Born Rigidity and Rigid Accelerated Motion

`aq-sr-born-rigidity` has already told you what a Born-rigid rod does. Its elements ride Rindler hyperbolae about a common vertex; the proper acceleration is graded as $\alpha(X) = c^2/X$ ; the back must be pushed harder than the front; the rod cannot be longer than $c^2/\alpha_{\text{nose}}$ or its tail would sit on the horizon; and Herglotz and Noether proved such motions have only three degrees of freedom, so a rigid disc can never be spun up from rest. All settled, and recalled here rather than repeated. What is not settled is what "rigid" means — as a tensor. "The proper distance between neighbouring points, measured in the momentarily comoving frame, stays constant" is a sentence with two frame-dependent words in it, and sixty years of Bell's-paradox arguments should have taught us that any definition with a frame in it can be attacked by asking whose. Born rigidity has a definition with no frame in it at all: the expansion tensor of the body's own four-velocity field vanishes.

\theta_{\mu\nu} \;\equiv\; h^{\alpha}{}_{\mu}\,h^{\beta}{}_{\nu}\,\nabla_{(\alpha}u_{\beta)} \;=\; 0 .

Expansion and shear, every component, everywhere on the wedge. That is what this article computes, and it is the definition Born actually wrote in 1909. And then it pays the bill. The tail's clock must run slower than the nose's, because $\alpha_{\text{tail}} > \alpha_{\text{nose}}$ — so a Born-rigid ship is rigid in length and not rigid in time. Its ends cannot stay synchronised. For a hundred-metre hull at $1\,g$ the nose gains 344 nanoseconds per year on the tail: $103$ metres of light travel, more than the length of the ship. Rigidity buys a body whose shape never changes and whose clocks never agree.

Key takeaways

Signature $(+,-,-,-)$ : $\mathrm{d}s^2 = c^2\mathrm{d}t^2 - \mathrm{d}x^2$ , timelike positive. Flat spacetime throughout.
Recalled, not re-derived. The profile $\alpha(X) = c^2/X$ , the rear must accelerate harder than the front, the causal bound $L_0 < c^2/\alpha_{\text{nose}}$ , and Herglotz–Noether are all owned by `aq-sr-born-rigidity`. The strain $\varepsilon = \gamma - 1$ is owned by `aq-sr-bell-spaceship-paradox`, the two-horizon structure by `aq-sr-bell-spaceship-paradox-s6`, the metric by `aq-sr-rindler-metric`. This article uses them and extends.
Owned result: the covariant definition. $\boxed{\theta_{\mu\nu} = h^{\alpha}{}_{\mu}h^{\beta}{}_{\nu}\nabla_{(\alpha}u_{\beta)} = 0}$ — the expansion tensor, not the scalar. Computed for the Rindler congruence: every one of the sixteen components vanishes identically on the wedge. Expansion $\theta = 0$ and shear $\sigma_{\mu\nu} = 0$ . No frame appears in the statement.
Owned result: a Born-rigid body IS the Rindler congruence, and the Rindler metric is its rest chart: $\mathrm{d}s^2 = (a_0 x/c^2)^2c^2\mathrm{d}T^2 - \mathrm{d}x^2$ , in which every element sits at fixed $x$ forever. Rigidity is the statement that the body is at rest in a static metric.
Owned result: the rod's ends cannot stay synchronised, and the ratio is EXACT. $\dfrac{\mathrm{d}\tau_{\text{nose}}}{\mathrm{d}\tau_{\text{tail}}} = \dfrac{x_{\text{nose}}}{x_{\text{tail}}} = \dfrac{\alpha_{\text{tail}}}{\alpha_{\text{nose}}} = 1 + \dfrac{\alpha_{\text{tail}}\,L_0}{c^2}$ — exact, not first order, provided $\alpha_{\text{tail}}$ is the tail's own proper acceleration.
Owned result: the SAME number grades the thrust and the clocks. At $1\,g$ with $L_0 = 100$ m, $\alpha_{\text{tail}}/\alpha_{\text{nose}} - 1 = \mathbf{1.0911\times10^{-14}}$ — and $\mathrm{d}\tau_{\text{nose}}/\mathrm{d}\tau_{\text{tail}} - 1$ is the same number. One number, two jobs. The nose gains 344 ns per year of tail proper time: $103$ m of light travel, longer than the ship.
Owned result: the length bound is the clock bound. $\mathrm{d}\tau_{\text{tail}}/\mathrm{d}\tau_{\text{nose}} = 1 - L_0\alpha_{\text{nose}}/c^2$ , which hits zero exactly when $L_0 = c^2/\alpha_{\text{nose}}$ . `aq-sr-born-rigidity` gets that bound from infinite acceleration at the horizon; it is the same bound as the tail's clock stops.
Key idea: rigid means "at rest in a static metric" — and static metrics have clock gradients.

The core insight

The everyday rigid body is defined by what its shape does. The relativistic rigid body must be defined by what its velocity field does, because shape is slice-dependent and velocity fields are not. Write the four-velocity of the material at every event, project out the flow direction, symmetrise the gradient — and demand zero. That single tensor equation, $\theta_{\mu\nu} = 0$ , contains the whole subject: it forces the hyperbolae, the graded thrust, the common vertex, the length bound, and the permanent disagreement of the body's ends about time. Rigidity is not a constraint on the body. It is a constraint on the congruence.

