Geometry Can Qualitatively Change
Rheological Effects on Propulsion
Yi Man (满怡)
Peking University
2026 IUTAM Symposium on
“Diversity-Driven Innovations in Complex Fluids for Advanced Manufacturing”
August 17, 2026 · Beijing, China
Active particles

Trevenen & Beltramo, J. Colloid Interface Sci., 2021
Convert energy into autonomous motion
Complex-fluid environments

Chrétien, Acta Obstet. Gynecol. Scand., 2003
Often exhibit non-Newtonian rheology
How does fluid rheology alter active propulsion?
Two distinct surface regions can enable self-propelled motion.
Chemical reaction
creates an uneven solute concentration
Surface concentration gradient
generates a thin surface flow (slip)
Asymmetric slip
produces net propulsion
Self-diffusiophoresis: asymmetric activity → solute gradient → surface slip → propulsion
What happens when these particles move in complex fluids?
Self-diffusiophoresis: asymmetric activity → solute gradient → surface slip → propulsion

Enhanced rotational fluctuations
Gomez-Solano et al., PRL, 2016

Reduced propulsion
Saad & Natale, Soft Matter, 2019

Suppressed active motion
Raman et al., ACS Phys. Chem. Au, 2023
Experiments reveal diverse responses, but the direct rheological effect on propulsion remains unclear.
Shear thinning

Méndez-Mora et al., Micromachines 12, 726 (2021)
Higher shear rate → lower viscosity
Viscoelasticity

Chrétien, Acta Obstet. Gynecol. Scand., 2003
Deformation → relaxation
These two rheological effects can coexist in the same complex fluid.
Shear thinning
Carreau model

Sphere: propulsion is reduced
Viscoelasticity
Second-order fluid

Half-coated sphere: leading-order correction vanishes
Datt et al., J. Fluid Mech., 2017
Sphere — correction vanishes at half-coating
?
Spheroid — geometric anisotropy
We vary two control parameters: eccentricity \(e\) (shape) and coverage \(\zeta_0\) (activity).
Surface flux \(\to\) Concentration field
\[ \frac{\partial C}{\partial t} + \boldsymbol{u}\cdot\nabla C = D\nabla^2 C, \quad \left. D\boldsymbol{n}\cdot\nabla C \right|_{\tau=\tau_0,\zeta<\zeta_0} = -A \]
\(C\): solute concentration; \(\boldsymbol{u}\): fluid velocity; \(D\): solute diffusivity; \(A\): surface activity / fixed solute flux on the active cap; \(\boldsymbol{n}\): unit normal.
Tangential gradient \(\to\) Phoretic slip
\[ \boldsymbol{u}_s = M(\boldsymbol{I}-\boldsymbol{n}\boldsymbol{n})\cdot \nabla C \]
\(M\): phoretic mobility; \((\boldsymbol{I}-\boldsymbol{n}\boldsymbol{n})\): tangential projection operator.
Slip velocity \(\to\) Force-free propulsion
\[ \boldsymbol{u}(\tau=\tau_0) = \boldsymbol{u}_s(\zeta)+\boldsymbol{U}, \]
\[ \int_S \boldsymbol{n}\cdot\boldsymbol{\sigma}\,\mathrm{d}S = \boldsymbol{0} \]

