Battery Cell Cooling

📖 New to Distributed-Generation Conduction (Classical Heat Transfer)? Read the guide: Battery Cell Cooling: Why Distributed Heat Generation Changes the Math — a plain-language explainer of the standard behind this calculator.

Heat generation is I²·R_internal (Joule heating only — the reversible/entropic term, typically ~10-15% of ohmic heating at moderate-to-high C-rate and sign-dependent on SOC/charge-discharge, is not modelled). Because a cell generates heat throughout its volume rather than at a single point, the internal conduction resistance uses distributed-generation forms, not naive point-source conduction: a radially-cooled cylinder gives R=1/(4π·k·L) (independent of radius — a useful self-check); a one-face-cooled slab gives R=t/(2·k·A); a both-faces-cooled slab gives R=t/(8·k·A) — four times lower than one-face, not two. Cylindrical cells' radial conductivity (~0.2-0.5 W/m·K) is reasonably well-constrained in the literature; axial conductivity is genuinely contested across measurement methods (published range roughly 2-30 W/m·K) and defaulted here toward the higher, more-cited value — disclosed as uncertain, and largely irrelevant for the recommended side-cooling path. Prismatic/pouch through-thickness (~0.5-0.8 W/m·K) and in-plane (~20 W/m·K) values are representative literature figures, not any specific manufacturer's datasheet. Base/top cooling is flagged with a warning — real packs almost always cool cylindrical cells from the can wall (side) and prismatic/pouch cells from their large flat faces (side, in this tool's convention), not end-to-end through the low-conductivity axis. Ambient (natural-convection) cooling reuses this site's existing Churchill-Chu correlations — horizontal cylinder for cylindrical cells, vertical flat plate for prismatic/pouch — solved by bisection on surface temperature.

Validated: the cylindrical radial formula reproduces its hand-derived value exactly and is confirmed independent of cell radius; the slab formula reproduces exactly half the naive point-source resistance when cooled from one face, and exactly one-eighth when cooled from both; a full series-stack case (21700 at 3C, side-cooled to water) reproduces a hand-worked 15°C core-to-coolant rise and correctly identifies internal conduction — not convection — as the dominant resistance, a genuinely useful design insight this tool surfaces directly. No published worked example exists in the literature for this exact end-to-end configuration, so validation here is via these self-consistent hand-derivations rather than a third-party number — verify against a cell datasheet and, for a final design, FEA or measurement.