Why abrasion rating misleads buyers
Martindale is a rubbing test. A fabric mounted on a machine is rubbed against a standard wool abradant under 12 kPa of pressure, and the cycle count at which two threads break is reported. A 40,000 cycle fabric survives roughly 40,000 rubs. In a restaurant, a seat does not experience rubbing in the way the test measures it: it experiences compressive load from a body, then a spill, then a cloth. The failure mode buyers actually see in year two is not a worn fibre, it is a stain that never came out, or a seam that opened.
The distinction matters commercially because the numbers on a spec sheet are not equally available. Martindale is quoted by every mill. Stain release is quoted by maybe a third. Cleaning code, which tells the housekeeping team which chemistry is safe, is quoted by almost none. A buyer who specifies 60,000 Martindale and stops there has specified the one property that a 700 gram per square metre polyester will pass anyway, and left the difficult ones to chance.
There is a rule of thumb from our own claim history that is worth more than any single test number. Across 1,900 upholstered chairs shipped in the last three years, the defects that came back were 61 percent staining, 24 percent seam or piping failure, 11 percent foam collapse and 4 percent frame. The fabric never wore through in a single case. Wearing through takes 8 to 12 years in a restaurant; staining shows up in the first six months.
The four numbers that predict cleaning performance
Stain release, tested to AATCC 130, is graded 1 to 5, and 3.5 is the floor for a dining seat. The test applies a measured volume of oily stain, waits, then removes it with a defined procedure and grades what is left. A fabric at grade 2 will shadow at the first olive oil spill; grade 4 releases cleanly enough that a 60 second response is enough. Ask for the report, not the word stain resistant.
Cleaning code is the operational number. W means water based cleaners only, S means solvent only, WS means either, and X means vacuum only, which is unusable in a restaurant and should be rejected on sight. A code W fabric cleaned with a solvent based spotter can delaminate a back coating, and a code S fabric hit with water can ring permanently. The kitchen and the front of house need the code written on the maintenance card, not buried in a technical data sheet.
Colourfastness to rubbing, graded 1 to 5 against both wet and dry cloth, matters more than buyers expect because a dark navy or black cover that crocks at grade 3 will transfer onto a cream linen napkin and onto the back of a white shirt. Grade 4 wet and dry is the minimum for a dining chair in a dark colour. Below that, the dye is not fixed properly, and cleaning will lift more of it.
Pilling, graded 1 to 5 after 2,000 rubs, is the fourth. Grade 4 or better. A grade 2 wool blend in a banquette shows visible bobbles at the seat front within 9 months, and no amount of cleaning removes them, because they are fibre ends and they have to be cut. This is the number most often left off a Chinese mill spec sheet, and it is the one that decides whether a booth still looks new at the end of year one.
Fibre by fibre, in restaurant conditions
Polyester at 300 to 350 grams per square metre, with a stain release finish and a WS code, is the default for good reason: it releases oily stains at grade 3.5 to 4, it does not support mildew, and it costs 6 to 9 US dollars per metre at 1.4 m width. Its weakness is heat: a cigarette end or a hot pan edge at 260 degrees melts a hole rather than scorching it, and that hole cannot be repaired invisibly.
Polypropylene, usually sold as an outdoor or contract fabric, releases stains better than polyester and is genuinely hydrophobic, so a spilled liquid sits on the surface for 30 to 60 seconds before wicking. It is also the weakest of the common fibres under sustained abrasion at 40,000 to 50,000 Martindale, and it furs at a cut edge, which shows at a seam allowance. Use it on banquette backs and booth panels, not on the seat cushion of the busiest table in the room.
Vinyl and polyurethane coated fabric are the only options that are genuinely wipe clean, and they are graded differently: a contract vinyl is quoted in 500,000 double rubs on the Wyzenbeek scale, which is not comparable to Martindale and should never be put in the same column. A 1.0 mm vinyl with a knitted backing is the right choice for a booth seat in a 200 cover room. The trade off is thermal: a vinyl seat at 22 degrees feels 3 to 4 degrees colder than a woven one, and in a warm room it feels clammy within 20 minutes, which guests notice even if they cannot name it. Between those two, most upholstered restaurant chairs in our own range are built with a finished polyester, and vinyl is reserved for the booths.
