Sliver Coiling: Can Geometry, Withdrawal and Quality Faults

A sliver can may look neatly filled and still feed poorly at the next machine. A loop can catch its neighbour, the top plate can tilt, or the strand can stretch on its way to the creel. The visible fault may appear long after the sliver left the card or drawframe.

Sliver coiling should place the strand in a can so that it can later leave in a controlled order. A useful check therefore follows the material through both stages. Looking only at the delivery trace, or only at the full can, leaves part of the process unseen.

This guide uses original examples to connect can geometry, fill weight, spring response and withdrawal faults. All trial values are illustrative. They are not machine settings or acceptance limits for a particular fibre blend.

Follow the strand beyond the delivery rollers

At a card or drawframe, the delivered sliver enters a guide or coiler tube. The coiling system lays it in loops while moving the loop pattern across the can. A support plate below the sliver moves as the load builds. The filled can then travels to another process, where the strand is withdrawn.

These stages have different jobs. The drafting system controls the delivered strand. The coiler arranges it. The can supports and carries it. The creel guides it into the next machine. A fault in one stage can be mistaken for a fault in another if those boundaries are not clear.

For example, a good mass trace before deposition does not establish that the strand remains sound after storage. Nor does a creel break prove that the original fibre blend was unsuitable. The useful question is where the change first becomes visible.

Record the locations of the actual sensors before interpreting their charts. Rieter's drawframe brochure describes delivery monitoring at the calender discs. That location measures the delivered sliver; it cannot directly observe all later events inside the can or during transport.

This leads to a practical comparison. Check the material near delivery using an approved sampling method, then check corresponding material after withdrawal. Keep the blend, test method and moisture conditions comparable. Take care that the sampling itself does not stretch the soft strand and create the very fault being investigated.

Understand the geometry before using a ratio

Coiling combines motions. In one arrangement, the coiler head rotates while the can turns more slowly. Other designs move parts of the coiling assembly over a stationary can. A drawing that shows one arrangement does not describe every machine.

The historical GB2084199A coiler patent describes a planetary system for a stationary can. It also distinguishes under-centre and over-centre coiling by comparing coil diameter with can radius. This is useful engineering background, not a current operating instruction.

Define the dimensions before comparing numbers:

• Can diameter, D: the width across the inside of the can.

• Can radius, R: half that diameter.

• Coil diameter, d: the diameter of the loop being laid.

Here, under-centre means d is smaller than R. Over-centre means d is larger than R. The diameter of the sliver strand itself is a different quantity.

A dimensional example

Take an imaginary can with an internal diameter of 800 mm. Its radius is 400 mm. Two possible loop diameters illustrate the naming rule:

The larger numerical ratio in this table goes with the smaller coil. That follows because d is the divisor. It is easy to reverse the labels if a ratio is copied without its definition.

These dimensions do not specify the offset of the loop centre, wall clearance or relative speeds. They therefore do not form a complete machine setup. Actual sliver width and the way successive loops overlap also matter. Use the machine maker's geometry and the intended can size when setting the equipment.

A central gap can be part of the intended pattern. Its appearance alone does not prove poor filling. Equally, a small gap is not proof of greater useful capacity. Judge the whole build and its withdrawal behaviour, including whether adjacent layers lift together.

Inspect the pattern as it grows

The finished top layer hides much of the can. Watch a safe, approved observation point during filling, then inspect accessible stages after the machine stops. Look for a loop pattern that remains consistent, rather than one that crowds against the wall or builds a persistent high area.

The arrangement of successive layers can matter as much as the first circle. Trützschler's TC15 brochure describes deliberate offsets between layers in its T-MOVE system. It also describes a particular can-change sequence. These are examples of how a maker controls deposition, not features to assume on every coiler.

Keep the tube and plate condition in the investigation. Fibre deposits, a rough contact surface or the wrong component can disturb the strand even when the main drive speeds appear normal. Compare the fitted parts with the specified assembly and inspect their condition during the approved maintenance stop.

