Multi-Point Optical Signal Separation for Localized Anode Strain Resolution in Fast-Charged Pouch Cells

Swell

Pouch cells charged at 3C or higher undergo severe internal physical stress. As lithium intercalates into high-capacity host matrices, the cell stack expands across its planar dimensions. In silicon-graphite composite anodes, this local volume expansion spikes well past the macro-scale swell visible externally.

Current density concentrates along electrode boundaries adjacent to collector tabs, where current crowding creates steep state-of-charge gradients and sudden dimensional shifts that generate localized shear forces inside the sealed pouch.

Global pouch expansion is typically measured with external displacement transducers, surface strain gauges, or laser profilometers. Because these instruments average displacement across broad surfaces, they miss the micro-scale strain spikes preceding failure. Local expansion gradients degrade structure, strip active material off copper current collectors, and induce separator creep.

Non-uniform strain patterns also directly trigger localized lithium deposition; once plating begins at a high-strain hotspot, dendrites quickly penetrate separator pores to cause internal shorts and thermal runaway.

Embedding optical fiber sensors inside a pouch cell offers a non-conductive, chemically inert way to monitor internal structural changes. Fibers containing Bragg gratings or Rayleigh backscattering systems can resolve micro-strain variations down to millimeter spatial scales. However, placing optical fibers inside an active electrochemical cell complicates signal interpretation.

The fiber responds to both mechanical strain and temperature. During fast charging, local ohmic heating and reaction entropy produce steep thermal gradients across the pouch plane. Without compensation, a spectral shift can easily mask a 12 degree Celsius temperature rise as mechanical deformation.

Resolving localized dimensional changes requires separating stress-induced wavelength shifts from thermal effects. Standard single-point readings conflate both into an ambiguous wavelength shift. At high charge rates, rapid heat buildup allows the thermo-optic effect to dominate the fiber’s response.

During C-rate escalation testing, spectral drift reaches 1.4 nanometers per percent of local silicon expansion, demonstrating how easily thermal signals mask strain data. Without multi-point signal separation, optical monitoring cannot supply reliable data for battery management system safety cutoffs.

Mechanical Strain Mapping inside Hermetic Pouch Enclosures

Internal deformation profiles differ sharply from external surface measurements. Flexible aluminum-laminated packaging bulges from gas generation and fluid pressure, smoothing out the sharp physical gradients within the electrode stack. Meanwhile, active anode layers undergo distinct structural phase transitions as lithium inserts.

In silicon-graphite blends, silicon particles expand up to 300 percent by volume at full lithiation, compared to roughly 10 percent for graphite up to the stage-one LiC6 structure. This volumetric mismatch builds concentrated stress pockets at particle boundaries.

Single-point displacement sensors on external pouch packaging yield averaged metrics that hide destructive internal stress concentrations.

Placing optical fibers directly along the anode surface captures particle-level mechanics long before swelling shows on the exterior. Placement requires precision: positioned between separator and anode material, the fiber must remain mechanically coupled to the electrode matrix without puncturing the polyolefin coating. Strain transfers from the lithiated host through the fiber’s polymeric coating to its silica core, with transfer efficiency governed by the coating’s shear modulus.

Polyimide coatings applied at 15 to 25 micrometers thick provide sufficient shear modulus to maintain linear strain transfer up to 15,000 micro-strain, whereas acrylate coatings deform under cyclic compression and distort optical signals.

Integrating optical fibers introduces potential leak paths along the pouch seal. Thermal impulse sealing must form a gas-tight bond around the silica cladding without crushing the optical waveguide or causing micro-bending loss. Fluoropolymer sealants at the exit point preserve hermetic integrity while dampening stress concentrations.

Although external compression plates are often assumed to suppress internal strain gradients entirely and render internal optical monitoring unnecessary for safety certification, physical teardowns and distributed strain maps reveal that compression plates merely convert out-of-plane buckling into internal planar compression, increasing shear stress at tab junctions.

Fringe

Reflected signals from multiplexed optical sensors form interference fringes across the detector array. During fast charging, localized thermal expansion and mechanical strain shift these fringes through optical frequency space. Isolating mechanical strain from thermal effects requires multi-point signal separation.

