A summary of ‘Jumping Random Walks: A Stochastic Description for the Modelling of Glass-like Materials.’ - Dr Bart Vorselaars, Delivered on 26th November 2025
Is glass a solid or a liquid? Most people would instinctively answer ‘solid’, but from a physicist’s perspective it’s a little more complicated than that. Although glass seems to be rigid and unchanging, its internal structure is completely disordered more like that of a liquid than a solid. Not only that, but it can also flow!! Just incredibly slowly
A good example of a material with both solid-like properties and liquid-like properties is pitch. The pitch drop experiment is the longest running experiment to date, dating back to 1927 and still running to this day. [1] Pitch acts as a solid in one sense: when struck with a hammer, it shatters. However, over long time scales it behaves like a liquid. Over the course of the whole experiment, it has formed 9 droplets: close to 1 droplet every 10 years.
This behaviour is quantified by a materials’ viscosity, which measures how easily a material flows. Water flows easily and therefore has a low viscosity. Obviously pitch, on the other hand, is extremely viscous. Glass is even more viscous, approximately 10,000 times more than pitch, meaning that its flow is essentially invisible on human timescales.
This behaviour can be explained on a molecular level. The atoms within glass are constantly jiggling due to thermal energy but is trapped by its neighbours with nowhere to go. Occasionally, a particle manages to rearrange slightly, changing the structure very slowly. This motion has no preferred direction and moves completely randomly. One could attempt to model individual particles, but this becomes very computationally difficult due to the large volume of atoms.
Dr Bart Vorselaars, in his seminar ‘Jumping Random Walks: A Stochastic Description for the Modelling of Glass-like Materials.’, delivered on 26th November 2025, explained that glass-like materials can be more effectively modelled using something called ‘jumping random walks’.
Let’s start with a random walk: it can be imagined as a person standing on a grid who takes steps of equal length in completely random directions. Because each step is equally likely to go forward as backwards, on average, a walker has a displacement of 0. However, the distance from the starting point on a specific iteration still increases over time. This distance is found to increase proportionally to the square root of the steps taken, which in simple terms means that quadrupling the number of steps taken would double the distance travelled.
This is important, as Dr Vorselaars derives that Brownian motion causes particles in a liquid to diffuse at the same rate! Meaning the distance particles travel in a liquid is proportional to the square root of the measured time, so after 4x the amount of time, the particles would spread double the distance. This means that we can use random walks as a mathematical model of liquids! But glass doesn’t exactly behave like a normal liquid. Its particles are mostly stuck in place, in tiny ‘cages’ formed by their neighbours. Every so often, they manage to ‘jump’ out, creating random motion over long timescales.
To capture this, Dr Vorselaars models the cages as a repeating, wavelike potential, where the particles rattle locally and occasionally ‘jump’. This sinusoidal model successfully shows the slow, long-term diffusion observed in glass-like materials. This combination of random walks and rare jumps is what Dr Vorselaars calls a ‘jumping random walk’.
This behaviour of glassy materials is more than just a theoretical curiosity, as it affects real-world engineering. As one review notes “Despite having many applications, current researchers still have difficulty in implementing coating challenges due to issues such as physical ageing, brittleness, etc.” [2]. This highlights why studying and modelling glassy polymers effectively is valuable: it helps scientists predict and improve the long-term performance of such materials used in coatings and other applications.
References
[1] Physics Museum (2026) 1927 Famous Pitch Drop Experiment. The University of Queensland. Available at: https://physicsmuseum.uq.edu.au/famous-pitch-drop-experiment (Accessed: 9 January 2026).
[2] Arya, R.K., Thapliyal, D., Sharma, J. and Verros, G.D. (2021) Glassy PolymersDiffusion, Sorption, Ageing and Applications. Coatings, 11(9), p.1049. Available at: https://www.mdpi.com/2079-6412/11/9/1049 (Accessed: 9 January 2026).
26801
Presentation: 14/15
You have followed the criteria of including the speaker’s name, title and date of the seminar, shown clearly at the start of the blogpost.
Content: 15/15
Conveyed the nature of glass by referencing the example in Bart’s Seminar (Pitch experiment), and covered the majority of the content in the seminar.
Context: 14/15
It connects the behaviour of glass to its potential applications in engineering and other sciences.
Style: 14/15
It is easy to read and explains concepts that are foreign to general public in relatively simple terminology, i.e. Brownian motion
Other sources: 14/15
Referenced additional sources for readers to explore for the factuality of the blogpost content, including the use of a direct quote.
Overall Mark: 14/15
Feedback: Repeated the date and title of the seminar in the blogpost, making the tone of post sound clunky at that part, it was not necessary to rewrite as the blog is already talking about Bart’s Seminar. The term stochastic is not directly defined, so I think the general public could inaccurately infer the benefits of a stochastic description.