Ambitious structures and required specific properties

A headline concept contains an engineering problem

Very long tensile structures, extreme thermal barriers or ultralight vehicles require more than one exceptional property. Self-weight, load transfer, joints, stability, fatigue and environment enter. A perfect microscopic constituent is not the same as a manufactured structure.

Imagine suspending a payload from a very long cable. Choosing a larger section reduces the payload force per area, but the added cable also weighs more. The upper connection must carry both. A material can pass the local stress calculation and still fail the connection or available-space requirement. These are coupled requirements, rather than three independent rankings of an exceptional constituent.

Strength and strength relative to density derives a short tie comparison with self-weight neglected. Here self-weight is retained and can dominate. All numerical materials, limits, payloads and body-force fields below are invented teaching inputs. The stress limit S is a stipulated screening value for the manufactured cable in the stated condition, not a measured ultimate strength, a certified allowable or a value inferred from an ideal atomic chain.

Start with a transparent reduced model

For an ideal vertical uniform cable of length L, constant density ρ and constant gravity g, the stress from its own weight at the top is ρgL. Real large systems can violate these assumptions. The model is a way to reveal requirements, not a feasibility certificate.

Define z upward from the cable’s lower end: z = 0 at the payload and z = L at its support. Let A(z) be the local section, ρ its uniform mass density, P > 0 the downward payload force, and T(z) the positive tensile force transmitted across a cut. Assume a straight vertical cable at static equilibrium, no bending or lateral load, negligible connection mass, and small enough strain that the stated geometry and density can be treated as fixed. There is no imposed thermal strain or dynamic acceleration in this first model.

Take the entire portion below a cut at z. The upward cut force balances P and the weight of all cable below it: T(z) = P + ∫₀ᶻ ρgA(s) ds. Here s is an integration coordinate with units of length. Equivalently, dT/dz = ρgA(z), with boundary T(0) = P. A larger z includes more suspended material, so the tension increases upward. The average local axial stress is σ(z) = T(z)/A(z). Local concentrations at a real fitting require another model.

For a uniform section A, integration gives T(z) = P + ρgAz and σ(z) = P/A + ρgz. The largest stress is at z = L because g, ρ and A are positive. With P = 0 this recovers the retained self-weight result ρgL. Multiplying the uniform section by two doubles both its weight and its force capacity; it does not reduce the self-weight stress.

Derive a requirement before inserting a promising property

Apply the stipulated local screen σ(z) ≤ S. For the uniform cable, P/A + ρgL ≤ S. If S − ρgL > 0, the smallest permitted section is A_req = P/(S − ρgL). If S − ρgL ≤ 0, no finite positive area can pass when P > 0. The statement is about this uniform, constant-gravity model; it is not a universal maximum length for every tapered structure.

Define q = S/ρ, the supplied specific stress limit. Dividing the stress inequality by ρ gives q ≥ gL + P/(ρA). Both sides have units m²/s² = J/kg. Those dimensions do not make q a measured fracture energy or toughness. With no payload, the uniform self-support screen is q ≥ gL; a positive payload requires a strict excess over gL at every finite A.

The cable mass at the minimum uniform section is m_req = ρLA_req = PL/(q − gL). Thus q ranks minimum cable masses if area is freely adjustable, P and L are common, and only this local stress screen applies. As q approaches gL from above, the required mass diverges. A ratio just above the self-support threshold need not be practical.

For the first candidate, use ρ = 1000 kg/m³, S = 30.0 MPa, L = 1000 m, g = 10.0 m/s² and P = 20.0 kN. The self-weight stress is 10.0 MPa. Therefore A_req = 1000 mm², cable mass is 1000 kg, and the top force is 30.0 kN. Its q = 30000 J/kg exceeds gL = 10000 J/kg, yet the top fitting still has to carry that 30.0 kN.

A stronger section can overload a fixed connection

Now stipulate an independent top-connection force screen T(L) ≤ C, with C = 28.0 kN. C is supplied for this teaching assembly; it does not scale when cable area is changed. Ignore the fitting’s mass and any extra local stress field only for this reduced comparison. Since T(L) = P + ρgAL, the connection requires A ≤ (C − P)/(ρgL), provided C > P. If C ≤ P, no positive cable section can pass here.