For a single Bragg grating, the response depends on its fundamental reflection wavelength, equal to twice the effective refractive index of the fiber core multiplied by the grating period.

Environmental changes alter both the effective refractive index and grating period. Applied mechanical strain stretches or compresses grating pitch while modifying refractive index via the photo-elastic effect. Temperature changes expand the silica core and alter the refractive index through thermo-optic response.

In standard single-mode silica fiber at 1550 nanometers, strain sensitivity is roughly 1.2 picometers per micro-strain, whereas temperature sensitivity is about 10.5 picometers per degree Celsius. In fast-charged cells, thermal effects dominate the total shift unless separated mathematically.

Deconvolution Mathematics for Decoupled Strain and Temperature Fields

Separating these environmental variables requires a multi-parameter sensor matrix. Two distinct sensing elements exposed to the same local conditions yield a solvable linear equation system. By using a dual-sensor node with differing sensitivity coefficients, matrix inversion can isolate strain from thermal shifts.

A hybrid node combining a Fiber Bragg Grating with an Extrinsic Fabry-Perot Interferometer provides one practical design, since an athermic Fabry-Perot cavity length shows near-zero temperature sensitivity.

Mathematically, the total optical wavelength shift vector correlates to the product of the sensitivity matrix and the environmental state vector. The system matrix takes the form:

Delta_Lambda_1 = K_strain_1 Strain + K_temp_1 Delta_T

Delta_Lambda_2 = K_strain_2 Strain + K_temp_2 Delta_T

Solving for strain and temperature requires a non-zero matrix determinant. Matrix conditioning dictates system susceptibility to optical measurement noise, so picking elements with distinct sensitivity ratios limits noise amplification during inversion. High-definition Optical Frequency Domain Reflectometry (OFDR) measures Rayleigh backscattering along a continuous fiber to reach spatial resolutions down to 20 micrometers.

However, fast charging generates steep thermal gradients above 15 degrees Celsius per millimeter near tab connections, causing continuous spectral broadening of Rayleigh patterns that degrades standard cross-correlation tracking.

Signal processing must filter structural crosstalk caused by fiber curvature and pouch swelling. Micro-bending from surface roughness on the compressed active layer introduces attenuation spikes, dropping the optical signal-to-noise ratio, broadening reflection peaks, and adding phase jitter in interferometric interrogators. Spatial frequency domain filtering isolates structural strain profiles from this high-frequency surface noise.

A dual-sensor optical node achieves matrix conditioning suitable for strain resolution only when the ratio of temperature-to-strain sensitivity coefficients between the two elements differs by more than 25 percent.

Crosstalk between sensor sites along a single fiber caps multiplexing density. Wavelength Division Multiplexing (WDM) assigns dedicated spectral windows to individual Fiber Bragg Gratings. Under severe local expansion, strain can exceed 4,000 micro-strain, shifting a reflection peak by nearly 5 nanometers.

If that shift crosses into an adjacent channel’s window, overlapping signals corrupt the demultiplexing algorithm. Guard bands between Bragg wavelengths must account for both maximum operational strain and peak thermal expansion during potential thermal runaway.

Optical signal separation modes fail under identifiable stress conditions during high C-rate operation:

• Spectral Window Overlap occurs when localized strain shifts a Bragg reflection peak beyond its allocated bandwidth into an adjacent sensor channel.
• Polarization Mode Dispersion arises from asymmetric transverse compression on the optical fiber cladding, splitting a single reflection peak into dual orthogonal polarization states.
• Micro-Bending Loss Attenuation drops reflected optical power below detector threshold limits when electrode particle movement pinches the optical cladding against the copper substrate.
• Thermo-Optic Phase Decoupling Failure develops under steep spatial thermal gradients where the sensor length experiences non-uniform heating across its optical length.

Modern optical interrogators use polarization-maintaining fibers and fast spatial sweeps to reduce polarization mode dispersion. Transverse mechanical loads induce birefringence, splitting the fundamental mode into fast and slow axes. Uncorrected, this birefringence produces false strain signals that algorithms misread as axial deformation.

