The Talus as a Mechanical Gearbox

A preliminary computational test of a mechanical transmission hypothesis in human movement

Research report

Aleksandr Myshkin
Vitality Audit

The talus is usually described as the central bone of the ankle joint, transmitting loads between the tibia and the foot. Its geometry is generally discussed in terms of joint congruence, stability and range of motion.

We propose a different mechanical hypothesis:
The talus, together with the surrounding ankle structures, may function as a mechanical gearbox — changing the relationship between force, moment and angular displacement as its position changes.

This report describes a minimal computational investigation of that hypothesis.

The starting observation was a mechanical difference between deadlifting and running. In a deadlift, increasing load appears to drive the ankle complex toward a more deeply engaged, mechanically constrained configuration. Running requires something different: loading, mechanical engagement, transfer and rapid release, repeated within each support cycle.

The first question was deliberately narrow:
Does the geometry of the talus contain enough spatial variation to support mechanically distinct transmission regimes?

Published 3D measurements of the talar trochlea provide a remarkably simple test. The anterior and posterior regions of the medial and lateral trochlea have substantially different radii of curvature. Using those values in a minimal geometric model produced differences of approximately 45% in the medial anterior-to-posterior radius ratio and 17% on the lateral side.

A cyclic model then showed that movement through these geometrically different regions can produce a reversible change between force-dominant and velocity-dominant mechanical states.

This does not establish that the human talus is literally a gearbox, nor does it reconstruct the actual contact trajectory during running. It does establish something narrower and, in our view, important:
The geometry of the talus has the mechanical prerequisites for functioning as a multi-regime transmission rather than merely as a passive hinge surface.

The next stage is to determine whether human movement actually exploits these mechanically distinct states.

1. The hypothesis

The hypothesis tested here is deliberately mechanical.

We are not proposing that the talus resembles an automotive gearbox visually, nor that it contains discrete anatomical gears.

We are proposing that it may perform an analogous mechanical function:

to change the relationship between transmitted force/moment and movement as its geometric position changes.

A gearbox does not create energy. It changes how available energy is expressed:
high force / low speed
can become
lower force / higher speed,
and vice versa.

This is a stronger claim than saying that the ankle has a variable lever arm.

A simple variable lever produces:

G = G(q)

where the mechanical relationship changes continuously with joint position q.

A gearbox-like mechanism would require something more interesting:
different geometrical regions — different mechanical regimes — transitions between those regimes.

The distinction is central to this investigation.

2. Why the talus suggested this hypothesis

The hypothesis did not begin with an abstract mathematical model.

It began with a mechanical observation.

During a heavy deadlift, the ankle is part of a closed kinetic chain connecting the bar, the body and the ground. As load increases, the system appears to seek a mechanically constrained configuration in which the talus becomes increasingly engaged within the ankle mortise.

The anatomical literature provides a solid basis for the geometric part of this observation. The talus is wider anteriorly than posteriorly. During dorsiflexion, its wider anterior portion enters the mortise and produces a more closely packed and stable configuration; during plantarflexion, the wider anterior portion moves out of the mortise and the joint becomes less tightly packed. (PubMed Central (PMC))

This suggested a working concept that we called the functional latch:

increasing mechanical demand may drive the talar complex toward a more engaged configuration.

But running raised a different question.

Running cannot simply lock the ankle into one mechanically advantageous configuration. The system must repeatedly:
accept load — transmit it — change mechanical state — release it — repeat.

That led to the next hypothesis:
Perhaps the same anatomical mechanism that acts as a latch under load can also act as a transmission mechanism when its geometry is traversed dynamically.

3. The crucial geometric observation

The hypothesis became testable when we looked at the geometry of different regions of the talar trochlea.

The talar trochlea is not a constant-radius cylinder.

A 3D study of 18 healthy subjects using weight-bearing CT divided the medial and lateral talar surfaces into anterior and posterior regions and fitted separate radii to them. The reported mean radii were: (PubMed Central (PMC))

The same study found that the anterior-to-posterior radius ratio was approximately 0.70 on the medial side and 0.87 on the lateral side, with the anterior and posterior radii significantly different. (PubMed Central (PMC))

Independent 3D studies also report significant regional differences in talar curvature and confirm that anterior and posterior regions cannot simply be treated as one uniform cylinder. (PubMed Central (PMC))

This is the key anatomical fact on which the model rests.
The talus does not present one mechanically uniform curved surface.

