Lessons · Lesson 3 of 6
- 01 · The garment is finished after it is sewn
- 02 · Specifying a measurement that does not exist yet
- 03 · Grading through a laundry, and the size set after wash
- 04 · Cloth or recipe: buying the same look twice
- 05 · The laundry's minimums, and the plan they rewrite
- 06 · Defects that are the product, and the ones that are not
Grading through a laundry, and the size set after wash
Get a trustworthy shrinkage figure without waiting for bulk, and stop a load of the wrong weight spending your whole shade band.
Lesson 3 of 6 · 24 min
The situation
Lesson 2 ended with a correction the pattern room could only make once it had garment shrinkage measured under bulk conditions. This lesson is about how you get that number, and why the obvious way of getting it produces a figure that is confidently wrong.
Ulvan did submit a size set — one garment in each size, sewn and washed together so the buyer can see the range. Six pairs, one of each size from 10 to 20, sewn in the bulk quality, washed at Sarband to the sealed recipe, measured and signed off on 21 May. Everything in that sentence is correct except one word nobody thought to ask about: where.
Six pairs of TK-905 weigh 4.44 kg. You cannot wash 4.44 kg in a 250 kg side-loader. The garments would sit at the bottom of the drum and never tumble. So Sarband did what every laundry does with a size set. It ran them in the 30 kg sample machine, made up to 22 kg with ballast — spare garments added only to bring the load to weight.
A sample machine is a different process
How hard the machine works each garment depends on how much stuff is moving against how much stuff, in a drum of a given diameter, turning at a given speed. Change any of those and the abrasion changes. On denim, that changes the shade, the surface and the shrinkage together.
Sarband washed the same twenty pairs both ways — same recipe, same cloth, same day:
| Where it was washed | Inseam shrinkage | Finished inseam | Deviation from 78.0 |
|---|---|---|---|
| 30 kg sample machine, 22 kg load | 3.0 percent | 78.73 | plus 0.73 |
| 250 kg side-loader, 180 kg load, 240 pairs | 3.9 percent | 78.00 | zero |
The tolerance on that measurement is 1.0 cm. The machine alone spent 73 percent of it, before a single thing had gone wrong.
So the rule that makes a size set worth signing is not about the size set at all:
A garment washed in a sample machine is evidence about a sample machine. It tells you nothing you can grade a pattern from.
Breaking the circle for USD 362.00
Lesson 1 named the circularity: a bulk-condition wash needs a production load, a production load needs bulk fabric and a cut, and the cut is what you are trying to release. Here is the break. It is small enough that no buyer refuses it.
- Cut forty pairs from the first roll of the bulk lot, across three sizes — the base size and both extremes. Not a size set. A shrinkage lot.
- Sew them to the production method, with the production thread, waistband interlining and topstitching. A shrinkage lot sewn a different way is a different assembly, and it answers a different question.
- Tag them and put them inside somebody else's production load running your recipe, so they experience a real load. Any style will do. What you are borrowing is the machine's behaviour, not the other order's cloth.
- Measure conditioned, at least four hours after drying, at every point on the specification, and record the spread rather than the average.
- If no load of your recipe is running that week, run your own load and make it up to weight with ballast garments. Then write on the report that it was a ballasted load, because it is not quite the same thing.
Forty pairs of cloth and CMT cost USD 362.00 and one day inside a week you were waiting anyway. Lesson 2 priced the alternative at USD 10,204.73.
The load is not a number of pairs
Now the second half. This is the part that catches people who have already learned the first.
A load is a weight. Everything about the wash — the liquor ratio, the chemical dosing, the mechanical action — is dosed against the mass of the goods. But a laundry's paperwork counts pairs, a factory's trolleys hold pairs, and a cut plan is organised by size. Sarband weighed ten pairs of each size:
| Size | Weight a pair | 240 pairs weigh | Percent of nameplate | Pairs in a 180 kg load |
|---|---|---|---|---|
| 8 | 0.63 kg | 151.2 kg | 60.5 percent | 285 |
| 12 | 0.75 kg | 180.0 kg | 72.0 percent | 240 |
| 20 | 0.91 kg | 218.4 kg | 87.4 percent | 197 |
Load 240 pairs of size 8 and the drum is barely more than half full: the garments fly, and the abrasion is high. Load 240 pairs of size 20 and it is nearly full: they compact, they tumble less, and the abrasion is low. Sarband measured both:
| Load | Inseam shrinkage | Finished inseam | Deviation |
|---|---|---|---|
| 151.2 kg, size 8 only | 4.2 percent | 77.76 | minus 0.24 |
| 218.4 kg, size 20 only | 3.5 percent | 78.33 | plus 0.33 |
A spread of 0.57 cm across the size range, both ends inside tolerance, and nobody has done anything wrong.
