "rationale": "Looks for evidence of sensor fusion and state estimation in materials processing, a prerequisite for the control loop." }, { "query": "digital twin manufacturing process control", "type": "cross_domain", "rationale": "Broad search for precedents of digital twin applications in manufacturing control outside aerospace." }, { "query": "model predictive control additive manufacturing", "type": "cross_domain", "rationale": "Examines another advanced manufacturing domain (additive) where advanced control paradigms from other fields might have been applied." }, { "query": "multi-physics simulation control transfer learning", "type": "cross_domain", "rationale": "Searches for methodological transfers of control/simulation frameworks between different engineering domains." } ] }, "sharpened": { "title": "Adaptation of Aerospace-Grade Ensemble Kalman Filter Digital Twins for Microstructure State Estimation in Twin-Roll Casting", "formal_statement": "If a high-fidelity, multi-physics digital twin of a twin-roll casting (TRC) process is integrated with an Ensemble Kalman Filter (EnKF) for data assimilation from in-line sensors, then the EnKF will estimate internal solidification state variables (e.g., mushy zone thickness, cooling rate) with a root-mean-square error (RMSE) ≤15% relative to direct experimental measurements, because the EnKF's capacity to handle non-linear, transient dynamics and assimilate noisy sensor data—proven in liquid rocket engine control—can be transferred to model the coupled thermal-fluid-mechanical fields governing rapid solidification.", "independent_variables": [ { "name": "Casting Speed", "type": "continuous", "range": "0.5-3.0", "unit": "m/min" }, { "name": "Roll Gap", "type": "continuous", "range": "1.0-5.0", "unit": "mm" }, { "name": "Melt Superheat", "type": "continuous", "range": "10-50", "unit": "K" }, { "name": "Roll Cooling Intensity", "type": "continuous", "range": "50-200", "unit": "kW/m²" } ], "dependent_variables": [ { "name": "Root-Mean-Square Error (RMSE) of Mushy Zone Thickness", "type": "continuous", "expected_direction": "decrease", "unit": "% (relative to measured value)" }, { "name": "Average Cooling Rate at Solidification Front", "type": "continuous", "expected_direction": "non-monotonic", "unit": "K/s" }, { "name": "Final Average Grain Size", "type": "continuous", "expected_direction": "non-monotonic", "unit": "µm" } ], "proposed_mechanism": { "causal_chain": [ "Step 1: An EnKF assimilates real-time sensor data (e.g., roll surface temperature, roll separating force) into a high-fidelity, transient TRC process model (coupled CFD-thermodynamics), reducing uncertainty in the estimated internal thermal and solidification fields.", "Step 2: The improved state estimation (e.g., of temperature gradient G and solidification rate R) allows for the accurate prediction of local cooling rates (≈ G*R) and undercooling within the roll gap.", "Step 3: The accurately predicted local solidification conditions determine the nucleation density and growth kinetics, which govern the final as-cast grain size and morphology according to solidification theory." ], "key_assumptions": [ "The core multi-physics phenomena in TRC (fluid flow, heat transfer, solidification, mechanical deformation) can be modeled with sufficient fidelity to capture first-order effects on microstructure.", "The EnKF algorithm is robust to moderate structural model errors (e.g., in the interfacial heat transfer coefficient) and sensor noise characteristics typical of industrial TRC lines.", "Key internal state variables (e.g., mushy zone profile) are experimentally accessible via a reference method (e.g., rapid quench interruption) for validation." ], "known_unknowns": [ "The exact bandwidth and phase lag of the primary TRC actuators (roll gap servo, cooling valve) under dynamic operating conditions.", "The proportion of final grain size variance attributable to macroscale process parameters versus stochastic nucleation events.", "The optimal covariance inflation parameters for the EnKF to maintain filter stability against unmodeled TRC dynamics." ] }, "falsifiable_predictions": [ { "prediction": "If the EnKF-based digital twin is deployed for open-loop state estimation on a pilot TRC line, the RMSE of the estimated mushy zone thickness at the centerline will be ≤15% of the thickness measured via quench-interruption experiments.", "quantitative_bound": "RMSE ≤ 15% across the operational range (casting speed 0.5-3.0 m/min, roll gap 1.0-5.0 mm).", "measurement_method": "Comparison of EnKF-predicted solid fraction field against metallographic analysis of a rapidly quenched and sectioned strip. Thickness defined as distance between 0.2 and 0.8 solid fraction isotherms.", "null_hypothesis": "H0: The RMSE of the EnKF-estimated mushy zone thickness is ≥15% relative to the quench measurement.", "statistical_test": "One-sample t-test against the 15% threshold (one-tailed, alpha=0.05), with n≥5 independent casting trials per major parameter set." }, { "prediction": "The sensitivity of the EnKF state estimate to a ±30% error in the model's interfacial heat transfer coefficient will be bounded, resulting in a less than ±25% change in the estimated average cooling rate.", "quantitative_bound": "Output variation < ±25% for a ±30% input parameter perturbation.", "measurement_method": "Monte Carlo simulation within the EnKF framework, varying the heat transfer coefficient parameter across the defined range while holding sensor inputs constant from a representative casting trial.", "null_hypothesis": "H0: A ±30% perturbation in the heat transfer coefficient causes a ≥±25% change in the EnKF-estimated average cooling rate.", "statistical_test": "Paired t-test comparing the estimated cooling rate distributions from the nominal and perturbed parameter ensembles (alpha=0.05)." }, { "prediction": "An analysis of variance (ANOVA) on pilot line data will show that the four primary macroscale actuators (casting speed, roll gap, superheat, cooling) explain ≥40% of the observed variance in the final grain size.", "quantitative_bound": "R² ≥ 0.40 from a full-factorial designed experiment.", "measurement_method": "Full-factorial or highly fractionated factorial design of experiments (DoE) across the defined IV ranges, with grain size measured via linear intercept method on SEM/EBSD maps for each produced strip.", "null_hypothesis": "H0: The macroscale process parameters explain <40% (R² < 0.40) of the variance in final grain size.", "statistical_test": "Multi-factor ANOVA, testing the significance of main effects and key interactions (alpha=0.05)." } ], "boundary_conditions": [ { "condition": "Process operates within the 'stable casting window' defined by the alloy's hot tearing susceptibility and roll sticking limits.", "justification": "Outside this window, defect formation dominates microstructure, invalidating the solidification kinetics model." }, { "condition": "Sensor data assimilation cycle time is ≤100 ms.", "justification": "The dominant thermal dynamics in the roll gap have time constants on the order of 100-500 ms; slower updates fail to track state evolution." }, { "condition": "Alloy composition is within the range of commercial AZ31 or a similar Mg/Al alloy.", "justification": "The solidification kinetics and phase diagrams for these alloys are well-characterized, providing a stable basis for the physics-based model component." } ], "theoretical_framework": "Control-Theoretic State Estimation and Data Assimilation, applied to Rapid Solidification Processing." }, "protocol": { "protocol_title": "Three-Phase Validation of EnKF Digital Twin for Microstructure Estimation in Twin-Roll Casting", "overall_timeline": "8-14 months", "overall_budget_estimate": "€25k-120k", "phases": [ { "phase_number": 1, "phase_name": "In Silico Validation", "objective": "To determine if the EnKF algorithm, when coupled to a simplified but representative TRC model, can achieve state estimation errors below 15% under simulated sensor noise and model errors, and to assess sensitivity to key unknown parameters.", "methodology": "1. Develop a 2D axisymmetric or 2.5D reduced-order model of TRC in COMSOL Multiphysics or OpenFOAM, capturing coupled heat transfer, fluid flow (VOF), and solidification (enthalpy-porosity). Use material properties for AZ31. 2. Implement an Ensemble Kalman Filter (EnKF) in Python (using libraries like `pyDA` or `filterpy`) to assimilate synthetic sensor data (roll surface temperature at 3-5 axial positions, roll separating force) generated by the model with added Gaussian noise (5-10% amplitude). 3. Run twin experiments: a 'truth' simulation and an 'estimation' simulation where the EnKF assimilates noisy data into a model with intentional errors (e.g., ±30% error in interfacial heat transfer coefficient, IHTC). 