Predicting the distribution of yield-stress fluids in branched pipe manifolds Journal: Chemical Engineering Science

Elliott Sutton, Waldo Rosales Trujillo, Adam Kowalski, Cláudio P. Fonte, Anne Juel · 2025

Read Predicting the distribution of yield-stress fluids in branched pipe manifolds Journal: Chemical Engineering Science by Elliott Sutton, Waldo Rosales Trujillo, Adam Kowalski, Cláudio P. Fonte, Anne Juel (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-2509-22192.txt.

Introduction

Elliott Sutton and four co-authors predict how thick industrial fluids split among the branches of a pipe manifold. The fluids are yield-stress materials: they sit like a solid until pushed hard enough, then flow like a paste. Manifolds that distribute them, in food, chemical, and construction plants, often deliver uneven shares to each outlet. Wall slip, where the fluid slides along the pipe wall instead of gripping it, complicates the pattern.

The team built a one-dimensional network model that accounts for friction losses between junctions. Slip enters through a power-law relation measured separately with a capillary instrument. The model's predictions matched both bench experiments and full three-dimensional flow simulations across many conditions. Slip proved to strongly shape how uniform the distribution is, by changing the relative resistance of the branches.

Catalog record

CREATORElliott Sutton, Waldo Rosales Trujillo, Adam Kowalski, Cláudio P. Fonte, Anne Juel
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE50 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-22192