For the first candidate the connection admits A ≤ 800 mm², whereas the cable stress requires A ≥ 1000 mm². Their intervals do not overlap. Making the uniform cable thicker cannot repair the assembly: it helps the material stress and worsens the connection force. Equality is admitted in these mathematical screens; it is not a physical design margin.

For the same P, L and g, compare the three fictional candidates in the table. Also impose a separate available-section cap A ≤ 500 mm². It represents a stated geometric envelope, not a density-dependent material property. Candidate C has the largest q and the smallest stress-limited cable mass, but needs 625 mm². Candidate B alone satisfies all three screens at its listed minimum section. No extrapolation from q supplies its fatigue, stiffness, connection durability or damage tolerance.

A supplied connection capacity can be relevant at a different endpoint from a cable tensile test. Changing grip geometry, bonding length, coating or environment can change load transfer and failure location. A stronger isolated tube does not automatically increase C. Specific strength and the scope of a reported result explains why area and mass must also describe the same physical object.

A taper changes the geometry available to the balance

Return to candidate A and remove the 500 mm² cap for this separate comparison; keep the same payload and 28.0 kN connection. Permit a smooth section A(z) that grows upward. To construct one fully stressed profile, impose T(z) = S A(z) at every z. Combining this condition with dT/dz = ρgA gives dA/dz = (ρg/S)A. This is a first-order differential equation for the local section, not an assumption that the tension is uniform.

The payload sets A(0) = P/S. Integrating dA/A from the bottom to z gives ln[A(z)/A(0)] = ρgz/S. Therefore A(z) = (P/S) exp(ρgz/S) and T(z) = P exp(ρgz/S). The exponential exp has a dimensionless argument. The taper ratio is A(L)/A(0) = exp(χ), where χ = ρgL/S = gL/q. Each additional part supports the payload and the material below, so the required section grows multiplicatively.

For constant g, integrate mass directly or use top force minus payload: m = ρ∫₀ᴸ A(z) dz = (P/g)[exp(χ) − 1]. For candidate A, χ = 1/3, bottom area is 666.667 mm², top area is 930.408 mm², cable mass is 791.225 kg and top force is 27.912 kN. The stated connection now passes. If the removed 500 mm² cap is restored, even the bottom section is too large and this profile fails that requirement.

This constructed profile has σ = S throughout under this one load, not spare capacity. A changed payload or body-force field changes stress in an existing profile; the cable does not retaper itself. For constant gravity and positive q the formula has no finite self-support-length cutoff, but its area and mass grow exponentially. That mathematical fact neither guarantees manufacturability nor removes limits from joints, permitted dimensions, deployment, elastic movement or damage. No general global optimum with all those constraints is claimed.

Changing gravity or rotation changes where the largest section occurs

Replace constant g by a specified downward acceleration field b(z), while retaining a straight tensile equilibrium, fixed density and the same constant target S. The balance is dT/dz = ρA b(z). A fully stressed construction then gives A(z)/A(0) = exp[Ψ(z)/q], with Ψ(z) = ∫₀ᶻ b(s) ds. Ψ has units J/kg and is a change in specific potential for the conservative examples below. The largest section occurs at the maximum of Ψ along the actual interval, not necessarily at its final endpoint.

For a simple original field, take b = 10.0 m/s² over the lower 500 m and b = 6.00 m/s² over the upper 500 m. Then Ψ(L) = 8000 J/kg and candidate A has top area 870.403 mm², lower than the constant-g value. The integral of acceleration is enough to obtain the endpoint taper ratio. It does not determine mass by dividing total weight by an arbitrary average acceleration: mass still requires ρ∫A dz.

A much longer ideal equatorial tether stationary in a frame rotating at angular speed Ω needs both gravity and rotation. Let r increase outward from a base radius r₀, and let μ = GM be the central body’s gravitational parameter. In a radial, conservative, straight-tether model the inward effective acceleration is b(r) = μ/r² − Ω²r. The corresponding change from the base is Ψ(r) = μ(1/r₀ − 1/r) − (Ω²/2)(r² − r₀²). The first term decreases in gradient with radius, while the outward rotational contribution grows.