4. Perform Monte Carlo sensitivity analysis: vary EnKF parameters (ensemble size, covariance inflation), IHTC, and sensor noise characteristics. Quantify RMSE of mushy zone thickness and cooling rate against the 'truth' simulation.", "required_resources": { "equipment": [ "High-performance workstation (32+ GB RAM, multi-core CPU)" ], "software": [ "COMSOL Multiphysics with CFD & Heat Transfer modules (or OpenFOAM)", "Python 3.x with NumPy, SciPy, Matplotlib, pandas, pyDA/filterpy", "Git for version control" ], "datasets": [ "Material property database for AZ31 Mg alloy (Thermo-Calc, JMatPro, or published literature)" ], "competences": [ "Computational Fluid Dynamics (CFD)", "Data Assimilation / Kalman Filtering", "Scientific Programming (Python)", "Solidification Modeling" ], "estimated_cost": "€0-2,000 (primarily software licenses if not available; OpenFOAM is open-source)", "estimated_duration": "4-8 weeks" }, "expected_outputs": [ "Output 1: A working Python-based EnKF coupled to a TRC process model.", "Output 2: Quantitative maps of state estimation error (RMSE%) for mushy zone thickness and cooling rate across the operational window.", "Output 3: Sensitivity matrices showing the impact of IHTC errors, sensor noise, and EnKF tuning parameters on estimation accuracy.", "Output 4: A preliminary assessment of the minimum sensor data frequency/quality required for stable filter performance." ], "success_criteria": [ { "metric": "RMSE of estimated mushy zone thickness", "threshold": "≤ 15% (relative to simulated 'truth') for ≥80% of the tested operational points (casting speed, roll gap).", "measurement": "Calculated from the EnKF state vector (solid fraction field) vs. the 'truth' simulation output at the same timestep. Thickness defined as distance between 0.2 and 0.8 solid fraction isotherms at the strip centerline." }, { "metric": "Sensitivity to IHTC error", "threshold": "A ±30% perturbation in the model's IHTC causes a < ±25% change in the EnKF-estimated average cooling rate.", "measurement": "Compare the distribution of estimated cooling rates from Monte Carlo runs with nominal vs. perturbed IHTC, using the paired difference analysis from the methodology." }, { "metric": "Filter stability", "threshold": "EnKF does not diverge (error remains bounded) over a simulated casting duration of 60 seconds for all tested tuning parameters within a reasonable range.", "measurement": "Visual inspection and analysis of estimation error time series." } ], "go_nogo_decision": { "go_if": "Success Criteria 1 AND 2 are met, demonstrating the core estimation concept is viable in silico and robust to a key model error.", "nogo_if": "Success Criterion 1 is not met (RMSE >15%) even under ideal simulated conditions (low noise, perfect model), indicating a fundamental flaw in the EnKF application or model observability.", "pivot_if": "Success Criterion 1 is met but Criterion 2 is not (high sensitivity to IHTC). Pivot to Phase 2 focusing on experimental IHTC characterization or to an EnKF variant more robust to parametric errors (e.g., dual estimation)." }, "risks": [ { "risk": "Simplified 2D model fails to capture essential 3D effects (edge effects, roll deflection) leading to misleadingly good/bad in silico performance.", "probability": "medium", "mitigation": "Include a 3D effect proxy in the 2D model (e.g., adjusted boundary conditions based on literature) and plan for a 3D model check in Phase 1 if results are borderline." }, { "risk": "EnKF tuning (ensemble size, inflation) is overly sensitive and difficult to stabilize, consuming excessive time.", "probability": "medium", "mitigation": "Start with established tuning heuristics from literature (e.g., ensemble size = 50-100, adaptive inflation). Use open-source DA libraries with built-in utilities." }, { "risk": "Lack of access to commercial simulation software (COMSOL).", "probability": "low", "mitigation": "Use open-source OpenFOAM. The 'chtMultiRegionFoam' solver with melting/solidification modifications is a viable, though more complex, starting point." } ] }, { "phase_number": 2, "phase_name": "Minimal Experimental Validation", "objective": "To validate the EnKF's core estimation mechanism using a single, well-instrumented pilot-scale TRC experiment with a quench-interruption, providing one direct comparison between estimated and measured internal state.", "methodology": "1. Select one 'nominal' operating condition within the stable window (e.g., 1.5 m/min, 2.5 mm gap, 30K superheat). 2. Instrument a pilot TRC line with key sensors: infrared pyrometers or embedded thermocouples for roll surface temperature (≥3 locations), load cells for roll separating force. Data acquisition at ≥10 Hz. 3. Conduct a casting run under nominal conditions. Simultaneously, run the EnKF digital twin (developed in Phase 1, now with a refined model) in real-time or post-process, assimilating the live sensor data. 4. At a predetermined steady-state moment, execute an emergency quench interruption (e.g., rapid roll separation and water jet immersion) to freeze the solidification structure in situ. 5. Section the quenched strip along the centerline. Perform metallographic preparation (mounting, polishing, etching). Measure the mushy zone thickness (distance between 0.2 and 0.8 solid fraction boundaries) via quantitative image analysis of the microstructure (e.g., using ImageJ with grain boundary contrast). 6. Compare this single, ground-truth measurement to the EnKF's estimated mushy zone thickness profile at the exact time of quench.", "required_resources": { "equipment": [ "Pilot-scale Twin-Roll Caster (e.g., laboratory unit)", "Data Acquisition System (NI cDAQ or similar, ≥10 Hz)", "Infrared Pyrometers (2-3) or embedded roll thermocouples", "Load Cells on roll bearings", "Quench system (fast-acting valves, water jets)", "Metallography lab (cutting, mounting, polishing, etching stations)", "Optical microscope with digital camera" ], "software": [ "Real-time capable EnKF code from Phase 1 (Python/C++)", "Data acquisition software (LabVIEW, Python DAQmx)", "Image analysis software (ImageJ, Olympus Stream)" ], "datasets": [ "Phase 1 simulation results for model initialization", "Single experimental dataset: sensor time series + quenched sample" ], "competences": [ "Pilot Plant Operation & Safety", "Sensor Integration & Data Acquisition", "Metallography & Microscopy", "Real-time Software Integration" ], "estimated_cost": "€8,000-15,000 (primarily for pilot line time, consumables (alloy), sensor rental/use, and metallography)", "estimated_duration": "6-12 weeks" }, "expected_outputs": [ "Output 1: A time-synchronized dataset of pilot TRC sensor readings and EnKF state estimates.", "Output 2: A metallographically prepared sample showing the frozen solidification front.", "Output 3: A direct, quantitative comparison: EnKF-predicted vs. measured mushy zone thickness at the quench moment.", "Output 4: Calibrated model parameters (especially IHTC) based on the inverse analysis of the quench experiment." ], "success_criteria": [ { "metric": "Single-point RMSE", "threshold": "The absolute difference between the EnKF-estimated and the measured mushy zone thickness at the quench moment is ≤15% of the measured value.", "measurement": "(|Estimated Thickness - Measured Thickness| / Measured Thickness) * 100%." }, { "metric": "Sensor-to-estimate correlation", "threshold": "The EnKF-estimated roll separating force shows a correlation coefficient R > 0.8 with the measured force during steady-state casting prior to quench.", "measurement": "Pearson correlation coefficient calculated over a 30-second steady-state period." } ], "go_nogo_decision": { "go_if": "Success Criterion 1 is met (≤15% error on the quench experiment), providing the first physical evidence supporting the hypothesis.", "nogo_if": "The error exceeds 30%, indicating a major discrepancy between the model+EnKF and reality that cannot be explained by measurement uncertainty alone.", "pivot_if": "Error is between 15% and 30%. Pivot to a detailed failure analysis: re-examine model assumptions (e.g., constitutive laws), IHTC calibration from the quench data, or sensor placement/fidelity before considering Phase 3." }, "risks": [ { "risk": "Quench interruption fails to perfectly freeze the microstructure, causing ambiguity in the 'true' mushy zone location.", "probability": "medium", "mitigation": "Use a highly aggressive quench (sub-zero water jet, rapid roll separation). Validate quenching effectiveness on a known sample first. Use multiple microstructural indicators (e.g., dendrite arm spacing, solute segregation) to triangulate the solidification front." }, { "risk": "Pilot line sensor data is too noisy or has unknown phase lags, corrupting the EnKF assimilation.", "probability": "high", "mitigation": "Perform pre-experiment sensor calibration and dynamic response tests. Implement data pre-processing (low-pass filtering) in the EnKF input pipeline. Use a robust EnKF variant (e.g., with adaptive noise estimation)." }, { "risk": "Single experiment is not statistically representative.", "probability": "medium", "mitigation": "Acknowledge this is a key limitation of Phase 2. The decision to proceed to Phase 3 will factor in the magnitude of the observed error and the quality of the diagnostic data from the experiment." } ] }, { "phase_number": 3, "phase_name": "Full Experimental Protocol", "objective": "To rigorously validate the hypothesis across the operational window, establish statistical significance, and quantify the explained variance in final grain size.", "methodology": "1. Design of Experiments (DoE): Implement a full or fractional factorial design across the 4 independent variables (casting speed, roll gap, superheat, cooling), with 3-5 center points for a total of ~20-30 casting trials. 