Interfaces, slippage, porosity and defects

Strong constituents can still slide

Tube-to-tube contacts transfer load through interfacial interactions. Slippage can limit how much intrinsic tube strength is used. Pores reduce bulk density and change connected load paths; their effects depend on packing and geometry.

A carbon nanotube wall contains strong covalent bonds along its connected carbon network. Bringing two tubes together does not automatically continue those bonds across the contact. Tangential forces between the contacting surfaces must transfer part of the load from one tube or bundle to another. A bundle is a group of tubes; a macroscopic fibre contains many such groups and their contacts. Strong constituents can remain poorly loaded if those contacts slide before much axial force reaches them.

Shear traction means force parallel to the interface per unit actual interfacial area. It differs from the axial stress across a section perpendicular to the fibre. Interactions at a contact can resist relative displacement; bridges or an added interphase can alter that resistance. Frictional resistance after sliding begins is a different state from an intact bonded contact. The surrounding tubes also deform, and contacts need not share force equally. A visible overlap is therefore not automatically a known load-bearing area.

For one deliberately simple contact, let S be the actual mobilised interface area, τ the uniform tangential traction and Q its resultant force. Adding traction over the area gives Q = τS. Equal and opposite forces act on the two contacting bodies. If traction is nonuniform, add the local traction contributions instead; one average is usable only when defined over the same S. This balance does not determine the displacement needed to mobilise Q or the breaking force of the whole fibre.

Load transfer through fibres and interfaces develops force build-up along a buried fibre. Alignment, tube length and load transfer follows the available paths in a nanotube assembly. Here we use a supplied contact state rather than derive a critical length or assume every contact reaches the same limiting shear.

Sliding can absorb work while restricting force transfer

A contact may resist relative movement and still slip. When two surfaces slide against a constant opposing resultant Q_s through a relative displacement δ, the external work needed for that sliding is W_slide = Q_s δ. The symbol δ measures motion of one contacting body relative to the other at this ideal contact, not extension of the entire fibre. If resistance varies, the work is the area under the resistance-versus-slip curve.

In the first figure, an already-sliding contact has Q_s = 0.100 N over δ = 0.500 mm, with constant contact area and resistance stipulated. Its work is 0.100 × 0.000500 = 0.0000500 J = 0.0500 mJ. In this ideal dissipative model that work is not recovered as elastic energy when the motion stops. Prior debond creation, tube deformation, changing overlap, heat transport and the full fibre’s work to fracture are outside this contact account.

Contrast a reversible contact spring that rises linearly from zero to the same 0.100 N at 0.500 mm. Its loading work is one half of the rectangle, 0.0250 mJ, and is stored and recovered on ideal unloading. Equal final force and displacement therefore do not identify equal dissipation. A measured loading–unloading path, with the surrounding specimen and instrument contributions addressed, is more informative than a single endpoint.

Sliding can permit redistribution and absorb energy while preventing a tube from reaching a high axial force. Suppressing it may raise transfer resistance, but cannot guarantee every performance measure improves: constituent damage, loss of deformability, added mass and a changed failure path can matter. Contact strength, tensile breaking force and energy absorption name different questions.

Define the envelope, solid and mass account

For a uniform reference gauge of length L and external-envelope area A, its outer volume is AL. The envelope includes the void space inside that boundary. Linear density λ = m/L is mass per length; bulk density ρ_bulk = m/(AL) = λ/A. Neither is the density of the solid walls alone. Use the geometry and mass of the same dry reference state, before tensile loading.

The textile unit tex means grams per kilometre: 1 tex = 10⁻⁶ kg/m. An area of 1 μm² is 10⁻¹² m². Thus a gauge with λ = 0.400 tex and A = 400 μm² has ρ_bulk = (0.400 × 10⁻⁶)/(400 × 10⁻¹²) = 1000 kg/m³. These are invented illustrative values, not a density specification for nanotube fibres.