At r_s = (μ/Ω²)^(1/3), b changes sign. If the tether spans this radius, Ψ and the fully stressed section are maximal there. Beyond it the section can decrease while remaining in tension. An anchor, upper termination or counterweight must still supply the boundary forces consistent with that equilibrium. This is a rotating-frame static construction, not a deployment or stability calculation. Lateral shape, moving payloads, oscillations, heating and impacts require further mechanics.

An unfamiliar dimensionless example uses u = r/r₀, μ/r₀ = 10000 J/kg and Ω²r₀² = 1250 J/kg over 1 ≤ u ≤ 3. These are invented parameters, not Earth data. Then Ψ(u) = 10000(1 − 1/u) − 625(u² − 1). Its maximum is 3125 J/kg at u = 2; the endpoint u = 3 gives only 1666.667 J/kg. At q = 30000 J/kg, the maximum area ratio is 1.109785, although the final ratio is 1.057128. Checking only the top endpoint would miss the largest required section.

Measured local strength and a manufactured load path answer different questions

Individual multiwalled carbon-nanotube tensile tests reported in 2000 found outer-layer strengths of 11–63 GPa in 19 tested tubes. Attachment primarily loaded the outer layer, and fracture followed by inner-shell pull-out was common in that geometry. The area convention and load-introduction mechanism belong with the result. It is not a measured allowable for an assembled kilometre cable. A calculation using that selected wall strength with a whole product’s different density would also mix physical bases.

Historical elevator concept work in 2003 treated very high ribbon strength as a required future input and discussed dispersion, alignment and stress transfer as challenges. Its production forecasts were proposals, not later measurements demonstrating those properties. Static taper calculations quantify a consequence of an assumed material; they do not close the synthesis, continuity or manufacturing gap.

To test transfer, measure force, extension, reference mass/length and failure location on representative continuous lengths and their actual terminations, retaining processing lot, rate, temperature and environment. A short specimen that breaks at the grip may test the fitting rather than the intended gauge. Conversely, a sound gauge result alone does not validate the fitting. Longer paths introduce more opportunities for defects and load-transfer interruptions; their effect must be observed or justified by a model rather than inferred from an ideal atomic strength.

Fatigue, creep, environmental cracking and local loss of section can change which limit is relevant. A deliberately simple uniform residual-area screen makes the distinction tangible: after a stipulated 10% area loss everywhere, keep the same remaining-material density and nominal S. For candidate A’s original 1000 mm² uniform cable, area becomes 900 mm² and self-weight also falls 10%. Top stress is nevertheless 20000/(900 × 10⁻⁶) + 10⁷ = 32.222 MPa, above S. Lower mass does not necessarily improve payload capacity.

That residual-area calculation is not a corrosion-rate or crack model. A localized notch can introduce a stress concentration and a different failure mechanism; neither is resolved by replacing A with a uniform fraction. Lifetime evidence, scatter and extrapolation explains why finite survival observations and accelerated tests cannot supply an unrestricted service life.

State which uncertainty a new observation would reduce

A feasibility statement needs both an explicit demand model and evidence for its inputs at the relevant object and conditions. For the fictional cable, the algebra can falsify a proposed combination of length, area, material screen and fitting capacity. Passing those inequalities establishes only internal compatibility of that combination. It leaves stiffness, dynamic response and the basis of S and C untested.

Choose the next observation according to the unresolved mechanism. If the bottleneck is a termination, raising an ideal bond-strength prediction is not a direct test; compare representative terminations and observe the failure path. If the concern is a weak continuous length, an average constituent property or a spectroscopic synthesis signal is insufficient. If the issue is environmental retention, a dry short-duration tensile maximum does not describe the combined history.

Predictions, targets and levels of evidence distinguishes numerical verification, model comparison and specimen identification. Here the target is a load-bearing system with defined geometry and boundary conditions. Preserve a narrower demonstrated achievement when that larger target remains unresolved: a verified taper equation, a measured individual wall, a continuous fibre and a jointed structure are useful but different pieces of evidence.

Worked examples

Try the reasoning

1. Does this establish a space-elevator design?

   Show the answer and reasoning

   No.

   Gravity, rotation, geometry, environment and dynamics require another model.

2. Why include connections?

   Show the answer and reasoning

   They transfer the actual forces.

