By Jae-Won Kim, CTO and Co-Founder
90 days separates a first cell sample from the data sheet an OEM partner needs for a go/no-go decision. Our answer assumes three specimens per layup variant and no process failures requiring a batch re-run. That timeline can seem long beside a conventional cell data sheet from an established manufacturer. For a structural design not yet tested under combined loads, however, 90 days is the minimum time needed for the degradation physics to run.
Days 0 to 14: Electrochemical Baseline
During the first two weeks, we establish the electrochemical baseline before placing a cell in the mechanical fixture. We run 50 formation cycles at C/10 charge and C/5 discharge to form the initial SEI layer. Formation efficiency, the ratio of first discharge capacity to first charge capacity, must exceed 92 percent. A cell below that mark indicates a formation process issue, not a structural cell issue, so we report it to the design team before continuing.
We then run 100 characterization cycles at C/5 charge and C/2 discharge, recording capacity at every cycle. A full electrochemical impedance spectroscopy (EIS) sweep runs at cycles 1, 10, 50, and 100. The 10 mHz to 100 kHz range separates charge transfer resistance, diffusion impedance, and ohmic resistance into distinct arcs. We use the cycle 100 EIS as the reference spectrum for future degradation measurements.
Coulombic efficiency at cycle 50 must exceed 99.8 percent. Failure points to electrolyte decomposition or lithium inventory loss that could cause early capacity fade under structural loads. We stop those cells at the electrochemical gate rather than begin mechanical testing, avoiding three weeks of combined-load fixture time for a design unlikely to pass final screening.
Days 15 to 30: Structural Property Baseline
Once the electrochemical baseline is clear, we characterize structural properties. At this stage, the aim is to measure the unaged cell's structural contribution, not to overload it. We use three-point bending across a 200 mm span on three specimens per batch, measuring flexural modulus and peak flexural strength. A short-beam shear test at 40 mm span measures interlaminar shear strength (ILSS) and probes fiber-to-matrix bonding at the electrode interface.
ILSS must exceed 35 MPa to clear the structural gate. Lower values indicate insufficient epoxy infusion in the composite layup or weak adhesion at the electrode-to-fiber bond interface. These are manufacturing process issues, not chemistry issues. Two of our fourteen completed characterization batches failed ILSS. Both failures traced to surface contamination of the CFC prepreg during assembly, rather than electrode formulation.
At this point, flexural modulus is a reference value, not a pass-fail gate. The OEM's structural design specification sets the requirement, and that requirement varies by application. By day 30, measured modulus must be within 5 percent of the design target declared at program start. Outside that band, we revise layup thickness or fiber orientation before combined-load testing.
Days 31 to 60: Combined Load Characterization
This is the longest phase and cannot be compressed. We run 500 cycles combining electrochemical cycling with mechanical loading, as described in our cycle life testing article. At the standard rate, 0.5C charge and 1C discharge with rest periods, the 500-cycle run takes approximately 18 to 22 days. We also schedule reference performance tests at cycles 100, 200, and 500.
At cycle 200, capacity retention must exceed 94 percent, while EIS high-frequency resistance must not exceed 110 percent of the cycle 1 reference. These early gates indicate whether degradation is tracking toward the cycle 500 requirement. A cell at 93 percent capacity retention at cycle 200 will almost certainly miss that gate. Stopping then avoids occupying the fixture for another three weeks.
The final combined-load gate comes at cycle 500: capacity retention above 90 percent, EIS high-frequency resistance below 120 percent of reference, and no differential voltage kink associated with cathode particle fragmentation. Cells meeting all three criteria become candidates for pilot progression. For any failure, we classify the mode as electrochemical or structural and return it to the design team for the next iteration.
Days 61 to 90: Data Sheet and Go/No-Go
The final 30 days combine longer-duration cycling, post-mortem work on at least one specimen from the combined-load batch, and data sheet compilation. We extend the two best-performing specimens from the cycle 500 gate to 800 cycles, improving projection accuracy at 2,000 and 3,000 cycles. Those additional 300 cycles add approximately 8 days of testing and reduce the cycle-life projection uncertainty band by roughly 40 percent versus a 500-cycle projection alone.
Post-mortem work includes scanning electron microscopy on cross-sections through the electrode-CFC bond interface, EDS mapping for electrolyte migration into the fiber matrix, and mechanical measurement of the aged specimen to quantify structural degradation. The OEM receives two separate data streams: electrochemical degradation over time and structural property change over time. Each supports a different part of the integration design.
The final data sheet includes initial electrochemical and structural properties, formation efficiency, a combined-load cycle life projection with uncertainty bounds, identified failure modes and the cycle count at which they become detectable, plus recommended BMS parameters. These include voltage limits, charge rate limits, and temperature operating range based on the characterization data. The OEM receives the data sheet on day 90 and decides whether to proceed to pilot integration against its application requirements.
Outside the 90-Day Window
This 90-day protocol is a screening process, not a complete qualification pathway. It excludes thermal abuse testing, vibration fatigue, humidity exposure, and accelerated aging at high temperature. Those tests remain necessary for OEM qualification and typically add 12 to 18 months to the full certification timeline. The data sheet gives an OEM enough information to fund the full qualification program, not to ship the cell in a production vehicle.
The protocol also excludes cell-to-cell variation within a production batch. We use hand-fabricated laboratory specimens with tightly controlled layup procedures. Cells from a scaled manufacturing process will have a different variation distribution, and combined-load performance will spread around the characterization mean according to process capability. Production batch variation is a separate program that begins after the design is locked and manufacturing is established.
For a team starting structural cell integration, 90 days to the first data sheet is the minimum calendar time when both electrochemical and structural gates clear. Budget for at least one 4-week iteration if either gate fails and requires a formulation or layup change. This program is also the right point to establish BMS baseline calibration, so the BMS team should join from day 15 rather than wait for the completed data sheet.