Where this lives today

The expansion-shear decomposition is the working language of relativistic continuum mechanics. $\theta$ , $\sigma_{\mu\nu}$ and the vorticity $\omega_{\mu\nu}$ are the standard kinematic invariants of any timelike congruence, and they are what Raychaudhuri's equation propagates — the same equation that, applied to geodesics, drives the Penrose–Hawking singularity theorems. The Born condition $\theta_{\mu\nu} = 0$ is the flat-space, non-geodesic corner of a very large building.

"Rigid" is the wrong word for real hardware. No spacecraft and no accelerator cavity is Born-rigid; all transmit stress at a sound speed far below $c$ , so the rigid programme is something you approach after transients decay, not something a material does. Relativistic elasticity — the covariant theory of deformable media — takes $\theta_{\mu\nu}$ as its strain rate and builds constitutive laws on top. Born rigidity is its zero.

And the clock gradient is not an idealisation. The relation $\mathrm{d}\tau \propto x$ along a rigidly accelerated body is exactly $\mathrm{d}\tau \propto \sqrt{g_{00}}$ along a static observer in a static gravitational field — and that one is measured daily. It is why GPS clocks are offset before launch, and why optical clocks a few centimetres apart tick at measurably different rates. The rigid ship's 344 ns per year is the same physics, made by an engine instead of a planet.

The physical picture

A body is a congruence: a family of timelike worldlines, one per material element, filling a tube of spacetime with no crossings and no gaps. Everything a body does — translate, rotate, stretch, shear, twist — is a statement about the four-velocity field $u^\mu(x)$ of that congruence, and nothing else. Now ask what "rigid" should mean. It cannot mean "the distance between the ends is constant" — distance on which slice? It cannot mean "no forces act" — a Born-rigid rod is under enormous thrust. It must mean: the material never gets closer to itself, or further from itself, or sheared relative to itself, as judged by the material. That judgement is made by the spatial projector

h_{\mu\nu} \;=\; \eta_{\mu\nu} - \frac{u_\mu u_\nu}{c^2},

the metric restricted to the 3-space orthogonal to the flow — the material's own rest space, defined event by event, with no global slice anywhere in sight. Take the gradient of $u$ , project both indices into that rest space, symmetrise, and you have the expansion tensor $\theta_{\mu\nu}$ : its trace is the expansion $\theta$ (rate of volume change), its trace-free part the shear $\sigma_{\mu\nu}$ (rate of shape change at fixed volume).

Born rigid $\iff$ $\theta_{\mu\nu} = 0$ . Nothing expands, nothing shears. And notice what has vanished from the definition: every reference to a frame, a slice, a simultaneity convention, an observer. That is the whole point.

Figure

The Born-rigid congruence. Every material element rides a Rindler hyperbola; every hyperbola shares the SAME vertex, so the whole family is asymptotic to the same light cone (gold) — that shared null line is the Rindler horizon, and it is what caps the rod's length. The dashed radial lines are the fan of common simultaneity slices: because they all pass through the vertex, EVERY element agrees about which events are simultaneous, and the proper gaps read along them never change. Small x means tight curvature means BIG proper acceleration (α = c²/x, owned by aq-sr-born-rigidity) — the tail is pushed hardest. The covariant content is the box: the expansion TENSOR of this velocity field is zero, all sixteen components, everywhere on the wedge. Nothing expands, nothing shears, and no frame appears anywhere in the statement

All the hyperbolae share one vertex, so they share one fan of simultaneity slices — the dashed radial lines. Every element of the rod agrees on which events are simultaneous, which is why a rigid rod has a well-defined length at all. Read the proper gaps along any slice in the fan: they are differences of the labels $x$ , constant forever. That is rigidity, drawn. And there is the cost. The slices fan out from the vertex, so a fixed step in the shared time coordinate is a shorter proper-time interval near the vertex than far from it. The tail is near the vertex. The tail's clock runs slow. There is no way to have the one without the other.

The core equations

\boxed{\;\theta_{\mu\nu} \;=\; h^{\alpha}{}_{\mu}\,h^{\beta}{}_{\nu}\,\nabla_{(\alpha}u_{\beta)} \;=\; 0\;,\qquad h_{\mu\nu} = \eta_{\mu\nu} - \frac{u_\mu u_\nu}{c^2}\;}

where: $u^\mu$ is the material's four-velocity field, $h_{\mu\nu}$ the projector onto the local rest space, $\nabla_{(\alpha}u_{\beta)} = \tfrac12(\nabla_\alpha u_\beta + \nabla_\beta u_\alpha)$ . Physical reading: this is the definition of Born rigidity. Its trace $\theta = \nabla_\mu u^\mu$ is the expansion (volume), its trace-free part $\sigma_{\mu\nu}$ the shear (shape). Both vanish.

