Does a Canvas Need a Backing Board? What Conservation Research Says
For centuries, artists have relied on cotton or linen stretched over wooden frames. While familiar, these materials come with well-known flaws: sagging, warping, moisture sensitivity, and eventual instability. Museums and conservators have been aware of these issues for decades, and modern research continues to reinforce what many artists experience in their studios. Traditional supports can fail the art they are meant to preserve.
The Box Board Canvas was our answer to part of that problem. This post sets out the research behind it, including the parts of that research that do not obviously agree with what we built.
What a national museum found when it surveyed 1,600 paintings
In 2019 the V&A had to move its entire paintings reserve collection out of Blythe House. Around 1,600 paintings, spanning 2,000 years. Every one was condition-surveyed first, and over 200 needed treatment before they could travel.
Carolina Jiménez Gray, the paintings conservator who ran that work, wrote it up in the V&A Conservation Journal. One sentence in it is worth more than any marketing claim we could write:
"We have found that unlined paintings without backings are generally the ones in the most fragile condition. The canvas is more brittle and has more planar deformations."
That is not a laboratory prediction. It is what 1,600 real paintings looked like when someone went through them one at a time.
She also gives the mechanism, which is more specific than the usual talk about protection. Dirt settling on the back of a canvas attracts minute water droplets. Those droplets turn acidic as carbon dioxide dissolves into them, and that acidity speeds the degradation of the canvas fibres. The weakened canvas then causes adhesion problems for the ground and paint on the front.
A backing keeps the reverse clean. That is the whole point of it.
And on their own long-standing policy of fitting backings when paintings are acquired, her verdict is four words: very beneficial.
Then there is the part we did not expect
When the V&A needs to protect the back of a historic painting, what do they reach for?
Polyester.
Stretcher bar lining, a technique devised at Tate, uses a polyester sailcloth backing fed beneath the cross bars and attached to the reverse of the stretcher. It creates an air cushion that substantially reduces vibration and keeps the reverse clean. For fragile unlined canvases with no cross-members, they fit padded inserts using polyester wadding against the back of the canvas.
Two national institutions, putting polyester in direct contact with irreplaceable paintings, because it is the material that does the job.
We do not need to argue that polyester belongs near serious art. That argument was settled by people with more at stake than us.
The oldest evidence: 300,000 cycles versus 100
The most persuasive argument for a rigid back is not a laboratory result either. It is a crack pattern.
Stefan Michalski of the Canadian Conservation Institute described a hundred-year-old painting with a pattern that turns up again and again: very little cracking in the areas sitting over the stretcher bars, extensive cracking everywhere else.
The explanation is buffering. Wood about a centimetre thick evens out humidity swings over roughly one to two days. So over a century, the middle of that canvas went through something like 300,000 daily humidity cycles. The strips sitting over the bars went through about 100 seasonal ones.
Same painting. Same room. The difference was whether something rigid sat behind the cloth.
Michalski points to the same effect at scale. Nineteenth-century Hudson River School paintings, which were made with wooden panels behind them as backing boards, are much less cracked than comparable paintings of the period without them.
That is a hundred and fifty years of accidental field testing, and it is the reason the Canadian Conservation Institute's guidance is that all paintings should have backing boards as a preventive measure.
What a modern study found
In 2020 a team led by Tim Padfield published Back protection of canvas paintings in Heritage Science. They built an instrumented test painting and ran it through a long series of experiments: open back, backed, gapped from the wall, insulated, buffered, glazed, with wall and room temperatures varied independently.
A back view of the experimental canvas stretched over a wooden frame. The fine thermocouples which spring against the painting canvas and the back plate are scarcely visible. Image from conservationphysics.org.
Their conclusion, in their own words:
"the best back protection for the painting amongst the examined experimental set-ups was an impermeable, non-hygroscopic plate, with spacers to separate it from the wall."
Elsewhere in the paper they put it more plainly: the addition of an impermeable, non-absorbent back plate is beneficial to a painting on canvas.