Rieter's brochure describes its shaped coiler tube, underside surface and model-specific start-up controls. It also cautions that changing coiler speed or tube size to address difficult materials can affect sliver quality. A quick change that clears one symptom therefore needs an output check before it becomes routine.

Do not judge a change solely by whether the can accepts more weight. The gain may be lost if the next machine produces more breaks or unusable bottom sliver. Useful capacity means material that can be handled and processed within the required quality limits.

Reconcile fill weight, length and time

A can's gross weight includes its shell, base, plate and other fitted parts. Subtract the actual empty weight before calculating the amount of sliver. Cans of the same nominal size need not have identical tare weights.

Suppose a full can weighs 55 kg and its measured empty tare is 15 kg:

Net sliver mass = 55 − 15 = 40 kg.

For this example, the sliver has a measured linear density of 5 ktex. Since 1 ktex equals 1 gram per metre, it contains 5 g in each metre. The estimated length is:

Length = 40,000 g ÷ 5 g/m = 8,000 m.

At a steady delivery speed of 400 m/min, with no stops:

Running fill time = 8,000 ÷ 400 = 20 minutes.

This is a mass-and-length balance, not a promise that the can will fill in exactly 20 minutes. Stops, start-up speeds, discarded material and measurement error can change the observed result. A variable-speed run needs the actual delivered-length record or a suitable integration of speed over time.

If the same run includes two minutes of stopped time, its elapsed fill time becomes 22 minutes. Dividing 8,000 m by 22 gives 363.64 m/min, the average over that elapsed period. It does not mean that the running speed was 363.64 m/min.

Compare calculated and recorded length before changing a fill target. A mismatch might come from the tare, linear-density basis, counter location or material removed during filling. ApparelDiary's yarn-count guide explains the mass-per-length units used in this calculation.

Keep moisture conditions with the measurements. A mass difference is not automatically a fibre loss. Use the same agreed condition for the weighed material and its linear-density sample, or explain the correction applied.

Treat the support plate and spring as an assembly

The plate needs to respond to the changing load without an abrupt drop, tilt or jam. The spring is part of that response, but wall contact, guides and the plate itself can also affect movement. Replacing the spring without checking the rest of the can may leave the fault in place.

Can makers supply different support systems. For example, Jumac's accessories information distinguishes box-spring and pantograph arrangements. Identify the fitted design and its intended duty before comparing two cans. Their nominal outside dimensions alone do not establish that their load response should match.

Compare force and travel on the same basis

A technician uses an approved static test to compare two example assemblies. The net applied force rises from 50 N to 250 N. Over that same interval, assembly A moves from 100 mm to 500 mm; assembly B moves from 100 mm to 600 mm.

The average change in force per unit travel is:

• Assembly A: (250 − 50) ÷ (500 − 100) = 0.50 N/mm.

• Assembly B: (250 − 50) ÷ (600 − 100) = 0.40 N/mm.

These equal 500 N/m and 400 N/m. Assembly B's measured slope is 20% lower than A's over this interval: (0.50 − 0.40) ÷ 0.50 × 100.

That result describes a difference. It does not establish the correct spring for the process or prove why the difference exists. The example is not a manufacturer tolerance, and the two assemblies may differ in more than spring stiffness.

Record the starting position, preload, direction of loading and any rubbing. Compare loading and unloading where the approved method requires it. A single slope can hide a sticking plate or a large difference between the two directions. Work on the stored-energy assembly through the maker's maintenance procedure.

Research supports taking this check seriously. Singh, Bhowmick and Vaz's 2019 study examined combed cotton using three spring-stiffness levels, three delivery speeds and different coil positions. Spring condition and coil position affected unevenness in that experiment; delivery-speed effects were not significant within its tested range. The finding supports investigating can condition. It does not supply a universal spring value or prove that speed never matters.