u^\mu \;=\; \frac{c}{\rho}\,\bigl(x,\; ct,\; 0,\; 0\bigr), \qquad \rho \equiv \sqrt{x^2 - c^2t^2}

where: $\rho$ is the invariant Rindler radius, constant along each worldline. The Rindler congruence written as a field on the wedge $x > |ct|$ — not a worldline, a field, which is what $\theta_{\mu\nu}$ needs. Check: $u^\mu u_\mu = c^2(x^2 - c^2t^2)/\rho^2 = c^2$ . ✓

\mathrm{d}s^{2} \;=\; \left(\frac{a_{0}\,x}{c^{2}}\right)^{2} c^{2}\,\mathrm{d}T^{2} \;-\; \mathrm{d}x^{2}

where: $T$ is the shared Rindler time, $x$ the element's position label (constant for each material point), $a_0$ the proper acceleration of the fiducial element at $x_0 = c^2/a_0$ , whose proper time is $T$ . Physical reading: the body is at rest in this chart — every element has $\mathrm{d}x = 0$ forever. Rigidity means being static in a static metric.

\frac{\mathrm{d}\tau_{\text{nose}}}{\mathrm{d}\tau_{\text{tail}}} \;=\; \frac{x_{\text{nose}}}{x_{\text{tail}}} \;=\; \frac{\alpha_{\text{tail}}}{\alpha_{\text{nose}}} \;=\; 1 + \frac{\alpha_{\text{tail}}\,L_{0}}{c^{2}} \qquad \textbf{(exact)}

where: $L_0 = x_{\text{nose}} - x_{\text{tail}}$ is the rod's proper length and $\alpha_{\text{tail}} = c^2/x_{\text{tail}}$ the tail's own proper acceleration. Physical reading: this is not a first-order expansion — it is exact, because $\alpha_{\text{tail}}$ is the emitter's acceleration and not some averaged $g$ . The same identity underwrites the exact clock-rate law of `aq-sr-rindler-metric`.

\frac{\mathrm{d}\tau_{\text{tail}}}{\mathrm{d}\tau_{\text{nose}}} \;=\; 1 - \frac{L_{0}\,\alpha_{\text{nose}}}{c^{2}} \;\;\xrightarrow[\;L_0 \to c^2/\alpha_{\text{nose}}\;]{}\;\; 0

where: $\alpha_{\text{nose}} = c^2/x_{\text{nose}}$ . Physical reading: the causal length bound $L_0 < c^2/\alpha_{\text{nose}}$ of `aq-sr-born-rigidity` is a statement about clocks. The rod may not be so long that its tail's clock stops — the same sentence as "its tail may not sit on the horizon", and now the same equation.

Derivation AThe expansion tensor of the Rindler congruence vanishes — all sixteen components

`aq-sr-born-rigidity` establishes the rigid profile by demanding constant proper separation on the comoving slice. That argument is correct and it is not repeated here. What is done instead is the thing that argument cannot do: state rigidity without a slice, and check it.

Step 1. Build the velocity field. Each element is labelled by the invariant $\rho = \sqrt{x^2 - c^2t^2}$ , constant along its hyperbola. At the event $(ct, x)$ the material has rapidity $\varphi$ with $\cosh\varphi = x/\rho$ , $\sinh\varphi = ct/\rho$ , so

u^\mu(ct,x) \;=\; c\bigl(\cosh\varphi,\; \sinh\varphi,\;0,\;0\bigr) \;=\; \frac{c}{\rho}\bigl(x,\; ct,\; 0,\; 0\bigr).

Why: a congruence is a field, not a curve. Everything below is done at a point, with derivatives, never referring to another point.

Step 2. Take the expansion scalar. In Cartesian coordinates $\nabla_\mu = \partial_\mu$ , with $x^0 \equiv ct$ :

\theta = \partial_\mu u^\mu = \partial_0\!\left(\frac{cx}{\rho}\right) + \partial_1\!\left(\frac{c\,x^0}{\rho}\right) = \frac{c\,x\,x^0}{\rho^{3}} \;-\; \frac{c\,x\,x^0}{\rho^{3}} \;=\; 0 .

Why: the two terms are equal and opposite for every $x$ and every $t$ on the wedge — because $\rho$ depends on $x$ and $x^0$ in exactly the way it does. The cancellation is the geometry of the hyperbola.

Step 3. Project and symmetrise — the full tensor. Form $h^{\alpha}{}_{\mu} = \delta^{\alpha}{}_{\mu} - u^\alpha u_\mu/c^2$ and contract:

\theta_{\mu\nu} = h^{\alpha}{}_{\mu}h^{\beta}{}_{\nu}\,\tfrac12\bigl(\partial_\alpha u_\beta + \partial_\beta u_\alpha\bigr) \;=\; 0 \quad\text{for every } \mu,\nu .

Not just the trace. Every component. So $\sigma_{\mu\nu} = \theta_{\mu\nu} - \tfrac13\theta\,h_{\mu\nu} = 0$ as well: the body does not shear. Why: $\theta = 0$ alone would permit a body that preserves its volume while distorting its shape — squashed in $x$ , stretched in $y$ . A shearing rod is not a rigid rod. Born rigidity is the vanishing of the whole tensor, and only the whole computation shows it.

