A weaving chart can tell you which strand passes above another at each crossing. It does not, by itself, tell you how much yarn lies inside a measured length of cloth. The missing information includes the path’s height, spacing and shape between the recorded crossings.
The short answer is that crossing order is a structural relationship, while yarn length is a geometric quantity. Two paths can follow the same over-under sequence and have different lengths. To calculate the difference, define a centreline model or measure a yarn length under stated conditions; do not infer it from the pattern alone.
This original paper lesson continues Weaving, Learning That Transfers into Life and How X Works. The ideal paths below are invented mathematical constructions, not measurements of a named fabric, a loom setup or a culturally specific textile design.
eduKateSG · Weaving
Find your next learning step
Separate crossing states from the distances and curves they leave unknown.
ROUTE 1 · CHAPTERS 1–7
Build a geometric counterexample
Keep the crossing record fixed while changing path lengths.
ROUTE 2 · CHAPTERS 8–14
Read ratios and model choices
Preserve scale, reference length and assumptions between nodes.
ROUTE 4 · CHAPTERS 24–29
Connect geometry to material evidence
Keep inverse claims, fibre paths, mass and conditions distinct.
ROUTE 5 · CHAPTERS 30–35
Practise and continue
Check a changed path and follow the existing making guides.
Chapter contents
Build a geometric counterexample · Chapters 1–7
1. Begin with the strand hidden inside the span
2. Keep crossing state separate from height
3. Declare a planar centreline model
4. Specify five original crossing nodes
5. Calculate the polygonal path length
Read ratios and model choices · Chapters 8–14
9. Relate the model to a declared crimp measure
10. Preserve the denominator when reporting a percentage
11. See what the polygonal path minimises
12. Keep the same nodes and move the bending
Derive useful bounds · Chapters 15–23
15. Check a small-slope approximation
16. Compare two bounds without solving the integral
17. Explain total variation through the first path
18. Check where those variation bounds are tight
19. See why a height envelope alone gives no finite upper bound
20. Add a slope bound and recover a useful upper limit
21. Allow a sideways path without changing the projected crossings
Connect geometry to material evidence · Chapters 24–29
24. Invert only within the declared shape family
25. Keep yarn geometry separate from fibre geometry
26. Use a length correction in a mass ledger carefully
27. Keep finishing stages in a separate record
Practise and continue · Chapters 30–35
30. Build a practical paper comparison
31. Practise with a changed segment geometry
32. Diagnose the first missing idea
33. Questions readers often ask
CHAPTER 1 OF 35 · Build a geometric counterexample
1. Begin with the strand hidden inside the span
Place two marks on a fictional piece of cloth, eight millimetres apart in the warp direction. A warp yarn travels between them, rising and falling around weft yarns. Its path can be longer than the straight distance between the marks.
The eight-millimetre span and the yarn’s path length answer different questions. One measures separation along the chosen fabric direction. The other follows the strand’s route through space. A ruler laid along the cloth cannot automatically recover the second quantity.
Begin by naming the endpoints and the direction being measured. The cloth may also be stretched, relaxed or finished under different conditions. A length comparison is meaningful only when the two quantities refer to the intended state and the same portion of material. This lesson first fixes those conditions within a simple geometric model.
CHAPTER 2 OF 35 · Build a geometric counterexample
2. Keep crossing state separate from height
Use a binary crossing record in which 1 means warp above weft and 0 means warp below weft on the stated reference face. The sequence 1,0,1,0,1 therefore records five alternating relationships.
That sequence contains no millimetres. It does not specify the vertical distance between neighbouring crossing levels, the diameter of either yarn, or how sharply the warp bends. It also does not specify tension or compression at contact points.
A structural drawdown can be correct while those geometric quantities remain unknown. This is not a defect in the chart; it is a limit of what its symbols encode. A useful reader asks which additional measurements or assumptions are needed for the next question. Copying more rows of the same binary pattern does not necessarily supply the missing length scale.
CHAPTER 3 OF 35 · Build a geometric counterexample
3. Declare a planar centreline model
Represent one yarn centreline in a cross-section by points (x,z). The x-coordinate runs along the fabric span; z describes height. For now the path does not move sideways, and x increases monotonically from one endpoint to the other.
