This simulator follows heat along a representative liquid-cooled battery thermal path: a cylindrical lithium-ion cell core, its case, a machined cold plate with serpentine coolant channels, a circulating coolant loop with pump and reservoir, and a finned radiator with an axial fan rejecting heat to ambient air.
• A real-time 3D cutaway of the cooling loop — cell module and vented cans, machined liquid cold plate with a visible serpentine flow path, coolant pump and reservoir, finned radiator and shrouded axial fan, plus a coolant inlet sensor and cell temperature sensor — with toggleable enclosure, auto-rotate, exploded view and selectable, labeled components. • Five controls: initial SOC for a new trial, pack current from −6 A (charge) to +12 A (discharge), ambient temperature (−10 to 45°C), coolant flow multiplier (0–3×) and radiator fan multiplier (0–2×). • Live readouts for cell core temperature, cell case temperature, coolant temperature, cell heat generation, radiator heat rejection and cell current. • Time controls: pause/run, advance 1 s, advance 60 s, and a playback selector from real time up to 1 hour per second, plus start/stop actions for the experiment. • An Operating sequence readout, a switch-state panel and a per-cell readings table. • An Analysis tab with temperature and heat-transfer charts, the underlying three-node lumped thermal model equations, and snapshot measurements. • A Tests tab with four guided experiments (normal cooling, pump failure, fan failure, hot day), a Verification bench of automated model checks, and a timestamped event log with report export. • A Learn tab with four lessons (three thermal stores, current produces heat, flow and airflow do different jobs, ambient sets the boundary), a knowledge-check quiz and a written model-scope statement with references.
The model tracks three separate thermal masses — cell core, cell case and coolant — each with its own temperature and its own thermal inertia, so a change in pack current does not show up as an instant temperature jump everywhere at once: Qloss = I²R₀ + Vp²/R₁ generates heat in the core, which conducts to the case, then to the coolant, then is rejected at the radiator. Because these are separate lumped nodes, the case and coolant temperatures visibly lag behind a fast current change while the core responds first.
The coolant flow multiplier and the radiator fan multiplier do different jobs in this chain: flow affects how efficiently the cold plate transfers heat from the case into the moving coolant, while fan speed affects how efficiently the radiator transfers heat from the coolant into moving air. Setting either to zero doesn't stop cooling outright — some heat still moves by weaker natural conduction and convection paths — but it starves that stage of the loop.
This is a passive-radiator cooling system: it has no refrigeration cycle, so it rejects heat by driving a positive temperature difference between the coolant and ambient air. That means the loop can never sustainably hold any node below the ambient temperature at steady state — raising ambient (as in the "Hot day" experiment) shifts every thermal node's baseline upward, and the system can only ever cool toward ambient, not past it.
This is a teaching model using three lumped thermal nodes for a representative cell-to-loop path. The flow and fan multipliers change modeled heat-transfer conductance directly; they do not solve an underlying pressure/flow network. The model excludes refrigeration, entropic (reversible) heat effects, cell aging, coolant boiling and thermal-runaway reaction kinetics.
The simulator models three separate thermal masses — core, case and coolant — each with its own thermal inertia. Heat generated in the core has to conduct through the case and into the coolant before the radiator can reject it, so a sudden current change shows up first at the core and only later, and more slowly, at the case and coolant readouts.
Coolant flow affects heat transfer from the cell case into the circulating coolant at the cold plate. The radiator fan affects heat transfer from the coolant into ambient air at the radiator. They act on different stages of the same loop, which is why the pump-failure and fan-failure experiments produce different symptoms.
No. This is a passive radiator system with no refrigeration cycle, so it can only reject heat by maintaining a positive temperature difference to ambient air. At steady state, none of the modeled thermal nodes can sit below the ambient temperature you set.
This is a teaching model using three lumped thermal nodes for a representative cell-to-cold-plate-to-radiator path. The flow and fan multipliers change modeled heat-transfer conductance rather than solving a real pressure/flow network. It excludes refrigeration, entropic heat effects, cell aging, coolant boiling and thermal-runaway reaction kinetics.