Mason Number: Contexts & Applications

• Mason number is a dimensionless measure that compares viscous forces to magnetic forces in various magneto-fluid dynamic models.
• It organizes critical regimes in high gradient magnetic separation, MR fluids, and magnetic micro-swimmers by balancing magnetic and hydrodynamic effects.
• In algebra and combinatorics, Mason number denotes degree bounds and log-concavity properties, offering invariant measures in polynomial and matroid theories.

Searching arXiv for relevant papers on “Mason Number” across magnetics, micro-swimmers, algebraic Stothers–Mason, and matroid Mason–Welsh contexts. “Mason number” denotes several distinct technical quantities whose meanings depend on context. In high gradient magnetic separation, magnetorheological suspensions, and magnetic micro-swimmers, it is a dimensionless measure of viscous drag or torque relative to magnetic forcing, and it organizes capture, rheology, and orientational dynamics (Eisenträger et al., 2014, Vagberg et al., 2017, Rüegg-Reymond et al., 2018). In the Stothers–Mason theorem, the same name is loosely attached to the radical-based invariant $M(a,b,c)=\deg(\mathrm{rad}(abc))-1$, and in the $q$-difference setting to $M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$ (Lu et al., 27 May 2026). In matroid theory, related terminology appears in the Mason–Welsh sequence $(I_0(M),I_1(M),\dots,I_r(M))$, although the cited paper does not introduce a separate explicit symbol for a “Mason number” (Giansiracusa et al., 2024). This suggests a family of domain-specific invariants rather than a single universal definition.

1. Magnetic-force and magnetic-torque interpretations

In magnetic transport and rheology, the Mason number is a nondimensional ratio comparing viscous and magnetic effects. The precise scaling depends on the governing model, but the common structure is a ratio of a characteristic viscous force or torque to a characteristic magnetic force or torque. Small values correspond to magnetic dominance; large values correspond to hydrodynamic or viscous dominance (Eisenträger et al., 2014, Vagberg et al., 2017, Rüegg-Reymond et al., 2018).

Context	Definition	Interpretation
HGMS	$\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$	typical viscous force / typical magnetic force
MR fluids	$\mathrm{Mn}=\dfrac{F_d}{F_m}$	viscous force scale / magnetic force scale
Magnetic micro-swimmers	$a=\dfrac{\alpha\eta\ell^3}{mB}$	viscous torque scale / magnetic torque scale

For high gradient magnetic separation, the relevant force balance is the Stokes drag on a small paramagnetic sphere moving through a periodic array of magnetized cylinders. After nondimensionalization, the particle equation takes the form

$$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$

so the magnetic contribution is weighted by $1/\mathrm{Mn}$ (Eisenträger et al., 2014). In dense non-Brownian magnetorheological fluids, the Mason number is written as $\mathrm{Mn}=F_d/F_m$, with $q$0 taken from the maximum magnetic attraction between two just-touching small particles and $q$1 chosen according to the dissipation law (Vagberg et al., 2017). For magnetic micro-swimmers in Stokes flow, the Mason number $q$2 compares the driving angular speed to the magnetic alignment rate, and it enters the autonomous orientation equations as the coefficient of the imposed rotation term (Rüegg-Reymond et al., 2018).

A common misconception is to treat these magnetic-fluid definitions as interchangeable. The papers instead show that the physical idea is shared, but the exact scale choice is geometry- and model-dependent: HGMS uses a particle-scale force balance in a wire array, MR fluids use microscopic particle-force scales under shear, and micro-swimmers use a torque balance for a rigid permanent-magnetic body.

2. High gradient magnetic separation

In “Particle Capture Efficiency in a Multi-Wire Model for High Gradient Magnetic Separation” (Eisenträger et al., 2014), the Mason number is the key dimensionless group controlling whether a paramagnetic particle is captured by, or escapes from, a periodic array of magnetized cylinders. The particle is modeled as a small paramagnetic sphere of radius $q$3 in a viscous fluid of dynamic viscosity $q$4, moving through a square lattice of cylinders of radius $q$5. Inertial effects are neglected, and the dimensional force balance is

$q$6

Using the maximum Stokes-flow velocity $q$7 as velocity scale and $q$8 as magnetic-force scale yields

$q$9

The sign distinguishes two magnetization setups, but the capture results depend only on $M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$0.

The central result is the existence of a critical absolute Mason number $M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$1. If $M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$2, every particle entering the array is captured irrespective of lateral entry position. If $M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$3, capture is only partial and depends on the inlet position. The particle that is hardest to capture enters on the centerline between two cylinders, and symmetry reduces that trajectory to a one-dimensional problem along the midline. On that line, the critical value is characterized by

$M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$4

where $M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$5 is the dimensionless axial fluid speed and $M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$6 the dimensionless magnetic-force component.

The geometry is controlled by $M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$7, the ratio of half-gap width to cylinder radius. The paper derives two asymptotic regimes. For closely packed cylinders, $M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$8, lubrication theory gives

$M_q(a,b,c)=\deg(\mathrm{rad}_q(abc))-1$9

For widely spaced cylinders, $(I_0(M),I_1(M),\dots,I_r(M))$0, the magnetic force decays as $(I_0(M),I_1(M),\dots,I_r(M))$1, and

$(I_0(M),I_1(M),\dots,I_r(M))$2

Numerical finite-element calculations connect these limits smoothly.

The same framework also quantifies single-pass efficiency. If $(I_0(M),I_1(M),\dots,I_r(M))$3 is the critical inlet position in a half-cell of width $(I_0(M),I_1(M),\dots,I_r(M))$4, then

$(I_0(M),I_1(M),\dots,I_r(M))$5

For $(I_0(M),I_1(M),\dots,I_r(M))$6, $(I_0(M),I_1(M),\dots,I_r(M))$7 and efficiency is $(I_0(M),I_1(M),\dots,I_r(M))$8. Above the threshold, efficiency decreases monotonically with $(I_0(M),I_1(M),\dots,I_r(M))$9. The paper further notes that operating at $\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$0 with multiple remixing cycles can reduce total separation time when full single-pass capture is unnecessary.

3. Magnetorheological fluids

In “On the Apparent Yield Stress in Non-Brownian Magnetorheological Fluids” (Vagberg et al., 2017), the Mason number is the central control parameter for dense two-dimensional MR suspensions. The model includes long-ranged magnetic dipole forces, finite-ranged elastic repulsion, and viscous damping. The magnetic force scale is defined from two just-touching small particles in the head-to-tail configuration of maximum magnetic attraction,

$\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$1

with $\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$2 the small-particle diameter, $\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$3 the fluid permeability, $\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$4 the applied field magnitude, and $\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$5. The viscous scale depends on the dissipation mechanism. For reservoir damping,

$\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$6

For contact damping,

$\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$7

so the Mason number depends explicitly on the mean contact number $\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$8.

This definition organizes the rheology. At high Mason number the dimensionless stress obeys

$\mathrm{Mn}=\pm \dfrac{6\pi\eta a R U}{\mu_0 m M}$9

which is the Newtonian regime. At low Mason number the behavior depends strongly on the dissipation law. For reservoir damping at $\mathrm{Mn}=\dfrac{F_d}{F_m}$0, the flow curves collapse over many decades in $\mathrm{Mn}=\dfrac{F_d}{F_m}$1, no low-$\mathrm{Mn}=\dfrac{F_d}{F_m}$2 plateau is visible down to $\mathrm{Mn}=\dfrac{F_d}{F_m}$3, and the paper finds

$\mathrm{Mn}=\dfrac{F_d}{F_m}$4

For contact damping at the same $\mathrm{Mn}=\dfrac{F_d}{F_m}$5, the low-$\mathrm{Mn}=\dfrac{F_d}{F_m}$6 regime crosses over to a clear yield-stress plateau,

$\mathrm{Mn}=\dfrac{F_d}{F_m}$7

corresponding effectively to $\mathrm{Mn}=\dfrac{F_d}{F_m}$8.

The Mason number also organizes the microstructure. In the reservoir-damped model, decreasing $\mathrm{Mn}=\dfrac{F_d}{F_m}$9 produces anisotropic chain-like clusters aligned with the field, and the size of the largest cluster collapses well when plotted against $a=\dfrac{\alpha\eta\ell^3}{mB}$0. In the contact-damped model, clustering already occurs at high $a=\dfrac{\alpha\eta\ell^3}{mB}$1 because inelastic collisions generate compact granular clusters before magnetism fully dominates; the largest-cluster data do not collapse when plotted against $a=\dfrac{\alpha\eta\ell^3}{mB}$2 alone. At sufficiently low $a=\dfrac{\alpha\eta\ell^3}{mB}$3, both models converge to nearly the same asymptotic contact number, with fits at $a=\dfrac{\alpha\eta\ell^3}{mB}$4 giving $a=\dfrac{\alpha\eta\ell^3}{mB}$5 and exponents $a=\dfrac{\alpha\eta\ell^3}{mB}$6 for reservoir damping and $a=\dfrac{\alpha\eta\ell^3}{mB}$7 for contact damping in

$a=\dfrac{\alpha\eta\ell^3}{mB}$8

The paper also relates Mason-number rheology to jamming. Increasing $a=\dfrac{\alpha\eta\ell^3}{mB}$9 toward $$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$0 is sufficient to induce a yield stress even in the reservoir-damped model. This leads to a qualified interpretation of “apparent yield stress”: at moderate $$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$1, the accessible low-$$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$2 range can show either a sub-Bingham power law or a plateau, depending on the dissipation mechanism, but the asymptotic $$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$3 limit is argued to be controlled by the magnetic-elastic energy landscape rather than by viscous details.

4. Magnetic micro-swimmers

In “Asymptotic Dynamics of Magnetic Micro-Swimmers” (Rüegg-Reymond et al., 2018), the Mason number is the nondimensional torque ratio governing a rigid permanent-magnetic body in a spatially uniform magnetic field rotating steadily in Stokes flow. The swimmer has magnetic moment $$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$4, the field has magnitude $$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$5 and angular speed $$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$6, and the defining parameter is

$$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$7

The paper states that this number “represents the balance between the drag and the magnetic load on the swimmer”; it is proportional to the angular velocity of the rotating field and inversely proportional to the field magnitude.

The orientation dynamics may be written in several equivalent forms. In the magnetic frame, with $$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$8, the autonomous equation is

$$\frac{\partial \hat{\boldsymbol{x}}_\mathrm{p}}{\partial \hat{t}} = \hat{\boldsymbol{u}} + \frac{1}{\mathrm{Mn}}\hat{\boldsymbol{F}}_\mathrm{m},$$9

where $1/\mathrm{Mn}$0 is the conical angle and $1/\mathrm{Mn}$1. The Mason number enters as the coefficient of the imposed rotation term.

Three asymptotic regimes are derived. In the low-Mason-number regime $1/\mathrm{Mn}$2, magnetic torque dominates and the magnetic moment tends to align with the instantaneous field. After rescaling time by $1/\mathrm{Mn}$3, the slow dynamics reduce to an ODE for an angular variable $1/\mathrm{Mn}$4,

$1/\mathrm{Mn}$5

with material angles $1/\mathrm{Mn}$6 and $1/\mathrm{Mn}$7. If

$1/\mathrm{Mn}$8

the system ունի stable and unstable relative equilibria; outside that interval it has a periodic solution. In the periodic case the slow-time period is

$1/\mathrm{Mn}$9

so the physical period scales like $\mathrm{Mn}=F_d/F_m$0.

In the high-Mason-number regime $\mathrm{Mn}=F_d/F_m$1, the field rotates too rapidly for the body to track its instantaneous direction. Averaging shows that the swimmer responds to the mean field

$\mathrm{Mn}=F_d/F_m$2

so the magnetic moment tends to align with the average field, parallel or antiparallel to the rotation axis according to $\mathrm{Mn}=F_d/F_m$3. A second-order calculation yields a slow residual rotation about this mean-field direction with period proportional to $\mathrm{Mn}=F_d/F_m$4.

The small-conical-angle regime $\mathrm{Mn}=F_d/F_m$5 bridges the low- and high-$\mathrm{Mn}=F_d/F_m$6 limits. In the magnetic frame, the magnetic moment moves on a small circle whose center shifts continuously from the instantaneous-field direction at low $\mathrm{Mn}=F_d/F_m$7 to the average-field direction at high $\mathrm{Mn}=F_d/F_m$8, while the radius tends to zero in both limits. This provides a uniform description of the crossover from synchronous field tracking to averaged-field alignment.

5. Algebraic Mason invariants in the Stothers–Mason theorem

In “$\mathrm{Mn}=F_d/F_m$9-difference analogue of the Stothers-Mason theorem” (Lu et al., 27 May 2026), the “Mason number” is not a dimensionless physical ratio but the radical-based degree invariant attached to a polynomial triple. For a nonzero polynomial $q$00, the radical $q$01 is the product of its distinct linear factors. If $q$02 are relatively prime and satisfy $q$03, the classical Stothers–Mason theorem states

$q$04

The quantity

$q$05

is described as the key number controlling degree versus radical, and in many contexts this is what people loosely call the Mason number of the triple $q$06.

The paper constructs a $q$07-difference analogue. Its basic notions are the Jackson derivative

$q$08

the $q$09-analogue of powers

$q$10

and the $q$11-weight of a zero, defined by the vanishing of $q$12 at a point. Theorem 2.1 shows that $q$13 has $q$14-weight $q$15 if and only if

$q$16

As $q$17, $q$18-weight reduces to ordinary multiplicity.

For a polynomial factorization

$q$19

the $q$20-difference radical is

$q$21

Its degree counts zeros by $q$22-chains while ignoring $q$23-weights. The $q$24-difference Stothers–Mason inequality then takes the form

$q$25

for relatively $q$26-prime polynomials $q$27 with $q$28 and $q$29. The corresponding $q$30-analogue of the Mason number is

$q$31

which reduces to $q$32 as $q$33.

The paper applies this invariant to $q$34-difference Fermat-type equations. For

$q$35

with the three terms relatively $q$36-prime, Theorem 5.1 gives $q$37, and if one of $q$38 is constant then $q$39. For the multi-term equation

$q$40

the bound

$q$41

is obtained, hence in particular $q$42. These results show that the Mason invariant in this setting bounds the complexity of polynomial solutions through the degree of a radical-like object adapted to the Jackson calculus.

6. Mason–Welsh sequences and related combinatorial terminology

In “Log-concavity for independent sets of valuated matroids” (Giansiracusa et al., 2024), the relevant object is the Mason–Welsh sequence rather than a single scalar invariant. For a matroid $q$43 of rank $q$44 on a finite ground set $q$45, the numbers

$q$46

form the sequence

$q$47

which the paper describes as the “Mason sequence,” sometimes identified with the Mason number sequence. The classical Mason–Welsh conjecture asserts log-concavity: $q$48 The ultra log-concave normalization is

$q$49

The paper generalizes this to valuated matroids. If $q$50 is a valuated matroid and

$q$51

then for $q$52

$q$53

The main theorem proves

$q$54

A parallel result holds for valuated discrete polymatroids: $q$55 For valuated bimatroids, the weighted counts

$q$56

are shown to be ultra log-concave.

The proof strategy uses Lorentzian polynomials and generic extensions that convert independent-set data into basis data on a larger ground set. The paper is explicit that it does not introduce a separate symbol for the “Mason number”; instead it works directly with the sequence $q$57 or its weighted analogues. A common source of confusion is therefore terminological: in this combinatorial literature, “Mason” usually names a sequence and its log-concavity properties, whereas in the magnetic-fluid and micro-swimmer literatures the Mason number is a single dimensionless ratio, and in the Stothers–Mason literature it is a single radical-based degree bound.