These restrictions simplify the geometry. A real yarn has thickness and can deform in three dimensions. Our path is a mathematical centreline, not a solid thread with a prescribed diameter or contact law.
We will first connect specified crossing nodes by straight segments. Such a polygonal path gives a transparent calculation using Pythagoras. Its sharp corners are an idealisation. They do not claim that real yarn bends through zero-radius corners. Later we compare other paths between the same nodes and identify which conclusions survive when the straight-segment assumption is removed.
CHAPTER 4 OF 35 · Build a geometric counterexample
4. Specify five original crossing nodes
Take x-coordinates 0, 2, 4, 6 and 8 millimetres. Assign heights +0.5, −0.5, +0.5, −0.5 and +0.5 millimetres. The reference weft level is zero at the named crossing locations.
The resulting above-below sequence is 1,0,1,0,1. Each neighbouring pair differs by two millimetres horizontally and one millimetre vertically.
The labels matter. A height of +0.5 is a coordinate in this model, not a universal offset for woven yarn. The example fixes five points so that a learner can reconstruct every segment without guessing material properties. Draw the nodes on perpendicular axes with a stated scale before calculating. A diagram drawn with exaggerated vertical scale should be labelled accordingly; its apparent angle is not an additional measurement.
CHAPTER 5 OF 35 · Build a geometric counterexample
5. Calculate the polygonal path length
Each of the four segments has length √(2² + 1²) = √5 millimetres. Total yarn-path length is therefore ℓ = 4√5, approximately 8.944 millimetres.
The fabric-direction span is L = 8 millimetres. The excess path length is approximately 0.944 millimetres. Expressing that excess relative to the fabric span gives c = ℓ/L − 1 = √5/2 − 1, approximately 0.1180.
Thus this model has about 11.8 per cent excess length relative to its span. This number follows from both the node spacing and the chosen straight-segment geometry. It does not follow from the alternating sequence alone. Check the calculation independently by dividing the total path length by four and recovering the length of one two-by-one right-triangle segment.
CHAPTER 6 OF 35 · Build a geometric counterexample
6. Change only the crossing heights
Keep the x-coordinates and above-below sequence unchanged, but use heights +1, −1, +1, −1 and +1 millimetres. Every segment now has horizontal difference two and vertical difference two.
Its length is √8 = 2√2 millimetres. Four segments give ℓ = 8√2, approximately 11.314 millimetres. The excess relative to the eight-millimetre span is √2 − 1, approximately 41.4 per cent.
| Model | Crossing sequence | Fabric span | Polygonal path length |
|---|---|---|---|
| Heights ±0.5 mm | 1,0,1,0,1 | 8 mm | 4√5 mm |
| Heights ±1 mm | 1,0,1,0,1 | 8 mm | 8√2 mm |
Both paths satisfy the same binary record. Their lengths differ by about 2.37 millimetres. This constructive counterexample demonstrates the information missing from the crossing sequence without needing to speculate about a real fabric’s exact shape.
CHAPTER 7 OF 35 · Build a geometric counterexample
7. Generalise one repeated segment
Let a be the positive horizontal distance between neighbouring crossing nodes and h the magnitude of their vertical difference. Suppose every segment has those magnitudes. For N segments, L = Na and ℓ = N√(a² + h²).
The ratio is ℓ/L = √(1 + (h/a)²). The number of identical segments cancels. Repeating the same ideal geometry increases total yarn length and fabric span together, leaving their ratio unchanged.
The important dimensionless input is h/a. Increasing vertical excursion while holding spacing fixed raises the ratio. Increasing spacing while holding excursion fixed lowers it. A crossing count supplies neither of those geometric changes by itself. Use this formula to identify which quantity a proposed comparison actually holds constant before predicting whether more or less yarn is required.
Double every x-coordinate and every z-coordinate in the first model. Each segment length doubles, the total span becomes sixteen millimetres, and the path length becomes 8√5 millimetres.
The ratio ℓ/L stays √5/2. The same relative geometry at another size has the same excess-length fraction. This is a useful dimensional check on the formula: a ratio should not change merely because all lengths are expressed at a uniform new scale.