Recognise thinning caused by unwanted draft

Sliver can stretch when fibres slide relative to one another. This unwanted draft changes the amount of material per unit length even if no fibre leaves the strand. A break is only one possible outcome; a thin place can remain continuous and travel onward.

Consider an idealised section that originally measures 1.00 m and weighs 5.00 g. Handling extends that same material to 1.04 m, with no mass loss. Its new linear density is:

5.00 ÷ 1.04 = 4.8077 g/m, or 4.8077 ktex, rounded.

The length rose by 4%, but the linear density fell by:

(5.00 − 4.8077) ÷ 5.00 × 100 = 3.846%, using the unrounded value.

The two percentages differ because they use different denominators. Multiplying the original count by 0.96 would be an approximation, not the exact mass balance for a 4% length increase.

This example assumes the extended length is the length being assessed. Real sliver may partly recover or change further during handling. Do not treat all observed extension as a permanent draft without checking the material under the agreed test conditions.

Look at the withdrawal path when investigating such thinning. A caught loop can raise the load on the strand. An unsupported span, poor alignment or a contact point can create another disturbance. Record where the strand catches or stretches before changing upstream drafting settings.

Observe the plate as the can empties. If it fails to rise as intended, the strand's free path may change. If adjacent coils lift as a group, note the level in the can and whether the event repeats. These details help distinguish a repeatable handling problem from an isolated break.

Read a fault's position in the correct order

In a normal top-withdrawal sequence, the last material laid down leaves first. The first material laid down is near the bottom and leaves later. This reverses the order of events between filling and withdrawal.

Return to the 8,000 m example. Assume orderly top-first withdrawal at 200 m/min, with no discarded length, breaks, joins or loops taken out of order. Total running withdrawal time is:

8,000 ÷ 200 = 40 minutes.

The first 100 m laid into the can will appear during the last 0.5 minute of withdrawal. The last 100 m laid will appear during the first 0.5 minute. A problem seen near an empty can may therefore relate to filling start-up, bottom-layer condition or the plate's final position.

Suppose an event occurred after 2,000 m had been laid during filling. On this idealised sequence, 8,000 − 2,000 = 6,000 m lies above it. The point reaches the withdrawal end after 6,000 ÷ 200 = 30 running minutes. At the earlier 400 m/min fill speed, it had occurred five running minutes into filling.

Real records need more care. Stops add clock time, and discarded tails change the length mapping. A loop pulled from below can also break the assumed order. Use the example to understand the direction of the sequence, then reconstruct an actual event from the can's length and stop records.

Check stops, deposits and storage separately

A small fibre deposit can enter the can at a stop rather than at a fixed location in every fill. Rieter's CLEANtube report discusses fibre accumulations called “mice” in a cotton application with substantial waste or noil. It explains that deposits can enter at interruptions within a fill, so they need not appear only at the top of a full can.

This gives a specific investigation route: compare the position of a lump with recorded interruptions and tube condition. Do not assume that every lump has this cause. Confirm its material and appearance, then check whether it existed before the coiler.

Storage time deserves its own comparison. A separate 2019 can-storage study compared stored and unstored combed-cotton material. Its public abstract reports effects associated with storage, spring condition and coil position. The study supports including these variables in a trial; it does not establish one safe storage period for all slivers.

Record fill completion, storage duration and transport route. Protect material from contact with dirty surfaces and other lots under the site's handling rules. Note jolts, a damaged rim or poor caster movement as observations to investigate, without assuming that every older can is defective.

Use a crossover trial to locate the cause

If can A repeatedly gives trouble, comparing it with can B on a different delivery head leaves two things changed at once. A more useful trial puts both cans through both heads. Keep the blend, nominal count, fill target, storage interval and withdrawal route as similar as practicable.

The following invented trial counts creel breaks during withdrawal. Each cell covers 24 km from repeated fills under its stated combination:

Across the two heads, can A has 13 breaks over 48 km. Its observed rate is 13 ÷ 48 × 100 = 27.08 breaks per 100 km, rounded. Can B has 3 over 48 km, or 6.25 per 100 km.

