Pack heat world: stop one bad cell from taking the pack
A 3D pack of 24 cells coloured by temperature. Load it, cool it, then short-circuit one cell and see whether thermal runaway spreads. Change chemistry, spacing, barriers and cooling until it stops at one cell.
How this lab works
Warm the pack with a load
Put a load on the cells and watch the colours. In a sealed box the middle cells run hottest. Try the fan and the liquid plate, and switch the BMS cut-off off and on.
Short-circuit one cell
Tap a cell to choose it, then short-circuit it. The cell heats from the inside until it goes into thermal runaway. The BMS cannot switch that off.
Change the design until it holds
Reset the pack, change one design choice and repeat the test. Every finished test is listed, so you can see which choices stop the fire at one cell.
Pack heat world: stop one bad cell from taking the pack
Voltesta, voltesta.com
This is a model for learning. Never short, heat or damage a real cell to see what happens.
Test the pack
The 3D picture needs WebGL, which is switched off in this browser. The controls, numbers and charts on this page still work.
Drag to turn the pack. Tap a cell to choose it.
Pick a design, add a load if you like, then short-circuit the chosen cell.
Temperature against time
Hottest cellChosen cellPack average
Reading the colours
- The ring marks the chosen cell. Tap another cell to move it.
- A glowing cell with smoke is in thermal runaway. A black cell has burnt out.
- You can change the cooling, the load and the air temperature while the test runs.
Your tests
Short-circuit a cell and let the test finish. Each finished test is listed here so you can compare designs.
| # | Design | Cells lost | Peak temperature | Time to spread |
|---|
Things to try
NMC cells, touching, sealed box: short one cell
The first cell vents in under a minute and its neighbours follow. In about four minutes all 24 cells are lost.
Same pack, but with the liquid plate
Cooling slows the spread and saves most of the pack, but it does not stop cells that touch each other from lighting their neighbours.
Add barrier sheets, or leave a gap between the cells
The shorted cell is still lost, but the fire stops there. Blocking the heat path does more than cooling it away.
Gap between cells, no barriers, sealed box, 45 °C air, 7.5 A load, BMS cut-off off: let the clock reach 10:00 at 100×, then short a cell
A design that held when cool now loses the whole pack. A hot, loaded pack has almost no margin left.
Switch to LFP cells, touching, sealed box
LFP starts runaway at a higher temperature and releases much less energy. The shorted cell is lost, the neighbours survive.
Sealed box, 7.5 A load, BMS cut-off off, clock at 100×
With no cut-off the load alone cooks the pack. Switch the cut-off on and the BMS stops the load at 60 °C and lets it back on at 50 °C. This is the fault the BMS is there for.
How the simulation works
Each cell is treated as one lump with a heat capacity of 41 joules per degree, which is about right for a small 18650 cell. Every second the model adds up the heat going in and out of each cell and moves its temperature by that much.
Heat comes from the load, as current squared times the cell's 30 milliohm resistance. Heat leaves to the surroundings, slowly in a sealed box, faster with a fan, fastest with a liquid plate. Heat also flows between neighbouring cells, easily when they touch and much less across a gap or a barrier sheet.
A short-circuited cell turns its own stored energy into heat at 150 watts. When any cell reaches its onset temperature, 150 °C for NMC and 220 °C for LFP in this model, it goes into runaway: it releases a burst of heat and sends hot gas and flame onto its neighbours. An NMC cell releases about three times the energy of an LFP cell.
The numbers are teaching values, picked so that each design choice shows its effect clearly. They are not test results for any real cell or pack. Real packs are proven by abuse testing to the standards listed below.
Every value here comes from a teaching model running in your browser, with typical figures. They are not measurements from real hardware.
Standards this lab refers to
- AIS-156
- Indian safety rules for traction batteries in L-category vehicles (two- and three-wheelers).
- IS 16893
- Indian Standard for lithium-ion cells that drive electric road vehicles, covering performance and reliability tests.
- UN 38.3
- Tests that lithium cells and batteries must pass before they can be shipped.
Named for learning only. Voltesta does not claim approval or certification from any of these bodies.