2. For each trial: run the EnKF digital twin in real-time for state estimation and record all data. Produce a strip long enough for microstructural characterization (~2-3 meters). 3. Post-process: For each strip, cut samples for metallography. Measure final average grain size using Electron Backscatter Diffraction (EBSD) or automated optical microscopy (linear intercept method, n≥500 grains per sample). 4. Validation: For a subset of trials (e.g., 5-8 spanning the operational space), perform rapid quench interruptions to obtain direct mushy zone thickness measurements for RMSE calculation as in Phase 2. 5. Data Analysis: a) Perform one-sample t-tests to evaluate if the RMSE across quench trials is ≤15%. b) Perform multi-factor ANOVA on the full dataset to determine the proportion of grain size variance (R²) explained by the four macroscale parameters and their interactions. c) Correlate the EnKF-estimated solidification conditions (cooling rate, G*R) with the measured grain size using regression analysis.", "required_resources": { "equipment": [ "Pilot-scale Twin-Roll Caster with full actuator control", "Comprehensive sensor suite (as Phase 2, plus possibly more)", "Quench system for subset of trials", "Metallography lab", "Scanning Electron Microscope with EBSD capability (or high-end automated optical microscope)" ], "software": [ "Robust, real-time EnKF digital twin", "Statistical analysis software (JMP, Minitab, or Python statsmodels)", "EBSD data analysis software (Oxford Instruments AZtec, EDAX OIM)" ], "datasets": [ "Full DoE dataset: process parameters, sensor logs, EnKF estimates, grain size measurements", "Quench validation subset dataset" ], "competences": [ "Design of Experiments (DoE) & Statistical Analysis", "Advanced Metallography (sample preparation for EBSD)", "EBSD Operation & Data Interpretation", "Project Management for multi-month experimental campaigns" ], "estimated_cost": "€40,000-100,000+ (major costs: extensive pilot line time, alloy material for ~30 trials, EBSD analysis time, labor)", "estimated_duration": "6-12 months" }, "expected_outputs": [ "Output 1: A comprehensive database linking TRC process parameters, EnKF-estimated internal states, and final microstructure (grain size).", "Output 2: Statistical validation (or rejection) of the three falsifiable predictions with p-values and confidence intervals.", "Output 3: A calibrated and validated EnKF digital twin ready for deployment in process monitoring or control.", "Output 4: A publication-ready manuscript with complete experimental data." ], "success_criteria": [ { "metric": "Prediction 1: RMSE of mushy zone thickness", "threshold": "One-sample t-test confirms RMSE ≤ 15% across validation trials (p < 0.05, one-tailed).", "measurement": "As defined in the hypothesis, using quench-interruption measurements as ground truth for n≥5 independent trials." }, { "metric": "Prediction 2: Sensitivity to IHTC error", "threshold": "Paired t-test confirms that a ±30% IHTC perturbation causes a < ±25% change in estimated cooling rate (p < 0.05).", "measurement": "Performed in-silico using the final calibrated model and representative sensor data from Phase 3 trials." }, { "metric": "Prediction 3: Explained variance in grain size", "threshold": "ANOVA shows R² ≥ 0.40 for the model containing the four macroscale parameters (p < 0.05 for the overall model).", "measurement": "Calculated from the full DoE grain size dataset." } ], "go_nogo_decision": { "go_if": "All three success criteria are met, providing strong, statistically significant evidence supporting the published hypothesis.", "nogo_if": "Prediction 1 (the core RMSE claim) is statistically rejected (RMSE >15%, p > 0.05), invalidating the main premise of the work.", "pivot_if": "Predictions 1 and 2 are met, but Prediction 3 is not (R² < 0.40). Pivot the narrative to focus on state estimation accuracy, acknowledging that stochastic effects dominate final grain size, and explore using the EnKF estimates for other purposes (defect prediction)." }, "risks": [ { "risk": "High cost and long duration of the full DoE campaign.", "probability": "high", "mitigation": "Use a highly fractionated factorial design (e.g., Resolution IV) to minimize the number of trials while still estimating main effects and two-way interactions. Seek external funding or industrial partnership based on positive Phase 2 results." }, { "risk": "EBSD analysis is a bottleneck, expensive and time-consuming.", "probability": "medium", "mitigation": "Use automated optical microscopy with etching as a primary method, reserving EBSD for a subset of samples to validate the optical measurements and provide crystallographic texture data." }, { "risk": "Process instability or strip defects (e.g., hot tearing) in some DoE runs, producing unanalyzable samples and wasting resources.", "probability": "medium", "mitigation": "Conduct preliminary trials to confirm the 'stable window' boundaries. Include process stability metrics (e.g., force oscillations) in real-time monitoring to abort unstable runs early." } ] } ], "phase_1_quick_start": { "can_start_today": true, "first_action": "Set up a Python virtual environment and install core libraries: `numpy`, `scipy`, `matplotlib`, `pandas`, and the Ensemble Kalman Filter library `pyda` (or `filterpy`). Download the OpenFOAM case tutorial 'solidificationMeltingFoam' or access COMSOL's 'Phase Change' tutorial model.", "tools_needed": [ "Computer with internet", "Python 3.9+", "Git client", "Text editor/IDE (VS Code, PyCharm)" ], "open_data_sources": [ "OpenFOAM Foundation (openfoam.org) for solver basics", "NIST Magnesium Alloy Database for AZ31 properties", "GitHub repositories for EnKF examples (e.g., 'pyda' or 'OpenDA')" ] } }, "panel": { "reviews": [ { "reviewer_persona": "methodologist", "overall_score": 8.2, "verdict": "accept", "strengths": [ "Le protocole en trois phases (in silico, validation minimale, validation complète) est un modèle de rigueur progressive. Il permet de détecter des défaillances conceptuelles à faible coût avant d'engager des ressources expérimentales importantes, respectant le principe de parcimonie.", "L'identification et la quantification a priori des critères de succès (GO/NO-GO/PIVOT) pour chaque phase sont exceptionnellement claires. Cela structure la démarche, limite les biais de confirmation en définissant des points d'arrêt objectifs, et améliore la reproductibilité.", "La stratégie pour aborder la validité interne est solide : utilisation d'expériences jumelles (twin experiments) en Phase 1, point de vérité terrain unique en Phase 2 (quench), et plan d'expériences factoriel avec points centraux en Phase 3. Le contrôle par quench-interruption est la méthode de référence appropriée pour la variable dépendante clé." ], "weaknesses": [ "La puissance statistique de la Phase 3 n'est pas calculée. L'affirmation '~20-30 essais' pour un plan factoriel à 4 facteurs avec interactions et points centraux est vague. Sans calcul de puissance préalable basé sur une taille d'effet attendue, on risque un test ANOVA sous-puissant (risque de Type II), compromettant la validation de la prédiction 3 (R² ≥ 0.40).", "Le risque de biais de mesure pour la variable clé (épaisseur de la zone pâteuse) est sous-estimé. La définition opérationnelle ('distance entre isothermes de fraction solide 0.2 et 0.8') dépend fortement du protocole de préparation métallographique (agent de révélation, contraste) et de l'algorithme de segmentation d'image. Aucun protocole d'étalonnage inter-opérateur ou de cécité de l'analyste n'est mentionné.", "La Phase 2 (validation minimale) présente un risque élevé de sur-interprétation due à un N=1. Un seul point de validation, même avec un bon accord, ne permet pas d'extrapoler la performance à toute la fenêtre opératoire. Le critère GO/NO-GO basé sur une seule mesure est trop permissif ; une erreur ≤15% pourrait être due à la chance." ], "critical_questions": [ "Pour la Phase 3, quel est le calcul de la taille d'échantillon (nombre d'essais) basé sur : une taille d'effet minimale détectable pour le R², une puissance statistique visée (e.g., 0.8), et un alpha de 0.05 ? Un plan fractionnaire de résolution IV ou V est-il suffisant pour estimer les effets principaux sans confusion avec les interactions d'ordre deux ?", "Comment comptez-vous contrôler et quantifier l'incertitude de mesure sur la 'vérité terrain' issue du quench ? Un protocole de mesure en aveugle (l'analyste ignorant la valeur estimée par l'EnKF) et des mesures répétées par plusieurs opérateurs sont-ils prévus pour établir un intervalle de confiance autour de la mesure expérimentale, contre laquelle l'erreur ≤15% sera