What cloth is

Five spins read the film a drained channel keeps

A spin empties a cotton's yarn channels from the inside of the load outward and leaves the rest full, so what it keeps above the swollen fibre's own water falls as one over the speed squared. A film left on the walls of the emptied channels does not fall at all. Plotted against one over the speed squared, five household spins lie on a line whose intercept is the film and whose slope is the full channels. Both are proportional to the fibres' surface, so their ratio reads the film's thickness with the fibre's surface cancelled out of it, and a half load or a coarser fibre moves the line without moving the reading.

Worth reading first: A spun cloth dries by its weight · A spin leaves the water a fibre swelled by · A cotton's own water is a twentieth of what a cloth holds.

A washing machine’s spin is a centrifuge, and a spin leaves the water a fibre swelled by: it empties a cotton cloth’s pores in the order they give up water and never touches the water inside the fibre. A spun cloth dries by its weight took that one step further. What a spin leaves in the yarn’s channels is set by the fibres’ surface and the drum’s speed, with no count, no packing and no thickness in it, so every plain cotton of one weight keeps the same share.

Both of those accounts took every drained pore to be empty. That is the arithmetic’s largest omission and both said so. A liquid that wets a wall leaves a film on it when the bulk drains away, and on a cotton fibre, which water wets completely, there is no reason to think the film is absent. How thick it is nobody here knows.

The last essay noticed something awkward about measuring it. A film is so many grams on every square metre of fibre surface, and so are the channels a spin leaves full, so changing the fibre’s fineness, the yarn’s packing or the cloth’s weight moves both together and cannot tell them apart. It suggested changing the speed instead. This essay works out what a series of speeds would show, and finds that it reads the film’s thickness without ever needing the fibre’s surface at all.

What a spin leaves above the swelling floor, with and without a film on the drained wallsThe water a spin leaves in a full load of 150 g/m² plain cotton of 40 tex, over and above the 30% the swollen fibre keeps at any speed, against the drum's speed, with a film on the drained channels' walls of no film: 33.2% at 800 rpm and 8.3% at 1,600; 0.300 µm: 36.6% at 800 rpm and 14.3% at 1,600. The channels a spin leaves full fall with the square of the speed; a film does not fall at all, so at fast spins it is most of what is left. The film's thickness is an input: nothing in the picture says how thick a drained cotton channel's film is.the kept channels fall with the speed's square and a film does not fall at all150 g/m² · 40 tex cotton · a 68 mm load · above the swelling floor40060080010001200140016000%10%20%30%40%50%60%70%spin speed, rpmwater above the floor, share of dry weightfilm 0.300 µmno filmfilm = h × the swollen fibre surface, on drained walls onlyfull load
Fig. 1 The water a spin leaves in a full load of 150 g/m² plain cotton above the 30 per cent the swollen fibre keeps at any speed, against the drum’s speed, with no film and with a film of three tenths of a micrometre on the drained channels’ walls; the slider sets the film from a twentieth of a micrometre to a half. Both curves fall steeply, and the film lifts the fast end most.

Two laws inside one residual

A spinning drum pins its load to the wall and drains it outward. The liquid at the perforated wall is at the air’s pressure, and the liquid a depth zz inside the load is pulled outward by the spinning column between it and the wall. That suction is ρω2(Rz−z2/2)\rho\omega^2(Rz - z^2/2), and a pore empties where it exceeds the pore’s entry pressure.

So the spin leaves a layer next to the wall in which the yarn’s channels stay full, and the share of the load’s mass in that layer is

Peρω2(RD−D2/2),\frac{P_e}{\rho\omega^2(RD - D^2/2)},

where PeP_e is the channels’ entry pressure, RR the drum’s radius and DD the load’s thickness. That is one over the speed squared, exactly, with no approximation in it. The holes between the yarns follow the same law at their own much lower entry pressure, 980 pascals against 52,000 for the channels, so very little hole water survives at any household speed.