Now stipulate only one solid phase of independently known density ρ_s and negligible void mass, with no retained liquid or binder. Solid volume is V_s = m/ρ_s. Its fraction of the envelope is V_s/(AL) = λ/(ρ_s A); therefore total void fraction is φ = 1 − λ/(ρ_s A). This is a mass–volume identity under those assumptions, not a fitted relation between porosity and strength. For ρ_s = 2000 kg/m³ in the toy account, the 400 μm² gauge has φ = 0.500.

The solid convention matters for hollow tubes. Here ρ_s describes the chosen wall solid and all space inside the envelope not occupied by it counts as void, including tube lumens. A measurement that instead assigns lumen volume to an effective tube solid uses a different reference. Total φ alone says neither how much pore volume is accessible from outside nor whether a connected load path crosses the gauge. An image of one section also need not represent the whole length.

For several retained solids with known densities and additive solid volumes, use V_s = Σ(m_i/ρ_i), or A_s = Σ(λ_i/ρ_i) per unit uniform length, and φ = 1 − A_s/A. Different compositions generally need different solid-density accounts. Processing, voids and property variation develops this bookkeeping for a fibre–matrix composite; here it prevents a binder or contaminant from being mistaken for densified nanotube walls.

Density changes need compatible comparisons

Densification can improve contacts while adding mass per volume and altering alignment. Compare actual force, mass per length, area convention and structure. A higher stress alone can partly reflect a changed denominator.

The table describes three fictional uniform dry fibre gauges, each 20.0 mm long. A is the starting route, B a compacted route and C a separately changed route. Destructive tensile tests require separate matched specimens: these are not three breaking tests on one surviving piece. Composition is stipulated as the same single solid, with ρ_s = 2000 kg/m³, and the outer-envelope area and linear density are measured in the unloaded reference state. Gauge, direction, temperature, loading history and endpoint definition are matched.

For this dossier the maximum recorded tensile force coincides with the breaking endpoint F_b. Nominal breaking stress is σ_b = F_b/A. Specific breaking force is R_b = F_b/λ; when the stated maximum-force endpoint defines tensile strength, it is also the corresponding specific tensile-strength ratio. In SI units R_b has units N·m/kg: 1 N/tex = 10⁶ N·m/kg. It is not work absorbed, which requires a force–extension path.

All three gauges have mass λL = 8.00 μg and solid cross-sectional area λ/ρ_s = 200 μm² in this chosen convention. A’s outer area is 400 μm² and B’s is 250 μm², so their void fractions are 50.0% and 20.0%. Their identical 0.400 N breaking forces give σ_b = 1.00 and 1.60 GPa, respectively. The nominal stress rises 60.0%, but R_b stays 1.00 N/tex and the actual breaking force does not rise.

C has the same 250 μm² area and 0.400 tex linear density as B, but F_b = 0.600 N. It gives σ_b = 2.40 GPa and R_b = 1.50 N/tex. Relative to B, force and force per linear density both rise 50.0%; that rise cannot come solely from changing the outer-area denominator. It is evidence for a changed mechanical response of these specimens, not unique proof of stronger contacts or a guarantee for every specimen from the route.

Using mutually compatible definitions gives σ_b/ρ_bulk = (F_b/A)/(λ/A) = F_b/λ. This explains why merely redrawing the outer boundary changes both stress and bulk density while their ratio cancels A. The cancellation does not remove uncertainty in λ, composition, the force endpoint or whether the gauge is representative. A single solid-density number paired with pore-inclusive stress would not be this compatible ratio.

Separate a changed result from its cause

A pore removes possible contacts or solid paths from part of an envelope and changes the neighbouring stress field. A long poorly connected region and many dispersed gaps need not interrupt the same paths, even at equal total void volume. Tube-wall defects, tube breaks, contact contamination and a nonuniform bundle arrangement are different structural changes. A processing step can alter several together. A total porosity number is not a universal strength law.

