An Unsolved Mystery in Physics That Could Transform Armor Design

The components of your body armor plate – from the ceramic strike face, to the adhesive layers, to the UHMWPE backing – contain a secret: We don’t fully understand what they are. Not the chemistry; that’s well-characterized. Not the manufacturing process; that’s been refined over decades. The secret is more fundamental. These materials typically incorporate amorphous phases alongside crystalline ones, and those amorphous phases exist in a state of matter that physics cannot properly explain.

The glass transition – the process by which a liquid becomes an amorphous solid – is listed among physics’ great unsolved problems, alongside questions about dark matter and quantum gravity. Yet unlike those cosmic mysteries, this one has immediate practical consequences. Every time we design transparent armor for a vehicle, optimize a ceramic composite, or evaluate the shelf life of polymer-backed plates, we’re working with materials whose fundamental nature remains mysterious.

What Is a Glass, Really?

First, let’s dispel a persistent myth: Glass is not a slow-moving liquid. You may have heard that medieval cathedral windows are thicker at the bottom because the glass has flowed downward over centuries. This is false. Those windows are thicker at the bottom because medieval glassmakers couldn’t produce perfectly uniform panes, and builders sensibly installed them heavy-side-down. While we don’t fully know how glass behaves over geological timescales, calculations indicate that at room temperature, the time required for observable flow far exceeds the age of the universe.

Glass is a solid – but a peculiar kind. When you cool most liquids slowly, they crystallize: Atoms arrange themselves into neat, repeating lattices. But cool certain liquids quickly enough, and they skip crystallization entirely. The atoms never find time to organize. They freeze in place in a disordered configuration, forming what we call a glass.

This isn’t limited to the silicate glass in windows. Polymers like the polyethylene in your UHMWPE backer are semi-crystalline: Perhaps 80-90% of the material forms ordered crystalline lamellae, but the rest – the amorphous regions between those lamellae – is amorphous and viscoelastic. The epoxy or polyurethane adhesives bonding your armor’s layers together? Those are entirely glassy. Even some high-performance ceramics contain glassy phases at their grain boundaries, depending on processing conditions. And the Amorphoid in Adept Armor’s Thunder plate is almost entirely amorphous, hence the name – it’s a very lightly recrystallized glass-ceramic, with small crystalline particles in an amorphous matrix.

The Glass Transition Temperature

The temperature at which a liquid becomes a glass is called T_g, the glass transition temperature. And here’s where things get strange: T_g isn’t a fixed property of the material. It depends on how fast you cool the liquid. Cool it faster, and the transition happens at a higher temperature. Cool it slower, and T_g drops. This is profoundly odd. The melting point of ice doesn’t change based on how quickly you freeze water. But the glass transition is different – it’s not a true phase transition at all, at least not in any conventional sense.

The Missing Theoretical Framework

To understand why this matters for armor engineering, we need to contrast how we model crystals versus glasses.

For crystalline solids, physicists have developed a beautiful and powerful theoretical framework, broadly under the umbrella of Density Functional Theory.  First, you start with the crystal structure, with the material’s atoms arranged in a repeating lattice. From that structure, you can derive the vibrational modes (phonons). From the phonons, you can calculate thermodynamic properties, thermal conductivity, and heat capacity. From the interatomic potentials and lattice geometry, you can compute elastic moduli, almost always to within 10%. Defects – dislocations, vacancies, grain boundaries – are well-defined objects whose behavior can be predicted. The theory is predictive: Given the atomic constituents and their interactions, you can calculate macroscopic properties without having to measure them first.

For glasses, there is absolutely no equivalent framework. This isn’t because no one has tried. Generations of physicists have worked on the problem. The difficulty is fundamental: Glasses are disordered and out of equilibrium and their properties depend on their history. The standard toolkit – equilibrium statistical mechanics, symmetry analysis, perturbation theory around ordered states – doesn’t straightforwardly apply.

What We Cannot Currently Predict

Consider what we cannot currently do:

We cannot predict T_g from molecular structure.  For a new polymer or metallic alloy, you have to synthesize it and measure the glass transition temperature.  There’s no reliable equation that takes atomic characteristics, molecular weight, chain stiffness, and interaction strengths as inputs and outputs T_g.

We cannot predict glass-forming ability. Why does one alloy composition readily form a bulk metallic glass while a nearby composition crystallizes?  Why do some polymers vitrify easily while others don’t?  Semi-empirical rules exist, but they’re unreliable guides for new systems.