A film is different. It sits on the walls of the channels that did empty, and it stays there whatever the speed. Its water is a thickness hh times the swollen fibres’ surface, which for a 150 g/m² cotton spun from 14-micrometre fibre is 34.2 square metres of fibre surface in every square metre of cloth. A film a tenth of a micrometre thick is 2.3 per cent of the cloth’s dry weight. Three tenths is 6.9 per cent.

So the water above the swelling floor has two parts, and they obey different laws. The kept channels fall as 1/ω21/\omega^2 and the film does not fall at all. At 800 rpm a cloth with no film keeps 33.2 per cent above its floor and a 0.3-micrometre film adds 3.4 points; at 1,600 rpm the bare cloth keeps 8.3 per cent and the film adds 6.0. The faster the spin, the larger the film’s share of what is left.

A line whose intercept is the film

Plot the same spins against 1/ω21/\omega^2 instead of against the speed. The kept channels become a straight line through the origin and the film lifts it by a constant.

Water above the floor against one over the speed squared, for three film thicknesses. The same spins plotted against one over the square of the speed, at 800, 1,000, 1,200, 1,400 and 1,600 rpm, with a straight line fitted to each and carried back to an infinitely fast spin: no film of film meets the axis at 0.00% with a slope that reads back a film of 0.000 µm; 0.100 µm of film meets the axis at 2.28% with a slope that reads back a film of 0.101 µm; 0.300 µm of film meets the axis at 6.85% with a slope that reads back a film of 0.325 µm. The intercept is the film and the slope is the kept channels. The picture cannot show a real channel's spread of sizes, which would bend the line at slow speeds.
Fig. 2 The same three cloths at five household speeds, 800 to 1,600 rpm, plotted against one over the speed squared with a straight line fitted to each and carried back to an infinitely fast spin. With no film the line meets the axis at nought. A tenth of a micrometre meets it at 2.3 per cent and three tenths at 6.9.

The five points lie on their lines, and carrying a line back to where 1/ω21/\omega^2 is nought, an infinitely fast spin, gives the water that no spin could remove. On the arithmetic that is the film. With no film the lines pass through the origin to within a few hundredths of a per cent, and with a film they meet the axis at its weight.

That is already a measurement of something, but in the wrong units. An intercept of 2.3 per cent of the dry weight says how many grams of film there are. Turning grams into a thickness needs the fibres’ surface, and the fibres’ surface is exactly the quantity that is hard to know for a real cotton, whose fibres are flattened, twisted ribbons with a lumen and a spread of widths.

The surface cancels

The slope of the line is the kept channels. Their water is the channels’ volume times the share of the load that stays full, and the entry pressure in that share is 2γcos⁡θ2\gamma\cos\theta over the channels’ hydraulic radius, which is their volume over the fibre surface that walls them. The volume cancels and the surface remains:

slope=2γcos⁡θ Sρ (RD−D2/2).\text{slope} = \frac{2\gamma\cos\theta\,S}{\rho\,(RD - D^2/2)}.

This is the identity the drying essay found: a spin’s kept channel water is the fibres’ surface over the drum’s suction. The intercept is hShS. Both are proportional to the same surface, so their ratio is not:

h=interceptslope⋅2γcos⁡θρ (RD−D2/2).h = \frac{\text{intercept}}{\text{slope}}\cdot\frac{2\gamma\cos\theta}{\rho\,(RD - D^2/2)}.

Everything on the right is either read off the line or known about the drum. The surface tension of water and its contact angle on cotton are the only material properties in it, and water wets cotton completely. A film’s thickness comes out of five spins with no measurement of the fibre at all. The yarn’s packing has dropped out too, as it did from a spin’s kept channels, and so has the stiffness of the wet yarn that a meniscus presses together, since nothing in the reading depends on the channel keeping its shape beyond its size.