The pattern directs attention toward the can or something that follows it. It does not yet prove a spring fault. The can may carry a different plate, follow a different route or be handled differently. The small trial also leaves sampling uncertainty, so do not present its rate difference as a guaranteed future saving.

Alternate the trial order to reduce the effect of a gradual change during the shift. Record each fill separately, including can level and break location. Repeated readings from one can are not independent evidence about every can of that design or age.

If the trouble follows the delivery head instead, inspect the head's coiler path and delivery condition. If it follows one withdrawal position, examine that creel route. If it appears mainly in bottom sliver across both cans, review the common load, storage and end-of-can conditions.

After a repair, repeat the relevant comparison with the same definitions. Keep breaks, thin-place findings and discarded mass as separate responses. One may improve while another worsens. ApparelDiary's practical SPC guide explains how stable definitions help distinguish a process change from ordinary variation.

Sample the can without hiding local changes

A whole-can average can conceal a fault concentrated near one end. Keep the top, middle and bottom observations separate during the investigation. State how those locations were identified: by delivered length, withdrawal length or another defined method.

Equal heights inside a can do not necessarily contain equal lengths of sliver. The lower layers can be packed differently from the upper ones, and the support plate moves. If the trial divides the material into length thirds, label them from the length record. Do not substitute a visual third of the can's height without explaining that change.

Where practical, collect the samples in their normal withdrawal sequence. Pulling a buried loop out through several layers can disturb the strand and leave the remainder in a different state. Follow the approved sampling method and record any material removed or discarded. That record also protects the event-to-length calculation used earlier.

Why a correct mean may be misleading

Consider two separately measured 10 m sections. One weighs 49 g and the other weighs 51 g. Their linear densities are 4.9 g/m and 5.1 g/m. Against a 5.0 g/m reference, their deviations are −2% and +2%.

If they are combined before weighing, the result is 100 g over 20 m, or exactly 5.0 g/m. The pooled mean looks correct, although neither section equals that value. This invented example illustrates lost location information; it does not set a tolerance or specify a laboratory sample length.

Keep those count measurements separate from a continuous evenness trace. A short weighed sample and an electronic trace describe different views of the material. Use each for its intended question, with its own method, settings and sampling basis.

For the coiling trial, the important link is the sample's place in the can and its handling history. If both before-can and after-withdrawal samples shift together, look upstream as well. If a new change appears only after handling, inspect the relevant deposition, storage and withdrawal steps before assuming that the drafting system caused it.

Close the investigation with usable material

The final check should connect the repair with what the next process receives. Confirm the intended coil pattern, adequate support and a stable withdrawal path. Then inspect the resulting material against the required quality measures.

Record enough detail to repeat the successful condition: can identity and assembly, material, fill basis, relevant machine settings and observed limits. A note that says only “coiling adjusted” does not explain which part of the problem was resolved.

Keep the conclusion as specific as the evidence. A repaired can that feeds well in a cotton trial has passed that trial. It has not established the correct support system for every blend, count or storage period. Sliver coiling works best when the deposited pattern and the later withdrawal are assessed as one material-handling process.

References and further reading

• GB2084199A: historical planetary sliver-coiler geometry and over-centre/under-centre definitions.

• Rieter: drawframes brochure, including delivery monitoring, coiler design and start-up controls.

• Trützschler: TC15 brochure, January 2022, with T-MOVE layer placement and can changing.

• Jumac: can accessories and support-system designs.

• Singh, Bhowmick and Vaz: Influence of Can-spring Stiffness, Delivery Speed and Sliver Coils Position on Unevenness, 2019.

• Rieter: RSB-D 50 CLEANtube report, February 2023.

• Singh, Bhowmick and Vaz: Effect of Can-storage Parameters of Finisher Drawframe on Combed Ring Spun Yarn Quality, 2019; public abstract.