Wool blends at 30 to 40 percent wool with polyester look the best and clean the worst of the wovens. Lanolin in the fibre holds oily soil, and a wool blend at grade 2.5 stain release will shadow permanently after the third olive oil spill even with correct cleaning. If the design calls for wool, put it on the back panel and the booth upholstery above seat height, and keep the seat in a finished polyester.
Bouclé, which is everywhere in design schemes at the moment, is a construction rather than a fibre, and its looped yarn traps crumbs at the base of the loop. A 6 mm bouclé pile holds rice and salt for the life of the cover. It is defensible on a lounge chair or a waiting bench; on a dining seat it is a decision that the operator will regret by month four.
Where the dirt actually lives: seams, piping and tufting
A seat cover has three zones of soil, and they fail differently. The front 80 mm of the seat edge takes 55 to 60 percent of the soiling, because that is where hands and cutlery touch. The seat centre takes the compression and the spills. The back takes 25 to 30 percent, mostly from shoulders and hair. Specifying one fabric across all three is normal, but specifying one cleaning regime across all three is not: the front edge needs a protectant refresh twice as often as the back.
Seam construction decides whether a spill becomes a stain inside the cushion. A 6 stitch per 25 mm seam with a 10 mm seam allowance resists wicking better than a 4 stitch seam, because the tighter stitch density closes the needle holes. Welt cord, or piping, is a soil trap by design: the cord sits in a channel and the crevice on each side holds liquid. On a dining chair, a top stitched seam without piping cleans 30 to 40 percent faster in our own cleaning trials, and it is the first thing we change when a client reports staining at the seat edge.
Tufting and channel stitching create the same problem in a different geometry. A deep button tuft is a 15 mm well that a cloth cannot reach, so the only effective method is extraction, and extraction on a fixed cover wets the foam. Channel stitching at 40 mm pitch is manageable because the channels can be wiped along their length. Button tufting at 100 mm centres on a dining seat is a maintenance liability that looks superb in a photograph on delivery day.
The frame underneath matters as well, because a cover that has to be removed for cleaning puts load on the joints. A chair that passes a 150 kg whole-chair load test on a static seat load will still loosen its corner blocks after 30 or 40 cover removals if the joints are only screwed. Glued and dowelled corners survive that cycle. This is one of the questions worth asking before the fabric is even chosen.
Detachable covers: the zip is a specification
A removable cover turns a 40 minute upholstery repair into a 4 minute swap, and it costs 4 to 7 US dollars per chair. It is the single highest return option on a restaurant chair, and it only works if four details are specified: a 5 mm nylon coil zip rather than a 3 mm metallic one, a zip that runs the full length of the back panel and around the seat deck, a 20 mm overlap flap over the zip so the teeth never touch a guest, and a cover that is dimensioned to the foam with 8 to 10 percent negative ease so it does not crease when fitted.
The common failure is a short zip. A cover with a 300 mm zip in the back panel cannot be fitted over a back that is 420 mm tall without stretching the seam, and the seam then fails at the top corner within 6 months. The second common failure is no overlap flap, which leaves a cold zip against the back of a seated guest and produces a complaint that is hard to diagnose from a photograph.
Detachable covers also solve the dye lot problem. On a 120 chair order, order 15 percent extra covers in the same dye lot rather than 15 percent extra fabric. The covers arrive cut and sewn, they store flat in a 600 by 400 mm box, and they need no upholsterer on site. On the JP-C023 upholstered dining chair we build, the removable version adds 22 mm to the seat deck height at the back edge unless the foam is reduced by the same amount, which is the kind of detail that should be agreed at drawing stage rather than discovered at installation.
A 20 minute trial you can run on swatches
Ask for three 300 by 300 mm swatches of each candidate, cut from the same roll, and run the same five stains on all of them: 5 ml of soy sauce, 5 ml of olive oil, 5 ml of tomato ketchup, a 30 mm ballpoint line, and a 10 mm bead of hand cream. Leave each in place for 10 minutes, which is roughly how long a spill sits before a server notices it, then blot with a white paper towel and clean with the chemistry the cleaning code allows.