Tracking both polarization modes independently resolves transverse stress components, yielding a complete 2D strain map across the anode surface.

Running multi-point signal separation in real-time control systems adds computational delay. Processing complex wavelet transforms requires significant computing power, raising control unit hardware costs. Low-latency state-space estimators bypass full spectral deconvolution by predicting thermal paths through electrochemical heat generation models, relying on optical measurements mainly to update states.

This hybrid approach cuts processing latency below one millisecond per channel while preserving micro-strain accuracy under fast-charging conditions.

Gas evolution creates local pockets along the embedded fiber path that introduce signal artifacts. An optical separation algorithm must distinguish whether strain relaxation stems from gas pressure forcing active material apart or from genuine mechanical stress relief in the electrode.

Plating

Lithium deposition on anode surfaces represents a key failure mechanism in fast-charged pouch cells. When local overpotentials fall below zero volts relative to lithium reference potential, metallic lithium precipitates onto the graphite surface instead of intercalating into the host lattice. This plating alters local electrode dimensions; early deposition forms a porous, high-volume layer that exerts sharp compressive forces against the separator.

Detecting these mechanical signatures requires high spatial strain resolution across the anode plane.

Lithium intercalation into graphite causes step-wise expansion corresponding to phase changes between intercalation stages. The transition from Stage 2L to Stage 1 creates a distinct expansion signature at roughly 50 percent state of charge. Metallic lithium plating alters this volumetric expansion rate: because lithium accumulates on particle surfaces rather than filling lattice vacancies, local thickness growth accelerates relative to normal intercalation.

Multi-point optical strain sensors pick up this rate inflection long before voltage signals register metallic deposition.

What Wavelength Separation Threshold Prevents Thermal Signal Bleed?

Detecting the rapid expansion caused by early metallic deposition demands strict wavelength separation between channels. Charging at 4C induces Joule heating that causes thermal spectral drift, masking subtle strain acceleration from initial plating. Maintaining at least 8 nanometers of channel separation between adjacent Bragg sensors prevents signal bleed during temperature rises up to 60 degrees Celsius.

Paired with polarization-maintaining dual-core fibers, this separation allows the system to resolve 5 micro-strain shifts within sub-second intervals, providing the resolution required for dynamic current throttling.

Blending silicon into graphite anodes alters the baseline strain response. Silicon expands continuously throughout charge, producing localized strain up to 10 times higher than graphite intercalation alone. Separating silicon swelling from localized lithium deposition requires evaluating spatial strain variance across neighboring sensor nodes.

Uniform electrode formulations expand evenly during silicon lithiation, whereas lithium plating accumulates at high-current regions near tab welds or electrode edges. Elevated spatial variance across a 5-millimeter sensor grid indicates non-uniform metallic deposition.

Structural integration of optical sensors during pouch cell manufacturing requires controlled physical protocols to prevent cell damage or optical signal distortion. The embedded fiber must lie completely flat against the anode active layer before electrolyte injection and vacuum sealing.

1. Align the coated optical fiber along the designated anode grid line using an automated optical alignment fixture.
2. Apply local tack welds using a polyacrylic ester binder to fix the fiber position without blocking electrolyte diffusion paths.
3. Guide the fiber ends through pre-formed fluoropolymer sleeve grommets positioned in the pouch seal allowance zone.
4. Perform z-fold or stack assembly while maintaining continuous optical continuity verification via back-reflection monitoring.
5. Heat-seal the pouch perimeter under partial vacuum using a dual-stage thermal bar set to 185 degrees Celsius for 3.5 seconds.

Electrolyte wetting changes mechanical strain coupling. Dry electrode stacks behave stiflly under compression, but once saturated with organic carbonate solvents, polymeric binders swell and active material matrices soften, altering strain transfer to the optical fiber coating. Consequently, fiber calibration parameters must be measured in an electrolyte-saturated state rather than dry.

Baseline strain shifts also occur during the first five formation cycles as the solid electrolyte interphase (SEI) forms and internal stack pressure settles.

