From Rock Tumbling to Telescope Mirror Grinding: How Optical Perfection Emerges from Random Motion
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Astrophysics. The day in astrophysics: observations, models and what the data forces.
Vera: We're looking at the paper: "From Rock Tumbling to Telescope Mirror Grinding".
Jocelyn: Here's the gist of the paper. When two glass disks are rubbed together by hand with an abrasive between them, their surfaces spontaneously converge to perfectly matching spherical curves, accurate to tens of nanometers Page 1.
Vera: This paper derives a first-principles explanation for why rubbing two glass disks together spontaneously produces matching spherical surfaces Page 8.
Jocelyn: Derivation from Abrasion Law The derivation starts from Preston’s law of abrasion Page 3.
Vera: The authors show that symmetries of the grinding motion make planes and spheres generic equilibria of the process Page 4.
Paper summary: Vera: So what they're saying is that this matching isn't just some random lucky outcome; it’s governed by a system involving exponential relaxation and diffusion >.
Jocelyn: They define the small net change per stroke using an equation derived from integrating Preston’s law over one stroke period >.
Subrahmanyan: This leads to a coarse-grained evolution where small differences between the surfaces become derivatives of each other over time >.
Vera: They identify equilibrium configurations as those surfaces that share the symmetries of the grinding stroke motion >.
Jocelyn: For generic strokes, planes and spheres are these equilibrium configurations >.
Subrahmanyan: But gravity breaks that symmetry, it drives the top surface concave and the bottom surface convex, which makes planar surfaces unstable >.
Vera: The key insight here is a hierarchy of timescales: Tstroke is much smaller than Tmatch which is similar to Tsmooth and Tgrind >.
Jocelyn: Then you have Tsphere larger than that, but smaller than Tdeform, meaning the surfaces reach their spherical equilibrium faster than gravity or edge effects can destabilize them >.
Subrahmanyan: This framework unifies mirror grinding, rock tumbling, and sanding as regimes governed by the same equations >.
Vera: They show how in rock tumbling, the very short stroke excursion produces a spherizing time that is orders of magnitude longer than the smoothing time >.
Jocelyn: But for sanding small flat objects on the other hand, the stroke excursion can be comparable to or even exceed the object size, making deformation and grinding times comparable >.
Subrahmanyan: They also look at how stroke length sets a trade-off between spherization and deformation >.
Vera: A longer stroke speeds up smoothing but shrinks that deformation time, eventually violating that final timescale inequality and producing aspherical surfaces if the stroke is too long >.
Conclusion: Jocelyn: So looking at the title, "From Rock Tumbling to Telescope Mirror Grinding: How Optical Perfection Emerges from Random Motion," it really captures the idea that we can get optical perfection from simple random motion >.
Vera: The authors are showing that this convergence, these perfectly matching spherical curves accurate to tens of nanometers, isn't just a trick or something special; it follows from fundamental physics like Preston’s law >.
Subrahmanyan: It means that the process is robust because it depends on symmetry and timescale separation rather than needing incredibly detailed modeling of every single contact point >.
Jocelyn: The conclusion is that planes and spheres are the attractors for this system, and exponential matching, diffusive spherization follow from those same principles >.
Vera: And the testing they did with a light source and a razor blade confirms that you can actually achieve this smoothness without needing super precise tools >.
Subrahmanyan: It suggests that we might be able to use these same principles to understand how surfaces evolve in other random physical processes, maybe even in cosmology or structure formation >.
Emmanuel Schaan
SLAC National Accelerator Laboratory · Kavli Institute for Particle Astrophysics and Cosmology Stanford University
astro-ph.IM, physics.class-ph, physics.ed-ph, physics.optics
Submitted: 2026-11-01
Updated: 2026-11-01
Comments: Conditionally accepted in American Journal of Physics; comments welcome
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 99/100
The gist: When two glass disks are rubbed together by hand with an abrasive between them, their surfaces spontaneously converge to perfectly matching spherical curves, accurate to tens of nanometers Page 1.
