What In-Situ Fiber Optic Strain Tracking Reveals About Residual Stress Evolution During Non-Isothermal Cure Cycles

This overview reflects widely shared professional practices as of May 2026; verify critical details against current official guidance where applicable.

Introduction: The Hidden Cost of Non-Isothermal Cure

In high-performance composite manufacturing, residual stress is the silent adversary. It warps wings, delaminates wind turbine blades, and microcracks automotive monocoques. For experienced process engineers, the challenge intensifies under non-isothermal cure cycles—where temperature gradients across the part create uneven cure kinetics and stress fields. Traditional post-cure inspection (e.g., ultrasonic C-scan, coordinate measuring machine) only reveals the aftermath, not the evolution. This leaves engineers blind to the dynamic interplay between chemical shrinkage, thermal expansion, and tool-part interaction during the cure itself. In-situ fiber optic strain tracking (FOST) changes that. By embedding optical fibers inside the layup, engineers can monitor strain in real time from gelation through vitrification and cooldown. This article unpacks what FOST reveals about residual stress evolution during non-isothermal cure cycles and how to leverage that data for robust process design.

Why Non-Isothermal Cycles Are a Stress Hotspot

Non-isothermal cycles—common in thick parts, tooling with thermal inertia, or out-of-autoclave processes—create a spatial gradient in degree of cure. The leading edge of the part cures faster than the trailing edge, setting up a time-dependent stiffness gradient. As the faster-curing region shrinks chemically while the slower region is still compliant, internal stresses lock in. Later, during cooldown, the coefficient of thermal expansion (CTE) mismatch between the composite and tool adds another layer. FOST captures these events in sequence, revealing when and where stress builds.

What This Guide Covers

We will walk through the core physics of FOST, practical embedding and data acquisition workflows, tool selection criteria (FBG vs. Rayleigh scattering vs. Brillouin), and how to interpret strain signatures for process optimization. We also discuss common mistakes—like fiber orientation mismatches or neglecting temperature compensation—and provide a decision framework for adopting FOST in production. By the end, you will understand how to convert real-time strain data into actionable cure cycle adjustments that reduce residual stress by up to 30%.

Core Physics: How Fiber Optic Strain Tracking Captures Stress Evolution

Fiber optic strain tracking relies on the principle that light traveling through an optical fiber changes its properties when the fiber is mechanically strained or thermally shifted. The most common implementation for composite cure monitoring uses fiber Bragg gratings (FBGs)—periodic refractive index modulations imprinted in the fiber core. When the fiber is strained, the Bragg wavelength shifts proportionally to the strain. By measuring this shift with an interrogator (typically an optical spectrum analyzer or a swept-source laser), engineers can resolve microstrain at multiple points along a single fiber. For non-isothermal cure cycles, where temperature and strain vary simultaneously, temperature compensation is essential. A separate, unstrained reference fiber (or a pair of FBGs with different sensitivities) isolates the mechanical strain component.

From Strain to Stress: The Constitutive Link

Strain alone is not stress. To infer residual stress, the engineer must combine strain data with the evolving mechanical properties of the curing resin—namely, the modulus and the coefficient of thermal expansion. These are functions of degree of cure and temperature. A common approach uses a cure kinetics model (e.g., Kamal-Sourour or autocatalytic) to compute degree of cure from the exotherm, then feeds that into a viscoelastic constitutive model. The measured strain is then decomposed into thermal strain (from CTE) and mechanical strain (from chemical shrinkage and external constraints). The mechanical strain, multiplied by the instantaneous modulus, yields the stress. This coupling is why FOST data is most powerful when combined with dielectric or DSC data for cure state verification.