Monitoring during high C-rate cycling reveals distinct strain signatures for different degradation mechanisms. Normal intercalation shows a smooth, reversible strain curve following state-of-charge. Inhomogeneous SOC distributions cause localized strain hysteresis, where charge expansion diverges from discharge contraction.

Metallic plating produces a permanent strain offset: plated lithium that reacts with electrolyte forms dead lithium on the anode surface, leaving a measurable thickness increase even at zero percent state of charge.

Stress relaxation during rest periods provides further diagnostic insight. Pausing charge current at 80 percent state of charge triggers thermal relaxation alongside mechanical relaxation as intercalated lithium redistributes through graphite particles. When plating occurs during charging, stress relaxation exhibits a slow, multi-stage decay as metallic lithium chemically re-intercalates into the underlying graphite structure.

Multi-point optical arrays capture this spatial relaxation rate, enabling real-time estimation of reversible intercalated versus irreversible plated lithium.

An irreversible structural strain baseline offset measured at zero percent state of charge serves as an unambiguous physical indicator of accumulated dead lithium and SEI thickening.

A basic constraint governs localized strain resolution: if sensor spacing exceeds electrode tab width, strain spikes from current crowding go undetected. Current density spikes near tabs decay exponentially with distance from the weld. An array spaced at 20-millimeter intervals misses 3-millimeter strain surges directly beneath the collector foil.

Detecting early plating during fast charging requires high-density arrays or continuous Rayleigh scattering.

Cell design directly affects strain monitoring feasibility. Ultra-thick electrodes exceeding 100 micrometers per side suffer severe lithium-ion diffusion limits across their depth, encouraging surface plating during fast charge. Fibers placed on the surface of thick electrodes respond mainly to surface expansion, missing internal shear stress near the copper current collector.

Integrating dual-depth optical fibers ~ at both the current collector interface and the electrode surface ~ resolves these 3D strain gradients across electrode thickness.

Proof

Validating fast-charging pouch cells equipped with optical fibers requires strict compliance with international transport safety codes. Adding optical waveguides to a sealed pouch cell is considered a material design change under UN 38.3 transport testing frameworks. Modifying internal mechanical boundaries, changing sealing gaskets, or adding external optical connectors requires complete re-qualification before cells can be shipped via commercial air or sea freight.

UN 38.3 requires eight physical and thermal tests to certify shipping safety. Embedded optical fibers directly affect cell behavior during mechanical testing. Under UN 38.3 Test T.3 (Vibration) and Test T.4 (Shock), internal fibers undergo high-frequency vibration and accelerations up to 150 G. If the fiber coating lacks sufficient fatigue resistance, the fiber can snap, leaving a sharp glass fragment inside the stack that risks puncturing the polyolefin separator and causing an internal short circuit during electrical evaluation.

Safety certification documentation must match the physical design of production units. Testing a baseline cell without fibers does not qualify an instrumented variant for transport. European Battery Regulation (EU) 2023/1542 sets performance, durability, and health-tracking requirements; instrumented cells supplying real-time strain and thermal data to a Battery Management System must undergo testing to prove embedded sensors do not shorten calendar or cycle life.

Assembling a complete safety dossier for customs and regulatory approval requires specific engineering and compliance documentation:

• UN 38.3 Test Summary Report covering the specific instrumented cell part number with documented optical feed-through sealing specifications.
• Helium Leak Rate Certificates proving pouch feed-through hermeticity below 1 times 10 to the power of minus 6 millibar-liters per second after thermal shock cycling.
• Chemical Compatibility Dossier demonstrating zero degradation of optical fiber coatings when exposed to alkyl carbonate solvents and lithium hexafluorophosphate salts.
• IEC 62133-2 Type Approval Certificate issued by an accredited ISO 17025 laboratory confirming internal short-circuit safety under mechanical stress.

IEC 62133-2 Clause 7.3.9 outlines forced internal short-circuit testing, using an L-shaped nickel particle pressed into the active layer to measure puncture resistance. With optical fibers inside the cell, the fiber acts as a rigid element under compression. Testing must confirm that external loads perpendicular to the cell plane do not drive the fiber through the separator, inducing a false short circuit during type approval.