Step 4. Contrast with Bell. `aq-sr-bell-spaceship-paradox-s6` computes the expansion of the equal-acceleration congruence and gets $\theta_{\text{Bell}} = \alpha\beta/c \neq 0$ . Why: "the string breaks" and "the rod is rigid" are, in this language, the same statement with opposite signs — and neither has a frame in it. That is the payoff. *Verified in verify_aq-sr-born-rigidity-s6.py: sympy builds $u^\mu$ , confirms $u^\mu u_\mu = c^2$ , computes $\theta = \partial_\mu u^\mu = 0$ symbolically, then builds $h^{\alpha}{}_{\mu}$ and the full $\theta_{\mu\nu}$ and asserts all sixteen components are identically zero on the wedge.*

Go deeper · Derivation B · 2 minThe rod's own rest chart — and the metric card that eats itself

Derivation BThe rod's own rest chart — and the metric card that eats itself

The Rindler chart is usually introduced as the coordinate system one accelerated observer builds (`aq-sr-rindler-metric` owns that). The goal here is the other reading: that a Born-rigid body is exactly a body at rest in that chart, every element pinned at fixed $x$ forever — and then to expose the notational collapse that destroys the chart if the symbol $\alpha$ is reused carelessly.

Step 1. Ask what coordinates the rod would choose. It wants labels in which it does not move. The elements sit at fixed $\rho$ , so use $\rho$ (call it $x$ ) as the space label, and use the rapidity as the time label. Set $\varphi = a_0T/c$ for some constant $a_0$ — the choice of $a_0$ is the choice of whose clock $T$ is.

Step 2. Map to Minkowski and pull back the interval, in $(+,-,-,-)$ where $\mathrm{d}s^2 = c^2\mathrm{d}t^2 - \mathrm{d}x^2$ :

ct = x\sinh\!\frac{a_0T}{c}, \qquad X = x\cosh\!\frac{a_0T}{c}

\Rightarrow\quad \mathrm{d}s^2 = \left(\frac{a_0 x}{c^2}\right)^{2}c^2\,\mathrm{d}T^2 \;-\; \mathrm{d}x^2 .

Why: derived, not transcribed. The $\mathrm{d}T\,\mathrm{d}x$ cross-terms cancel identically — which is exactly the statement that the fan of slices is orthogonal to the worldlines, i.e. that the chart is the body's own.

Step 3. Read the lapse. $\sqrt{g_{TT}}/c = a_0x/c^2$ , so $\mathrm{d}\tau = (a_0x/c^2)\,\mathrm{d}T$ : proper time is proportional to $x$ , and the constant of proportionality is fixed by declaring that the element at $x_0 = c^2/a_0$ keeps $T$ as its own proper time. Every other element's clock runs at a different rate, forever. Why: gravitational time dilation, in flat spacetime, with no gravity anywhere. `aq-sr-rindler-metric` owns the general statement; owned here is the consequence for a material body — a rigid rod cannot carry a single clock.

Step 4. Now the trap. The base card prints the lapse as $(\alpha x/c^2)$ with the same symbol $\alpha$ it has already defined as $\alpha(x) = c^2/x$ . Substitute:

\frac{\alpha(x)\,x}{c^2} = \frac{(c^2/x)\cdot x}{c^2} \;=\; 1 \qquad \text{identically.}

The lapse collapses to unity, the metric becomes Minkowski, and the entire clock-rate gradient — the reason the card exists — is annihilated by its own notation. The $\alpha$ in the lapse is a constant, the acceleration of one chosen fiducial element; it is emphatically not $\alpha(x)$ . Why: not pedantry. A metric whose lapse is identically 1 predicts that nose and tail clocks agree, which is the exact opposite of the truth. *Verified in verify_aq-sr-born-rigidity-s6.py: sympy pulls back $\mathrm{d}s^2$ from the map and reproduces $(a_0x/c^2)^2c^2\mathrm{d}T^2 - \mathrm{d}x^2$ exactly; it then evaluates $\alpha(x)\cdot x/c^2$ with $\alpha(x) = c^2/x$ and returns $1$ , and asserts this article prints the signature $(+,-,-,-)$ and contains no mostly-plus line element.*

Go deeper · Derivation C · 2 minA rigid ship cannot keep a shipwide clock — 344 ns per year, on a hundred metres

Derivation CA rigid ship cannot keep a shipwide clock — 344 ns per year, on a hundred metres

The goal is to turn the lapse gradient of the previous derivation into a number a crew would notice; to show that the same number grades the thrust; and to show that pushing it to its limit reproduces, from clocks alone, the causal length bound that `aq-sr-born-rigidity` derives from thrust alone.

Step 1. Put the rod in the chart. Nose at $x_{\text{nose}}$ , tail at $x_{\text{tail}} = x_{\text{nose}} - L_0$ , both fixed forever. Divide the two proper-time rates; the shared $\mathrm{d}T$ cancels:

\frac{\mathrm{d}\tau_{\text{nose}}}{\mathrm{d}\tau_{\text{tail}}} = \frac{x_{\text{nose}}}{x_{\text{tail}}} = \frac{\alpha_{\text{tail}}}{\alpha_{\text{nose}}} .