Browse the full library · About the Tower of Babel

Full text

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

Predicting the distribution of yield-stress fluids in branched pipe manifolds Journal: Chemical Engineering Science Elliott Sutton, Waldo Rosales Trujillo, Adam Kowalski, Cláudio P. Fonte, Anne Juel Abstract We develop a one-dimensional network model to predict the steady-state distribution of yield-stress fluids in branched pipe manifolds under wall-slip conditions. The model accounts for major friction losses between junctions and incorporates wall slip through a power-law relation calibrated independently via capillary rheometry. Predictions from the model are validated against both bench-scale experiments and fully resolved computational fluid dynamics simulations, showing excellent agreement across a range of flow conditions. Our results demonstrate that wall slip strongly influences the uniformity of fluid distribution by modifying the relative resistance between outlet branches. Furthermore, we show that the problem can be inverted: measured distribution profiles can be used to estimate slip parameters, offering a practical method for slip characterisation without pressure measurement. This modelling framework is computationally inexpensive, robust, and adaptable to various network configurations, making it a valuable tool for the design and analysis of industrial manifold systems involving viscoplastic fluids. † † highlights: A reduced-order network model accurately predicts yield-stress fluid distribution in pipe manifolds. Wall slip is incorporated via a power-law model calibrated using capillary rheometry. Model predictions show excellent agreement with both experiments and 3D flow simulations. Wall slip enhances distribution uniformity by reducing resistance disparities between branches. The manifold system can be used to extract slip parameters without direct pressure measurements. 1 Introduction In many industrial filling operations, fluids are delivered from the process to a series of filling nozzles via a pipe manifold located at the end of the manufacturing line. The challenge for designers of such processes is to devise a manifold configuration that can achieve the desired distribution (typically uniform) amongst many outlets. Maldistribution arises from dissimilar resistances to flow in the outlet branches ( Miller 1978 ; Yazici et al. 2024 ) , as well as dynamical effects that arise due to inertial ( Chen and Sparrow 2009 ) , thermal ( Baikin et al. 2011 ; Aka and Narayan 2022 ) , or elastic instabilities ( Varshney et al. 2016 ) in different sections of the pipeline such as bends and bifurcations. Previous studies have focused on Newtonian fluid flow through manifolds and have successfully applied network model methodologies to predict fluid distribution ( Miller 1978 ; Majumdar 1980 ; Kee et al. 2002 ) . The distribution profile in the manifold outlets depends on the resistance to flow throughout the different paths of the network, determined by the total pressure drop across each branch. The Hagen-Poiseuille equation provides an analytical description of major friction losses in straight pipe segments of a manifold, and loss factors describe minor friction losses in junctions and bends; minor losses are empirically derived factors specific to particular flow configurations. Additional loss mechanisms cause minor losses in Newtonian laminar flows and arise due to inertia, which forms secondary flow structures in bends ( Dean 1928 ) and bifurcations ( Chen et al. 2017 ) . By contrast with the vast literature on Newtonian fluids, the flow of non-Newtonian fluids through pipe manifolds is sparse. There have been some attempts to use network models to describe distribution systems, such as the work by Brod 2003 for polymer melts. They aimed to minimise residence times by selecting optimal tube diameters for a given system design and imposed pressure gradient. They report that ideal designs maintain a uniform apparent wall shear rate throughout the network. Network models can also describe the flow through porous media, which distribute fluid through a network of pores. Fraggedakis et al. 2021 could accurately predict the first fluidised path of viscoplastic fluid through a porous medium utilising a Dijkstra algorithm, avoiding computationally expensive flow simulations. Similarly, shear-thinning ( Castro and Goyeau 2021 ) and viscoplastic ( Balhoff and Thompson 2004 ; Balhoff et al. 2012 ; Liu et al. 2019 ) fluid flows through porous media were described by network models, reporting modifications to Darcy’s law. These models are akin to the network model of Miller 1978 , calculating flow resistance to predict preferential routes and flux through a medium. To the best of our knowledge, no studies have directly investigated yield-stress fluid flow through pipe manifolds. Yield-stress fluids are prone to slipping on solid surfaces. Wall slip, when present, drastically changes the behaviour of yield-stress fluid flow by reducing friction between the fluid and a solid boundary and permits displacement from imposed stresses insufficient to induce yielding ( Kamdi et al. 2021 ) . As such, a network model based on frictional resistance should account for this phenomenon, though previous models have typically neglected it. Many yield-stress fluids exhibit apparent wall slip due to the formation of a thin lubrication layer near the boundary, typically with a thickness on the order of 10 nm ( Seth et al. 2008 ; Zhang et al. 2017 ) . This layer, composed of the continuous phase or solvent, has a significantly lower viscosity than the bulk material. Consequently, the bulk material appears to slip at the wall, although this effect arises from flow within the low-viscosity lubrication layer rather than actual motion at the solid boundary. In soft jammed suspensions, Péméja et al. 2019 identified two distinct slip regimes, depending on the magnitude of the wall shear stress: a linear regime , in which the slip velocity is directly proportional to the applied shear stress, and a nonlinear regime , in which the slip velocity scales quadratically with stress. In the linear regime, dissipation is described by a Couette flow within the continuous-phase lubrication layer, while the particulate bulk behaves as a rigid solid sliding over it. The formation of this lubrication layer can be attributed to repulsive/attractive forces between the occlusions of the material and solid walls, and to the inability of the material to retain the bulk jammed structure in the immediate vicinity of the wall due to geometric confinement ( Zhang et al. 2017 ) . The elastohydrodynamic lubrication model proposed by Meeker et al. 2004b ; Meeker et al. 2004a accounts for the nonlinear regime, wherein hydrodynamic lift depletes further the near-wall region of suspended particles, setting the effective thickness of the lubrication layer. This thickness is governed by a balance between the hydrodynamic lift force, the osmotic pressure of the bulk suspension, and the elastic deformation of the particles. Both slip mechanisms become more prominent at low stresses, below the yield stress. Above the yield stress, apparent slip persists but gradually become negligible as the material’s microstructure is disrupted under increasing deformation. This transition has been observed in both steady-shear ( Buscall et al. 1993 ) and oscillatory-shear ( Walls et al. 2003 ) measurements. The lack of literature characterising the flow of yield-stress materials through pipe manifolds and the inability of process engineers to accurately predict the fluid distribution from a manifold design is a significant hindrance to industry. Furthermore, computational fluid dynamics simulations (CFD) of yield-stress fluid flow in such large geometries converge notoriously slowly ( Saramito and Wachs 2017 ) and are prohibitively expensive. Thus, in this study, we develop a reduced-order network model capable of modelling the flow with a fraction of

Read the complete text