Environmental durability testing verifies long-term chemical compatibility. Fiber coatings must withstand exposure to electrolyte solutions containing lithium hexafluorophosphate (LiPF6) and trace hydrofluoric acid (HF) from moisture ingress. Polyimide-clad silica glass resists HF corrosion at room temperature, but high temperatures during fast charging accelerate chemical attack on exposed fiber ends.

Uncoated tips or damaged coatings suffer etching, creating micro-cracks that grow under strain and break the fiber.

Helium leak testing conducted per IEC 60068-2-17 Clause 4.2 confirms that fluoropolymer pouch feed-through bushings prevent electrolyte solvent evaporation over 10-year operational life targets.

Liabilities escalate if cell alterations invalidate dangerous goods declarations. Unauthorized cell variants risk seizure at ports alongside heavy fines for the shipper. Under maritime dangerous goods rules (IMDG Code Special Provision 188), small lithium cells are exempt from certain Class 9 transport restrictions only when backed by UN 38.3 test summaries covering all structural design changes.

Contractual agreements between cell buyers and original equipment manufacturers must define precisely who holds liability for embedded sensor failures. A contract clause governing optical sensor qualification reads as follows:

“The Cell Supplier warrants that embedded optical fiber sensing components do not alter the UN 38.3 safety certification status of the primary cell, and the Supplier assumes full liability for all carrier rejection costs, freight demurrage, and regulatory fines resulting from hermetic seal failures at optical feed-through points during transport.”

This language assigns financial exposure to the manufacturer, ensuring engineering modifications for optical signal separation pass safety audits before volume shipping begins.

Yield

Integrating optical fibers into high-speed assembly alters cell manufacturing economics. Production lines operating above 30 ppm incur cycle-time penalties when placing fiber components during stacking or folding. Handling delicate waveguides between 80 and 125 micrometers in diameter demands specialized robotic pick-and-place systems, raising assembly line capital costs.

Manual placement introduces human error, increasing scrap during initial production ramps.

Yield losses during fiber integration stem from three main failure modes: fiber breakage during automated stacking, connection loss during heat sealing, and spectral calibration failures during electrolyte wetting. A single broken fiber inside a stack forces the rejection of the entire unsealed assembly. With material costs averaging $45 per cell for a 60 Ah fast-charge automotive pouch, a 3 percent drop in stacking yield adds more than $1.35 in scrap cost per finished cell.

Interrogation hardware adds significantly to installed equipment costs. Industrial optical interrogators built with tunable lasers and high-speed array detectors run between $8,000 and $25,000 per unit. Laboratory settings tolerate these costs, but commercial deployment in vehicle packs or stationary storage requires low-cost ASIC interrogators.

Photonic Integrated Circuits (PICs) on silicon nitride platforms consolidate lasers, splitters, and spectrometers onto a single chip, pushing target interrogator hardware costs down toward $150 per pack system.

Cell sorting and grading routines must track baseline strain metrics alongside standard capacity, internal resistance, and K-value data. During formation, interrogators record spatial expansion signatures across initial charge-discharge cycles. Cells showing spatial strain variance above 15 percent are flagged for grade-B classification, since uneven expansion points to physical misalignments that accelerate degradation under fast charge.

Incorporating strain metrics helps catch latent manufacturing defects before cells move to final pack assembly.

Warranty risk calculations underline the commercial utility of real-time strain monitoring. Accelerated wear from fast charging creates substantial liability for system integrators. Operating a fast-charging pack without localized strain tracking invites premature capacity fade and lithium plating failures, driving warranty replacement claims up to $12,000 per vehicle pack.

Real-time strain monitoring lets control units adjust charge current dynamically when local strain thresholds are met, extending cycle life by up to 40 percent under aggressive 4C charging.

Financial risks multiply if unmonitored local strain causes a safety failure in the field. A single thermal runaway incident triggers recalls, litigation, and brand damage that dwarf cell manufacturing costs. Cutting multi-point optical separation hardware to save minor cell BOM costs exposes integrators to uncapped liability if concentrated local strain triggers an internal short circuit.