What the Strain Signal Tells You

During a typical non-isothermal cycle, the strain signal shows distinctive phases. First, a slight compressive strain as the resin heats and expands before gelation. Then a sharp tensile spike as chemical shrinkage begins while the resin is still rubbery—this is the critical window for stress buildup. After vitrification, the strain plateaus or slowly increases as the glassy composite cools. If the part is constrained by the tool (e.g., aluminum or steel mold), a large tensile strain develops during cooldown due to CTE mismatch. FOST reveals not just the magnitude but the spatial variation of these strain events. For example, a 2 m long fiber with 10 FBGs can show that the leading edge of a thick laminate experiences 200 microstrain more shrinkage than the trailing edge, pointing to an asymmetric cure front.

Implementation Workflow: Embedding, Data Acquisition, and Analysis

Successful FOST deployment requires careful planning across three phases: sensor integration, data acquisition, and signal interpretation. Here is a step-by-step workflow based on production-level experience.

Step 1: Sensor Selection and Placement

Choose between FBG arrays (multiplexed on a single fiber) or distributed Rayleigh scattering sensors. FBGs are cost-effective for discrete points; Rayleigh offers continuous strain profiles but requires higher-end interrogators. Embed the fiber between plies, typically at the mid-plane or near the tool surface where stress gradients are largest. Use a polyimide-coated fiber for compatibility with cure temperatures up to 250°C. Avoid sharp bends—minimum bend radius 5 mm—to prevent signal loss. For non-isothermal cycles, place at least two fibers: one inside the part and one on the tool surface to measure tool-part interaction.

Step 2: Temperature Compensation Setup

Without temperature compensation, the strain signal is dominated by thermal effects. The standard method uses a reference FBG placed in a loose tube that sees the same temperature but no mechanical strain. Subtract the reference wavelength shift from the active FBG to isolate mechanical strain. A more advanced method uses dual-parameter sensing (FBG with different cladding diameters) to simultaneously measure temperature and strain. Calibrate the temperature sensitivity of the reference fiber using a thermocouple embedded nearby. For non-isothermal cycles with fast ramps (5°C/min), ensure the interrogator sampling rate is at least 1 Hz to capture transient strain peaks.

Step 3: Data Acquisition and Real-Time Monitoring

Use a commercial interrogator with a wavelength resolution of 1 pm or better (1 pm corresponds to ~1 microstrain). Connect the fiber via a patch cable and record wavelength data throughout the cure cycle. Many interrogators offer software that plots strain vs. time in real time. On a typical project, the team monitors three key milestones: gelation (sudden strain rise), vitrification (strain rate change), and cooldown (linear thermal strain). Any unexpected strain spike above 500 microstrain during the rubbery phase may indicate incipient microcracking.

Step 4: Post-Processing and Stress Calculation

After cure, export the wavelength-time data. Apply temperature compensation using the reference fiber. Convert wavelength shift to strain using the gauge factor (typically 1.2 pm/microstrain for FBG at 1550 nm). Then, using a cure kinetics model, compute the degree of cure and the instantaneous modulus. Multiply the mechanical strain by the modulus to obtain stress. For a thick laminate, the stress profile through the thickness can be reconstructed if multiple FBGs are placed at different depths. Compare the computed stress with the material's strength to assess cracking risk.

Tools, Stack, and Economics: Choosing the Right FOST System

The choice of FOST hardware and software significantly impacts data quality, cost, and scalability. Below we compare three common approaches: FBG arrays, Rayleigh distributed sensing, and Brillouin optical time-domain analysis (BOTDA). Each has distinct trade-offs for non-isothermal cure monitoring.

Which System Fits Your Application?

For most aerospace and automotive cure monitoring, FBG arrays offer the best balance of cost, resolution, and thermal range. A typical setup with 8 FBGs per fiber and 4 fibers yields 32 measurement points—sufficient to capture stress gradients in a complex part. Rayleigh systems excel when continuous strain profiles are needed (e.g., to detect edge effects), but their higher cost and lower temperature tolerance (standard polyimide fiber limits to 200°C) may restrict use. Brillouin systems are rarely used for cure monitoring due to poor spatial resolution (1–5 m) and high strain noise, but they can be useful for monitoring large tool structures.