It presents different geometrical regions.

4. The minimal model

We deliberately did not attempt to model the whole human ankle.

The purpose of this first experiment was not to reproduce human movement in detail. It was to test whether the proposed mechanism was mechanically possible at all.

Included:
anterior and posterior talar geometry;
medial and lateral radii;
a changing working radius;
an idealized mechanical transmission relationship;
cyclic movement through different geometric regions.

Excluded:
muscles;
neural control;
conscious movement control;
tendon elasticity;
soft-tissue deformation;
individual anatomy;
detailed ligament mechanics;
full 3D contact mechanics;
a predefined gait pattern.

This was intentional.

If a gearbox-like effect could not appear in such a minimal model, adding physiological complexity would not rescue the basic hypothesis.

Conversely, if it could appear, the result would justify a more detailed investigation.

5. First test: does the geometry produce a meaningful transmission difference?

For the first test, we used a simple geometric proxy.

If a mechanical element operates through different effective radii, an idealized rolling/lever relationship gives a corresponding change in the relationship between angular movement and transmitted force.

We therefore used the ratio:
G_speed = R2 / R1
and its reciprocal as a force/moment proxy:
G_force = R1 / R2

This should be understood precisely for what it is:
a geometric transmission proxy, not a measured ankle transmission ratio.
It deliberately isolates the effect of radius.

5.1 Medial trochlea
Anterior radius:
18.3 mm
Posterior radius:
26.6 mm
Therefore:
26.6 / 18.3 = 1.454
The geometric velocity ratio changes by approximately:
+45.4%
The reciprocal force/moment proxy changes to:
18.3 / 26.6 = 0.688
or approximately:
−31.2%

Thus, in the idealized model, moving between these two regions changes the mechanical relationship substantially.

5.2 Lateral trochlea
Anterior radius:
21.5 mm
Posterior radius:
25.1 mm
Therefore:
25.1 / 21.5 = 1.167
The corresponding velocity ratio changes by approximately:
+16.7%
and the reciprocal force/moment proxy becomes:
21.5 / 25.1 = 0.857
or approximately:
−14.3%

6. What this first calculation means

The result is larger than a trivial anatomical variation.

The medial anterior-to-posterior difference produces a geometric ratio of approximately 1.45.

The lateral difference produces approximately 1.17.

This immediately rules out the simplest mechanical representation of the talus:
a constant-radius cylindrical hinge.

It also gives us something more interesting than a single continuously variable lever.

The medial and lateral surfaces have different geometrical transmission characteristics, and the anterior and posterior regions within each side are different again.

In other words:
the talus contains multiple geometrically distinguishable mechanical regions.

That is the minimum structural condition required for a gearbox hypothesis to become mechanically plausible.

It is not yet proof of a gearbox.
But it is no longer merely an analogy.

7. From geometry to a cyclic transmission

The next question was:
What happens if the system moves cyclically through these different geometrical regions?

We constructed the simplest anatomically motivated cycle:
posterior — anterior — posterior

This corresponds to the broad mechanical sequence of:
plantarflexion — dorsiflexion — plantarflexion.

This is not an arbitrary movement pattern. In vivo studies of walking, running and hopping show a dorsiflexion phase followed by a plantarflexion phase, with the talocrural joint providing the majority of sagittal-plane motion of the ankle complex in these tasks. In running, approximately 83–86% of sagittal-plane motion is attributed to the talocrural joint depending on phase. (PubMed Central (PMC))

Importantly, this does not mean that we have reconstructed the exact contact path of the talar surface. We have not.

The model asks a more limited question:
If the working region of the talus changes from posterior to anterior and back again, what mechanical pattern follows from the measured geometry?

8. The modelled running cycle

1.077 — 0. Using the medial geometry:
posterior: 26.6 mm
anterior: 18.3 mm
posterior: 26.6 mm

we obtain the normalized sequence:
1.185 — 0.815 — 1.185
for the velocity proxy.

The reciprocal force/moment proxy is:
0.846 — 1.225 — 0.846

The important point is not the absolute numerical values. Those depend on normalization.

The important point is the reversible change in mechanical state:

velocity-dominant — force-dominant — velocity-dominant

with a maximum geometric ratio of approximately:
1.45 : 1
between the two extreme states.

The lateral geometry produces the same topology, but with a smaller amplitude:
923 — 1.077
with a corresponding ratio of approximately:
1.17 : 1

9. The first important interpretation

This is where the gearbox hypothesis becomes mechanically interesting.

A constant-radius model would produce:
G(t) = constant

A single smoothly varying lever would produce:
G(t) = continuously varying

Our two-region model produces:
one mechanical state — transition — another mechanical state — return.

The same physical structure therefore has the potential to operate differently depending on which part of its geometry is functionally engaged.

This is much closer to the mechanical principle of a transmission than to a simple hinge.

But there is an important qualification:

The model has demonstrated the possibility of different transmission regimes arising from talar geometry. It has not demonstrated that the human ankle actually uses those regions as discrete gears during running.

That distinction is retained deliberately.

10. The latch and the gearbox may be the same mechanism

The most interesting consequence of the calculation may be that our original concept of the talus as a functional latch and the new concept of the talus as a gearbox may not describe two different mechanisms.

They may describe two aspects of the same geometry.

Consider the anterior part of the talus.

It is:
wider than the posterior part;
differently curved;
associated with dorsiflexion;
driven more deeply into the ankle mortise during dorsiflexion;
associated with increased bony congruence and stability. (PubMed Central (PMC))

Thus the same movement can simultaneously produce:
greater engagement
greater geometric constraint
different effective radius
different mechanical transmission characteristics.

This suggests a possible chain:
geometry — engagement — constraint — transmission
rather than separate mechanisms for "locking" and "transmitting".

That is a potentially important conceptual shift.

11. Deadlift versus running

This model also gives us a more precise way to compare the two movements that originally motivated the hypothesis.

Deadlift
The mechanical problem is predominantly:
load — engagement — stabilization — force transmission

The system can remain in a highly loaded mechanical configuration.

The hypothetical function of the talus here is therefore predominantly:
to establish and maintain a force-dominant mechanical state.

Running
The problem is different:
load — engagement — transmission — release — next cycle

The system cannot simply remain in the mechanically constrained state.

It must repeatedly move between states.

Therefore:
The deadlift may emphasize the latch function; running may expose the transmission function.

This is not a claim that the same exact talar position occurs in every deadlift or running stride. It is a statement about the mechanical logic of the two tasks.

12. The unexpected comparison: walking

One of the most interesting outcomes of the broader modelling work was initially counterintuitive.

Walking looks like a close relative of running.

Mechanically, however, our minimal model suggested that ordinary walking is closer to the load-bearing logic of the deadlift than to the transmission-switching logic of running.

The distinction can be expressed schematically:

Deadlift
load — engage — hold

Ordinary walking
load — partial engagement — support — release

Running
load — engagement — transmission — release — repeat

The important distinction is therefore not simply:
slow movement versus fast movement.

It is:
How much of the available mechanical transmission range is actually being used?

This provides a possible mechanical interpretation of what we have called functional narrowing.

A person can walk normally while using only a relatively restricted portion of the available mechanical geometry.

That condition could remain invisible if assessment asks only whether the person can perform the movement.

13. What the experiment does — and does not — show

This distinction is essential to the integrity of the project.

What the model supports
The calculations support the following propositions:
1. The talar trochlea is geometrically heterogeneous.
2. Anterior and posterior regions have substantially different radii.
3. Medial and lateral regions have different geometrical characteristics.
4. Differences are large enough to produce substantial changes in a simple geometric transmission proxy.
5. A cyclic passage between different regions can therefore produce reversible changes between force-dominant and velocity-dominant mechanical states.
6. The geometry is therefore compatible with a multi-regime transmission mechanism.

What remains unproven
We have not yet demonstrated:

• the exact three-dimensional contact trajectory of the talus during running;
• the actual instantaneous torque ratio of the human talocrural joint;
• that the anatomical regions behave as discrete gears;
• the exact number of possible "gears";
• the quantitative contribution of ligaments;
• how muscles and tendons modify the mechanism;
• whether individual anatomical differences alter the transmission map;
• whether trained dancers actually exploit a wider transmission range.

They define the next experiments.

14. Why we stopped short of a full 3D reconstruction

A complete test would require combining:

• 3D talar geometry
• 3D tibiofibular mortise geometry
• in-vivo six-degree-of-freedom talar kinematics
• contact reconstruction
• force and moment data.

Such a model would be valuable eventually.

But it would also introduce many additional assumptions and parameters.

For the present question, that level of detail is unnecessary.

Our first objective was not:
"Can we reproduce the human ankle?"

It was:
"Is there enough mechanical structure in the talus for the gearbox hypothesis to be worth testing?"

The answer from the minimal model is:
Yes.

15. The strongest result so far

The strongest result is therefore not the numerical value of 45%.

It is the change in the type of mechanical description that is now justified.

Before the calculation, the talus could reasonably be represented as:
a joint surface with variable curvature.

After the calculation, a more specific hypothesis becomes mechanically legitimate:
a geometrically heterogeneous mechanical element capable, in principle, of changing transmission characteristics as different regions become functionally engaged.

That is the point at which the metaphor of a gearbox becomes a testable mechanical hypothesis.

16. Predictions

The gearbox hypothesis now makes several experimentally testable predictions.

Prediction 1
The mechanical relationship between ankle moment and angular movement should not be completely uniform across the talocrural range.

Prediction 2
The transition between dorsiflexion- and plantarflexion-dominant phases should be associated with a change in the effective mechanical transmission characteristics.

Prediction 3
Running should exploit a broader transmission range than ordinary level walking.

Prediction 4
High-load closed-chain tasks such as the deadlift should bias the system toward a more deeply engaged, force-dominant configuration.

Prediction 5
Different movement practices may use different portions of the available talar transmission range.

This last prediction is particularly interesting for future comparison with highly trained movers, including professional dancers.

17. A working vocabulary

For the purposes of this research, we propose the following terminology.

Talar latch
The load-dependent mechanical engagement of the talus within the ankle mortise.

Talar transmission
The change in mechanical relationship produced by movement through different regions of talar geometry.

Gear state
A mechanically distinguishable range of operation associated with a particular geometric configuration.

Gear switching
A transition between such states.

Functional transmission range
The portion of the mechanically available range that a person actually uses during a movement task.

These are working terms, not claims that anatomy textbooks should adopt them.

Their purpose is to make the hypothesis precise enough to test.

18. Conclusion

The talus is not a constant-radius mechanical element.

Its trochlea contains distinct anterior, posterior, medial and lateral geometries. Published 3D measurements show differences large enough that, even in a deliberately minimal geometric model, movement between these regions produces substantial changes in a radius-based transmission proxy. (PubMed Central (PMC))

When this geometry is placed into a simple cyclic model motivated by the dorsiflexion–plantarflexion structure of running, the result is a reversible sequence of mechanically different states:
force-dominant — velocity-dominant — force-dominant
or, depending on the direction in which the geometric path is traversed:

velocity-dominant — force-dominant — velocity-dominant.
This does not prove that the talus is a gearbox.

It does something more useful at this stage:

It shows that the gearbox hypothesis is mechanically plausible, quantitatively testable, and grounded in measurable anatomical geometry.

The next question is therefore no longer:

Could the talus possibly work this way?

The more interesting question is:
Does human movement actually use the mechanically distinct states made available by talar geometry?
That is the experiment that remains.

Sources and data used in this report

Bi-radial curvature of the healthy talus. The four radii used in the minimal model — 18.3, 26.6, 21.5 and 25.1 mm — come from this weight-bearing CT study of 18 healthy subjects. (PubMed Central (PMC))

Three-dimensional talar trochlea morphology. Independent CT-based work confirms distinct anterior/posterior radii and asymmetric medial/lateral geometry. (PubMed)

Broader 3D curvature data. A larger CT-derived dataset likewise found significant differences among regions of the talar trochlea. (PubMed Central (PMC))

Ankle geometry and anterior talar width. Anatomical and biomechanical literature describes the wider anterior talus entering the mortise during dorsiflexion and producing a more closely packed, stable configuration. (PubMed Central (PMC))

In-vivo locomotor kinematics. Recent biplanar-radiography work shows that talocrural motion provides most of the sagittal-plane motion of the ankle complex during walking and running and identifies distinct dorsiflexion and plantarflexion phases. (PubMed Central (PMC))