Short answer: A museum-grade case is defined by four numbers, not by how it looks. Air exchange rate at or below 0.1 per day, UV transmission under 1% at 380 nm, an Oddy Test pass on every material that touches the object cavity, and a documented load and security rating for the frame. Everything else - finish, lighting, mounting - sits on top of those four.
This page is the specification sheet we hand to curators and exhibition contractors when they ask what actually goes into a vitrine. It is written from our own test records at the Shenzhen plant, where we build museum cases, jewellery showcases and watch cabinets for clients in more than 150 countries.
1. Air exchange rate (AER): the single most useful number
AER is the number of complete air changes inside the case volume per day, measured with a tracer gas method (typically CO2 decay) at a defined internal/external pressure difference. A sealed case with an AER of 0.1 exchanges its full air volume once every ten days. A typical shop-fitting cabinet with a sliding door and no gasket sits somewhere between 5 and 20.
| Case type | Typical AER (d⁻¹) | Buffer life with silica gel | Where it is acceptable |
|---|---|---|---|
| Display cabinet, sliding glass doors, no gasket | 5 – 20 | 2 – 4 weeks | Retail watch and jewellery counters |
| Retail showcase, magnetic gasket, single seal | 0.5 – 2.0 | 3 – 6 months | Luxury boutique, short-term exhibition |
| Museum vitrine, double gasket, cable-free seal | 0.1 – 0.3 | 12 – 18 months | Permanent gallery, low-sensitivity collections |
| Hermetic conservation case, welded frame, conditioned plinth | ≤ 0.1 | 24 months and above | Metal artefacts, polychrome, organic material |
ISO 11799:2023 is the reference most European institutions write into tenders. It does not fix one AER for every object; it asks the supplier to state a measured value and the test method, and to prove the value holds after the doors have been opened and closed 500 times. That last clause is where cheap cases fail - the gasket takes a compression set and the number drifts upward within a season.
2. Glass: what "UV protection" should actually mean
Almost every supplier claims UV filtering. The useful questions are: at what wavelength is the cut measured, is the interlayer laminated or is it a surface film, and what is the visible transmission loss you pay for that cut.
| Build-up | Thickness | UV cut at 380 nm | Visible transmission | Notes |
|---|---|---|---|---|
| Clear float, tempered | 6 mm | 25 – 40% | 89% | Cheapest option, unsuitable for textiles or pigments |
| Low-iron float, tempered | 6 mm | 30 – 45% | 91% | Removes the green edge cast, no UV benefit by itself |
| Low-iron laminated with UV interlayer | 6.76 mm (3+0.76+3) | ≥ 99.9% | 88 – 90% | Our standard for museum vitrines |
| Laminated + anti-reflective coating | 8.76 mm | ≥ 99.9% | 96 – 98% | Used where glare across a painting is the main complaint |
Two points that come up in every tender review. First, the UV interlayer degrades: it is a PVB or EVA film with absorbers, and under continuous 200 lux gallery lighting the cut drops by roughly 1 percentage point per decade. Ask for the interlayer batch certificate, not only the glass certificate. Second, laminated glass does not automatically give you security. If the case must resist impact, the specification is EN 356, and for museum work the usual minimum is P5A (a 3.8 kg steel ball dropped from 9 m without penetration) or P6B where the object is high value.
3. Oddy Test and the material list behind it
The Oddy Test exposes a material to polished copper, silver and lead coupons in a sealed jar at 60 °C and high humidity for 28 days, then grades corrosion. A pass means the material releases no sulphur, chloride or acid vapour at a level that attacks metal. It is the single test that separates a conservation case from a pretty box.
What we test and what we refuse, based on our last audit cycle:
| Component | Accepted | Rejected in object cavity |
|---|---|---|
| Panel substrate | Powder-coated steel with polyester topcoat, anodised aluminium, FSC birch ply sealed with water-based acrylic | Untreated MDF, PVC foam board, chipboard with urea-formaldehyde binder |
| Adhesive | Water-based acrylic, hot-melt polyolefin | Solvent-based contact adhesive, cyanoacrylate, silicone with acetic cure |
| Gasket | Silicone sponge, neutral-cure EPDM | PVC gasket, sulphur-cured rubber |
| Fabric and lining | Untreated linen, washed cotton, inert polyethylene foam (Plastazote) | Flame-retardant treated fabric with brominated compounds, dyed felt with sulphur dyes |
| Mount hardware | 316 stainless steel, brass with clear lacquer, acrylic | Galvanised steel, nickel-plated zinc alloy in direct contact |
We keep the Oddy reports on file for three years and issue them with the case documentation pack, together with the interlayer and powder-coat certificates. Museums that run their own conservation lab will usually retest a sample anyway - that is expected, and it is faster when the supplier can name the exact batch.
4. Sealing, doors and cable management
AER is decided at three places: the door gasket, the panel joints, and every hole you drill for a cable or a sensor. On our hermetic cases the frame is welded rather than screwed, panel joints sit on a continuous EPDM bead, and cables leave the cavity through a potted gland rather than an open slot. Doors run on a compression latch, not a magnetic catch: a magnet holds the glass closed but does not compress the gasket, so the seal depends on how carefully the last person shut it.
Where a case needs active humidity control, we add a conditioned plinth with a silica gel cassette or a Peltier unit and a sensor port. The port is the weak point. Specify a double-seal port with a silicone plug and a blanking cap, otherwise the AER figure measured at the factory is meaningless.
5. Lighting heat and colour
LED has solved the heat problem in the beam but not on the driver. A 12 W/m LED strip running inside a closed case raises the internal temperature by 3 to 6 °C above ambient depending on volume, and that shift moves RH by several points. Two rules we apply: mount drivers outside the object cavity, and keep the total connected load under 15 W per cubic metre of case volume for sensitive material.
For colour, CRI alone is no longer enough for a conservation spec. Ask for TM-30 Rf and Rg values as well. Our museum fixtures run CRI 95+, Rf 92, Rg 100 at 3000 K or 4000 K, with a dimming range down to 1% and no measurable flicker at camera shutter speeds used for documentation photography.
6. What to put in your RFP
If you are writing the tender document, these five lines do more work than a page of adjectives:
- Measured AER ≤ 0.1 d⁻¹ by CO2 decay, with the result stated after 500 door cycles.
- Laminated low-iron glass, UV cut ≥ 99.9% at 380 nm, interlayer certificate supplied.
- Oddy Test pass for all materials in contact with the object cavity, reports supplied with the case.
- Frame load rating verified at 500 kg static with the glass fitted, report supplied.
- EN 356 class stated, with the test house named.
Send us a drawing or a room schedule and we will come back with a case specification against those five points, a 3D render in seven days, and a material sample set by courier. Our plant in Shenzhen covers 20,000 m² across six workshops - glass, sheet metal, low-VOC coating, woodwork, hardware and final assembly - which is why we can hold a tolerance on a welded frame and still ship a full gallery in one container schedule.