Published assembled-fibre evidence includes a treatment comparison with nearly unchanged linear density and increased breaking load. That distinguishes a force increase from an area-only change. It does not by itself separate changes in alignment, packing and interfacial chemistry. Conversely, the bridge experiment cited in the references measures a bundle bending-modulus change; that endpoint cannot be silently relabelled as an axial tensile-strength increase. The original macroscopic wet-spun fibre evidence also concerns assembled specimens rather than isolated ideal tubes.

A useful next comparison measures the intended mechanism as well as the final quotient. If contact transfer is proposed, seek matched evidence of relative sliding or interface response at a stated load, while checking alignment, bundle length, wall damage, composition and where failure occurred. A lower observed slip at equal force is consistent with a changed transfer state; it does not determine every contact’s shear law. A changed Raman signal, external diameter or fracture image alone does not uniquely select the governing failure mechanism.

Linear density itself needs a justified measurement model. A hierarchical fibre need not vibrate as one uniform string, so a resonance-based mass estimate can be biased even when a force trace is sound. Area measurements also need a defined boundary, spatial sampling and specimen condition. Wet material, retained binder, local taper or a long-piece average paired with a different short section can defeat the simple account before any strength calculation is made. Specific strength and the scope of a reported result develops denominator matching and uncertainty.

These records justify only their stated specimen comparison. Replicates, location and processing-history records are needed to examine scatter and reproducibility. Greater breaking force, lower total porosity or larger R_b alone supplies neither a component lifetime nor evidence that a particular joint, environment or repeated-loading history is satisfactory.

Change the assumptions before reusing the quotient

Consider an additional fictional route retaining 0.400 tex carbon walls and 0.200 tex binder, with A = 500 μm² and F_b = 0.600 N. Stipulate carbon density 2000 kg/m³, binder density 1000 kg/m³, additive volumes and negligible void mass. The carbon contributes 200 μm² of solid area per unit length and the binder another 200 μm². Total solid area is 400 μm²; φ = 1 − 400/500 = 20.0%.

Total λ is now 0.600 tex and ρ_bulk = 1200 kg/m³. Using all that mass with the carbon-only density would infer only 300 μm² solid area and falsely report 40.0% void. The true toy-account value remains 20.0%. The discrepancy is a composition-model error, not evidence of a new pore geometry.

This route’s nominal breaking stress is 0.600 N/(500 × 10⁻¹² m²) = 1.20 GPa. Its whole-fibre specific breaking force is 0.600/0.600 = 1.00 N/tex. Compared with A, breaking force is 50.0% higher but mass per length is also 50.0% higher; whole-fibre R_b is unchanged. Dividing whole-fibre force by carbon mass alone gives 1.50 N/tex on a different basis and must not be presented as the whole-fibre result.

The same care applies to location. A local outer area and a long-length average λ establish a local void fraction only if that λ describes the same location or uniformity is justified. If the binary formula produces φ < 0, the assumptions or input compatibility have failed: solid volume cannot exceed the outer volume in this account. Resolve composition, density convention, local mass/geometry or measurement error rather than reporting a negative physical void fraction.

Worked examples

Try the reasoning

1. Does lower porosity guarantee every performance improvement?

   Show the answer and reasoning

   No.

   Trade-offs and structural changes remain. Equal total void fractions can have different contact connectivity and defect placement. A treatment can also change wall damage, alignment, mass and the failure path.

2. Can sliding dissipate energy while limiting strength?

   Show the answer and reasoning

   Yes.

   Different performance measures can favour different mechanisms. A resisting contact can perform dissipative sliding work while transmitting too little force to load a strong tube fully. One endpoint does not establish total specimen work or dissipation.

3. A different already-sliding contact has uniform traction 8.00 MPa over actual area 0.0120 mm² and slips 0.250 mm at that constant resistance. Find resultant and sliding work. May its outer fibre area replace the given contact area?

   Show the answer and reasoning

   0.0960 N and 0.0240 mJ. No.

   Use 8.00 N/mm² × 0.0120 mm² = 0.0960 N; multiply by 0.000250 m to get 2.40 × 10⁻⁵ J. The interface area carries tangential traction; an outer section carries a different axial-stress account.