Software Integration and Data Pipelines

Beyond hardware, the software stack matters. Many interrogators come with basic data acquisition software that exports CSV files. For advanced analysis, teams use MATLAB or Python scripts to apply temperature compensation, run cure kinetics models, and compute stress. Some vendors offer turnkey packages that combine interrogator, compensation algorithms, and stress visualization. Open-source libraries (e.g., Pyfiber) can reduce development time. Consider total cost of ownership: interrogator + fibers + software licenses + training. A full FBG setup for a pilot line typically costs $25k–$50k, while a Rayleigh system can exceed $100k. For high-volume production, the cost per part may still be prohibitive, but for R&D and process optimization, the investment pays for itself through reduced scrap and rework.

Growth Mechanics: Using FOST Data to Optimize Cure Cycles for Reduced Residual Stress

Once you have reliable strain data, the next step is to use it to adjust the cure cycle parameters—heating rate, dwell temperature, and cooling rate—to minimize residual stress. This is where FOST transitions from a diagnostic tool to a process control feedback system. The key growth mechanism is iterative cycle refinement: run a baseline cycle with FOST, identify stress hotspots, modify the cycle, and verify improvement in the next run.

Identifying Stress Hotspots from Strain Signatures

In a typical non-isothermal cycle, the strain-time curve shows a peak during the chemical shrinkage phase. If that peak exceeds the matrix's strain-to-failure at the corresponding degree of cure, microcracks form. FOST reveals the exact moment and magnitude. For example, in a 10 mm thick carbon/epoxy laminate cured with a 2°C/min ramp, the leading edge (near the tool) may show a 300 microstrain tensile spike at 80% degree of cure, while the trailing edge shows only 150 microstrain. This indicates a cure gradient that should be reduced by lowering the ramp rate or adding a hold step. By correlating the strain peak with the degree of cure from a dielectric sensor, you can set a maximum allowable strain rate (e.g., 50 microstrain/min) to keep stress below the critical threshold.

Cycle Adjustments Based on FOST Feedback

Four common adjustments emerge from FOST data: (1) Reduce heating rate when the strain rate exceeds a threshold—slower ramp gives the resin more time to relax viscoelastically. (2) Add an intermediate dwell at a temperature where chemical shrinkage is nearly complete but the resin is still rubbery, allowing stress relaxation before vitrification. (3) Adjust the cooldown rate based on CTE mismatch—a slower cooldown reduces thermal stress. (4) Modify the tool surface (e.g., apply a release agent or use a compliant insert) to reduce tool-part constraint. In one anonymized project, a team reduced residual stress by 25% by adding a 30-minute dwell at 120°C (just above Tg of the partially cured resin) after observing that 70% of the stress accumulated during the final ramp. The FOST data guided the dwell duration: they held until the strain rate dropped below 10 microstrain/min.

Scaling to Production: Real-Time Feedback Loops

In a production environment, FOST can be integrated into a closed-loop control system. The interrogator feeds strain data to a PLC that adjusts the oven temperature setpoints in real time. For instance, if the strain rate exceeds a safety threshold, the ramp is paused. This requires robust algorithms and fail-safe modes. While still emerging, this approach has been demonstrated in research pilot lines for automotive CFRP components. The key challenge is fiber survival—embedded fibers must survive demolding and handling. For production, consider using sacrificial fibers that are removed after cure, or semi-permanent fibers with reinforced coatings.

Risks, Pitfalls, and Mitigations: What Can Go Wrong with FOST

Even with careful planning, FOST deployments encounter pitfalls that can compromise data quality or lead to misinterpretation. Here are the most common issues and how to avoid them.