The spacers matter as much as the plate. With the back plate flat against the wall, its relative humidity climbed above 80% once the wall dropped to 16°C. Cold outside wall, warm room, moisture collects on the inside of the board, runs down, and damages the bottom edge of the painting. Their gap was 10mm.
They also removed the usual objection to a sealed back. There is no reason to ventilate the space behind the painting, and perforating it does not help, because gallery air often carries enough moisture to condense at the backboard temperature anyway.
The awkward part, which we are not going to skip
Two of those papers raise a problem with what we actually make.
Padfield's recommendation is an impermeable, non-hygroscopic plate. Our back board is 3mm HDF, which is compressed wood fibre, and wood fibre is hygroscopic.
Michalski goes further and names the material. Porous materials rise steeply in moisture permeability above 60% RH, with plywood and wood up to twenty to fifty times more permeable at 100% RH than at 50%, and he expects paper and hardboard to behave the same way. Damp then gets in faster than it gets out. A painting cycling between a 50% day and a 90% night can settle well above 70% RH. His mould timings are about a year at 65% RH, a hundred days at 70%, and two days at 100%.
That is a real warning, and it is aimed squarely at a board like ours.
Here is why we think a Box Board sits outside it. You can judge the reasoning yourself.
1. The chain depends on the canvas moving
Padfield's warning runs as a chain. A hygroscopic back exchanges moisture with the air inside the enclosure. A small temperature difference between board and canvas produces a large humidity swing at the canvas surface. The canvas takes up that moisture, swells, loses tension, and the paint on it cracks.
Every link has to hold. The last two are about what the cloth does when the humidity around it changes.
Cotton and linen are cellulose. They are hygroscopic. They take up moisture, swell, and let go of tension, and that movement is what the conservation literature spends chapters documenting.
Polyester PET is not hygroscopic. It does not take up moisture. It does not swell. It does not let go of tension when the room changes.
The humidity swing still happens. The canvas does not respond to it.
2. Mould needs something to eat
Michalski is specific about where the mould risk bites hardest: where the canvas carries especially vulnerable materials, such as old glue and starch paste linings. The mould grows on the organic substrate.
Our canvas has no animal glue size and no starch paste. Polyester is a poor meal.
Worth adding that the V&A found the same thing about animal glue from the other direction. Paper facings applied with sturgeon glue thirty years ago had become brittle and much less soluble with age, and in some cases were very difficult to remove at all. Any animal glue does this.
3. There is no cavity behind the canvas
Both papers model an enclosed air space between the painting and a separate backing panel fitted behind it. That space is where the damp collects and the humidity swings.
On a Box Board the canvas is stretched directly onto the board. There is no cavity between them for damp to gather in.
4. Nothing hygroscopic touches the wall
Both papers put the wall at the centre of the risk. Padfield's condensation formed on a plate against a cold wall. Michalski's advice is to keep the painting several centimetres off it.
Our canvas wraps around to the back, so the surface that meets the wall is coated, waterproof polyester. No timber, no board edge, and no part of the strainer makes contact with the wall at any point. There is nothing there to draw moisture in.
5. The gap is bigger than the one they tested
Padfield used 10mm and found it important. Michalski says several centimetres. A 40mm gallery depth leaves roughly 37mm between the back board and the wall.
To be clear about what all this is. Padfield's team tested aluminium plates and never tested wood-fibre board, and neither paper was written with a non-hygroscopic canvas in mind. This is our reading of their mechanisms applied to a different material. It is not a finding of theirs, and we are not going to dress it up as one. The field is not settled either: the Canadian Conservation Institute recommends hygroscopic backs, and Padfield's team disagree with that advice in print. We would rather write these paragraphs than quietly leave the citations off.
What the V&A actually tested about the material itself
We used to summarise this badly, so here it is properly.
Capucine Korenberg, working in the V&A's Science Section, exposed two polyester fabrics to light in a box at roughly 9 klux for 69 days, calculated as equivalent to about 80 years on permanent museum display at 50 lux for ten hours a day. The tensile strength did not vary significantly. The samples were then heat-aged at 60°C in the dark for three months, and when that produced no change, at 80°C for a further two months. The strength was still unchanged. Tensile testing was done on an Instron machine using 19mm strips, three replicates in both warp and weft. Her conclusion was that polyester fabrics should be very stable in the museum environment.