Key concepts
- Preston’s Law of Abrasion
- This law describes the relationship between the motion of grinding or rubbing and the resulting material removal. The authors use this law as a starting point to model how surfaces change during the rubbing process, providing a mathematical basis for surface evolution.
- Timescale Hierarchy
- The paper establishes a specific order of events: Tstroke (the time for one stroke) is much shorter than Tmatch (time to match), which is shorter than Tsmooth (time to smooth), and so on. This separation shows that the surfaces reach their final spherical shape much faster than external factors like gravity can cause them to become unstable.
- Equilibrium Configurations
- The surfaces that are perfectly matched spheres or planes are the stable states, or attractors, of the grinding process. For generic motions, planes and spheres are these stable states. The analysis shows that these configurations are robust because of the underlying symmetries in how the rubbing motion is performed.
Terminology
Summary
When two glass disks are rubbed together by hand with an abrasive between them, their surfaces spontaneously converge to perfectly matching spherical curves, accurate to tens of nanometers Page 1. p. 0 This paper derives a first-principles explanation for why rubbing two glass disks together spontaneously produces matching spherical surfaces Page 8. p. 6
Derivation from Abrasion Law
The derivation starts from Preston’s law of abrasion Page 3. The authors show that symmetries of the grinding motion make planes and spheres generic equilibria of the process Page 4. p. 0 A hierarchy of timescales governs how surfaces smooth, match, and spherize Page 8. p. 0
Dynamics of Surface Evolution
The evolution is governed by a system that includes exponential relaxation and diffusion Page 5. The small net change per stroke is defined by an equation derived from integrating Preston’s law over one stroke period Page 4. p. 3 This leads to a coarse-grained evolution where small differences become derivatives Page 5. p. 3
Equilibrium and Stability
Equilibrium configurations are precisely those surfaces that share the symmetries of the stroke motion Page 4. For generic strokes, planes and spheres are the equilibrium configurations Page 4. p. 2 Gravity breaks this symmetry, driving the top surface concave and the bottom convex, thus making planes unstable Page 8. p. 4
Timescale Hierarchy
The key insight is a hierarchy of timescales: Tstroke ≪ Tmatch ∼ Tsmooth = Tgrind ≪ Tsphere ≪ Tdeform Page 8. The first inequality places us in the grinding regime, where each stroke has only a small cumulative effect Page 8. p. 6 The final inequality (Tsphere ≪ Tdeform) ensures that surfaces reach their spherical equilibrium faster than gravity and edge effects can destabilize them Page 8. p. 6
Unification of Phenomena
The framework unifies mirror grinding, rock tumbling, and sanding as regimes of the same equations Page 8. In rock tumbling, the very short stroke excursion produces a spherizing time orders of magnitude longer than the smoothing time Page 8. p. 7 Sanding small flat objects on the other hand, the stroke excursion typically exceeds the object size, making Tdeform comparable to Tgrind Page 8. p. 7
Role of Stroke Length
Stroke length sets a trade-off between spherization and deformation Page 7. Longer strokes speed up smoothing but shrink Tdeform, eventually violating the final inequality and producing aspherical, mismatched surfaces Page 8. p. 7 The normal stroke (one-third diameter travel) sits in a sweet spot where Tsphere ≪ Tdeform Page 7. p. 6
Numerical Estimates
Numerical estimates for a 25 cm mirror show the hierarchy of timescales Page 8. The first, sagitta-digging stage takes roughly ∼ 15 Tgrind ∼ 10 hours Page 8. p. 5 The second stage, spherizing with the normal stroke, is controlled by Tsphere ∼ 30 Tgrind ∼ 10 hours Page 8. p. 5 The characteristic timescales are summarized in Fig. p. 6 7 Page 8.