By contrast, doubling only the horizontal spacing changes h/a and therefore changes the ratio. A larger drawing is not always a physically scaled version of the same structure. Ask whether the vertical and horizontal dimensions were transformed together. This distinction helps when comparing illustrations, microscope images or samples photographed at different magnifications.
CHAPTER 9 OF 35 · Read ratios and model choices
9. Relate the model to a declared crimp measure
A common fabric-yarn crimp measure compares straightened yarn length with its span in fabric, using the fabric span as denominator. The public description of ISO 7211-3 specifies that relationship and notes that straightening tension depends on the yarn.
Our ideal geometric c = ℓ/L − 1 resembles that ratio when ℓ represents the corresponding straightened, unextended centreline length. That correspondence is an assumption, not something a binary crossing chart verifies.
The full laboratory procedure requires its actual method and conditions. This lesson uses only the public definition to explain why the two lengths differ; it does not reproduce the standard or claim that the paper calculation constitutes a compliant test. Keep mathematical path length and a measured straightened-yarn result distinct until the measurement model connects them.
CHAPTER 10 OF 35 · Read ratios and model choices
10. Preserve the denominator when reporting a percentage
For a fictional span L = 80 millimetres and a corresponding yarn length ℓ = 100 millimetres, the excess relative to fabric span is 20/80 = 25 per cent. The reduction from yarn length to span is 20/100 = 20 per cent.
Both use the same twenty-millimetre difference. They differ because their reference quantities differ. A record saying only “twenty per cent take-up” would be ambiguous without its equation and stage definition.
The existing Yarn Crafts lesson on stage-specific take-up and shrinkage develops that accounting problem in depth. Here the main task is earlier: explain why a yarn path can be longer even when the crossing order is known. Once a path-length or measurement result enters a material budget, carry its denominator and conditioning state into the separate stage ledger.
CHAPTER 11 OF 35 · Read ratios and model choices
11. See what the polygonal path minimises
Suppose all five crossing-node coordinates in the first model are known exactly. Any rectifiable path connecting each neighbouring pair has length at least their straight-line distance.
Adding those four lower bounds gives total length at least 4√5 millimetres. Our polygonal model attains that lower bound. A smoother curved path through the same nodes can be longer; knowing the nodes does not force the yarn to travel along the shortest connection between them.
This is a geometric lower bound under the planar ordered-node description. It does not establish that the shortest path is physically attainable for a yarn of finite thickness with contact and bending constraints. A bound can remain useful even when unattainable physically, provided it is labelled as a bound rather than presented as the predicted actual yarn length.
CHAPTER 12 OF 35 · Read ratios and model choices
12. Keep the same nodes and move the bending
For each two-millimetre interval with vertical change one millimetre, replace the direct diagonal segment by a one-millimetre horizontal section followed by a diagonal that changes height by one over the remaining one millimetre.
Each interval now has length 1 + √2, rather than √5. Across four intervals, total length is 4 + 4√2, approximately 9.657 millimetres. The original five crossing nodes and their binary sequence remain unchanged.
The example isolates a second missing ingredient: even complete crossing-node coordinates do not identify the path between nodes. Moving where the height change occurs alters the length. The piecewise construction again has ideal sharp corners, but it is enough to show that a node record and a continuous-path record are different kinds of information.
CHAPTER 13 OF 35 · Read ratios and model choices
13. Use an arc-length integral for a smooth path
For a differentiable planar centreline z(x) over span zero to L, arc length is ℓ = ∫₀ᴸ √(1 + z′(x)²) dx. The derivative describes local slope, and the square root converts a horizontal increment into a small path-length increment.
A constant height has z′ = 0 and length L. A constant slope h/a over interval a gives √(a² + h²), recovering the straight-segment calculation.
This is a useful bridge from the concrete triangle to calculus. The integral accumulates local contributions; it does not average heights and then draw one line through the average. Two paths can have the same mean height or the same endpoint heights while their slope distributions differ. Length responds to those slopes, so height summaries alone are generally insufficient.