4. Two matched fictional gauges give before F_b = 0.300 N, A = 300 μm², λ = 0.300 tex; after F_b = 0.360 N, A = 200 μm², λ = 0.450 tex. Compare nominal stress and whole-fibre R_b. Does higher stress ensure better force per mass per length?

   Show the answer and reasoning

   Stress rises 1.00→1.80 GPa; R_b falls 1.00→0.800 N/tex.

   Force rises 20.0%, outer area falls by one third, and their quotient rises 80.0%. Linear density rises 50.0%; the ratio factor is 1.20/1.50 = 0.800, a 20.0% decrease. These changes do not identify which constituent carries the added force.

5. A 100 mm fibre has 0.200 tex in its first 50 mm and 0.600 tex in its second 50 mm. At a first-half location A = 200 μm². Under the single-solid model with ρ_s = 2000 kg/m³, calculate the whole-length average λ and the first-half local φ. What would using that average falsely imply?

   Show the answer and reasoning

   Average λ = 0.400 tex; first-half local φ = 50.0%. Using the average with that local area would falsely give 0%.

   The two halves weigh 10.0 and 30.0 μg, so 40.0 μg over 100 mm gives 0.400 tex. The actual first-half λ gives solid area 100 μm² inside 200 μm². The average gives 200 μm² solid only because it mixes mass from the heavier second half with the first-half area.

6. The binary uniform-gauge inputs are A = 200 μm², λ = 0.500 tex and ρ_s = 2000 kg/m³. What φ does the formula give, and what is the physical conclusion?

   Show the answer and reasoning

   φ = −25.0%; the stated single-solid, geometry and mass inputs cannot all describe that same uniform envelope.

   Solid area λ/ρ_s = 250 μm² exceeds 200 μm². Check density convention, composition, local matching and measurements. Negative physical pore volume is not a material discovery.

7. An intertube-bridge study reports a larger bending modulus. Can that number be used as the same factor multiplying axial breaking force of a manufactured fibre?

   Show the answer and reasoning

   No; the object, loading mode and endpoint differ.

   Suppressed sliding can affect a bending response. Axial failure also depends on how force enters the full specimen and on competing rupture or slip paths. A direct matched tensile comparison is required for that separate claim.

8. Two gauges have the same mass, area, solid density and total φ. One has distributed gaps; the other has a thin, poorly connected region across many bundle paths. Must their breaking forces match?

   Show the answer and reasoning

   No. The scalar volume account does not specify the connected load paths or local defect geometry.

   No new total void volume is required to rearrange where contacts are lost. Nor does this description alone quantify a strength difference: spatial structure, loading and an appropriate response model or matched tests are still needed.

9. A reversible linear contact spring reaches 0.100 N at 0.500 mm. Compare its loading work with the constant-resistance sliding contact and state the ideal unloading distinction.

   Show the answer and reasoning

   The spring stores 0.0250 mJ, half the sliding work of 0.0500 mJ.

   The spring’s triangular loading area is ½Qδ. Its ideal reverse path returns that energy. The stipulated sliding rectangle is dissipative; equal final coordinates do not make the paths equivalent.

10. C carries 50.0% more breaking force than B. Images show denser contacts, but alignment and wall defect indicators also changed. A report says contact improvement alone caused the entire force gain. Is that established?

    Show the answer and reasoning

    No. The force gain is established for the supplied records; the exclusive causal attribution is not.

    Several candidate causes changed together. A more discriminating comparison holds alternatives matched and measures contact/slip behaviour, constituent state and failure location. Even matched central values need repeated specimens before a population statement.

What to carry forward

• Inspect contacts and pores.
• Retain forces and denominator definitions.
• Tangential interface force, axial breaking force and dissipated work require different areas or histories.
• An area-only change can raise nominal stress without raising force or force per linear density.
• A solid–void balance needs matched geometry, composition and density conventions; total void fraction does not specify connectivity.
• A changed breaking-force record can exclude a denominator-only explanation while leaving its mechanism unresolved.