Modeling and Sensing Shear Dependent Fluid Dynamics

17 concepts14 papers1 proposal

When we think of fluid flow, we often default to the behavior of water. Water is a Newtonian fluid: its viscosity remains constant regardless of the forces acting upon it. If you stir it twice as fast, it resists twice as hard, maintaining a linear relationship between stress and strain.

However, many fluids in nature and industry—such as blood, wastewater, pancake batter, and culinary sauces—defy this simplicity. These are non-Newtonian fluids, and their defining characteristic is a shear-dependent or stress-dependent viscosity. When subjected to mechanical force, their internal structures rearrange, causing the fluid to thin out or thicken dynamically. Modeling, simulating, and perceiving these fluids requires moving past the assumption of constant viscosity, a challenge currently being addressed across fluid mechanics, computer science, and robotics.

Physical Manifestations: Blunted Flows and Deformed Bubbles

To understand why classical fluid dynamics models fail, we can look at how shear-dependent viscosity alters basic flow behavior. In a standard Newtonian fluid flowing through a pipe, the velocity profile forms a perfect parabola—fastest in the center and zero at the walls.

But in micro-scale biological environments, this assumption breaks down. For instance, Wi-Han Kim et al. demonstrated in Particle Image Velocimetry of the Blood Flow in a Micro-channel Using the Confocal Laser Scanning Microscope that blood flow exhibits a distinctively "blunt" velocity profile rather than a parabolic one. Because blood viscosity changes in response to varying shear rates across the vessel diameter, the fluid layers slide past one another non-linearly. This effect is closely tied to cell distribution; as shown by Soo-Hee Lim et al. in Flow Visualization of Blood Cell and Detection of Cell Depleted Layer Using a Confocal Laser Scanning Microscope, the interaction between red blood cells and the vessel walls creates a cell-depleted layer, driving complex shear-thinning behaviors that dictate micro-scale hemodynamics.

Newtonian Flow (Parabolic)         Non-Newtonian Flow (Blunted)
      =======>                           =========>
    ===========>                         =========>
  =============>                         =========>
    ===========>                         =========>
      =======>                           =========>

This variable viscosity also complicates multiphase interactions, such as how gases disperse within liquids. In industrial wastewater treatment, the non-Newtonian rheology of concentrated waste significantly alters how gas bubbles are retained and how flow patterns develop, as detailed by Daolin Du et al. in The Effect of the Spargers Design on the Wastewater Treatment of Gas-Liquid Dispersion Process in a Stirred Tank.

In a comprehensive review, Yiyi Ma et al. explained in Bubble mass transfer in fluids under gravity: a review of theoretical models and intensification technologies in industry that shear-rate-dependent viscosity directly alters bubble rise velocity, drag forces, and shape deformation. Because classical mass transfer models assume a constant viscosity, they exhibit poor accuracy when applied to these complex fluids, making real-world industrial scaling highly unpredictable.

The Computational Frontier: Simulating Variable Rheology

To simulate these materials accurately, computational frameworks must explicitly track the fluid's internal stress state over time. Many foundational mathematical models are strictly limited to the Newtonian regime. For example, the unique inverse-kinetic theory developed by Massimo Tessarotto et al. in Unique representation of an inverse-kinetic theory for incompressible Newtonian fluids was explicitly restricted to incompressible Newtonian fluids. Similarly, advanced interface models, such as the coupled Cahn-Hilliard Navier-Stokes system preconditioned by Jessica Bosch et al. in Preconditioning of a Coupled Cahn-Hilliard Navier-Stokes System, or the nonequilibrium two-phase model proposed by Ilya Peshkov et al. in Nonequilibrium model for compressible two-phase two-pressure flows with surface tension, require significant architectural extensions to handle non-Newtonian dynamics.

