Phase change material inside a sealed enclosure: pressure and venting

Short answer: yes — the melt expansion belongs in the pack's volume budget before you seal anything. Published measurements put phase change material volume change between about 2% and 24%, and only your material's datasheet pins down where in that span it sits. On a 10 L pack filled with PCM, a 5% melt expansion is 0.5 L [30.5 in³] of liquid that has to go somewhere. A fully rigid, completely filled IP67 housing turns that into roughly 750 bar [10,900 psi] of internal pressure. Leave gas headspace of at least twice the expansion — about 1 bar [15 psi] of pressure — or fit an internal diaphragm sized for 1.5 times it, and the seal stays dry.

How much does a phase change material expand when it melts?

Most PCMs change volume by a few percent to about a quarter on melting. A 2020 Journal of Energy Storage survey measured melt expansions from 2.44% to 23.53% across chemistries, so the wax in your pack sits in that span until the datasheet pins it down. The span is wide because the chemistry sets it: hydrated salts, fatty acids and paraffin blends all move differently between solid and liquid.

Paraffin-type waxes, the usual choice beside lithium-ion cells, start melting near 37 °C (99 °F) and boil above 370 °C [700 °F]. A pack working near the top of its allowed temperature range therefore melts and refreezes its PCM on every duty cycle. Contraction on freezing mirrors the expansion, so the volume budget needs the same liters in both directions.

Get two numbers from the vendor: solid density and liquid density at each end of the working range. The measured difference between them is the expansion you size the housing with. Survey values screen materials, and datasheet values size seals.

What does that expansion do inside a rigid, fully filled housing?

A fully filled rigid housing cannot absorb the melt, so the liquid itself compresses. On the example pack that is 0.5 L of trapped expansion, about 750 bar [10,900 psi] of pressure — roughly 7,600 times the 0.1 bar [1.4 psi] water head at the bottom of the IPX7 test. A liquid resists compression stiffly: the bulk modulus of a typical organic liquid is about 1.5 GPa, and water's is 2.2 GPa — textbook values. Assume a rigid wall and a complete fill, and the pressure rise equals the modulus times the expansion fraction: 1.5 GPa × 0.05 = 75 MPa = 750 bar. A real wall bulges a little, so the true figure lands somewhat lower — still hundreds of bar, and still enough to burst a clamped gasket or split a seam.

This is the wax-motor effect: trapped paraffin that melts pushes hard enough to open a car thermostat. A leftover air bubble smaller than the expansion barely helps. A 50 mL [3 in³] void in the 10 L pack still leaves 450 mL of expansion for the liquid itself to absorb, so the pressure stays in the hundreds of bar. The joint that gives way becomes the assembly's real rating. The 5% expansion and the 1.5 GPa modulus here are worked-example inputs; replace both with the values from your PCM datasheet.

Where should the melt expansion go?

Give the melt a gas headspace or an internal diaphragm. Headspace of twice the expansion holds pressure near 1 bar [15 psi], three times holds it near 0.5 bar [7 psi], and a diaphragm sized at 1.5 times the expansion holds it close to zero. The table comes from Boyle's law on the headspace gas: gauge pressure equals atmospheric pressure times expansion divided by headspace minus expansion — 1.01 bar × 0.5 L ÷ 1.0 L = 0.5 bar for the 3× case. Warming the gas from 20 °C to 50 °C adds about 10%, and the fill cooldown puts the housing under suction before the first melt, so the headspace absorbs both.

Decide by orientation. Pick 3× headspace — 1.5 L on the example pack — when the pack always mounts upright and the top space can stay above the cells: simplest, no moving parts. Use the diaphragm when the pack tilts, ships on its side, or the cells need full immersion in wax for thermal contact. A rolling diaphragm or metal bellows inside the sealed volume sweeps 0.75 L at well under 0.1 bar of its own stiffness. Either way the expansion volume stays inside the housing, and nothing crosses the wall. The same gas-law sizing governs rapid temperature swings in air-filled boxes, where the full arithmetic is worked through in the condensation and pressure guide.

Why doesn't a membrane vent solve this?

A membrane vent moves air, and your 0.5 L of melt expansion is wax. A hydrophobic membrane that passes air will clog with warm paraffin or slowly weep it, so the expansion volume has to be absorbed inside the housing and never exchanged with the water outside. Vents belong to air-filled boxes, where a little air crosses to equalize weather and altitude while water stays out. Choosing a pressure-equalization vent is its own sizing task. Your box holds wax at 50–60 °C [122–140 °F] near the top of a pack's allowed range, and a membrane built to repel water droplets wets and clogs with warm hydrocarbon.