Two spin series with a 0.200 µm film, differing only in fibre width. Spins at 800 to 1,600 rpm of a 150 g/m² cotton spun from fibres of two widths, each with a 0.200 µm film, plotted against one over the speed squared with the fitted lines carried back: 14 µm fibre, intercept 4.57%, slope 2171, film read as 0.209 µm; 21 µm fibre, intercept 3.04%, slope 1505, film read as 0.201 µm. A wider fibre has less surface, so both the kept channels and the film shrink together and the ratio that reads the film does not move. The picture assumes one film thickness on every drained wall.
Fig. 3 Two cloths of the same weight and count with a 0.2-micrometre film, one spun from 14-micrometre fibre and one from 21. The coarser fibre has two thirds the surface, so both its intercept and its slope fall by a third: 4.57 per cent and a slope of 2,171 against 3.04 per cent and 1,505. Read as a ratio, both lines give a film of 0.21 and 0.20 micrometres.

That is the cancellation in practice. Spin two cloths that differ only in their fibre’s width, and the lines differ in every respect that a reading of grams would see. The coarser cloth keeps less in its channels and less in its film. The thickness read from each is the same to within four per cent, because the ratio of film to channels is fixed by the film’s thickness against the channels’ hydraulic radius, not by how much surface either has.

A half load moves the slope and not the film

The drum offers a second handle the fibre does not. The slope depends on the load’s thickness, through RD−D2/2RD - D^2/2, because a thin load keeps a larger share of itself in the layer next to the wall. The intercept does not depend on it at all, since a film on a drained wall does not know how deep in the load the wall is.

Two spin series with a 0.200 µm film, differing only in load. Spins at 800 to 1,600 rpm of a 150 g/m² cotton as a full load and as a half load, each with a 0.200 µm film, plotted against one over the speed squared with the fitted lines carried back: full load, intercept 4.57%, slope 2171, film read as 0.209 µm; half load, intercept 4.57%, slope 4025, film read as 0.209 µm. A half load keeps a thicker share of its mass against the wall, so its slope nearly doubles while its film, which does not depend on the speed or the load, stays where it was. The picture assumes one film thickness on every drained wall.
Fig. 4 One cloth with a 0.2-micrometre film, spun as a full 68-millimetre load and as a half load of 34. The half load’s slope is 1.85 times the full load’s, and both lines meet the axis at 4.57 per cent. The film read from each is the same.

Halving the load nearly doubles the slope and leaves the intercept exactly where it was. That is the most direct test of the account available on a domestic machine. A residual that does not move with the load is not water held in the kept layer, whatever else it is. A residual that does move is. If the two disagree with the film model’s prediction, the model has the wrong picture of where a drained channel’s water sits, and the check needs nothing but a scale and a drum run twice.

Where a film stops being thin

The reading assumes the film is thin beside the channel it lines. Two small effects move it, in opposite directions.

The film a household spin series reads, over the film that is there. The film thickness read from five spins at 800 to 1,600 rpm, as a multiple of the film actually present, against that film: 0.050 µm reads as 0.049 µm; 0.200 µm reads as 0.209 µm; 0.500 µm reads as 0.588 µm; 1.00 µm reads as 1.49 µm. Thin films read within a few per cent either way: the holes add a little to the slope and pull the reading down, the film missing from the full channels pulls it up. Thick ones read high, by about the film over the channels' hydraulic radius of 2.8 µm, because a full channel holds no film of its own and the line takes that absence for part of the channels. The picture assumes one channel size.
Fig. 5 The film a household series reads, five spins from 800 to 1,600 rpm, as a multiple of the film that is there, against that film. Up to about a fifth of a micrometre the reading is within a few per cent. A half-micrometre film reads as 0.59 and a one-micrometre film as 1.49, following the grey curve of one plus the film over the channels’ 2.8-micrometre hydraulic radius.

The holes add a little to the slope. They follow the same 1/ω21/\omega^2 law at their own entry pressure, so their kept water looks like extra channel water and lowers the reading by a per cent or two. And a full channel has no film of its own. The film stands on the drained channels only, so as the kept layer thickens at slower speeds, the film’s water falls a little with it, and that falling part is taken by the line for channel water. It raises the reading by about the film’s thickness over the channels’ hydraulic radius, which in this yarn is 2.8 micrometres.