comparée ?", "Le biais de sélection est-il adressé pour les essais de quench dans la Phase 3 ('un sous-ensemble de 5-8 essais') ? La sélection de ce sous-ensemble sera-t-elle aléatoire ou stratifiée pour couvrir l'espace des paramètres ? Un biais systématique (e.g., ne réaliser le quench que sur les essais les plus stables) invaliderait l'inférence statistique sur le RMSE." ], "recommendation": "Approuvez le protocole avec des modifications obligatoires avant la Phase 3. Exigez : 1) un calcul formel de puissance statistique et un plan d'expériences définitif (e.g., plan de Box-Behnken ou factoriel fractionnaire de résolution V) pour la Phase 3 ; 2) l'ajout d'un protocole de métallographie standardisé avec validation inter-opérateur pour la mesure de l'épaisseur de la zone pâteuse ; 3) la modification du critère de la Phase 2 pour inclure au moins 3 répétitions au point nominal afin d'estimer la variabilité expérimentale. La structure fondamentale est excellente mais nécessite ce renforcement quantitatif.", "confidence": 0.9 }, { "reviewer_persona": "domain_expert", "overall_score": 7.8, "verdict": "accept", "confidence": 0.82, "strengths": [ "L'hypothèse est théoriquement cohérente et bien structurée. La chaîne causale allant de l'assimilation de données à la prédiction de la microstructure suit une progression logique, fondée sur la physique, qui s'accorde avec la théorie de la solidification (G*R → nucléation/croissance → structure de grains).", "Le positionnement est judicieux. Exploiter la robustesse éprouvée du filtre de Kalman d'ensemble dans des systèmes aérospatiaux fortement non linéaires et transitoires (tels que les moteurs-fusées à ergols liquides) pour traiter la « boîte noire » que constitue l'entrefer des cylindres en coulée entre cylindres est un transfert conceptuel convaincant et novateur. La base de preuves identifie correctement la littérature à l'appui, tant directe qu'analogue.", "La reconnaissance explicite des « inconnues connues » (dynamique des actionneurs, nucléation stochastique, inflation de covariance) témoigne d'une compréhension approfondie des défis pratiques du passage de la théorie à l'application, ce qui renforce la crédibilité de la proposition." ], "weaknesses": [ "Le postulat central concernant la « fidélité suffisante » du modèle multi-physique constitue un obstacle majeur. La capture des effets de premier ordre sur la microstructure dans le TRC exige une modélisation exceptionnellement précise du coefficient de transfert thermique interfacial (IHTC) et de la déformation dans la zone pâteuse, lesquels sont notoirement difficiles à caractériser et hautement dynamiques. La robustesse de l'EnKF face à des « erreurs structurelles modérées » dans ces termes est optimiste et nécessite une démonstration rigoureuse.", "Le postulat de validation est un élément critique du cheminement. L'obtention d'un accès expérimental aux variables d'état internes (par exemple, le profil de la zone pâteuse) par trempe rapide pour un procédé à grande vitesse comme le TRC est extrêmement difficile et pourrait ne pas fournir les données spatialement et temporellement denses requises pour une validation robuste du RMSE par rapport aux estimations de champ complet du jumeau numérique.", "L'hypothèse suppose implicitement l'observabilité. L'assimilation proposée de la température de surface des cylindres et de la force de séparation pourrait ne pas fournir suffisamment d'informations pour estimer de manière unique les champs thermiques et de solidification internes (par exemple, pour dissocier G et R indépendamment). Il s'agit d'une préoccupation fondamentale de la théorie du contrôle qui n'est pas abordée dans le mécanisme." ], "critical_questions": [ "Étant donné que l'incertitude dominante dans les modèles TRC provient de la condition aux limites (IHTC) plutôt que des conditions initiales, comment l'EnKF serait-il spécifiquement configuré (par exemple, l'augmentation d'état pour l'estimation de paramètres) afin d'estimer simultanément à la fois l'état thermique interne *et* ce paramètre critique variant spatialement ?", "La cible de RMSE de ≤15 % est relative à quelle mesure expérimentale ? S'il s'agit d'une mesure ponctuelle (par exemple, un thermocouple à un emplacement), elle peut être atteignable. S'il s'agit d'une mesure de champ (par exemple, le profil complet de l'épaisseur de la zone pâteuse), il s'agit d'une cible exceptionnellement ambitieuse compte tenu des difficultés de validation. Pouvez-vous définir la métrique de validation et la méthode de référence avec plus de précision ?", "L'article cité de 2024 sur le jumeau numérique TRC se concentre sur la corrélation. Quelles preuves spécifiques ou quels résultats préliminaires suggèrent que l'erreur *dynamique* d'estimation d'état (que l'EnKF minimise) est le facteur limitant pour la prédiction de la microstructure, par opposition à l'erreur *statique* de structure du modèle ?" ], "recommendation": "Il s'agit d'une hypothèse prometteuse et intellectuellement rigoureuse qui mérite d'être étudiée. Je recommande de procéder selon une approche par phases : premièrement, une analyse détaillée de l'observabilité et de l'identifiabilité de la suite de capteurs proposée avec le modèle haute fidélité ; deuxièmement, une étude EnKF « hors ligne » utilisant des données expérimentales de haute qualité issues d'une ligne pilote TRC instrumentée afin de quantifier la réduction réelle de l'erreur d'estimation d'état avant de revendiquer une capacité de prédiction de la microstructure. La nouveauté réside dans le transfert spécifique de techniques d'estimation de qualité aérospatiale, et c'est sur cela que devrait porter l'attention." }, { "reviewer_persona": "contrarian", "overall_score": 4.5, "verdict": "weak_reject", "confidence": 0.85, "strengths": [ "L'hypothèse identifie correctement le filtre de Kalman d'ensemble (EnKF) comme un outil théoriquement adapté à l'estimation d'état non linéaire et transitoire, et son succès antérieur dans des environnements aérospatiaux contrôlés constitue un point de départ plausible.", "La méthode de validation proposée (interruption par trempe) pour la prédiction centrale est directe et physiquement fondée, dépassant le cadre d'une validation purement numérique." ], "weaknesses": [ "FAIL REASON #1: Le postulat de « fidélité suffisante » pour le modèle multi-physique est probablement entaché d'un vice rédhibitoire. La coulée entre cylindres implique des phénomènes extrêmes et couplés — écoulement turbulent à surface libre avec solidification, contact interfacial dynamique et transfert thermique, et déformation viscoplastique — notoirement difficiles à modéliser même hors ligne. L'erreur du modèle structurel sera probablement si grande et non gaussienne que la capacité de l'EnKF à « gérer » l'erreur de modèle sera submergée, conduisant à une divergence du filtre ou à des estimations d'état physiquement dépourvues de sens.", "FAIL REASON #2: Les données de capteurs disponibles pour l'assimilation sont d'une insuffisance critique pour contraindre l'état interne de haute dimension. La température de surface et la force de séparation sont des sorties intégrales à faible bande passante, liées de manière non univoque au front de solidification interne. Il s'agit d'un problème d'observabilité sévère. L'EnKF ne peut pas créer par magie de l'information sur les gradients thermiques internes à partir de ces quelques mesures indirectes et bruitées.", "FAIL REASON #3: Le postulat de validation lui-même est instable. La « vérité terrain » issue des expériences d'interruption par trempe n'est pas une mesure directe et synchrone de l'état du procédé. L'événement de trempe lui-même perturbe l'histoire thermique et de solidification. La cible de RMSE ≤15 % est arbitraire et probablement inatteignable compte tenu des incertitudes composées dans le modèle, les capteurs et la méthode de validation." ], "critical_questions": [ "Compte tenu de l'incertitude connue de plusieurs ordres de grandeur sur le coefficient de transfert thermique interfacial (IHTC) dans les contacts de laminage/solidification — hautement dynamique et dépendant de la pression — de quelles preuves spécifiques disposez-vous que l'inflation de covariance de l'EnKF peut compenser cela, plutôt que de produire simplement une estimation précise mais inexacte d'un état erroné ?", "L'hypothèse postule que la chaîne causale allant des paramètres macroscopiques à la taille de grain est dominante (≥40 % de variance expliquée). Et si la source de bruit dominante était à l'échelle microscopique (par exemple, la distribution des particules d'inoculant, les bi-films d'oxyde) ou provenait de la préparation du bain en amont ? Cela ne rendrait-il pas l'ensemble du jumeau numérique, aussi précis soit-il pour les champs thermiques, non pertinent pour l'objectif de microstructure énoncé ?" ], "recommendation": "Avant de tester l'hypothèse complète, démontrer qu'un modèle CFD-thermodynamique autonome (sans EnKF) à haute fidélité du procédé TRC peut prédire le profil de la zone pâteuse avec un RMSE inférieur à 25 % par rapport aux données de trempe