Why: the same ratio grades the thrust and the clocks. $\alpha = c^2/x$ makes the acceleration ratio the inverse position ratio; the lapse $\propto x$ makes the clock ratio the position ratio. One number, two hats — and that is the equivalence principle, before anyone has mentioned gravity.

Step 2. Make it exact. Write $x_{\text{tail}} = c^2/\alpha_{\text{tail}}$ and $x_{\text{nose}} = x_{\text{tail}} + L_0$ :

\frac{x_{\text{nose}}}{x_{\text{tail}}} = \frac{c^2/\alpha_{\text{tail}} + L_0}{c^2/\alpha_{\text{tail}}} = 1 + \frac{\alpha_{\text{tail}}L_0}{c^2}\qquad\textbf{— exactly, with no expansion taken.}

Why: textbooks (and this article's base card) write this with a " $+\cdots$ " and call it first-order. It is not; it is closed. The condition is that $\alpha_{\text{tail}}$ be the acceleration of the emitting end; use the nose's and you do get a series. Same exactness `aq-sr-rindler-metric` establishes for $1 + g_1h/c^2$ — the "approximate" redshift formula is only approximate because people evaluate $g$ in the wrong place.

Step 3. Put a ship in it. Take $L_0 = 100$ m and let the nose hold $1\,g$ , so $x_{\text{nose}} = c^2/g = 9.1648\times10^{15}$ m and $x_{\text{tail}} = x_{\text{nose}} - 100$ m:

\frac{\mathrm{d}\tau_{\text{nose}}}{\mathrm{d}\tau_{\text{tail}}} - 1 \;=\; \frac{L_0}{x_{\text{nose}} - L_0} \;=\; \mathbf{1.0911\times10^{-14}} .

Why: this is the same $1.0911\times10^{-14}$ as the thrust grading — $\alpha_{\text{tail}} - \alpha_{\text{nose}} = 1.0700\times10^{-13}\ \mathrm{m/s^2}$ , and $1.0700\times10^{-13}/9.80665 = 1.0911\times10^{-14}$ . One number. Two jobs.

Step 4. Integrate it. Over one year of the tail's proper time ( $3.15576\times10^7$ s):

\Delta\tau \;=\; 1.0911\times10^{-14} \times 3.15576\times10^{7}\ \mathrm{s} \;=\; \mathbf{344\ ns} \qquad\Longrightarrow\qquad c\,\Delta\tau \;=\; \mathbf{103\ m}.

Why: the light-travel equivalent of the drift exceeds the length of the ship. Synchronise nose and tail at launch; a year later a ranging system trusting that synchronisation misplaces the far end by more than the whole hull. The shape is rigid to the last atom; the time is coming apart.

Step 5. Close the loop with the length bound. Invert:

\frac{\mathrm{d}\tau_{\text{tail}}}{\mathrm{d}\tau_{\text{nose}}} = \frac{x_{\text{nose}} - L_0}{x_{\text{nose}}} = 1 - \frac{L_0\,\alpha_{\text{nose}}}{c^2},

which reaches zero precisely at $L_0 = c^2/\alpha_{\text{nose}}$ — the maximum rigid length of `aq-sr-born-rigidity`. Why: 1914 derives that bound from the tail needing infinite proper acceleration at the horizon. Here it falls out of the tail's clock stopping. Same fact — and the bound is not about how hard the engines can push, but about how slowly the tail is allowed to age. *Verified in verify_aq-sr-born-rigidity-s6.py: every number is recomputed at 50 decimal digits — $x_{\text{nose}} = 9.1648\times10^{15}$ m, ratio $-1 = 1.091137\times10^{-14}$ , $\Delta\alpha = 1.0700\times10^{-13}$ m/s², $344.3$ ns/yr, $103.2$ m — and each is grepped out of this file and asserted against the computed value.*

Figure

ONE NUMBER, TWO JOBS — and the bill it presents. The lapse of the rod's own rest metric is proportional to x, so proper time runs SLOWER toward the vertex; the proper acceleration goes as c²/x, so thrust runs HARDER toward the vertex. Both gradings are the same geometry read in opposite directions, and for a 100 m hull at 1 g both come out to the identical 1.0911e-14. LEFT: the rod, with the tail deep in the strong-acceleration, slow-clock region and the nose out in the weak-acceleration, fast-clock region. RIGHT: the bill. Integrate that fractional rate difference over one year of tail proper time and the nose has gained 344 nanoseconds — which is 103 metres of light travel, LONGER THAN THE SHIP. Born rigidity holds the hull's shape to the last atom and lets its time come apart. And push the length out toward c²/α(nose) and the tail's clock rate falls all the way to zero: the causal length bound of aq-sr-born-rigidity IS the clock bound