So a film of 0.05 micrometres reads as 0.049 and one of 0.2 as 0.209. A film of half a micrometre reads 18 per cent high, and a full micrometre half as high again. At that thickness the film is more than a third of the channel’s hydraulic radius, it fills about a third of the channel’s section, and calling it a film has stopped meaning much. The reading’s own bias says when that point has been reached: if a series reads a film above a few tenths of a micrometre, the model of an emptied channel with a skin on its wall is the thing to doubt.

Whether a household drum can see one at all

A film has to be a fair part of the residual for five spins to read it well, and a household drum is not fast.

The spin at which a film holds as much water as the kept channels, against the film's thickness. For a full load of 150 g/m² plain cotton, the drum speed above which a film on the drained channels' walls holds more water than the channels and holes the spin leaves full, against the film's thickness: 0.050 µm, 4355 rpm; 0.100 µm, 3105 rpm; 0.200 µm, 2232 rpm; 0.500 µm, 1479 rpm. A household drum spins at 400 to 1,600 rpm, the shaded band; a film thinner than 0.42 µm is never the larger part of what such a spin leaves. The picture cannot say which thickness a real cotton channel keeps.
Fig. 6 The drum speed above which a film on the drained walls holds more water than the channels and holes the spin leaves full, against the film’s thickness. A twentieth of a micrometre needs 4,355 rpm and a fifth 2,232. Household drums spin at 400 to 1,600 rpm, shaded; only a film thicker than 0.42 micrometres outweighs the kept channels inside that band.

At 1,600 rpm the drum’s wall is at about 715 times gravity and a full load’s kept layer holds 12.6 per cent of its mass. A film outweighs that layer’s channels only if it is thicker than 0.42 micrometres. A tenth of a micrometre is outweighed until 3,100 rpm and a twentieth until 4,400.

That does not make a thin film unreadable. It makes it the intercept of a line whose points are all well above it, so the reading depends on extrapolating, and the precision of the weighing sets the smallest film the series can resolve. At 1,600 rpm a 150 g/m² cloth’s residual above its floor is 8.3 per cent with no film, which is 12.5 grams on a square metre. A tenth of a micrometre of film adds 3.0 grams to that, about a quarter, and is plainly there. A hundredth of a micrometre adds a third of a gram, and would be lost in the variation from one spin to the next. A spin at 3,000 rpm in the same drum would leave the channels about a quarter of what 1,600 leaves, and would see such a film nearly four times as clearly.

What this does to the drying arithmetic

A cotton’s own water is a twentieth of what a cloth holds when the cloth is saturated, and a spin takes the cloth down to its swollen fibre plus a little. A film is part of that little. It changes the drying essay’s central number, that every 150 g/m² cotton keeps 40.9 per cent of its weight at 1,400 rpm, by hShS, so a 0.2-micrometre film raises it to about 44.7.

Crucially, it changes it in the same proportion for every count, since the film, like the channels, goes with the fibre’s surface and not with the yarn. The drying essay’s finding that a spun cloth dries by its weight survives a film intact, and so does its corollary that a finer fibre leaves more water behind. What a film moves is the number, not the rule.

The same holds for the time a cloth takes to dry. A drying cloth loses water at a rate the room sets until its surface stops being wet, and a drying cloth cannot lift what a sealed tube can because the channels empty from the face in. A film is the last water on a drained wall, so it is part of what the falling-rate period at the end of drying has to remove, and it is there in every count in the same proportion.

What water does to a thread began this account with the fibre swelling a fifth wider. The film is the one piece of water in the account that sits outside the fibre and is held by the fibre’s surface alone, and it has the curious property of being invisible to every measurement of a soaked cloth and visible only when the cloth is drained, which is why none of the earlier essays could see it.

It also connects to a question about the same fibres at a different scale. The hairs decide the sign of the wetting because a fibre wets or does not by its contact angle, and here the same angle sits in the reading as cos⁡θ\cos\theta. On a polyester, which water wets poorly, the film would be thinner or absent and the kept channels themselves would hold less, so a spin series on a polyester cloth is a reading of its contact angle as much as of a film. On cotton the angle is nought and the series reads the film alone.