No cotton or linen was tested, and the property measured was tensile strength. That study is a statement about how stable polyester is on its own. It is not a comparison against natural fibres, and we are not going to present it as one.
What the standard reference text says about the material
The comparison against cotton and linen comes from somewhere else, and it is about as authoritative as a source gets.
"In the early twenty-first century, polyester (PET form) exhibits the best combination of properties and provides the most promising starting point to improve the performance of artists' canvas even though it has yet to match linen or cotton kinaesthetically or aesthetically."
Christina Young, "History of Fabric Supports", chapter 5, in The Conservation of Easel Paintings (Stoner & Rushfield, eds., Routledge, 2012).
Best combination of properties. Most promising starting point. And an open admission that, as of 2012, nobody had matched how traditional canvas feels to work on. That last clause is the gap we spent years on.
How ACF put this into practice
In 2017, three years before the Padfield paper was published, we introduced the Box Board Canvas.
- 3mm HDF back board. Rigid and stiff enough that the surface does not flex or dent under a palette knife, a marker, or charcoal pressure.
- 100% Polyester PET canvas. Dimensionally stable, drum-tight, and non-hygroscopic, which is what makes the rigid back work rather than fight.
- Wrapped to the back. The coated polyester continues around the board, so nothing hygroscopic touches the wall.
- 40mm gallery depth. Roughly 37mm of separation between the back and the wall, against the 10mm Padfield's team found made a difference.
- No keys. The V&A found that over the years many wedges work loose from their stretchers, get lost, or fall behind the bars, where they cause bulges in the canvas and eventually paint loss. Our strainer does not use them.
The canvas is stretched over the board, not glued to it. It can be unstapled and restretched years later, which is not true of a panel.
Why this matters to artists
There is one more line in that V&A article worth sitting with.
"Unfortunately, we do not have the resources within the current project to provide rigid backings as many frames would need additional joinery before a backboard could be fitted."
They wanted rigid backings. They knew the paintings without them were in the worst condition. And they could not afford to retrofit them, because retrofitting means joinery on hundreds of frames one at a time.
That is the honest problem with back protection. It is easy to recommend and expensive to add later.
Most artists never fit one. It means sourcing the right material, cutting it to size, sealing it, and choosing correctly for the climate. Galleries do not require it. Suppliers do not push it.
We built it in at manufacture instead, which is the only point at which it costs almost nothing.
Find out what you actually know
We built a 10-question quiz on what your canvas is made of and what it does over time. Every question is drawn from named conservation science sources, and every source is linked in the quiz so you can check us.
Most artists get at least one wrong. Some get a lot more than one wrong.
Everyone who completes it earns a discount. Score higher, earn more.
Or just paint on one
Both are 100% Polyester PET in warp and weft, pre-coated with our own gesso, on a strainer that holds its shape.
Sources
- Jiménez Gray, C. (2020) "The paintings reserve collection is moving", V&A Conservation Journal 67, Victoria and Albert Museum. Link
- Michalski, S. (2005) "Risk analysis of backing boards for paintings: damp climates vs cold climates", in Minimo intervento conservativo nel restauro dei dipinti, Il Prato, Saonara. Paper given at CESMAR7, Thiene, October 2004.
- Padfield, T., Padfield, N., Sang-Hoon Lee, D., Thøgersen, A., Valbjørn Nielsen, A., Krarup Andersen, C. & Scharff, M. (2020) "Back protection of canvas paintings", Heritage Science 8, Article 96. Link
- Korenberg, C. (2003) "How fast do polyester fabrics age in the museum environment?", V&A Conservation Journal, Issue 44, Victoria and Albert Museum. Link
- Young, C., "History of Fabric Supports", chapter 5, in Stoner, J.H. & Rushfield, R. (eds.), The Conservation of Easel Paintings, Routledge / Taylor & Francis, 2012.
- Canadian Conservation Institute, CCI Notes 10/10, Backing Boards for Paintings on Canvas.