When simulators do attempt to model these complex materials, they must capture the "conformation tensor"—a mathematical representation of the fluid's internal polymeric or cellular alignment. Young Ju Lee et al. addressed this in A stable and scalable hybrid solver for rate-type non-Newtonian fluid models by combining higher-order flow approximations with a refined mesh specifically designed to track this conformation tensor.

Without these specialized solvers, engineering designs optimized in idealized environments fail in the real world. As Qiang Gu et al. demonstrated in Optimization of oxygen transfer and power consumption in aerobic bioprocess by designing disc turbine impeller based on CFD-Taguchi method, bioreactors optimized using water perform unreliably when filled with thick fungal fermentation broths, because the variable viscosity under shear fundamentally alters oxygen transfer and power consumption dynamics.

The Perception Challenge: Real-Time Robotic Interaction

If simulating non-Newtonian fluids in a computer is difficult, perceiving and manipulating them with physical robots in real time is even harder. When a robot interacts with a liquid, it often uses simplified physics models to infer fluid properties. For instance, a robot might shake a container and use a damped-oscillations model to estimate viscosity.

However, as noted by Hung-Jui Huang et al. in Understanding Dynamic Tactile Sensing for Liquid Property Estimation, these simplified models fail for non-Newtonian fluids because the viscosity changes during the oscillation as the stress varies. To bridge this gap, researchers have turned to computer vision and machine learning. In Fluid viscosity prediction leveraging computer vision and robot interaction, J. Park et al. utilized a 3D convolutional autoencoder to predict viscosity from video profiles, though they highlighted that adapting such models to non-Newtonian fluids remains a major hurdle due to the force-dependent nature of the material.

[Robot Action: Squeeze/Stir] ──> [Variable Stress Applied]
                                          │
                                          ▼
[Static Model Fails] <── [Viscosity Dynamically Changes]

In high-speed biophysics, characterizing these fast-changing properties is equally critical. Matthew G. Smith et al. demonstrated in Machine learning opens a doorway for microrheology with optical tweezers in living systems that combining deep learning with optical trapping can dramatically accelerate microrheology, yet capturing materials whose viscosity changes under stress remains an open frontier.

When robots must actively dispense or manipulate these fluids—such as in automated cooking—they must account for non-linear flow rates. In Kitchen Artist: Precise Control of Liquid Dispensing for Gourmet Plating, Hung-Jui Huang et al. built a system that performs haptic exploration to sense a sauce's properties, explicitly noting that ignoring shear-thinning behaviors under varying squeeze pressures leads to severe dispensing inaccuracies.

Similarly, Xinyuan Luo et al. tackled this in An Intelligent Robotic System for Perceptive Pancake Batter Stirring and Precise Pouring. Because pancake batter's viscosity changes based on ingredient ratios and the mechanical action of stirring, the authors used real-time force-torque data to dynamically adjust pouring speeds. By closing the loop between physical interaction and rheological change, such systems demonstrate that mastering non-Newtonian fluids requires treating viscosity not as a fixed constant, but as a dynamic variable.

Key Foundational Work

Particle Image Velocimetry of the Blood Flow in a Micro-channel Using the Confocal Laser Scanning Microscope (2010) — contributed 2 concept elements to this direction.
Unique representation of an inverse-kinetic theory for incompressible Newtonian fluids (2006) — contributed 1 concept elements to this direction.
Flow Visualization of Blood Cell and Detection of Cell Depleted Layer Using a Confocal Laser Scanning Microscope (2010) — contributed 1 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_non_newtonian_fluid_modeling_1e032446,
  title = {non-Newtonian fluid modeling},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/non-newtonian-fluid-modeling/rdr_1e0324461e2f4b8da9169973f8225d3d},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-22},
}

This exact version (version 1, frozen text):

@misc{lacuna_non_newtonian_fluid_modeling_1e032446_v1,
  title = {non-Newtonian fluid modeling},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/non-newtonian-fluid-modeling/rdr_1e0324461e2f4b8da9169973f8225d3d/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-22},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake