Pressure Waves During Granular Flows in Varying Gravity Environments

\firstname Abigail \lastname Tadlock \fnsep, \firstname Lori \lastname McCabe \fnsep, \firstname Kerstin \lastname Nordstrom \fnsep · 2025

Read Pressure Waves During Granular Flows in Varying Gravity Environments by \firstname Abigail \lastname Tadlock \fnsep, \firstname Lori \lastname McCabe \fnsep, \firstname Kerstin \lastname Nordstrom \fnsep (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-14525.txt.

Introduction

Abigail Tadlock and co-authors simulated sand-like flows under different gravity in this 2025 paper. They poured 30,000 soft spheres through a vertical silo inside a computer. Gravity strength, grain stiffness, and outlet size all varied between runs.

Waves of pressure rose upward through the falling grains in every setup. The waves moved far faster than the grains themselves. The authors argue these are nonlinear shock waves, like those seen in jammed materials. Wave speed followed a power law that breaks from the pattern of ordinary sound in loose grains. The results matter for handling powders on Earth and for guessing how grains behave on the Moon or Mars.

Catalog record

CREATOR\firstname Abigail \lastname Tadlock \fnsep, \firstname Lori \lastname McCabe \fnsep, \firstname Kerstin \lastname Nordstrom \fnsep
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE14 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-14525

