In a cold room, the panel is rarely the component that fails. The fixing holding it is.

Start with a clarification that saves a lot of confusion, because the term cold room panel is used loosely.

The insulated structural panel that forms the envelope of a cold room is a sandwich of rigid foam between metal skins. A PVC foam board is not that product and does not replace it. What PVC foam board actually does in a cold room is the internal lining, the door body, the corner profiles, the coving, the thresholds and the trim. It is a detail and finish material working on someone else's structure.

That role is precisely why the fixings matter so much. A lining has to be attached to a substrate, the substrate and the lining move by different amounts, and the connection between them is a row of small rigid points. Everything in this article follows from that geometry. The composition of the board itself is covered in our guide to what is actually in a PVC sheet.

A mechanical fixing in ordinary construction has one job, which is to hold something in place.

In a cold room, the same fixing acquires a second job that nobody specified. It becomes the point at which the lining's dimensional movement is resisted. Because the lining moves further than the substrate it is fastened to, every fixing in the field has to absorb part of that difference, and a screw in a round hole with no clearance cannot absorb anything. It can only resist.

The distinction between a restraint and a guide is the whole subject of this article. A fixing that holds a component firmly at one point and allows it to slide elsewhere is a guide. The same fixing repeated across a field with no clearance at any of them is a restraint. Cold room linings are usually detailed as the second, because that is how linings are detailed in ordinary rooms where the differential is small enough not to matter.

The reason the ordinary approach stops working here is arithmetic rather than anything exotic, and it is worth doing the calculation for a typical case.

Different materials change dimension by different amounts for the same temperature change. The property governing this is the coefficient of linear thermal expansion.

Now apply that to a realistic installation. A lining panel four metres long, installed in a room at ambient temperature around twenty degrees, operating at minus twenty-five. That is a forty-five degree swing.

The PVC panel wants to shorten by a little over ten millimetres. A steel substrate of the same length wants to shorten by about two. The two are fastened together, so they have to shorten by the same amount, and the difference - close to nine millimetres across the panel - has to be accommodated somewhere.

The lining and the structure behind it move by different amounts. The difference has to go somewhere, and where it goes is decided by the fixing detail.

This is the same mechanism that produces the cracking, sagging and gapping described in our piece on PVC thermal expansion in trim, fencing and ceilings, and the underlying behaviour of the material under restraint is covered in our guide to why PVC sheet warps. Cold rooms differ only in the size of the temperature swing, which is larger than in almost any other building application.

A useful way to hold the figures is that a cold store imposes roughly twice the dimensional range of an exterior application, on materials that are already the most dimensionally active in the assembly.

The differential above would be manageable on its own. It is not on its own, because the same temperature that drives it also removes the material's ability to cope with it.

Impact-modified PVC works because the modifier introduces a second phase that absorbs energy when the material is struck. Those modifier particles have their own low-temperature transition, and below it they stop absorbing energy. The material does not gradually become less tough. It reaches a point and drops off, and past it the board behaves like unmodified material, which is brittle.

So at operating temperature, a cold room lining is both at its most dimensionally active relative to its substrate and at its least able to absorb localised stress. Those are two consequences of one cause, and they arrive in the same place.

Cold affects two separate properties at once. The material moves more and absorbs less, and the second effect is the one that turns movement into damage.

There is a third consideration that belongs here rather than in a section of its own. A panel that has been machined has a cut core, and a foamed core is a different material from the face. In a cold room those cut edges are also the places where vapour condenses. The structural reason is covered in our guide to skin and core in PVC foam board, and in this application the edge detail is both a mechanical and a moisture control decision.

A fixing hole is the worst possible place in a panel to ask anything difficult of the material, and it is where every difficult thing happens at once.

Placing a brittle material inside a stress concentration and then loading it cyclically is a well understood recipe for crack initiation, and that is what a cold room fixing detail frequently amounts to.

The fixings are where the assembly's geometry, load and material limitations all meet. That is why the panel is usually still intact when the connection is not.

One further characteristic of this failure deserves attention because it misleads inspectors. A fixing that has torn through the panel leaves the surrounding face in perfect condition. The board is not discoloured, not warped and not delaminated. It looks exactly as it did on the day it was installed, apart from a hole that has become a slot with a crack running from it. Panels are replaced on the assumption that the material was at fault when the detail was.

There is a second mechanism operating in freezers that turns small gaps into large ones, and it does not exist in a chiller at all.

Water expands when it freezes, by roughly nine percent of its volume. That expansion generates significant pressure, and it operates in whatever shape the water happens to occupy. Any void in a cold room assembly that can collect moisture will therefore be wedged apart the first time it freezes, and it will be wedged again every time the room is defrosted and refrozen.

The places where this happens are predictable. Between the lining and the substrate wherever the two are not in contact. Around fixings, where the hole and the fastener are never perfectly matched. At the floor and wall junction, which is the lowest point and collects everything. Inside a machined edge that was left unsealed, where the cellular core provides an enormous number of small cavities.

That last one is the reason unsealed edges are a different class of defect in a freezer than in an ordinary room. A cut core in a dry interior is a mechanical weakness. The same cut core in a freezer is a set of containers for water, each of which will expand against the surrounding material on every cycle.

Every gap is a potential reservoir. In a freezer, a reservoir that freezes is a force applied to the joint from the inside.

The practical conclusion is that sealing in a freezer is not a finishing operation. It is part of the structural detailing, and it has to be specified with the movement of the assembly in mind rather than selected from whatever the contractor has on the van.

Everything in this section adds cost. That is the honest position, and the reason to state it plainly is that the alternative is a lining that has to be replaced in five years.

Detailing for movement is the whole difference between a lining that lasts the life of the room and one that needs replacing within a few years.

The comparison with flooring is instructive and worth a note, because the same differential appears there in a different form. The expansion gaps required in a cold-store floor, and the reasoning behind their size, are covered in our guide to expansion gaps in flooring. The principle is the same in both cases: the movement is not optional, so the only decision available is where it is allowed to happen.

Temperature band decides most of this, and the honest answer is that PVC foam board has a comfortable range and an uncomfortable one.

Density selection for a cold room follows the same logic as anywhere else, covered in our piece on why density is not strength, with the addition that a denser board also carries a higher expansion coefficient in absolute terms and therefore generates more movement for the fixings to accommodate.

The flooring in these rooms is a separate subject that overlaps at the expansion gap and nowhere else. Where the floor rather than the wall is under consideration, the core behaviour, seam performance and installation consequences for cold-store conditions are covered in our piece on SPC flooring for cold storage and freezers.

The pattern across the whole of this range is worth restating once, because it applies to cold rooms more sharply than to anything else in it. Panels are selected on material properties and fail on connection details. In a room operating through a fifty-degree cycle, the connection detail is not a secondary consideration that follows the material choice. It is the design decision that determines whether the material choice works.