Pitfall 1: Fiber Breakage During Layup or Cure

Optical fibers are brittle. During layup, sharp edges of carbon fibers or tight corners can fracture the fiber. During cure, thermal expansion of the tool can pinch the fiber if it is routed through a gap. Mitigation: Use a protective sleeve (e.g., PTFE tube) at entry and exit points. Embed the fiber between two layers of resin-rich veil to cushion it. Perform a continuity test (OTDR) before and after layup and after cure. If the fiber breaks, you lose all downstream sensors. Redundancy (multiple fibers) is wise.

Pitfall 2: Temperature-Strain Decoupling Errors

The reference fiber method assumes the reference FBG sees the exact same temperature as the active FBG. In non-isothermal cycles, temperature gradients across the part can cause errors of ±5°C, translating to ±10 microstrain uncertainty. Mitigation: Place the reference fiber as close as possible to the active fiber (within 1 mm). Use a dual-parameter FBG (e.g., with a regenerated grating that has different temperature sensitivity) for direct compensation. Alternatively, use a thermocouple at each FBG location and compute thermal strain using the CTE of the composite (which itself varies with degree of cure). This adds complexity but improves accuracy.

Pitfall 3: Misinterpreting Strain During Gelation and Vitrification

During gelation, the resin transitions from liquid to rubber. The measured strain includes both mechanical strain from chemical shrinkage and hydrostatic pressure from the tool. Novices often attribute the entire strain to shrinkage, overestimating stress. Mitigation: Correlate with degree of cure from DSC or dielectric analysis to separate chemical shrinkage from pressure effects. Use a viscoelastic model that accounts for the evolving bulk modulus. Also, note that the strain signal can appear noisy during gelation due to local stress relaxation events—filter with a moving average (window of 10–30 seconds) to reveal the trend.

Pitfall 4: Cost Overruns and Underutilized Data

Investing in a high-end Rayleigh system but only using it for a few cure cycles leads to poor ROI. Mitigation: Start with an FBG system for initial process development. Reserve Rayleigh for specific investigations (e.g., edge effects or through-thickness gradients). Also, allocate time for data analysis—raw wavelength data is useless without post-processing. Ensure team members are trained in both measurement and interpretation. A common mistake is collecting gigabytes of data but never converting it to stress. Build a standardized analysis pipeline from day one.

Pitfall 5: Ignoring Tool-Part Interaction

The tool constrains the part during cure, creating a biaxial stress state that FOST alone may not capture fully. If the tool is not instrumented, you miss the frictional shear at the interface. Mitigation: Embed a fiber near the tool surface (within 2 mm) to measure tool-part interaction. Compare the strain on the tool side vs. the free surface. If the tool-side strain is significantly higher, consider reducing tool constraint (e.g., using a release film with lower friction) or adjusting the cycle to allow more relaxation.

Mini-FAQ: Common Questions from Experienced Practitioners

Based on discussions with composites process engineers, here are answers to the most frequent questions about FOST for residual stress monitoring.

Can I use FOST in an autoclave environment?

Yes, but with precautions. Autoclaves subject fibers to high temperature and pressure (up to 200°C and 7 bar). Standard polyimide-coated fibers withstand these conditions, but the feedthrough must be sealed to prevent pressure leaks. Use a high-pressure gland designed for optical fibers. The interrogator must be placed outside the autoclave; a long patch cable (up to 50 m) connects to the fiber inside. Pressure itself does not affect the FBG wavelength significantly (pressure sensitivity is ~0.01 pm/bar), so compensation is usually unnecessary.

How many FBGs do I need per fiber, and what spacing?

Typical FBG arrays have 2–20 gratings per fiber, spaced 2–50 cm apart. For a 300 mm long part, 8 FBGs at 40 mm spacing provides good resolution. Too many FBGs (>20) can cause overlapping spectra if the strain range is large (e.g., ±3000 microstrain). Use a wide spectral bandwidth (e.g., 40 nm) and ensure each grating has a unique central wavelength. In practice, 8–12 FBGs per fiber is a safe starting point.

What is the minimum detectable stress change?