CHAPTER 14 OF 35 · Read ratios and model choices
14. Give a sinusoidal model its own assumptions
For another original family, take z(x) = A sin(2πx/λ) over one wavelength λ, with A nonnegative. Its derivative is q cos(2πx/λ), where q = 2πA/λ.
The length ratio becomes (1/2π)∫₀²π √(1 + q²cos²u) du. The ratio depends on A/λ, but it is not the polygonal formula with the same amplitude inserted automatically.
A sinusoid is a smooth mathematical model. It is not a universal description of yarn inside a plain weave. Real contacts, yarn flattening and material response can produce different centreline shapes. Choosing a smooth function makes the calculation possible while adding a shape assumption. A good explanation states that choice before presenting a numerical result as though it came directly from the weaving chart.
When |q| is small, expand the integrand: √(1 + q²cos²u) ≈ 1 + q²cos²u/2. The average of cos² over a full period is one half, giving excess ratio c ≈ q²/4.
Keeping the next term gives c ≈ q²/4 − 3q⁴/64. The correction is negative. The leading expression therefore should not be treated as exact when slopes are no longer small.
The inequality √(1 + t) ≤ 1 + t/2 for nonnegative t also gives c ≤ q²/4 for this sinusoidal family. That is an upper bound valid beyond the approximation’s small-slope range, although it may become loose. A formula can serve as an approximation and as a one-sided bound for different reasons; state which use the current comparison requires.
CHAPTER 16 OF 35 · Derive useful bounds
16. Compare two bounds without solving the integral
For the sinusoid, total vertical variation over one period is V = 4A. A geometric lower bound is ℓ ≥ √(λ² + V²). A separate upper bound from concavity is ℓ/λ ≤ √(1 + q²/2).
For A/λ = 0.1, these give excess ratio at least √1.16 − 1, approximately 0.0770, and at most √(1 + 0.02π²) − 1, approximately 0.0943.
These bounds constrain the original sinusoidal model without evaluating its arc-length integral numerically. They do not identify the crimp of an actual fabric. Their usefulness is methodological: first ask whether a justified range already answers the comparison. A long numerical calculation is unnecessary when independently derived bounds settle the question at the needed precision.
CHAPTER 17 OF 35 · Derive useful bounds
17. Explain total variation through the first path
Total vertical variation adds every absolute rise and fall. For the first five-node path, there are four changes of magnitude one millimetre, so V = 4 millimetres. The net endpoint height change is zero.
Net change and total variation are therefore different. Cancelling rises against falls would discard the movement that contributes to path length. A yarn ending at its original height can still travel a longer route than the fabric span.
For an x-monotone planar path, the length satisfies √(L² + V²) ≤ ℓ ≤ L + V. The lower bound follows by accumulating vectors (dx, |dz|); the upper follows from √(dx² + dz²) ≤ |dx| + |dz|. These bounds retain more information than endpoints alone, while still leaving the detailed arrangement of slopes unresolved.
CHAPTER 18 OF 35 · Derive useful bounds
18. Check where those variation bounds are tight
Our original polygonal path has L = 8 and V = 4 millimetres. The lower bound is √80 = 4√5 millimetres, exactly its length, because every segment has the same absolute slope one half.
The changed path with horizontal pieces has the same L and V but length 4 + 4√2, larger than that lower bound. Concentrating vertical movement into part of each interval raises the length relative to spreading it at constant absolute slope.
The upper bound is twelve millimetres. Paths with increasingly steep short rises and falls separated by horizontal travel can approach it, subject to the mathematical family. A real yarn’s bending radius and contact geometry would restrict those choices. A statement about an infimum or supremum within ideal paths should not be mistaken for a physical construction instruction.
CHAPTER 19 OF 35 · Derive useful bounds
19. See why a height envelope alone gives no finite upper bound
Consider a float section with centreline z(x) = b + A sin(2πnx/L), where b > A > 0 and n is an integer. Every path stays strictly above the reference level, and its height remains between b − A and b + A.
Yet its total vertical variation is 4An. Its length is at least that variation, so it grows without bound as n increases. A fixed span and bounded height have not bounded length when arbitrarily rapid oscillations are allowed.