sur plusieurs points de fonctionnement. Ensuite, réaliser une analyse d'observabilité à l'aide de la suite de capteurs proposée pour prouver que l'état thermique interne est théoriquement récupérable. Sans ces résultats fondamentaux, l'intégration de l'EnKF est une solution en quête d'un problème soluble." }, { "reviewer_persona": "industrialist", "overall_score": 6.5, "verdict": "weak_accept", "strengths": [ "Adresse un besoin critique de l'industrie des métaux légers (Al, Mg) pour le contrôle en ligne de la microstructure, avec un marché adressable immédiat chez les producteurs de bandes laminées à chaud comme Novelis, Constellium, Hydro, et les équipementiers de fonderie (Danieli, SMS group).", "Avantage compétitif potentiellement disruptif : passer d'un contrôle rétrospectif (post-mortem métallographique) à un contrôle prédictif et en temps réel de la solidification, permettant un réglage dynamique pour des propriétés mécaniques cibles et une réduction des rebuts." ], "weaknesses": [ "Barrière majeure au déploiement : l'intégration sur lignes industrielles existantes nécessite des capteurs in-situ robustes (thermographie, ultrasons) dans un environnement hostile (haute température, contraintes mécaniques), et une infrastructure informatique temps-réel non triviale.", "Le modèle économique est flou : vente de licence logicielle (CAPEX lourd) vs service d'optimisation (subscription) ? La valeur perçue par l'opérateur doit surpasser le coût et la complexité de mise en œuvre." ], "critical_questions": [ "Quelle est la stratégie de propriété intellectuelle ? Le cœur de valeur est-il dans l'algorithme EnKF (difficile à breveter seul) ou dans le couplage spécifique avec le modèle multi-physique et les données propriétaires de calibration ? Une approche par secret industriel sur les modèles calibrés est-elle plus viable ?", "Quel est le ROI concret pour un client ? Peut-on quantifier la réduction de rebuts (%), l'économie d'énergie (via un superheat optimisé), ou le gain de parts de marché sur des alliages à haute valeur ajoutée (ex : alliages d'aluminium pour batterie) pour justifier un investissement >500k€ ?" ], "recommendation": "Lancer la Phase 1 (budget minimal) pour valider le principe en silico et affiner la proposition de valeur. En parallèle, initier des discussions exploratoires non-disclosing avec un partenaire industriel pilote (ex : institut de recherche appliquée comme l'Institut de Recherche Technologique Jules Verne ou un métallurgiste avancé) pour co-définir les spécifications des capteurs et le cas d'usage prioritaire, condition sine qua non pour un passage aux phases 2 et 3.", "confidence": 0.75 }, { "reviewer_persona": "funding_strategist", "overall_score": 6.5, "verdict": "weak_accept", "strengths": [ "Hypothèse très bien formulée avec un protocole de validation incrémentale clair et réaliste, montrant une compréhension fine des étapes de maturation technologique (TRL).", "Approche de transfert de technologie (aerospace vers manufacturing) originale et potentiellement disruptive pour le contrôle de procédés métallurgiques complexes." ], "weaknesses": [ "Le budget et la timeline proposés (€25k-120k, 8-14 mois) sont atypiques et ne correspondent pas aux standards des grands programmes européens, indiquant un TRL initial très bas (TRL 2-3).", "Le consortium est actuellement implicite ou mono-partenaire. Les appels ciblés nécessitent un consortium multi-acteurs (industrie, académique, pays multiples) pour être compétitifs." ], "critical_questions": [ "Quelle est la stratégie pour constituer un consortium incluant un partenaire industriel propriétaire d'une ligne de TRC pilote ou industrielle, essentiel pour la crédibilité des phases 2 et 3 ?", "Comment justifier le saut d'échelle entre la validation sur un modèle simplifié (Phase 1) et l'application sur un procédé physique pilot-scale avec toutes ses incertitudes non modélisées ?" ], "recommendation": "Cibler d'abord un financement de type 'Proof of Concept' ou 'Seed' pour exécuter la Phase 1 et affiner l'hypothèse. En parallèle, structurer un consortium académique-industriel pour viser un appel collaboratif de l'EU sur la digitalisation des procédés industriels, en présentant les résultats préliminaires comme un atout majeur.", "confidence": 0.8, "funding_programs": [ { "program": "ERC Proof of Concept Grant", "agency": "European Research Council", "fit_score": 0.7, "typical_budget": "€150,000", "success_rate": "~35-40%", "next_deadline": "16 octobre 2024 (prochaine fenêtre prévue)", "rationale": "Ce programme est idéal si le PI détient déjà un ERC (Starting/Consolidator/Advanced) sur des sujets liés (data assimilation, modélisation multi-physique). Il permet de financer l'exploration du potentiel d'innovation et de transfert de la recherche fondamentale, correspondant parfaitement au budget et à la timeline des Phases 1 et 2. Il valide la faisabilité avant un projet plus large." }, { "program": "M-ERA.NET 3 Call 2024 (Thème 1 : Modeling for manufacturing)", "agency": "M-ERA.NET (Réseau d'agences nationales et européennes)", "fit_score": 0.85, "typical_budget": "€300k - €800k par partenaire (projet total ~€1-2M)", "success_rate": "~15-20%", "next_deadline": "Novembre 2024 (pour pré-propositions)", "rationale": "Appel transnational parfaitement aligné sur le développement de modèles numériques pour les procédés de fabrication. L'hypothèse sur le jumeau numérique avec assimilation de données pour le coulage est un cas d'école. Nécessite un consortium d'au moins 3 entités issues de 3 pays participants. L'ANR est partenaire pour les équipes françaises." }, { "program": "Horizon Europe - Cluster 4 'Digital, Industry and Space' - Destination 7 : Innovative Digital Technologies for Manufacturing (DIGITAL-2024-CLOUD-AI-07)", "agency": "European Commission", "fit_score": 0.9, "typical_budget": "€3M - €5M par projet", "success_rate": "~5-10%", "next_deadline": "20 novembre 2024 (pour le topic 7.1 : AI-powered modelling and prediction)", "rationale": "Le topic spécifique 'AI-powered modelling and prediction for manufacturing processes' vise exactement ce type d'innovation : utiliser l'IA/le data assimilation pour créer des jumeaux numériques prédictifs et réduire l'écart modèle-réalité. Le projet doit démontrer un impact industriel significatif, nécessitant un consortium fort incluant un équipementier métallurgique (ex: Primetals, Danieli) et un producteur d'aluminium (ex: Constellium, Hydro)." } ] } ], "meta_review": { "consensus_score": 6.72, "verdict": "publish_brief", "key_consensus": [ "Le protocole en trois phases est reconnu comme une structure méthodologique rigoureuse et progressive, représentant un modèle de validation incrémentale.", "L'hypothèse est jugée théoriquement cohérente et présente une approche de transfert de technologie (aérospatial vers la métallurgie) originale et prometteuse.", "La nécessité d'un renforcement quantitatif du protocole, notamment sur la puissance statistique de la Phase 3 et la robustesse des mesures expérimentales, est un point d'accord unanime.", "L'importance cruciale de s'associer à un partenaire industriel disposant d'une ligne pilote pour les phases de validation expérimentale est soulignée par plusieurs experts." ], "key_disagreements": [ "La faisabilité fondamentale de l'hypothèse est contestée. Le Contrarian et le Domain Expert doutent fortement de la capacité du modèle multi-physique à atteindre une fidélité suffisante et de l'observabilité de l'état interne avec les capteurs proposés, tandis que le Méthodologiste et les autres acceptent le principe sous réserve d'améliorations techniques.", "L'évaluation du risque diffère radicalement : le Contrarian identifie des « raisons d'échec » potentiellement fatales (erreur structurelle du modèle, observabilité), alors que les autres reviewers considèrent ces défis comme des obstacles techniques à surmonter dans le cadre d'une approche par phases.", "L'ambition du critère de validation (RMSE ≤15%) est perçue soit comme un objectif ambitieux mais réalisable avec un protocole robuste (Domain Expert, Méthodologiste), soit comme arbitraire et probablement inaccessible compte tenu des incertitudes cumulées (Contrarian)." ], "critical_path": "La démonstration préalable, en Phase 1, que le modèle numérique standalone peut prédire le profil de la zone pâteuse avec une erreur raisonnable (<25% RMSE) face à des données de quench, couplée à une analyse rigoureuse de l'observabilité du système avec la suite de capteurs proposée. Sans cette validation fondatrice, l'intégration de l'EnKF risque d'être un « solutionnisme technologique » appliqué à un problème mal contraint.", "final_recommendation": "Le panel recommande la publication du brief, reconnaissant le mérite intellectuel et le potentiel de l'hypothèse, mais avec des réserves substantielles. La poursuite du projet est conditionnée à l'exécution stricte de la Phase 1, qui doit servir de test décisif des prémisses fondamentales (fidélité du modèle et observabilité). Les phases ultérieures devront intégrer les exigences de robustesse méthodologique (calcul de puissance, protocole