Assuming a modulus of 10 GPa and a strain resolution of 1 microstrain, the stress resolution is 10 kPa. For most applications, 10–50 kPa resolution is sufficient to track stress evolution. If higher resolution is needed (e.g., for very low-stress processes like optical lenses), use Rayleigh sensing with 0.1 microstrain resolution.

How do I calibrate the FBG to strain?

Standard FBGs have a strain sensitivity of ~1.2 pm/microstrain at 1550 nm. However, the exact gauge factor depends on the fiber's photoelastic coefficient. Calibrate by applying a known strain (e.g., using a cantilever beam with a strain gauge) and measuring the wavelength shift. Most fiber vendors supply a calibration certificate. For temperature compensation, measure the temperature sensitivity of the unstrained fiber (typically 10 pm/°C) using a thermocouple.

Can FOST detect microcracking in real time?

Yes, but only if the fiber is near the crack location. When a microcrack forms, it releases energy that causes a sudden strain jump (positive or negative) of 50–500 microstrain, followed by a relaxation. Distributed Rayleigh sensing can locate the crack within millimeters. FBG arrays may miss cracks that occur between gratings. For crack detection, use a high sampling rate (≥10 Hz) and monitor for abrupt events. However, be aware that fiber breakage can mimic a crack signal; cross-check with acoustic emission if possible.

Synthesis and Next Actions: Turning Insights into Process Improvements

In-situ fiber optic strain tracking is a powerful technique for revealing the hidden dynamics of residual stress evolution during non-isothermal cure cycles. By providing real-time, spatially resolved strain data, it enables engineers to identify stress hotspots, optimize cycle parameters, and reduce warpage and microcracking. The key takeaways from this guide are: (1) FOST, particularly FBG arrays, offers a cost-effective way to monitor strain from gelation through cooldown. (2) Temperature compensation is non-negotiable for accurate stress inference; use a reference fiber or dual-parameter sensors. (3) Combine strain data with cure kinetics and a viscoelastic model to compute stress, not just strain. (4) Common pitfalls—fiber breakage, decoupling errors, and misinterpretation—can be mitigated with careful design and training. (5) The economic case for FOST strengthens when used iteratively: each cycle refinement reduces scrap and rework costs, often paying for the equipment within a year.

Your Action Plan

To get started, follow this three-phase approach: Phase 1 (Pilot): Select one representative part with a history of warpage. Embed two FBG fibers (8 gratings each) at critical locations. Run a baseline cycle and record strain data. Apply temperature compensation and compute stress using a simple model (e.g., assuming constant modulus for a first estimate). Identify the peak stress point. Phase 2 (Optimization): Based on the strain signature, modify the cycle—e.g., reduce the ramp rate by 50% during the chemical shrinkage window. Run the modified cycle and compare the stress profile. Iterate until the peak stress is below your material's strength. Phase 3 (Transition): Once the optimized cycle is validated, develop a standard operating procedure for FOST deployment on new parts. Train your team on fiber embedding, data acquisition, and analysis. Consider integrating FOST into a closed-loop system for automated process control.

Limitations and Next Frontiers

FOST is not a silver bullet. It requires careful setup, and the stress calculation depends on accurate material models. For thick parts (>20 mm), three-dimensional stress states may require multi-axial strain measurement (e.g., using fibers oriented at 0°/45°/90°). Emerging technologies like distributed Rayleigh sensing with femtosecond-laser-written FBGs promise higher resolution and temperature tolerance. Additionally, combining FOST with other in-situ sensors (dielectric, ultrasonic, or acoustic emission) provides a more complete picture. As the composites industry moves toward Industry 4.0, FOST will play a key role in digital twins and real-time process control.

By embracing in-situ fiber optic strain tracking, you transform the cure cycle from a black box into a transparent, data-driven process. The result: higher quality parts, fewer rejects, and a deeper understanding of the material behavior that governs performance.