This is an information-limit construction, not a proposal for a physical yarn. Real thickness, curvature and contact conditions prevent arbitrary fine oscillations. The example shows exactly why those additional constraints matter. A crossing-state record or a height envelope cannot silently supply a bound on slope or on how often the path bends.
CHAPTER 20 OF 35 · Derive useful bounds
20. Add a slope bound and recover a useful upper limit
If the centreline is x-monotone and satisfies |z′(x)| ≤ k throughout its span, then ℓ ≤ L√(1 + k²). Unlike a height envelope alone, this condition limits the local path length per unit horizontal distance.
For L = 8 millimetres and k = 1/2, the upper limit is 4√5 millimetres. The original alternating polygonal path has that absolute slope on each segment and reaches the limit within its piecewise model.
The slope bound must come from evidence or a justified modelling restriction. It cannot be deduced merely from a photograph showing that yarn stays within a thin layer. A thin layer can contain many bends in a mathematical construction. Keep the source of each geometric restriction visible; otherwise the resulting bound can look rigorous while relying on information the observation never supplied.
CHAPTER 21 OF 35 · Derive useful bounds
21. Allow a sideways path without changing the projected crossings
A three-dimensional centreline can be written r(x) = (x,y(x),z(x)). Its length is ∫₀ᴸ √(1 + y′(x)² + z′(x)²) dx. A cross-section that shows only x and z omits the nonnegative sideways contribution.
For a straight segment with horizontal change two, vertical change one and sideways change one, the actual length is √6. Its x-z projection has length √5. The above-below relationship can remain the same while the unobserved sideways motion increases length.
This gives a useful evidence boundary. A projected path can supply a lower bound on spatial length when it genuinely represents the corresponding projection. It does not supply equality unless the omitted motion is absent or otherwise accounted for. A physical assumption that yarns stay in one plane should therefore be stated and checked, rather than inferred from a two-dimensional drawing chosen for convenience.
CHAPTER 22 OF 35 · Derive useful bounds
22. Carry geometric uncertainty through the ratio
Suppose the repeated polygonal model remains appropriate, but neighbouring horizontal spacing a is only known to lie between 1.9 and 2.1 units, and vertical difference h between 0.9 and 1.1. Both are positive.
The ratio increases with h and decreases with a. Therefore the excess fraction lies between √(1 + (0.9/2.1)²) − 1 and √(1 + (1.1/1.9)²) − 1. This interval follows directly from the allowed parameter rectangle; no probability distribution has been assigned.
If the two quantities share measurement errors or must satisfy another geometric constraint, not every corner of that rectangle may be possible. The interval remains an outer bound when the true possibilities lie inside it, but calling its endpoints equally likely would add an unsupported claim. Keep a geometric range distinct from a statistical confidence interval. Both can be useful, provided their meanings remain attached to the reported result.
CHAPTER 23 OF 35 · Derive useful bounds
23. Preserve curve length when changing the viewpoint
A rigid rotation of a three-dimensional curve changes its coordinate components but preserves its actual length. The dot product of each small tangent vector with itself is unchanged under an orthogonal rotation, so its accumulated arc length is unchanged too.
Its projected image can nevertheless become much shorter. A segment viewed nearly end-on has a small projected span even though the material segment has not shortened. This is another reason that an image ratio cannot automatically be read as a physical yarn-length ratio.
For a practical comparison, identify whether two images show the same viewing geometry and scale before treating their apparent slopes as measurements. A change of viewpoint, a change of material shape and a change of magnification are distinct transformations. The mathematical record should preserve which one occurred. When that information is missing, report a conditional interpretation or seek an appropriate additional view instead of pretending that the hidden depth has already been measured.
CHAPTER 24 OF 35 · Connect geometry to material evidence
24. Invert only within the declared shape family
For the constant-absolute-slope polygonal family, c = √(1 + (h/a)²) − 1. Rearranging gives h/a = √((1 + c)² − 1), taking a nonnegative excursion magnitude.
If c = 0.25, this gives h/a = 0.75. With a = 2 millimetres, h = 1.5 millimetres. Four segments then have length 4√(4 + 2.25) = 10 millimetres across an eight-millimetre span.