de mesure standardisé) et s'appuyer sur un consortium incluant un partenaire industriel pilote. L'accent doit rester sur la validation incrémentale et la quantification des incertitudes à chaque étape.", "brief_quality_gate": true, "revision_guidance": [], "llm_verdict": "publish_brief", "llm_consensus_score": 6.7 } }, "vulgarization_fr": { "title_fr": "Un jumeau numérique piloté par un filtre aérospatial pour voir l'intérieur des bandes coulées", "hypothesis_in_brief": "La coulée entre cylindres produit des bandes métalliques en quelques secondes, mais l'intérieur du métal en train de se solidifier reste invisible. Cette hypothèse propose d'estimer cet état interne en temps réel en combinant un modèle physique détaillé du procédé et un filtre mathématique conçu à l'origine pour piloter des moteurs de fusée. L'objectif est de prédire l'épaisseur de la zone pâteuse et la vitesse de refroidissement avec une erreur inférieure à 15 % par rapport à des mesures directes.", "why_it_matters": "Dans la production de bandes d'aluminium ou de magnésium, la taille des grains formés pendant la solidification détermine la résistance et la ductilité du produit final. Aujourd'hui, cette taille de grains n'est connue qu'après coup, par analyse métallographique en laboratoire, quand la bande est déjà fabriquée. Estimer en direct les conditions de solidification permettrait d'ajuster la vitesse de coulée, l'écartement des cylindres ou la surchauffe du métal pour viser une microstructure cible, et de réduire les rebuts. Cela concerne les producteurs de bandes laminées et les fabricants d'équipements de coulée continue.", "imagine_that": "Imaginez un boulanger qui enfourne des centaines de pains par jour sans jamais pouvoir ouvrir le four. Il règle la température et le temps selon son expérience, mais il ne sait si la mie est bien alvéolée qu'en coupant un pain après cuisson. Maintenant, imaginez qu'il dispose d'un thermomètre à pâte planté dans chaque four et d'un modèle mathématique qui, à partir de la température en surface et de la résistance de la pâte, reconstitue en direct l'état de la mie à cœur. C'est exactement ce que cette hypothèse tente de faire pour le métal en fusion entre les cylindres.", "concretely": { "intro": "L'approche se teste en trois étapes, de la simulation pure jusqu'à la validation sur une ligne pilote instrumentée.", "phase1": "Un modèle informatique du procédé est construit et couplé au filtre mathématique. On vérifie, en simulant du bruit de capteurs et des erreurs de modèle, si l'erreur d'estimation reste sous les 15 %.", "phase2": "Une seule expérience sur une ligne pilote, avec des capteurs de température et de force, permet de comparer une estimation à une mesure directe obtenue en interrompant brutalement la solidification par trempe.", "phase3": "Un plan d'expériences balaie plusieurs vitesses de coulée, écartements, surchauffes et refroidissements pour vérifier que les quatre paramètres principaux expliquent au moins 40 % de la variance observée sur la taille des grains." }, "reviewers_say": "Le panel reconnaît une hypothèse théoriquement cohérente et un transfert de technologie original, de l'aérospatial vers la métallurgie. Le protocole en trois phases, avec des critères d'arrêt clairs, est jugé rigoureux et reproductible. Les réserves portent sur deux points : la capacité du modèle physique à atteindre une fidélité suffisante pour capturer les phénomènes couplés dans l'entrefer, et l'observabilité de l'état interne à partir de capteurs de surface peu nombreux et bruités. Un reviewer conteste même la faisabilité fondamentale, évoquant un risque de divergence du filtre. Le verdict global est de publier le brief, mais en conditionnant la suite à la réussite de la Phase 1, qui doit démontrer que le modèle seul prédit correctement l'état interne avant d'y ajouter le filtre." }, "vulgarization_en": { "title": "An aerospace filter-driven digital twin for seeing inside casting strips", "hypothesis_in_brief": "Twin-roll casting produces metal strip in a matter of seconds, yet the interior of the solidifying metal remains invisible. This hypothesis proposes to estimate that internal state in real time by combining a detailed physical model of the process with a mathematical filter originally designed to steer rocket engines. The objective is to predict the thickness of the mushy zone and the cooling rate with an error below 15 % relative to direct measurements.", "why_it_matters": "In the production of aluminium or magnesium strip, the grain size formed during solidification determines the strength and ductility of the final product. At present, this grain size is known only after the fact, through metallographic analysis in the laboratory, once the strip has already been manufactured. Estimating solidification conditions in real time would allow the casting speed, roll gap or metal superheat to be adjusted towards a target microstructure, and would reduce scrap. This concerns producers of rolled strip and manufacturers of continuous casting equipment.", "imagine_that": "Imagine a baker who loads hundreds of loaves a day into the oven without ever being able to open it. He sets the temperature and the time from experience, but he only knows whether the crumb is well aerated by cutting a loaf after baking. Now imagine he has a dough thermometer inserted into each oven and a mathematical model that, from the surface temperature and the resistance of the dough, reconstructs the state of the crumb at the core in real time. This is exactly what this hypothesis attempts to do for molten metal between the rolls.", "reviewers_say": "The panel recognises a theoretically coherent hypothesis and an original technology transfer, from aerospace to metallurgy. The three-phase protocol, with clear stop criteria, is judged rigorous and reproducible. Reservations concern two points: the capacity of the physical model to reach sufficient fidelity to capture the coupled phenomena in the air gap, and the observability of the internal state from few and noisy surface sensors. One reviewer even contests the fundamental feasibility, evoking a risk of filter divergence. The overall verdict is to publish the brief, but conditioning the continuation on the success of Phase 1, which must demonstrate that the model alone correctly predicts the internal state before adding the filter.", "concretely": { "intro": "The approach is tested in three stages, from pure simulation through to validation on an instrumented pilot line.", "phase1": "A computational model of the process is constructed and coupled to the mathematical filter. By simulating sensor noise and model errors, it is verified whether the estimation error remains below 15%.", "phase2": "A single experiment on a pilot line, with temperature and force sensors, allows an estimate to be compared against a direct measurement obtained by abruptly interrupting solidification through quenching.", "phase3": "An experimental design sweeps several casting speeds, spacings, superheats and cooling rates to verify that the four principal parameters account for at least 40% of the variance observed in grain size." } }, "panel_en": { "reviews": [ { "reviewer_persona": "methodologist", "overall_score": 8.2, "verdict": "accept", "confidence": 0.9, "strengths": [ "The three-phase protocol (in silico, minimal validation, full validation) is a model of progressive rigour. It enables conceptual failures to be detected at low cost before substantial experimental resources are committed, in keeping with the principle of parsimony.", "The a priori identification and quantification of success criteria (GO/NO-GO/PIVOT) for each phase are exceptionally clear. This structures the approach, limits confirmation bias by defining objective stopping points, and improves reproducibility.", "The strategy for addressing internal validity is robust: the use of twin experiments in Phase 1, a single ground truth point in Phase 2 (quench), and a factorial experimental design with centre points in Phase 3. Quench-interruption control is the appropriate baseline method for the key dependent variable." ], "weaknesses": [ "The statistical power of Phase 3 is not calculated. The claim of '~20–30 trials' for a 4-factor factorial design with interactions and centre points is vague. Without a prior power analysis based on an expected effect size, an underpowered ANOVA is risked (Type II error), compromising the validation of prediction 3 (R² ≥ 0.40).", "The risk of measurement bias for the key variable (paste zone thickness) is underestimated. The operational definition ('distance between solid fraction isotherms of 0.2 and 0.8') depends strongly on the metallographic preparation protocol (etching agent, contrast) and on the image segmentation algorithm. No inter-operator calibration protocol or analyst blinding is mentioned.", "Phase 2 (minimal validation) presents a high risk of overinterpretation owing to an N=1. A single validation point, even with good agreement, does not permit the performance to be extrapolated to the entire operating window. The GO/NO-GO