Unlevelness: reading the pattern to find the cause and save the batch

Is unlevelness a fault or a symptom?

Unlevelness on its own is not a fault. It tells you that one part of the cloth dyed differently from another. It does not tell you why. The same appearance comes from residual size, from a steep heating ramp, and from a blocked filter. So the diagnosis does not start with "which levelling agent should we add". It starts with "how is the difference distributed".

The first useful split is the shape of the pattern. There are four, and each points somewhere different:

• Random, scattered patchiness. Light and dark areas with no order to them.
• Directional difference. The variation runs warp-wise, weft-wise, or along the length of the piece.
• Edge-to-centre difference. A systematic shade gap between selvedges and the middle of the cloth.
• Repeating, periodic marks. Bands or lines returning at a regular interval.

The words used to describe it matter too. "Came out unlevel" in a shift report tells the next person nothing. "Shade drifting from light to dark along the piece" tells them where to look.

Record this before the cloth dries and before it leaves the machine. Once the batch is dry and cut, the pattern information is gone and all that remains is "it came out unlevel", which diagnoses nothing.

How do you recognise unlevelness that came from pre-treatment?

Pre-treatment unlevelness has one signature: whatever the machine does, the pattern stays in the same place. The dye went on evenly. The substrate was not even.

There are four main sources.

Uneven absorbency. One area takes up liquor quickly, another slowly, and dye follows the water. The drop test is pre-treatment's exit check. It is the one number worth looking at before the cloth goes into a dyebath.

Residual size. Size left on the yarn sits as a film and delays the dye reaching the fibre. It usually shows warp-wise, because that is where size was applied.

Residual alkali. Caustic not fully rinsed out after scouring or bleaching shifts dyebath pH locally. In reactive dyeing that is a fixation difference, directly.

Residual peroxide. It oxidises reactive and vat dyes. The appearance is usually pale, washed out and scattered. A test strip settles it in thirty seconds.

A fifth source is the fibre itself. Immature cotton and neps take dye differently, and patchy mercerising gives the same result. Chemistry does not close this group. What the dyehouse can do is see it on the greige cloth before the batch is opened.

Which dyeing variable hides the cause?

In dyeing, unlevelness is almost always a speed problem. The dye reached the fibre faster than the liquor could reach every part of the cloth.

Heating ramp. Every degree of a steep ramp passes through the range where dye strikes hard but has not yet distributed. Slowing the ramp through the critical zone does more than adding levelling agent, and it costs nothing.

The typical HT working temperature quoted in the literature for disperse dyeing is 125 to 135 °C. That is a class range, not a product specification. The critical zone where the ramp needs slowing sits below that band, where the dye strikes hardest.

Salt and alkali dosing. Adding salt in one shot applies the electrolyte effect that drives dye onto the fibre as a shock. The same goes for alkali: migration stops the moment fixation begins, so levelness has to be achieved before that point. Dose in steps, or feed continuously with a dosing pump.

Liquor ratio. A short liquor ratio means a high dye concentration and less margin for the machine to distribute it. If the machine was not built for short liquor, the risk rises immediately.

Circulation. A blocked filter, a worn nozzle, low pump speed, fabric jamming inside the machine. All of these produce the same result and none of them appears in the recipe. Checking the filter after an unlevel batch should be routine.

Dye dissolution. Undissolved dye entering the bath is a mass of particles, and it lands on the cloth as specks. With disperse dyes the choice of dispersing agent is precisely this job.

None of those five has anything to do with dye selection. Changing the dye is the most expensive intervention available and the slowest to give an answer, so it goes last in the queue.

What do mechanical causes look like?

Overload the machine and the rope rides on itself, so liquor reaches the trapped part late. The appearance is soft-edged light and dark areas with no clear boundary.

Fabric running in rope form that folds in the same place gets both a crease and a shade difference there. A thin dark line along the edge of a crease mark means the diagnosis is mechanical, not chemical. Adding levelling agent will not change the outcome; look at liquor ratio, loading and the anti-creasing side instead.

A third mechanical source is how the batch was put together. Cloth from different yarn lots, different piece lengths or different weights loaded into one machine produces a difference that belongs to planning, not to dyeing. No recipe closes it.

There is also the transfer problem between machines. A recipe proven on an overflow does not behave the same way in a high-speed jet, and the reverse is equally true. The recipe belongs to the machine.

What does a levelling agent actually do?

A levelling agent works two ways, and products are usually built with one of them dominant.

The first is retarding. The agent either competes with the dye for sites on the fibre, or forms a temporary complex with the dye and lowers the free dye concentration. Uptake slows down and the liquor gets time to distribute.

The second is migration. The agent helps dye that has already gone onto the fibre move from the heavily dyed area to the lightly dyed one. Correcting an unlevel batch in the machine depends on this mechanism, and in reactive dyeing it only works before fixation.

More is not better. As levelling agent dosage rises, exhaustion falls. Colour yield drops, the same shade costs more dye, reproducibility gets worse and the washing-off load goes up. The "let's add a bit extra to be safe" reflex produces a shade that is level and wrong.

Reactive and disperse dyeing do not offer the same room either. On reactives the window is narrow: once alkali goes in, the dye bonds covalently and migration ends. On disperse dyes migration continues through the hold at top temperature, so the chance to correct stays open longer. Treating both processes with the same product logic means always being late on the reactive side.

Product choice follows the fibre and the dye class. On cotton, DK EGP 100, DK EGP 200 and DK EGP NEW work on the levelling side, while DK EG 111 A carries dispersing and levelling functions together. In polyester dyeing the real work sits with the dispersing agent, which is where DK EG 111 and DK EG 111 PLUS belong. Cationic dyeable polyester has a separate penetration problem, and DK SYNTHEDIF is used there. Dosage is product and process specific: pending from the client.

Diagnostic table: what is the pattern pointing at?

The table is a shortcut, not evidence. Once a row matches, go and check what that row points at; two different causes can produce the same appearance.

First response when a batch comes out unlevel

1. Decide while the cloth is still in the machine. Once the batch is unloaded, the correction options narrow.
2. Record the pattern. Photograph it, mark the direction on the cloth, note the piece number and its position in the machine.
3. Measure the difference. Take colour readings from several points across the batch; the acceptance limit depends on the customer standard, so build your own limit from recorded data.
4. Place the origin in one of three groups: pre-treatment, dyeing, mechanical. Matching the appearance against the table narrows it down in a few minutes.
5. Try migration first. On reactives, if alkali has not gone in yet, running at temperature with a levelling agent is often enough. On disperse dyes, extending the hold at top temperature does the same job.
6. If migration does not work, two options remain: dye over the top to a darker shade, or strip. Stripping costs strength and handle, and on dark shades it rarely ends cleanly.
7. Do not start a second batch on the same recipe until the root cause is closed. A corrected batch is not a corrected process.