We cannot predict how processing affects properties from first principles.  Two glass samples with identical composition but different thermal histories will have different densities, different elastic moduli, different fracture toughness.  We can measure these differences.  We can fit empirical models to the data.  But we cannot start from the molecular structure and the processing parameters and calculate what will come out.

We cannot define “defects” in glasses the way we do in crystals.  In a crystal, a dislocation is a topologically distinct object.  In a glass, what’s a defect versus just… structure?  Ramsey Theory teaches that there’ll be islands of order, which in this case can approximate defects, in any sufficiently large chaotic structure.  Machine learning has recently identified “soft spots” in amorphous materials – regions likely to rearrange – but we don’t understand what makes them soft in terms of fundamental physics.

The models we do have are descriptive, not predictive.  The Tool-Narayanaswamy-Moynihan model for physical aging uses a “fictive temperature” concept and can fit experimental data quite well.  The Williams-Landel-Ferry and Vogel-Fulcher-Tammann equations describe how viscosity changes near the glass transition. These are useful engineering tools.  But the parameters must be measured for each material; they don’t emerge from first principles.  And extrapolating these models beyond their fitting range is unreliable – which matters when you’re trying to predict 20-year shelf life from accelerated testing.

Why the Fundamental Physics Matters for Armor

You might reasonably ask: Can’t we just build models using assumptions?  Engineers don’t wait for philosophers to settle debates before building bridges.

The answer is that the unsolved physics has at least two possible solutions, and we don’t yet know which is correct.

Here’s the crux  of it: In 1948, chemist Walter Kauzmann noticed that if you extrapolate the entropy of a supercooled liquid to lower temperatures, it would eventually drop below the entropy of the corresponding crystal – an apparent impossibility. This “Kauzmann paradox” suggests that something dramatic must happen before that point is reached. Either the liquid crystallizes, or there’s a true thermodynamic phase transition – an “ideal glass transition” – that we never observe because kinetic arrest always intervenes first.

RFOT vs Dynamical Facilitation: Two Competing Theories

This has spawned two competing families of theories:

Random First-Order Transition (RFOT) theories hold that there is a true thermodynamic transition underlying the glass transition. In this picture, the glass is a “mosaic” of different local configurations, and the key physics involves configurational entropy and the growing size of cooperatively rearranging regions. The mathematical framework involves replica theory borrowed from spin glass physics.

Dynamical facilitation theories hold that the glass transition is purely kinetic – there’s no hidden thermodynamic transition, but physical dynamics in the material become extraordinarily slow. In this picture, mobility is the key variable: Relaxation in one region enables (“facilitates”) relaxation in neighboring regions, creating complex spatiotemporal patterns. The mathematical framework involves kinetically constrained models and dynamic phase transitions.

These competing theories suggest different mathematical structures for building predictive models.  RFOT points toward free energy landscapes, configurational entropy as the control variable, and mosaic length scales.  Facilitation points toward dynamic heterogeneity, kinetic constraints, and mobility fields. Betting on the wrong framework means building on a foundation that will eventually crack.

The situation is analogous to trying to build electromagnetic devices in the early 1800s, when some physicists thought electricity and magnetism were separate phenomena and others suspected a deeper connection.  You could build motors and generators either way; empirically, by trial and error.  When Maxwell published a unified framework, he satisfied scientific curiosity, to be sure – but his discovery also enabled radio, radar, and wireless communications, whereas the wrong framework would have been a dead end.

For glasses, we’re still waiting for our Maxwell.  We have competing frameworks, each with successes and failures, and no consensus on which is correct – or whether the true theory combines elements of both.  Until this is resolved, our ability to build predictive (as opposed to descriptive) models for amorphous materials will remain limited.

The Strange Physics: Four Surprises

If this were merely a gap in theoretical elegance, it would matter little to engineers. But the physics of glasses produces genuinely strange effects that matter for real materials.

1. Dynamic Heterogeneity: Your Armor Is a Frozen Patchwork

At any given moment in a glass or supercooled liquid, some microscopic regions are relaxing – rearranging their configurations – while others remain frozen. Different regions of the same material, micrometers apart, can have relaxation rates differing by orders of magnitude. This “dynamic heterogeneity” has been directly imaged in colloidal glasses and confirmed in molecular simulations.

What this means: The amorphous regions in your armor aren’t uniform. There’s a hidden patchwork of mobility. How this affects response to ballistic impact – where strain rates are extreme and stress concentrations intense – is not well understood.

There does appear to be an amorphous-to-stishovite transition in certain silicate glasses under certain types of ballistic impact, but the nature of this phenomenon is still exceedingly mysterious, and most experts find it surprising when they first encounter it.

2. Physical Aging: The Plate Is Different Tomorrow

Glasses are never truly in equilibrium. They relax toward equilibrium forever, changing density, stiffness, brittleness, and permeability along the way. The armor plate in a warehouse today is measurably different from what it will be in five years.

NIST researchers studying UHMWPE fibers have documented complex aging involving both oxidative degradation and physical relaxation. The physics is entangled: crystalline and amorphous regions age differently, and they interact. Our shelf-life predictions carry uncertainties we cannot eliminate without better fundamental understanding.

3. The Gardner Transition: A Phase Transition Inside the Glass

Recent work has identified the Gardner transition – a phase transition within the glassy state. Below this transition, the energy landscape fractures from a single basin into a hierarchy of sub-basins. The glass enters a “marginally stable” state where small perturbations trigger avalanches of rearrangements.

The implications for armor under impact – where the material experiences extreme, rapid loading – are potentially significant but almost unexplored. We don’t even know if the Gardner transition occurs in amorphous solid materials under conditions relevant to armor.

4. Machine Learning Reveals Hidden Order

For decades, physicists sought a structural “order parameter” for glasses – some quantity that would predict dynamics the way crystalline order predicts behavior in solids. They mostly failed; a glass looks structurally almost identical to the liquid it came from.

Then an AI system, poring over vast amounts of data, found something. Researchers trained algorithms on simulations and discovered that local structure does predict which particles rearrange – just not in ways humans were able to perceive. The machine-learned quantity, called “softness,” captures subtle patterns invisible to traditional analysis.

This is both hopeful and sobering. Hopeful in that structure-property relationships exist; we just couldn’t see them. Sobering in that the order parameter is high-dimensional and non-intuitive. Understanding glasses may require mathematical frameworks that haven’t been developed yet.

Implications for Armor Engineering

Consider ceramic armor.  High-performance ceramics like silicon carbide and boron carbide derive their ballistic effectiveness from complex microstructures.  Some ceramics – particularly those processed with oxide sintering aids, and liquid-phase sintered SiC – contain glassy intergranular phases. The properties of these phases depend sensitively on processing: Sintering temperature, cooling rate, atmosphere, trace impurities. Small changes can produce meaningful differences in ballistic performance.  Why?  We don’t fully know, because we lack a predictive theory for how processing determines the properties of those amorphous regions.

(To be clear: Not all armor ceramics contain glassy phases. Hot-pressed boron carbide and solid-state sintered silicon carbide can have clean grain boundaries. But the general point stands, that where amorphous phases exist, our predictive capability is limited.)

Next, as mentioned previously, consider polymer backing materials. UHMWPE is semi-crystalline. Its ballistic properties emerge from highly aligned crystalline lamellae, but the amorphous regions govern critical behaviors: Impact response, creep under load, aging over time.  We can characterize these behaviors empirically.  We cannot predict them from the molecular structure and processing history.

Then consider adhesive layers. The epoxies and polyurethanes bonding armor systems are entirely glassy. How will a new adhesive formulation perform over 15 years of storage in variable conditions? We can run accelerated aging tests and fit empirical models. But we’re extrapolating, and the uncertainty in those extrapolations reflects our incomplete understanding of glass physics.

Armor engineers reading this know that adhesive failure is the most unpredictable, most vexing reason plates fail – and one which seems to have contributed to high-profile struggles in the sector, such as those faced by Ceradyne.

Lastly, consider transparent armor – glass by definition – and metallic glasses, which offer extraordinary hardness and elasticity but whose glass-forming ability we cannot reliably predict.  In each case, development proceeds by empirical optimization: Make samples, test them, adjust. This process, one not unfamiliar to ancient alchemists, works. Armor gets fielded. But it’s slow, expensive, and definitely leaves performance on the table.  In fact, I’ll say outright that metallic glasses are still very rare and exotic simply because there’s no physical theory for them.

Science as Strategic Advantage

Unlike the deep mysteries of the universe, the glass transition problem represents a missing chapter in our understanding of matter – one that directly limits our ability to design and optimize militarily-relevant materials.

This argues for sustained investment in fundamental condensed matter physics.  The nation that develops a predictive framework for amorphous materials gains the ability to design what others can only discover by laborious trial and error. In an era where materials often determine the limits of military capability, that’s a decisive advantage.

The mystery of glass is hiding in plain sight – in the armor plates, vehicle windows, and helmet shells we use every day.  The field of materials science is waiting for our Maxwell.