Limiting cases

Case	Result	Meaning
$L_0 \ll c^2/\alpha$	$\alpha_{\text{tail}}/\alpha_{\text{nose}} \to 1$ , clock ratio $\to 1$	the Newtonian rigid body: one acceleration, one clock. Everything here is a $10^{-14}$ correction — and every bit of it is real
$L_0 \to c^2/\alpha_{\text{nose}}$	$\mathrm{d}\tau_{\text{tail}}/\mathrm{d}\tau_{\text{nose}} \to 0$	the tail reaches the horizon: its clock stops and its required thrust diverges. The two limits coincide — the length bound and the clock bound are one bound
$\alpha \to 0$	$x \to \infty$ , chart degenerates	no vertex, no horizon, no gradient. An unaccelerated rigid body is trivially rigid and its clocks agree exactly
$\theta = 0$ but $\sigma_{\mu\nu} \neq 0$	not rigid	volume preserved, shape distorted — why the definition needs the whole tensor, not just its trace
Bell's congruence	$\theta = \alpha\beta/c \neq 0$	equal thrust everywhere: the tensor is non-zero and the thread snaps (`aq-sr-bell-spaceship-paradox-s6`)
rotation switched on	forbidden	Herglotz–Noether: 3 functional degrees of freedom, not 6 (owned by `aq-sr-born-rigidity`; recalled, not re-derived)

Common trapThe lapse

\alpha x/c^2

is a CONSTANT

\alpha

— write

\alpha(x)

there and the metric annihilates itself

This is the sharpest notational trap in the subject, and this article's own base card falls straight into it. Within Born rigidity the symbol $\alpha$ is already spoken for: $\alpha(x) = c^2/x$ . Write the Rindler lapse as $(\alpha x / c^2)$ with the same symbol and substitution gives $(c^2/x)\cdot x/c^2 = 1$ identically — the metric collapses to Minkowski and the clock gradient, the whole physical content, vanishes without trace. The $\alpha$ in the lapse must be a fixed number $a_0$ : the proper acceleration of one nominated fiducial element, the one whose proper time you have chosen to call $T$ . Nominate a different element and $a_0$ changes and $T$ rescales; the ratios of clock rates, which are what is physical, do not. Any Born-rigidity metric whose lapse is identically one is not a Born-rigidity metric.

Common trapWhat this article does not own — and the "first order" that is not first order

The rigidity profile $\alpha(X) = c^2/X$ , the result that the rear must be pushed harder than the front, the causal length bound $L_0 < c^2/\alpha_{\text{nose}}$ , and Herglotz–Noether are all owned by `aq-sr-born-rigidity` (serial 1914) and are recalled here by name, never re-derived. The strain $\varepsilon = \gamma - 1$ belongs to `aq-sr-bell-spaceship-paradox`; the two-horizon structure and $\theta_{\text{Bell}} = \alpha\beta/c$ to `aq-sr-bell-spaceship-paradox-s6`; the metric and the exact clock-rate law to `aq-sr-rindler-metric`; the horizon itself to `aq-sr-rindler-horizon`. What is owned here is the covariant definition ( $\theta_{\mu\nu} = 0$ , expansion and shear, computed), the identification of the rigid body with the Rindler congruence in its own rest chart, and the desynchronisation of the ends. And one correction. The base card and many textbooks write the clock ratio as $x_{\text{nose}}/x_{\text{tail}} = 1 + gh/c^2 + \cdots$ , calling it a weak-field expansion. It is exact, with no higher terms at all, provided $g$ is the proper acceleration of the tail — the emitting end. The series appears only if you evaluate $g$ somewhere else. There is nothing approximate about a Born-rigid rod's clock gradient.

Worked examples

1 · A hundred-metre ship at one gravity, for one year

Step 1 — the geometry. The nose holds $\alpha_{\text{nose}} = g = 9.80665\ \mathrm{m/s^2}$ , so it sits at $x_{\text{nose}} = c^2/g = \mathbf{9.1648\times10^{15}\ m}$ from the vertex. The tail, $L_0 = 100$ m astern, sits at $x_{\text{nose}} - 100$ m.

Step 2 — the thrust grading (recalled from `aq-sr-born-rigidity`). $\alpha_{\text{tail}} = c^2/x_{\text{tail}}$ , so

\Delta\alpha = \alpha_{\text{tail}} - \alpha_{\text{nose}} = \mathbf{1.0700\times10^{-13}\ m/s^2}, \qquad \frac{\Delta\alpha}{g} = \mathbf{1.0911\times10^{-14}} .

Compute this stably as $L_0/(x_{\text{nose}} - L_0)$ , never as $1/(1 - L_0/x_{\text{nose}}) - 1$ : the latter is a catastrophic cancellation in double precision and gets the last three digits wrong (see `aq-sr-bell-spaceship-paradox-s6`, worked example 2).

Step 3 — the clock grading (owned here). The same number:

\frac{\mathrm{d}\tau_{\text{nose}}}{\mathrm{d}\tau_{\text{tail}}} - 1 = \frac{x_{\text{nose}}}{x_{\text{tail}}} - 1 = \frac{L_0}{x_{\text{nose}} - L_0} = \mathbf{1.0911\times10^{-14}} .

Step 4 — integrate over a year. One Julian year of tail proper time is $3.15576\times10^7$ s, so the nose gains

\Delta\tau = 1.0911\times10^{-14} \times 3.15576\times10^{7}\ \mathrm{s} = \mathbf{344\ ns}, \qquad c\,\Delta\tau = \mathbf{103\ m}.