criterion based on a single measurement is too permissive; an error ≤15% could be due to chance." ], "critical_questions": [ "For Phase 3, what is the sample size calculation (number of trials) based on: a minimum detectable effect size for R², a target statistical power (e.g., 0.8), and an alpha of 0.05? Is a fractional design of resolution IV or V sufficient to estimate the main effects without confounding with two-order interactions?", "How is the measurement uncertainty on the ground truth derived from the quench to be controlled and quantified? Are a blinded measurement protocol (the analyst being unaware of the value estimated by the EnKF) and repeated measurements by several operators planned, in order to establish a confidence interval around the experimental measurement against which the error ≤15% will be compared?", "Is selection bias addressed for the quench trials in Phase 3 ('a subset of 5-8 trials')? Will the selection of this subset be random or stratified to cover the parameter space? A systematic bias (e.g., performing the quench only on the most stable trials) would invalidate the statistical inference on the RMSE." ], "recommendation": "Approve the protocol with mandatory modifications before Phase 3. Require: 1) a formal statistical power calculation and a definitive experimental design (e.g., Box-Behnken design or resolution V fractional factorial) for Phase 3; 2) the addition of a standardised metallography protocol with inter-operator validation for measurement of the mushy zone thickness; 3) modification of the Phase 2 criterion to include at least 3 replicates at the nominal point in order to estimate experimental variability. The fundamental structure is excellent but requires this quantitative strengthening." }, { "reviewer_persona": "domain_expert", "overall_score": 7.8, "verdict": "accept", "confidence": 0.82, "strengths": [ "The hypothesis is theoretically coherent and well structured. The causal chain from data assimilation to microstructure prediction follows a logical, physics-based progression that aligns with solidification theory (G*R → nucleation/growth → grain structure).", "The positioning is judicious. Exploiting the proven robustness of the ensemble Kalman filter in highly non-linear and transient aerospace systems (such as liquid-propellant rocket engines) to address the \"black box\" constituted by the gap between rolls in roll casting is a convincing and novel conceptual transfer. The evidence base correctly identifies the supporting literature, both direct and analogous.", "The explicit recognition of \"known unknowns\" (actuator dynamics, stochastic nucleation, covariance inflation) demonstrates a thorough understanding of the practical challenges of moving from theory to application, which strengthens the credibility of the proposal." ], "weaknesses": [ "The central postulate concerning the \"sufficient fidelity\" of the multi-physics model constitutes a major obstacle. Capturing first-order effects on the microstructure in the TRC requires exceptionally accurate modelling of the interfacial heat transfer coefficient (IHTC) and of the strain in the mushy zone, which are notoriously difficult to characterise and highly dynamic. The robustness of the EnKF in the face of \"moderate structural errors\" in these terms is optimistic and requires rigorous demonstration.", "The validation postulate is a critical element of the pathway. Obtaining experimental access to internal state variables (for example, the mushy zone profile) by rapid quenching for a high-speed process such as the TRC is extremely difficult and may not provide the spatially and temporally dense data required for robust validation of the RMSE against the digital twin's full-field estimates.", "The hypothesis implicitly assumes observability. The proposed assimilation of the roll surface temperature and of the separating force may not provide sufficient information to estimate the internal thermal and solidification fields uniquely (for example, to dissociate G and R independently). This is a fundamental concern of control theory that is not addressed in the mechanism." ], "critical_questions": [ "Given that the dominant uncertainty in TRC models arises from the boundary condition (IHTC) rather than from the initial conditions, how would the EnKF be specifically configured (for example, state augmentation for parameter estimation) so as to estimate simultaneously both the internal thermal state *and* this critical spatially varying parameter?", "The RMSE target of ≤15 % is relative to which experimental measurement? If it concerns a point measurement (for example, a thermocouple at a single location), it may be attainable. If it concerns a field measurement (for example, the full profile of the mushy-zone thickness), it constitutes an exceptionally ambitious target given the difficulties of validation. Can the validation metric and the baseline method be defined with greater precision?", "The cited 2024 paper on the TRC digital twin focuses on correlation. What specific evidence or preliminary results suggest that the *dynamic* state-estimation error (which the EnKF minimises) is the limiting factor for microstructure prediction, as opposed to the *static* model-structure error?" ], "recommendation": "This is a promising and intellectually rigorous hypothesis that merits investigation. A phased approach is recommended: first, a detailed analysis of the observability and identifiability of the proposed sensor suite with the high-fidelity model; second, an \"offline\" EnKF study using high-quality experimental data from an instrumented TRC pilot line in order to quantify the actual reduction in state-estimation error before any claim of microstructure-prediction capability is made. The novelty lies in the specific transfer of aerospace-grade estimation techniques, and it is on this that attention should be focused." }, { "reviewer_persona": "contrarian", "overall_score": 4.5, "verdict": "weak_reject", "confidence": 0.85, "strengths": [ "The hypothesis correctly identifies the ensemble Kalman filter (EnKF) as a tool theoretically suited to non-linear and transient state estimation, and its prior success in controlled aerospace environments constitutes a plausible starting point.", "The proposed validation method (quench interruption) for the central prediction is direct and physically grounded, exceeding the scope of a purely numerical validation." ], "weaknesses": [ "FAIL REASON #1: The assumption of \"sufficient fidelity\" for the multi-physics model is probably vitiated by a fatal flaw. The casting between rolls involves extreme and coupled phenomena — turbulent free-surface flow with solidification, dynamic interfacial contact and heat transfer, and viscoplastic deformation — which are notoriously difficult to model even offline. The structural model error will probably be so large and non-Gaussian that the capacity of the EnKF to \"handle\" model error will be overwhelmed, leading to filter divergence or to physically meaningless state estimates.", "FAIL REASON #2: The sensor data available for assimilation are critically insufficient to constrain the high-dimensional internal state. Surface temperature and separation force are low-bandwidth integral outputs, related non-uniquely to the internal solidification front. This is a severe observability problem. The EnKF cannot magically create information about internal thermal gradients from these few indirect and noisy measurements.", "FAIL REASON #3: The validation assumption itself is unsound. The \"ground truth\" derived from quench-interruption experiments is not a direct and synchronous measurement of the process state. The quench event itself perturbs the thermal and solidification history. The RMSE target of ≤15 % is arbitrary and probably unattainable given the compounded uncertainties in the model, the sensors and the validation method." ], "critical_questions": [ "Given the known uncertainty of several orders of magnitude in the interfacial heat transfer coefficient (IHTC) at rolling/solidification contacts — highly dynamic and pressure-dependent — what specific evidence is available that covariance inflation in the EnKF can compensate for this, rather than merely producing a precise but inaccurate estimate of an erroneous state?", "The hypothesis posits that the causal chain from macroscopic parameters to grain size is dominant (≥40% of variance explained). What if the dominant noise source is at the microscopic scale (for example, the distribution of inoculant particles, oxide bi-films) or originates from upstream melt preparation? Would this not render the entire digital twin, however accurate for thermal fields, irrelevant to the stated microstructural objective?" ], "recommendation": "Before testing the full hypothesis, it must be demonstrated that a standalone high-fidelity CFD-thermodynamic model (without EnKF) of the TRC process can predict the mushy-zone profile with an RMSE below 25 % relative to quench data across several operating points. An observability analysis must then be performed