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Pressure Waves During Granular Flows in Varying Gravity Environments \firstname Abigail \lastname Tadlock \fnsep, \firstname Lori \lastname McCabe \fnsep, \firstname Kerstin \lastname Nordstrom \fnsep Abstract We present results of LAMMPS Molecular Dynamics simulations of 2D gravity-driven flows of 30,000 soft uniform spheres through a vertical silo. We vary the gravitational field ( g g ), elastic modulus of the particles ( E E ), and silo outlet diameter ( D D ). We present results on upwards pressure waves observed in the system. We compare our results with previous work on granular acoustics. Despite the typical particle speed being substantially less than the measured wave speed, we posit these are nonlinear shock waves, as observed in other systems near jamming. We demonstrate that the wave speeds in all systems appear to follow a power law that is distinct from linear wave expectations. 1 Introduction The study of granular discharge flow from a silo is of great importance to many industries. It is also a simple system with inherent complexity, with non-uniform stresses and particle movements within the same system that may be quasistatic (“slow") through inertial (“fast"). The mass discharge rate in granular silos is usually adequately described by the empirical Beverloo equation beverloo_flow_1961 , which depends on the ratio of outlet diameter ( D ) to particle diameter ( p_{d} ). The equation also includes gravity as a factor of \sqrt{g} . Previous experimental work has tried to investigate this scaling mathews_model_2016 using a large centrifuge, but had limited dynamic range. Understanding the role of gravity in basic granular processes is important, especially with the advent of increased planetary/asteroid surface exploration. Pulsatile behaviors have been observed in industrial silos for years, and sometimes create unsettling shock waves at frequencies humans can hear roberts_flow_2002 . The mechanism of this is stick-slip at the wall, and the actual discharge rate with time may be affected. In contrast, we observe more subtle pressure waves that do not significantly impact the discharge rate itself, but are nonetheless present, suggesting these waves are a universal feature of silo flow. We find dependencies on elastic modulus and gravity, and compare to other work on granular acoustics. 2 Simulations Simulating granular flow allows us to easily vary conditions that are difficult to experimentally alter, such as gravity. We use the granular package of LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator) to create our system thompson_lammps_2022 . This package uses the Hertz-Mindlin model of particle interaction. Some parameters we keep constant for all simulations, such as the coefficient of restitution of 0.66, coefficient of static friction of 0.5, and a Poisson’s ratio of 2/7 , similar to other work using this model silbert_granular_2001 . We simulate a quasi-2D system of monodisperse spheres in a flat-bottomed silo that is 63.1 particle diameters ( p_{d} ) wide x 700 p_{d} tall x 1.1 p_{d} deep. Quasi-2D systems are frequently used in silo flow research and serve as analogues to 3D systems hong_clogging_2017 ; hong_clogging_2022 ; endo_obstacle-shape_2017 . The p_{d} = 1 m, and the density, \rho , is 1 kg/m 3 . The simulation uses timesteps of 0.0001 s, with data writeout every 1000 timesteps (every 0.1 seconds). We systematically vary three elements of the simulation: gravity, elastic modulus, and outlet diameter. Gravity is varied between 0.1 and 20 times Earth’s gravity, g = 9.8 m/s 2 , the elastic modulus is varied between 10^{4} - 10^{7} Pa, and the outlet diameter is varied between 6 p_{d} and 20 p_{d} . We define a dimensionless constant \Gamma in equation 1 to navigate within this phase space and compare different simulations. \Gamma=\frac{\rho g}{p_{d}E}\propto\frac{g}{E} (1) In our simulations, since the particle diameter p_{d} and the density \rho are both 1 in their respective units, \Gamma is numerically equivalent to g/E . Throughout our simulation phase space, many simulations have the same values of \Gamma and we observe some similarities in behavior in these systems’ mass flow rate and other characteristic measurements, although that is not the focus of this paper. Figure 1 shows our phase space with dashed lines denoting simulations that have the same values of \Gamma . Figure 1 : Phase space of \Gamma as a function of gravity and elastic modulus. In our simulations, \rho and p_{d} are 1, allowing these different systems to be directly compared. Figure 2 : Pseudoimage of one simulation ( g=1 g, D_{0}=15 pd, E=10^{4} Pa) at a particular time. The colorscale is instantaneous vertical velocity. The dark blue “bubbles" are the pressure waves. They travel upwards even though the constituent particles are moving down at high velocity. The waves also have a lower density than the surrounding particles (visible from the blank white spots in the blue regions). We use the LAMMPS create_atoms command to insert particles into the silo. All of the particles are created at once, located at random positions with random velocities within the silo, and allowed to settle for an extended period of time such that the kinetic energy is minimized. The random initial positions and velocity depend on a seed specified in the code, which we did not vary. As a result, the initial packings for particular g / E combinations are identical. After the packing has settled, we remove the “plug" at the outlet instantaneously (in a single data time-step) and allow the particles to flow with no further modifications to the simulation. Figure 3 : Spacetime diagrams for two systems with the same gravity, g=1 g, and outlet, D_{0}=9\,p_{d} , but different values of particle stiffness, (a) E=10^{4}\,Pa (b) E=10^{7}\,Pa . The slope of the line is the velocity of the wave and the separation gives both frequency and wavelength, as shown in (a). 3 Pressure Waves We see pressure waves develop in all systems, illustrated in Figure 2 . These waves travel upward in the silo, opposite the gravitational flow of the particles. Generally, a softer system will have less frequent waves and a slower wave speed than a stiffer system with the same value of g . The softer systems also have more attenuation. We do not see any dependence on outlet size. In some simulations, the initial wave pulse was clear to the eye: as a void formed near the outlet, particles along the sides would rush into the empty space, catalyzing an upward pressure wave of lower density. We have separately verified that these pulses have a slightly lower particle density than the packing structure of the surrounding particles. Figure 4 : The raw (a) wave speed and (b) frequency of each system, and the normalized by \Gamma (c) wave speed and (d) wave frequency, fitted with a -1 power law fit. We note the velocity had multiple datapoints of the same value, so they overlap. The collective dataset is the same size in all subfigures. Figure 5 : Wave speed versus gravity (a) raw data and (b) scaled by the expected speed of sound for a continuous medium, c_{0} . The collapsed data has two potential power law comparisons, of 1/5 (dashed) and 1/6 (dotted). To analyze these pressure waves, we generate spacetime diagrams by averaging the instantaneous velocity of all particles at a given height. This creates a vertical stack

Read the complete text