biological propulsion analogy
Stokes flow:
\[ \nabla\cdot\boldsymbol{\sigma}=0, \qquad \nabla\cdot\mathbf{u}=0, \qquad \boldsymbol{\sigma}=-p\mathbf{I}+\mathbf{T} \]
\[ \dot{\boldsymbol{\gamma}} = \nabla\mathbf{u}+(\nabla\mathbf{u})^T \]
Shear thinning: Carreau model
\[ \mathbf{T} = \left[ \beta+(1-\beta) \left( 1+\frac{Cu^2}{2} \dot{\boldsymbol{\gamma}}:\dot{\boldsymbol{\gamma}} \right)^{(n-1)/2} \right] \dot{\boldsymbol{\gamma}}. \]
\(Cu\) — flow shear-rate / characteristic shear-rate ratio
\(\beta\) — high-/low-shear viscosity ratio
\(n\) — shear-thinning index
Viscoelasticity: Second-order fluid
\[ \mathbf{T} = \dot{\boldsymbol{\gamma}} -De\left( \mathbf{G} +b\,\dot{\boldsymbol{\gamma}}\cdot\dot{\boldsymbol{\gamma}} \right). \]
\(De\) — elastic / viscous stress ratio
\(b\) — normal-stress ratio
\(\mathbf{G}\) — upper-convected derivative
Diffusion-dominated limit: \(Pe=\dfrac{MAa}{D^2}\) — advection / diffusion
\[ Pe\ll1 \;\Rightarrow\; \nabla^2 c=0, \qquad \left.\boldsymbol{n}\cdot\nabla c\right|_{\tau=\tau_0} = \begin{cases} -1, & \zeta\le \zeta_0 \quad \text{active},\\ 0, & \zeta>\zeta_0 \quad \text{inert}, \end{cases} \qquad c|_{\tau\to\infty}=0 \]
\(c\): dimensionless relative solute concentration; \(\tau=\tau_0\): particle surface.
Series solution and induced slip
\[ c(\tau,\zeta) = \sum_{n=0}^{\infty} \rho_n Q_n(\tau)P_n(\zeta) \]
\[ \implies u_s(\zeta) = \tau_0\sum_{n=1}^{\infty} B_n\frac{P_n^1(\zeta)}{\sqrt{\tau_0^2-\zeta^2}} \]
\(P_n,Q_n\): Legendre functions; \(\rho_n\): coefficients set by the flux boundary condition; \(B_n=-\rho_n Q_n(\tau_0)\).
Note: finite-\(\mathrm{Pe}\) coupling
Finite \(Pe\): advection couples the solute and flow fields, allowing rheology to modify the concentration field and hence the phoretic slip.
Weakly non-Newtonian: expand in \(\epsilon\ll 1\)
Shear thinning (Carreau): \(\epsilon = 1-\beta\)
Viscoelasticity (second-order fluid): \(\epsilon = \mathrm{De}\)
\[ \{\mathbf u,p,U\} = \{\mathbf u_0,p_0,U_0\} + \epsilon\{\mathbf u_1,p_1,U_1\} + O(\epsilon^2) \]
The zeroth order is the Newtonian problem, valid for any eccentricity, with a known speed:
\[ U_0 = -\frac{\tau_0}{2} \int_{-1}^{1} u_s \frac{\sqrt{1-\zeta^2}}{\sqrt{\tau_0^2-\zeta^2}} \,\mathrm{d}\zeta \]
Poehnl et al., J. Phys.: Condens. Matter, 2020; Popescu et al., Eur. Phys. J. E, 2010

Linearity makes the Newtonian speed symmetric about half-coating: \(U_0(\zeta_0)=U_0(-\zeta_0)\).
Non-Newtonian rheology enters as a forcing set by the Newtonian flow:
\[ \boldsymbol{\sigma}_1 = -p_1\mathbf I + \dot{\boldsymbol{\gamma}}_1 + \mathbf A \]
Shear thinning (Carreau)
\[ \color{#1f5f8f}{\mathbf A_{\mathrm{ST}}} = \left[ -1+ \left( 1+\frac{Cu^2}{2} \dot{\boldsymbol{\gamma}}_0:\dot{\boldsymbol{\gamma}}_0 \right)^{(n-1)/2} \right] \dot{\boldsymbol{\gamma}}_0 \]
Viscoelasticity (second-order)
\[ \color{#a95724}{\mathbf A_{\mathrm{VE}}} = -\left( \mathbf G_0 + b\,\dot{\boldsymbol{\gamma}}_0\cdot \dot{\boldsymbol{\gamma}}_0 \right) \]
An auxiliary towing problem + the reciprocal theorem give \(U_1\) directly, without solving the first-order flow:
\[ U_1 = -\frac{(\tau_0^2+1)\operatorname{coth}^{-1}\tau_0-\tau_0}{4\tau_0^2} \int_{\tau_0}^{\infty}\int_{-1}^{1} (\boldsymbol{A}:\nabla\hat{\boldsymbol{u}})(\tau^2-\zeta^2)\,\mathrm{d}\zeta\,\mathrm{d}\tau \]
\(\nabla\hat{\boldsymbol{u}}\): velocity gradient of the auxiliary (towing) flow.
Half-coated particles: \(\zeta_0=0\)

G. Zhu, B. vanGogh, L. Zhu, O.S. Pak and Y. Man. JFM., 2024.
Sphere

Complementary coatings → same speed
Spheroid

Complementary coatings → different speeds
Geometry breaks the complementary-coating symmetry.

Symmetry breaking requires both geometric anisotropy and nonlinear rheology.
Half-coated particles: \(\zeta_0=0\)
Q. Zhao, G. Zhu and Y. Man. (submitted).
Propulsion vs Coverage
Enhancement vs Reduction
Eccentricity breaks the coverage antisymmetry, and the enhancement regime broadens with elongation.
Sphere — 𝒦 antisymmetric, cancels exactly ⟹ \(U_1=0\)
Spheroid — curvature localizes 𝒦, antisymmetry broken ⟹ \(U_1\neq0\)
Geometry can qualitatively change non-Newtonian propulsion.
Acknowledgements
Qiang Zhao (赵强)
Peking University
Guangpu Zhu (朱光普)
Nanjing University of Aeronautics and Astronautics
Prof. On Shun Pak
Santa Clara University