Step 5 — the boundary number. The hull is $100$ m long; the accumulated clock error, as a distance, is $103$ m. After one year of flight the ship can no longer locate its own nose to within its own length using a clock synchronised at launch. A full second takes $2.9\times10^6$ years to accumulate — but a shiplength takes one. Born rigidity holds the hull together perfectly and lets its time fall apart.

2 · How long may a rigid ship be before its tail stops ageing?

Step 1 — set a tolerance. Demand the tail age at no less than $99\%$ of the nose's rate: $\mathrm{d}\tau_{\text{tail}}/\mathrm{d}\tau_{\text{nose}} \geq 0.99$ .

Step 2 — solve. $1 - L_0\alpha_{\text{nose}}/c^2 \geq 0.99$ , i.e.

L_0 \;\leq\; 0.01 \times \frac{c^2}{\alpha_{\text{nose}}} \;=\; 0.01 \times 9.1648\times10^{15}\ \mathrm{m} \;=\; \mathbf{9.16\times10^{13}\ m} \;\approx\; \mathbf{613\ AU}\ \text{at } 1\,g .

Step 3 — read the bound. Nothing stops you making the ship longer; the geometry stays consistent all the way out to $L_0 \to c^2/\alpha_{\text{nose}} = 9.16\times10^{15}$ m, about $0.97$ light-years. What happens as you approach it is that the tail's clock rate falls to zero and its required proper acceleration rises without bound.

Step 4 — the boundary number. $L_0 = c^2/\alpha_{\text{nose}}$ is both the point at which the engines must deliver infinite thrust and the point at which the tail stops ageing, because the same $x_{\text{tail}} \to 0$ does both. `aq-sr-born-rigidity` states the bound from the thrust side. From the clock side: a Born-rigid body may not be so long that its own tail falls out of time. Same bound, same number, a different reason — and arguably the better one, because it is the one that survives when the engines are replaced by a gravitational field and there is no thrust to speak of.

Origins · 2 minRigid body and how this was discovered

Rigid body

1687–1910 · Newton's unbreakable idealisation · Born asks what it means at speed · Herglotz and Noether count the cost

The classical fiction. A rigid body is the most useful lie in mechanics: particles whose mutual distances never change, with six degrees of freedom, an inertia tensor and Euler's equations. Newton needed it, Euler formalised it, and every statics course still runs on it. Taken literally it is also a device for signalling faster than light — push one end and the other moves now — which is why relativity cannot have it. What Born asked. In 1909, Max Born asked the honest question: what is the nearest thing to a rigid body that relativity permits? His answer: demand not that distances be constant in some frame, but that the proper distances between neighbouring elements, measured in the material's own rest space, be constant. That is a local, tensorial condition, and it is the definition this article computes. Born's rigid body does not signal superluminally — every element's motion is prescribed in advance, and no push propagates anywhere. What it cost. Within a year, Gustav Herglotz and Emmy Noether had counted the degrees of freedom and found three, not six. You cannot spin up a Born-rigid disc from rest; you cannot deform one; you can do nothing to one except prescribe the worldline of a single point and let the rest follow. The rigid body survives in relativity only as a limiting idealisation — and, as this article shows, even the idealisation cannot keep its own clocks together.

History

1687

— Newton's Principia takes the rigid body as primitive; distances between the parts of a body are simply assumed constant.

1765

— Euler, Theoria motus corporum solidorum, gives the rigid body its six degrees of freedom and its equations of motion.

1909

— Born, Ann. Phys. 30, 1, defines rigid motion relativistically: constant proper distance between neighbouring elements in the momentarily comoving rest space.

1909

— Ehrenfest poses the rotating-disc paradox, which is what makes the rigidity question urgent rather than academic.

1910

— Herglotz (Ann. Phys. 31, 393) and Noether (Ann. Phys. 31, 919) independently prove that Born-rigid motions form a 3-parameter family: no smooth spin-up from rest.

1959

— Dewan and Beran's two-rocket problem shows what happens when you fail to grade the thrust: the connecting string breaks (`aq-sr-bell-spaceship-paradox`).

1966

— Rindler, Am. J. Phys. 34, 1174, relates the uniformly accelerated frame to the Kruskal extension, fixing the chart in which a Born-rigid body is simply at rest.

Going further

Texts. Born, Ann. Phys. 30, 1 (1909) · Herglotz, Ann. Phys. 31, 393 (1910) · Noether, Ann. Phys. 31, 919 (1910) · Rindler, Relativity: Special, General and Cosmological, 2nd ed., §§ on rigid and hyperbolic motion · Giulini, arXiv:0802.4345, on Herglotz–Noether · Ellis, Maartens & MacCallum, Relativistic Cosmology, ch. 4, for the expansion–shear–vorticity decomposition in its native habitat.

Provenance. Signature $(+,-,-,-)$ . Every equation and number is recomputed in verify_aq-sr-born-rigidity-s6.py, which reads this file, regexes the printed values out, and asserts each against what it computed at fifty digits. Base defects repaired: (1) the base's metric card prints $\mathrm{d}s^2 = -(\alpha x/c^2)^2c^2\mathrm{d}T^2 + \mathrm{d}x^2$ — wrong signature, and degenerate, because the same card defines $\alpha(x) = c^2/x$ , which makes the lapse identically $1$ ; the corrected card names the fiducial constant $a_0$ . (2) The base's rigidity-condition card prints $\partial_{(\mu}\xi_{\nu)} + \dot u_{(\mu}\xi_{\nu)} = 0$ , not a well-formed tensor equation; the correct condition is $\theta_{\mu\nu} = 0$ , derived here. (3) The base's third derivation contains a literal unfinished formula, $\mathrm{d}\tau = (x/c^2)\alpha\,\mathrm{d}T = \tfrac{x}{x}\dots$ — discarded. (4) The base calls the clock ratio a first-order expansion; it is exact when $g$ is the tail's own proper acceleration.

Scope. $\alpha(X) = c^2/X$ , rear-pushes-harder, $L_0 < c^2/\alpha_{\text{nose}}$ and Herglotz–Noether are owned by `aq-sr-born-rigidity` and are recalled, never re-derived. The Unruh effect (`aq-sr-unruh-effect`) is gestured at only: the Rindler wedge this congruence sweeps out is the arena in which it is computed, and nothing further is claimed about it here.

Isn't "the proper distance between neighbouring points stays constant" already a perfectly good definition?

It is correct, and it is the one Born wrote. It is just not a safe one: it contains the words "proper distance" and "neighbouring points", both of which require a rest space to be nominated at each event before they mean anything, and every wrong resolution of Bell's paradox in sixty years has been an argument about whose rest space is right. The expansion tensor removes the argument: $\theta_{\mu\nu}$ is built from $u^\mu$ and its derivatives alone, transforms as a tensor, and is zero or is not. The two definitions are equivalent — $\theta_{\mu\nu} = 0$ is exactly the statement that the Lie derivative of $h_{\mu\nu}$ along the flow vanishes, which is Born's sentence in symbols — but only one of them can be checked without a conversation.

Why does the shear have to vanish separately? Isn't a rod one-dimensional?

The rod is; the congruence is not. A four-velocity field lives on the whole wedge and has a full $3\times3$ spatial gradient available to it, so a body could keep its volume exactly constant ( $\theta = 0$ ) while being squashed along $x$ and stretched along $y$ at compensating rates ( $\sigma_{\mu\nu} \neq 0$ ). That body is deforming, and nobody would call it rigid. Demanding only $\theta = 0$ under-specifies rigidity; demanding $\theta_{\mu\nu} = 0$ does not. For the Rindler congruence the shear does vanish along with the expansion — but that is a result, obtained by computing all sixteen components, not something one may assume.

Does the tail's clock really run slow, or does it just look that way from the nose?

Really. This is not a signalling artefact and not a Doppler shift. Both ends are permanently at rest in the same static chart and permanently at rest relative to each other, and the ratio $\mathrm{d}\tau_{\text{nose}}/\mathrm{d}\tau_{\text{tail}} = x_{\text{nose}}/x_{\text{tail}}$ compares two proper times accumulated between the same pair of slices from the shared fan — a comparison every element of the rod agrees on, because they share the fan. Bring the clocks back together after a year and one has logged $344$ ns more than the other. That is a fact about the worldlines, not about the observing.

If the rigid rod's ends can't stay synchronised, in what sense is it rigid at all?

In precisely the sense the definition says, and no other. $\theta_{\mu\nu} = 0$ constrains the spatial geometry of the congruence: proper distances between material elements, in the material's own rest space, never change. It says nothing about the rate at which the elements' clocks run, and no reason it should — clock rates are a statement about $g_{00}$ , rigidity a statement about the projector $h_{\mu\nu}$ , which is orthogonal to exactly that direction. A Born-rigid body's shape is eternal. Its time is not, cannot be, and never was promised to be.

Is the Rindler metric here a different metric from the one in the Rindler-metric article?

No — the same metric in the same chart, arrived at by a different route and for a different purpose. `aq-sr-rindler-metric` derives it as the coordinate system a single accelerated observer builds, and owns the clock-rate law and the flatness proof ( $R^{\lambda}{}_{\mu\nu\kappa} = 0$ ). Here it arrives as the rest chart of an extended body: a Born-rigid body is exactly a body that is static in this metric, every element pinned at fixed $x$ forever. That reframing is what makes the length bound and the clock bound the same bound.

Could you build a Born-rigid ship?

Not exactly, and the reason is instructive. Born rigidity is not a material property; it is a prescribed programme of motion. Nothing about the ship enforces it — you must feed every element its own precomputed proper acceleration, graded as $c^2/x$ , and if you get it wrong the hull takes up the difference as stress. Real materials transmit stress at their sound speed, far below $c$ , so a real ship rings after ignition and only settles toward the rigid programme once the transient dies. What you can build is a ship Born-rigid to within its elastic tolerance: for a $100$ m hull at $1\,g$ that means grading the thrust by one part in $10^{14}$ , about $10^{-13}$ m/s² between the ends. No engine control system on Earth resolves that, and no hull would notice if it did.