The measurement is the smaller half of this problem. The bigger half is shade. Less tumbling means less abrasion, which means a darker garment. So the size-20 loads come out at the dark limit of the sealed band and the size-8 loads at the light limit. The band exists to absorb the things you cannot control — a new drum of enzyme, a different stone lot, a fabric lot change. Load composition can spend all of it before any of those happen. And it spends it in a way that maps onto size, which is the one variable a shop hangs side by side on the same rail.
The fix costs almost nothing. Load by weight, and never let a load be single-size. Mixing at the trolley before the machine adds about 25 minutes a load, which across 53 loads is USD 90.54 for the whole order.
Grading, and where the correction actually is
Shrinkage is a percentage. A buyer's grade step — the fixed gap between one size and the next — is a constant number of centimetres. So the cut grade step is not the graded step. It is the graded step divided by what survives the wash:
| Measurement | Graded step after wash | Measured shrinkage | Cut step | Error at the largest size if you cut the graded step |
|---|---|---|---|---|
| Waist half | 2.0 | 0.6 percent | 2.01 | 0.05 |
| Thigh half | 1.1 | 4.3 percent | 1.15 | 0.20 |
| Front rise | 0.7 | 4.4 percent | 0.73 | 0.13 |
| Leg opening half | 0.3 | 1.9 percent | 0.31 | 0.02 |
Read that honestly. The largest correction here is 0.20 cm at the extreme size, on a tolerance of 0.6. It is real, but it is an order of magnitude smaller than the 0.85 cm the base size was out by in lesson 2. Fix the base size first. Grade correction is a refinement, and a reader who chases it before the base-size shrinkage is right has spent effort on the small number.
What makes it worth doing at all is the shape of the error rather than its size. It is zero at the base size and grows outward, because it accumulates one step at a time. Every sample you approve is a base size or next to it. So this defect is structurally invisible in the approval process, and it only ever appears in the sizes at the ends of the range. Those are also the sizes you have fewest of, so it shows up as a scatter of complaints rather than a batch failure. It is worth doing wherever the range is wide or the tolerance is tight.
The measurement that arrives after the ship
Leg twist is the last dimensional property of a jean to settle, and it does not settle in your building. It is measured after three domestic washes, on a garment that has already been through the industrial one. So the number reaches you weeks after the container has sailed. Tarnwick's limit is 2.0 cm at the hem. The shrinkage lot measured 1.4 cm.
Why a leg twists at all is course 2.4's subject — yarn twist and twill direction pulling the same way. What belongs here is the consequence for planning: the only moment at which leg twist is actionable is on the shrinkage lot, because after that you are either shipping or remaking. So the forty pairs are not only a shrinkage test. Send half of them for the home-wash sequence at the same time. It is the same forty pairs and one more laboratory line.
Check yourselfYour laundry offers to wash your six-pair size set tomorrow so you can release the cut. Do you take it?Show the answer
Take the wash, but do not release the cut on it. Six pairs cannot make a production load, so they will run in a sample machine or in a heavily ballasted small load, and the mechanical action per garment there is not the action in a 180 kg side-loader. On this style that difference was 0.9 points of inseam shrinkage, which is 0.73 cm, or 73 percent of the tolerance. Use the size set for what it is genuinely good for — proportion, balance, the look of the wash across the range, whether the pocket sits right on the largest size — and get the dimensional numbers from a sewn lot run inside a real load. If the calendar cannot carry both, run the shrinkage lot and submit the size set late, because the size set is not the thing the pattern is built from.
Check yourselfYour laundry loads 240 pairs a load because that is what the trolley holds. Why does that matter, and what do you write instead?Show the answer
Because a load is a weight and 240 pairs is not one. Across this size range a fixed count of 240 varies from 151.2 kg to 218.4 kg, which is the difference between a drum 60.5 percent full and one 87.4 percent full. The light load tumbles freely and abrades hard. The heavy one compacts and abrades little. That shows up as 0.57 cm of inseam spread, and more seriously as shade: the big sizes come out dark, the small sizes light, and your whole sealed band is consumed by something that is not a fault. Write the instruction as a weight — loads made to 180 kg, no load single-size — and put a scale at the trolley. The sort costs about 25 minutes a load.