using the proposed sensor suite to establish that the internal thermal state is theoretically recoverable. Without these foundational results, the integration of the EnKF is a solution in search of a tractable problem." }, { "reviewer_persona": "industrialist", "overall_score": 6.5, "verdict": "weak_accept", "confidence": 0.75, "strengths": [ "Addresses a critical need of the light-metals industry (Al, Mg) for in-line microstructure control, with an immediate addressable market among hot-rolled strip producers such as Novelis, Constellium and Hydro, and among foundry equipment manufacturers (Danieli, SMS group).", "Potentially disruptive competitive advantage: a shift from retrospective control (post-mortem metallography) to predictive, real-time control of solidification, enabling dynamic adjustment towards target mechanical properties and a reduction in scrap." ], "weaknesses": [ "A major barrier to deployment: integration into existing industrial lines requires robust in-situ sensors (thermography, ultrasound) in a hostile environment (high temperature, mechanical stress), together with a non-trivial real-time computing infrastructure.", "The business model is unclear: software licence sales (heavy CAPEX) versus an optimisation service (subscription)? The value perceived by the operator must outweigh the cost and complexity of implementation." ], "critical_questions": [ "What is the intellectual property strategy? Does the core value reside in the EnKF algorithm (difficult to patent on its own) or in the specific coupling with the multi-physics model and the proprietary calibration data? Is an approach based on trade secret protection of the calibrated models more viable?", "What is the concrete ROI for a client? Can the reduction in scrap (%), the energy savings (via an optimised superheat), or the gain in market share on high-value-added alloys (e.g. aluminium alloys for batteries) be quantified to justify an investment >€500k?" ], "recommendation": "Launch Phase 1 (minimal budget) to validate the principle in silico and refine the value proposition. In parallel, initiate exploratory non-disclosing discussions with a pilot industrial partner (e.g. an applied research institute such as the Institut de Recherche Technologique Jules Verne or an advanced metallurgist) to co-define the sensor specifications and the priority use case, a sine qua non condition for progression to Phases 2 and 3." }, { "reviewer_persona": "funding_strategist", "overall_score": 6.5, "verdict": "weak_accept", "confidence": 0.8, "funding_programs": [ { "program": "ERC Proof of Concept Grant", "agency": "European Research Council", "fit_score": 0.7, "typical_budget": "€150,000", "success_rate": "~35-40%", "next_deadline": "16 octobre 2024 (prochaine fenêtre prévue)", "rationale": "Ce programme est idéal si le PI détient déjà un ERC (Starting/Consolidator/Advanced) sur des sujets liés (data assimilation, modélisation multi-physique). Il permet de financer l'exploration du potentiel d'innovation et de transfert de la recherche fondamentale, correspondant parfaitement au budget et à la timeline des Phases 1 et 2. Il valide la faisabilité avant un projet plus large." }, { "program": "M-ERA.NET 3 Call 2024 (Thème 1 : Modeling for manufacturing)", "agency": "M-ERA.NET (Réseau d'agences nationales et européennes)", "fit_score": 0.85, "typical_budget": "€300k - €800k par partenaire (projet total ~€1-2M)", "success_rate": "~15-20%", "next_deadline": "Novembre 2024 (pour pré-propositions)", "rationale": "Appel transnational parfaitement aligné sur le développement de modèles numériques pour les procédés de fabrication. L'hypothèse sur le jumeau numérique avec assimilation de données pour le coulage est un cas d'école. Nécessite un consortium d'au moins 3 entités issues de 3 pays participants. L'ANR est partenaire pour les équipes françaises." }, { "program": "Horizon Europe - Cluster 4 'Digital, Industry and Space' - Destination 7 : Innovative Digital Technologies for Manufacturing (DIGITAL-2024-CLOUD-AI-07)", "agency": "European Commission", "fit_score": 0.9, "typical_budget": "€3M - €5M par projet", "success_rate": "~5-10%", "next_deadline": "20 novembre 2024 (pour le topic 7.1 : AI-powered modelling and prediction)", "rationale": "Le topic spécifique 'AI-powered modelling and prediction for manufacturing processes' vise exactement ce type d'innovation : utiliser l'IA/le data assimilation pour créer des jumeaux numériques prédictifs et réduire l'écart modèle-réalité. Le projet doit démontrer un impact industriel significatif, nécessitant un consortium fort incluant un équipementier métallurgique (ex: Primetals, Danieli) et un producteur d'aluminium (ex: Constellium, Hydro)." } ], "strengths": [ "The hypothesis is very well formulated, with a clear and realistic incremental validation protocol, demonstrating a refined understanding of the technology readiness level (TRL) maturation stages.", "An original technology-transfer approach (aerospace to manufacturing) with potentially disruptive implications for the control of complex metallurgical processes." ], "weaknesses": [ "The proposed budget and timeline (€25k–120k, 8–14 months) are atypical and do not correspond to the standards of major European programmes, indicating a very low initial TRL (TRL 2–3).", "The consortium is currently implicit or single-partner. Targeted calls require a multi-stakeholder consortium (industry, academic, multiple countries) to be competitive." ], "critical_questions": [ "What is the strategy for constituting a consortium that includes an industrial partner owning a pilot or industrial TRC line, essential for the credibility of Phases 2 and 3?", "How is the scale-up between validation on a simplified model (Phase 1) and application to a physical pilot-scale process, with all its unmodelled uncertainties, to be justified?" ], "recommendation": "A 'Proof of Concept' or 'Seed' funding type is to be targeted first in order to execute Phase 1 and refine the hypothesis. In parallel, an academic-industrial consortium is to be structured with a view to a collaborative EU call on the digitalisation of industrial processes, presenting the preliminary results as a major asset." } ], "meta_review": { "consensus_score": 6.72, "verdict": "publish_brief", "brief_quality_gate": true, "llm_verdict": "publish_brief", "llm_consensus_score": 6.7, "key_consensus": [ "The three-phase protocol is recognised as a rigorous and progressive methodological structure, representing a model of incremental validation.", "The hypothesis is judged to be theoretically coherent and presents an original and promising approach to technology transfer (aerospace to metallurgy).", "The need for quantitative strengthening of the protocol, notably regarding the statistical power of Phase 3 and the robustness of the experimental measurements, is a point of unanimous agreement.", "The crucial importance of partnering with an industrial partner possessing a pilot line for the experimental validation phases is emphasised by several experts." ], "key_disagreements": [ "The fundamental feasibility of the hypothesis is contested. The Contrarian and the Domain Expert strongly doubt the capacity of the multi-physics model to achieve sufficient fidelity and the observability of the internal state with the proposed sensors, whereas the Methodologist and the others accept the principle subject to technical improvements.", "The assessment of risk differs radically: the Contrarian identifies potentially fatal \"reasons for failure\" (structural model error, observability), whereas the other reviewers regard these challenges as technical obstacles to be overcome within the framework of a phased approach.", "The ambition of the validation criterion (RMSE ≤15%) is perceived either as an ambitious but achievable objective with a robust protocol (Domain Expert, Methodologist), or as arbitrary and probably unattainable given the cumulative uncertainties (Contrarian)." ], "revision_guidance": [], "critical_path": "The prior demonstration, in Phase 1, that the standalone numerical model can predict the mushy zone profile with a reasonable error (<25% RMSE) against quench data, coupled with a rigorous analysis of the observability of the system with the proposed sensor suite. Without this foundational validation, the integration of the EnKF risks being a \"technological solutionism\" applied to a poorly constrained problem.", "final_recommendation": "The panel recommends publication of the brief, recognising the intellectual merit and potential of the hypothesis, but with substantial reservations. Continuation of the project is conditional upon strict execution of Phase 1, which must serve as a decisive test of the fundamental premises (model fidelity and observability). Subsequent phases will be required to incorporate methodological robustness requirements (power analysis, standardised measurement protocol) and to rely on a consortium including a pilot industrial partner. The emphasis must remain on incremental validation and quantification of uncertainties at each stage." } } }