Even if such a vent passed its ingress test when new, the assembly rating would then rest on that single part — the weakest part sets the rating, and here the weakest part spends its life coated in wax. Cutting a through-wall hole for any vent also ends the immersion rating unless the vent itself is immersion-rated. For air-filled outdoor boxes the vent-or-sealed choice is a real decision worth working through in the outdoor enclosure guide. With molten material inside, keep the expansion internal and skip the vent entirely.

What does the IP67 rating actually test?

IEC 60529 proves the digit 7 by submerging the enclosure 1 m [3.3 ft] under the surface for 30 minutes — a 0.1 bar [1.4 psi], 9.8 kPa external head. It tests the box as delivered, not the pressure and vacuum cycles your PCM adds in service. The full code and its digits sit in the IP rating chart. The test covers one soak: room-temperature water, 30 minutes, the housing in its as-delivered state. It does not cover thousands of melt–freeze cycles, years of warm wax against the gasket, or a seal that swings from about +0.5 bar outward to about −0.25 bar inward on every cycle with 3× headspace.

The suction stroke is the dangerous one. Negative internal pressure pulls water through a gasket nick, a pinhole or a badly seated fitting, which is how sealed boxes drown in service, so design the joint for both directions. Sealed also means sealed against liquid water, not against water vapor — moisture that rides in on assembly day works inward with every suction stroke.

If the pack sits under water for hours rather than minutes, that is continuous immersion: an IP68 build with depth, duration and fluid agreed between manufacturer and user. In NEMA terms that pairs with Type 6 for temporary submersion and Type 6P for prolonged immersion, as the NEMA and IP comparison sets out. For a battery project weighing IP67 against IP68, the battery-specific tradeoffs are worked through separately.

What else changes the volume inside the pack?

Add the cells' own growth to the same volume budget. Lithium-ion cells swell as they cycle and age, the cell datasheet states the expected growth, and pouch and prismatic formats usually need more allowance than cylindrical ones. Cell growth never reverses and comes out of the headspace permanently — one more reason to start at 3× rather than 2×.

Gas temperature is the second term: the table above assumes the headspace near ambient, and a 30 K rise adds about 10% to the melt pressure (293 K to 323 K). The fault case is the third: a cell that vents is a safety event with its own gas volume and its own path, and the battery system's instructions govern that design separately from this normal-operation budget. Humidity rides along with every pressure swing, so dry the assembly and desiccate the headspace before final closure.

What goes on the drawing and the order?

Put the melt expansion in liters, the headspace or diaphragm volume, and the gasket material on the drawing. Then order the housing as a custom build rated for the immersion case you actually have, because the stock IP66 wall-mount line is not immersion-rated. Work the list in the vendor's units and hand the liters to whoever builds the housing.

• PCM datasheet: melt expansion at your working range; expansion volume = fraction × liquid volume, so 0.05 × 10 L = 0.5 L on the example pack.
• Headspace: minimum 2× the expansion (1.0 L), preferred 3× (1.5 L), kept uppermost in every mounting attitude and marked on the drawing so filling leaves it in place.
• Diaphragm route: swept volume of at least 1.5× the expansion (0.75 L [46 in³]), mounted inside the sealed volume, no through-wall port.
• Gasket: nitrile against wax — the standard choice for oil-like media, good from about −30 to 120 °C [−22 to 248 °F]; move to silicone only if the range must reach −60 to 200 °C [−76 to 392 °F].
• Heat: the PCM flattens temperature spikes, but the steady heat load still needs the usual sealed-box sizing from the enclosure sizing guide before you fix the depth.
• Rating wording: the immersion number on the drawing is a manufacturer rating; ours are manufacturer-rated and not UL listed.
• Material: paraffin is benign to stainless, so 304 or 316L suits the fluid; our stainless range spans outside sizes from 6 x 4 x 4 in to 60 x 36 x 16 in (H x W x D) — see the stainless enclosure lines.
• Order: an immersed or liquid-filled build goes through the custom enclosure route. Send the depth, duration, fluid and penetration list for the design review; feasibility comes back with the drawing, and the finished housing is delivered in 10 business days from your approval of the drawing.

Catalog dimensions to compare

Compare these catalog dimensions with your component layout. Open each enclosure for its rating, material and usable space, then check those specifications against your application.

Questions