Record three things per stain: how far the liquid wicked beyond the drop in 10 minutes, how many blotting passes it took to stop transferring colour, and whether a ring remained after drying for 24 hours. A finished polyester at grade 4 stain release typically wicks 2 to 3 mm and clears in 4 to 6 passes with no ring. An unfinished wool blend wicks 8 to 12 mm and leaves a permanent tide mark on the oil.
The trial costs one afternoon and it will disagree with the spec sheet often enough to be worth doing. Mills test on a flat, clean, conditioned sample; a chair has seams, a foam substrate that draws liquid inward, and a 20 mm crown that pools a spill at the centre. If the swatch behaves badly flat, it will behave worse fitted.
Run the same trial on a finished chair if you can get one. Our own sampling routine includes a 48 hour soil and clean cycle on the first off tool sample, and it has changed the specified fabric on 4 of the last 11 projects. A sample chair that has been through that cycle tells you more than any certificate.
Specify by zone and by cover life, not by swatch
A restaurant does not have one upholstery specification, it has four. Booths and banquettes take the highest soil load per square metre because they cannot be pulled away from the table, and they are the hardest to clean because the seat meets the back at a fixed angle. Dining chairs take impact soiling at the seat edge and are easy to move. Lounge and waiting seats take far less food soil and can carry the wool or the bouclé. Bench seating in a high volume room is a vinyl decision, not a woven one.
Set a cover life target in years and buy to it. A woven polyester at 320 grams per square metre in a 1,200 cover per year room gives 3 to 4 years before the seat edge shows pile flattening, and 5 to 6 years if the chair is one of 60 rather than one of 20. A contract vinyl gives 7 to 9 years but changes the acoustics and the feel of the room. Neither is wrong; the error is buying a 6 year cover for a 2 year design cycle, which many operators do without noticing.
Then write it into the order as data rather than adjectives. Six lines do the job: fibre and weight, stain release grade with the test method, cleaning code, colourfastness wet and dry, pilling grade, and whether the cover is fixed or removable with the zip specification. That is a page a mill cannot argue with, and it is also the page that decides a claim two years later. The full field list we work to is in our restaurant furniture specification notes, and it can be attached to a purchase order unchanged.
Two of our own builds show how the choice lands in practice. The solid wood armchair with fabric seat is specified with a 340 gram finished polyester at grade 4 stain release for dining rooms above 150 covers, with a removable cover as standard. The solid wood dining chair with velvet seat uses a 280 gram velvet at grade 3, which is the right call for a 60 cover room where the seat is cleaned nightly, and the wrong call for a 300 cover room where it is cleaned weekly.
The operational half of the problem
Even a perfectly specified cover fails if the cleaning regime is wrong. The three failures we see repeatedly are: a solvent spotter used on a water cleanable fabric, which leaves a ring at the edge of the spot; a steam cleaner held too close, which drives soil into the foam and delaminates a back coating; and a generic all purpose cleaner at the wrong dilution, which leaves a sticky residue that attracts soil faster than the original surface did.
Write a one page card per fabric and pin it in the housekeeping store. It needs the cleaning code, the dilution ratio, the dwell time, the tool, and the two things not to do. Training takes 20 minutes and it prevents most of the staining claims that would otherwise be blamed on the fabric. A restaurant furniture supplier who cannot supply that card is selling metres, not seats.
Finally, plan for the first re-cover before opening. On a 90 cover room, budget 10 percent of the seat covers as a replacement reserve in the original dye lot, and 2 percent of the frames as spare parts. When the first four seats at the pass are marked in month 14, the fix is a 4 minute cover swap rather than a 6 week order. Cleanability is not only a fabric property; it is a decision about how quickly a soiled seat can be put back into service.
If you want the numbers checked against your own floor plan and cover count, request a restaurant furniture quote with the room size, the seat count and the cleaning regime you can actually sustain. The recommendation that comes back will be a fabric and a cover construction, with the test reports attached, rather than a colour card.