   A constituent property cannot replace joint performance.

3. Keep candidate A and its uniform 1000 m geometry, but reduce P to 12.0 kN while C stays 28.0 kN. Find the admitted area interval and minimum cable mass. Is the 500 mm² section cap compatible if it is imposed?

   Show the answer and reasoning

   Without the cap, 600 ≤ A ≤ 1600 mm² and minimum cable mass is 600 kg. A cap of 500 mm² makes the interval empty.

   The local screen gives 12000/(20 × 10⁶) = 600 mm²; the fitting gives (28000 − 12000)/(10 × 10⁶) = 1600 mm². Both must hold. Payload reduction can open a joint window while leaving another constraint unsatisfied.

4. A new material is assigned S = 12.0 MPa and ρ = 1200 kg/m³ on the same L = 1000 m, g = 10.0 m/s² span. Can any finite uniform area carry a positive 1.00 N payload under σ ≤ S? What does P = 0 change?

   Show the answer and reasoning

   No finite area can carry the positive payload. With P = 0, self-weight stress equals S for every positive uniform area in the reduced model.

   Here S − ρgL = 0. Adding any positive P/A exceeds S; it cannot be removed at a finite area. The equality-only self-weight case is not a margin or a real structure approval.

5. Use B and C from the table. Relax the section cap from 500 to 700 mm², retaining all other inputs. Which minimum-mass uniform candidate passes, and why was its q not sufficient under the old cap?

   Show the answer and reasoning

   Both B and C now pass the stated screens; C has the smaller cable mass, 500 kg versus 666.667 kg.

   C needs 625 mm² and top force 25.0 kN. It fits the new cap and connection. Its greater q correctly ranks freely adjustable stress-limited masses, but did not make its required area fit a 500 mm² envelope.

6. For the fully stressed taper of candidate A, replace the constant acceleration by 10.0 m/s² over 0–500 m and 6.00 m/s² over 500–1000 m. Find the top tension and area. Does dividing the added top force by the length-average acceleration 8.00 m/s² give the exact cable mass?

   Show the answer and reasoning

   Top tension 26.1121 kN and area 870.403 mm². No: that division gives 764.013 kg, while direct section integration gives 776.870 kg.

   The integrated field is Ψ = 8000 J/kg, so the force ratio is exp(8000/30000). For mass use A₀ = P/S and integrate A₀ exp(z/3000) on the lower half and A₀ exp(1/6) exp[(z − 500)/5000] on the upper half. The area is not constant, so the mass-weighted acceleration differs from the length average.

7. A uniform candidate-A cable initially uses 1000 mm². A hypothetical change removes 10% of area everywhere without changing remaining-material density or S. P stays 20.0 kN. Find the new mass, top force and stress. Does its lower mass ensure the original stress screen still passes?

   Show the answer and reasoning

   Mass 900 kg, top force 29.0 kN and top stress 32.222 MPa. The stress screen fails.

   The weight falls from 10.0 to 9.00 kN, but the unchanged payload is divided by only 900 mm². This is a uniform area-loss account, not a local-crack stress or damage-rate law.

8. A hypothetical report supplies a large strength/density ratio from individual supported tubes but no continuous-length tests, joint tests or operating-environment data. Could inserting it into the taper equation establish an operational elevator? Name one observation that would address a specific missing load-transfer claim.

   Show the answer and reasoning

   No. For example, representative continuous fibre-and-termination tests with observed failure location could test whether the intended axial force actually enters the manufactured gauge.

   The equation is conditional on a cable-level supplied limit and density describing the same object. A synthesis signal, an ideal wall calculation and a local tensile result do not provide that operating limit or qualify a jointed system.

What to carry forward

• Derive explicit requirements and assumptions.
• Keep constituent promise separate from system feasibility.
• Uniform-cable self-weight stress does not shrink when the whole area increases; payload stress does.
• A connection can impose an upper area bound while cable stress imposes a lower one. Check that all admitted intervals overlap.
• A fully stressed taper follows a local force balance and grows exponentially with integrated body force divided by specific stress limit.
• A changing body-force field can place the largest required area inside the span. Endpoint checks and arbitrary average gravity can miss different requirements.
• Calculated requirements, measured specimen properties and demonstrated system operation require different evidence.