This does not mean a measured 25 per cent crimp identifies a real crossing height of 1.5 millimetres. A different path family or unequal slopes can produce the same length ratio with different heights. The inversion is conditional on the model, just as the forward calculation is. One measured scalar cannot generally reconstruct an entire spatial curve.
CHAPTER 25 OF 35 · Connect geometry to material evidence
25. Keep yarn geometry separate from fibre geometry
The path discussed here follows a yarn centreline as it passes through fabric. Individual fibres within that yarn can themselves follow inclined, twisted or irregular paths. Those are different geometric levels.
A centreline length therefore need not equal the sum of fibre lengths, nor does it determine their orientations or the yarn’s breaking behaviour. Conversely, knowing a yarn’s twist does not identify its out-of-plane route through a finished weave.
The existing Yarn articles develop fibre-path and test-length questions separately. Use those connections when moving from cloth structure into yarn construction or strength. Keeping the levels distinct helps avoid double counting a length allowance: one factor may describe yarn travel through fabric, while another describes fibres within yarn. Each should be attached to the material quantity it actually measures.
CHAPTER 26 OF 35 · Connect geometry to material evidence
26. Use a length correction in a mass ledger carefully
For a small fictional example, suppose one hundred equal warp yarns each span one metre of fabric and each has centreline length 1.10 metres within that span. Their total yarn length is 110 metres.
If the relevant yarn linear density is twenty grams per kilometre under compatible conditions, their model mass is 110 × 20/1000 = 2.2 grams. Ignoring the path excess would give two grams. This calculation includes only the named warp yarns and excludes weft, ends, waste and other components.
The example shows why geometry can matter to quantities. It is not a complete fabric-weight prediction. Linear density, conditioning and length reference must be compatible, and the geometry should describe the portion being counted. An accurate multiplication cannot repair an input whose definition belongs to a different yarn state.
CHAPTER 27 OF 35 · Connect geometry to material evidence
27. Keep finishing stages in a separate record
A fabric can change dimensions during finishing, while yarn properties and its path can also change. The path-to-span ratio in one state cannot automatically be applied to another state.
Record which dimensions belong to the loom state, relaxed state or specified finished state. The existing stage-specific take-up and shrinkage lesson explains how to reverse explicitly defined stage relationships without adding percentages that use different bases.
Here the important boundary is conceptual. The centreline calculation explains excess route length within a specified geometry. A later finishing allowance describes another transformation with its own evidence. If both are used in one budget, define how the states connect. Otherwise the same shortening effect might be counted twice, or a real change might be omitted because two unrelated percentages have been given the same label.
CHAPTER 28 OF 35 · Connect geometry to material evidence
28. Treat straightening tension as part of measurement
A removed yarn does not necessarily become a unique straight length independent of how it is handled. Too little tension may leave waviness; different tension can also change extension of the yarn itself.
The public ISO description explicitly recognises that the straightening tension varies with yarn characteristics. That is enough to show why a measured length needs its method and conditions. The abstract alone is not a complete laboratory instruction.
For a classroom record, distinguish an ideal unextended centreline length from a hypothetical measured length at a named tension. If those disagree, do not automatically declare one calculation wrong. The geometric model and the physical measurement may be targeting different states. A useful investigation identifies which reference length is intended and what evidence connects the measurement to it.
CHAPTER 29 OF 35 · Connect geometry to material evidence
29. Understand what a photograph can and cannot reveal
An image can show visible crossing relationships and projected spacing when its scale and viewpoint are known. It does not automatically show the complete three-dimensional centreline hidden under other yarns.
Perspective, compression, surface texture and occlusion can affect what is visible. A colour boundary may also reflect yarn colours rather than a crossing’s full shape. A smooth-looking strand in a photograph need not be geometrically straight through its hidden portion.
For an evidence check, label every quantity as directly observed, inferred with a model, or still unknown. A calibrated image might support one projected spacing while leaving vertical height unresolved. The correct response is a conditional length calculation or a bound, rather than a precise reconstructed path drawn from information absent from the image. More decimal places do not recover hidden geometry.