How the spins were computed

The cloth is a plain cotton of 150 g/m² woven from 40-tex yarn packed at 0.6, from the same weight line the drying essay used. Its water is in the three places the three-reservoir account put it: inside the swollen fibre, in the yarn’s channels with a hydraulic radius of 2.8 micrometres, and in the holes between the yarns. The spin is a 250-millimetre drum with a full load 68 millimetres deep, or a half load of 34, at five speeds from 800 to 1,600 rpm; each pore empties where the suction at its depth exceeds its entry pressure. The film is a thickness times the swollen fibres’ surface, standing only on the channels that emptied.

The reading is an ordinary least-squares line through the five points against 1/ω21/\omega^2, and the thickness is its intercept over its slope times 2γ/ρ(RD−D2/2)2\gamma/\rho(RD - D^2/2).

Required of it, and not shown: that with no film the line passes within a fraction of a per cent of the origin; that films of a twentieth, a tenth and a fifth of a micrometre read back within five per cent; that the same films read the same from a fibre half as wide again and from a half load, while the slope moves by more than a fifth with each; that the half load leaves the intercept where it was; that a half-micrometre film reads high by between ten and thirty per cent; and that no faster spin ever leaves more water.

What a single channel size cannot say

Every channel here is one size. A real yarn’s channels have a spread of sizes, so they empty over a range of depths rather than at one, and the kept water is a sum of 1/ω21/\omega^2 terms with different coefficients. That is still exactly 1/ω21/\omega^2 in total, as long as every channel size is emptying somewhere in the load. At slow speeds, where the coarsest channels are still full throughout, the line would bend, which is why the series starts at 800 rpm and not at 400.

The film is one thickness on every wall. A real film’s thickness depends on the suction holding it, through the forces between the liquid and the wall, and would thin at faster speeds. That would tilt the line and put part of the film into its slope. The reading would then be the film at an infinitely fast spin, which is not the same as the film at a household one, and nothing here says how large the difference is.

And nothing here says how thick a film cotton keeps. Every number in this essay takes the thickness as an input. The account says how a measurement would read it and how far it can be trusted; it does not supply it.

Who worked out which part

The centrifuge method for a porous medium’s capillary pressure curve is Hassler and Brunner’s, from 1945, and the end effect at the outlet face is part of it. Residual films and pendular rings in drained pores are standard in soil physics and petroleum engineering, where the irreducible water saturation they leave is measured routinely. What is added here is the application to a spun cloth: that the kept channels and a wall film obey 1/ω21/\omega^2 and a constant; that both are proportional to the fibre surface, so a series of spins reads the film’s thickness with the surface cancelled; that the reading survives a change of fibre width and a change of load while the line itself moves; that it is biased by about the film over the channel’s radius; and that a household drum sees a film directly only above 0.42 micrometres.

Still open: whether a towel’s own film decides how much it takes

The same film sits on the walls of every cotton channel that has drained, including a towel’s. A wet cloth rolled in a dry towel gives its hole water to the towel’s empty channels, and the towel’s capacity was counted as its channels’ volume plus its fibre’s swelling. If the towel’s channels were never fully emptied in the first place, because the towel was spun and hung and kept a film, its capacity is smaller by that film’s water.

The prediction is that a towel taken from a drier takes measurably more than one taken from a line, by the film on its walls, and that the difference grows with the towel’s fibre surface rather than its weight. The calculation needs the film’s thickness on a towel’s own cotton, which is what the spin series above would measure, and so the two questions are one experiment apart: spin the towel at five speeds, read its film, and the towel’s capacity follows. Nothing in how high a cloth wicks or where the wicking pore leaks has yet asked what a film does to a channel that has to fill again.

What links here

Computed from the collection rather than written here: the essays that point at this one.

Shares its objects with

Essays naming at least two of the same things, that neither author linked.

Named objects

A flat tag is an object no other essay names yet.

Contact angleInter fibre porePore radiusSwellingTwo pore systemsWater