Pressure Field Modeling
The pressure field is modeled using an elastic (Winkler) foundation model where P = k (zB − zT) Page 3. The key quantity to model is the pressure field P(⃗x), which is shaped by surface irregularities, finite disk geometry, and gravity Page 4. p. 2 The ring-averaged pressure shows how grinding efficiency varies with radial position for different overhangs Page 9. p. 9
Edge Effects and Gravity
Gravity drives the top surface concave and the bottom convex, thus making planes unstable Page 8. The geometry of overlap and pressure asymmetry from gravity contribute to this effect Page 9. p. 4 Extreme overhangs lead to extreme edge pressure on the bottom surface Page 9. p. 4
Neglected Effects
The model neglects friction and horizontal forces from the stroking motion Page 8. It also neglects the operator’s hand pressure distribution and the viscoelastic pitch tool used in polishing Page 8. p. 7 The main results follow from symmetry and timescale separation rather than modeling details Page 8. p. 7
Conclusion on Robustness
The main results, planes and spheres as attractors, exponential matching, diffusive spherization follow from symmetry and timescale separation Page 8. These findings suggest that the attractors are robust to these idealizations Page 8. p. 7 The testing of the finished mirror surface confirms smoothness and figure accuracy using nothing more than a light source and a razor blade Page 8. p. 7
Acknowledgments
The author is deeply grateful to the members of the Telescope Makers’ Workshop at Chabot Space & Science Center in Oakland, CA, who generously shared their time and expertise in teaching him how to grind, polish, and figure a telescope mirror Page 8. Special thanks go to Gert Gottschalk, Richard Ozer, Dave Barosso, David Lawson, Phil Okada, Henner Zeller, Rolph, Nate, and Alan Page 8. p. 7 The author also thanks Patrick Jelinsky for lending him a wire tester Page 8. p. 7 Megan deBettencourt supported the process throughout Page 8. p. 7 The author performed this work independently; it does not represent the views of SLAC National Accelerator Laboratory or the U. p. 7S. Department of Energy Page 8. p. 7 The references are listed at the end Page 8. p. 0 The author is grateful to the editor and reviewers for their thoughtful feedback and suggestions, which improved the clarity of the paper Page 8. p. 7 The author performed this work independently Page 8. p. 7 The testing of the finished mirror surface is as physically rich as the grinding itself Page 8. p. 7
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved simulation of material removal processes by incorporating Preston's law (Eq 1) into a multi-scale analysis framework allows AI systems to predict grinding outcomes based on surface irregularities and stroke geometry. This enables AI to model how small-scale irregularities decay exponentially on the grinding timescale Tgrind and how large-scale deviations diffuse away on the spherizing timescale Tsphere.
-
Enhanced ability to distinguish between different surface evolution regimes by analyzing the hierarchy of timescales (Tstroke << Tgrind << Tsphere << Tdeform). AI systems could use this framework to determine whether a given grinding process is operating in the smoothing regime, matching and spherizing regime, or the destabilization regime driven by gravity and edges.
-
Development of a predictive model for optimal manual grinding strokes by calculating the required stroke parameters based on the predicted timescale hierarchy. Specifically, AI could identify which stroke (chordal, normal 1/3 diameter, MOT/TOT alternation) is best suited to achieve a desired final shape within practical time constraints defined by Tdeform and Tsphere.
-
Implementation of diffusion equation solvers to accurately predict the large-scale evolution of surface irregularities during the matching phase. This allows AI to quantify how quickly a surface deviates from its equilibrium shape based on the stroke excursion radius Rstroke and the relevant diffusivity kappa.
-
Creation of an automated system for analyzing rock tumbling and sanding processes by mapping their physical parameters (Rrock/Rstroke ratio) onto the derived timescales. This would allow an AI to predict whether these processes will produce spherical results or rounded edges based on whether Tsphere is orders of magnitude longer than Tgrind, as noted in the comparison with rock tumbling.
Related papers
- On-sky demonstration of a vector Zernike wavefront sensor in a cascaded adaptive optics system
- Euclid Data Release 1 (DR1): Calibration strategy and observations
- The General Antiparticle Spectrometer (GAPS) Antarctic Balloon Payload
- Detection and excision of radio-frequency interference in radio arrays
- Unified matched-filter and excess-power pipeline for gravitational-wave detector characterization
- Ushering in sub-arcsecond angular resolution observatories in the far-infrared: The case for a double-Fourier interferometer