Use two copies of the same five-node crossing record. Give one learner the heights ±0.5 millimetres and another the heights ±1 millimetre, with both using two-millimetre horizontal spacing and straight connecting segments.
Ask each to predict whether the above-below sequence changes before calculating lengths. Then compare the paths, ratios and units. The intended discovery is that the pattern stays fixed while the geometric quantity changes.
For a second round, keep the node coordinates fixed and let one model use direct diagonals while the other inserts the horizontal pieces described earlier. This removes height as the changed variable and isolates shape between nodes. The exercise uses original mathematical paths, so it can teach the distinction without borrowing a traditional motif or implying expertise in a particular weaving practice.
CHAPTER 31 OF 35 · Practise and continue
31. Practise with a changed segment geometry
Try four segments with horizontal distance three units and alternating height difference four units. Keep a consistent planar polygonal model. Find the total span, total path length and excess relative to span.
Each segment is a three-four-five triangle. Span is twelve units and path length is twenty. The excess fraction is eight divided by twelve, or two thirds. This relatively large value is a teaching construction, not a typical fabric parameter.
Now halve both horizontal and vertical dimensions. Span and path length halve while the fraction stays two thirds. Finally halve only the height difference: each segment becomes √13 units long, total path 4√13, and ratio √13/3. Explain why the binary alternating sequence can remain unchanged through all three cases even while the geometric ratios differ.
If a learner equates yarn length with fabric span, return to one right triangle rather than assigning a more complex weave chart. If they count over-under symbols but never use distances, ask which input supplies millimetres.
For stabilisation, compare two paths with equal crossing sequences and different heights, then equal node coordinates and different intervening shapes. Require a sentence naming the changed assumption in each comparison.
For extension, provide a slope bound, a total-variation bound or a measured length ratio and ask what can actually be inferred. The learner should distinguish a forward model, a bound and an inverse claim. Progress appears when they can identify the evidence needed to narrow the answer, rather than forcing every fabric image into one convenient formula.
Does an alternating crossing sequence determine crimp?
No. It records which strand is above the other, while a length ratio also depends on geometry and measurement conditions.
Does a straightened-yarn measurement reveal the whole path?
No. One length is a scalar summary. Several centreline shapes can have that length, and the measurement conditions determine which physical reference is represented.
Is a sinusoid more correct than a polygonal model?
It is smoother, but smoothness alone does not validate it for a particular fabric. The appropriate model needs evidence about the actual geometry and intended question.
Does this lesson predict strength or comfort?
No. Those outcomes require material properties, structure, loading and other evidence beyond the ideal path-length calculation.
CHAPTER 34 OF 35 · Practise and continue
34. Follow the connected making and measurement routes
Read the crossing structure
Return to Weaving for warp, weft, edges and material purpose. Reconstructing weave interlacements from threading and lift plans develops the structural crossing map treated here as a supplied input.
Carry quantities through their actual stages
Reconstructing starting length from stage-specific take-up and shrinkage develops the separate accounting needed to recover a starting length from a finished span. Keep that stage problem distinct from identifying the spatial curve within one stated fabric condition.
Compare material evidence with a model
The core aim of weaving mastery explains sampling, sett, tension and finishing as connected making decisions. Use a real sample and an appropriate method when the question concerns actual cloth, rather than treating a fictional centreline as a measured material result.
CHAPTER 35 OF 35 · Practise and continue
35. Sources, limits and an independent final check
ISO 7211-3’s public description of yarn crimp in fabric provides the stated length comparison and the role of straightening conditions. The official page, checked on 26 October 2026, identifies the 1984 edition as confirmed in 2023. Our node coordinates, polygonal alternatives, smooth-path comparisons and bounds are original educational constructions, not a reproduction of the standard’s procedure.
For a final independent check, use five nodes with two-unit spacing and alternating heights ±0.75. Calculate the polygonal length and ratio, then explain why those same nodes admit a longer path. State one additional constraint that would supply an upper bound.
Continue through Learning That Transfers into Life and How X Works by preserving what each representation actually contains. A crossing chart, a spatial curve and a measured length can support each other while still answering different questions.