A quantitative evaluation of viscosity regularization in predicting transient flows of viscoplastic fluids

Ahmadi, Amir Hossein; Karimfazli, Ida

doi:10.1016/j.jnnfm.2020.104429

Cited by 7 publications

(11 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is the ratio of the maximum apparent viscosity η max and plastic viscosity µ because it defines the order of magnitude and therefore the dominance of the numerical matrices in which the apparent viscosity η is used. This approach of limiting the ratio to 1000 (or an equivalent regularisation parameter) is in line with the philosophy by Schaer et al [21], Bercovier and Engelman [22], Ahmadi and Karimfazli [29], Papanastasiou [47]. Moreover, this is exactly the same approach and recommendations as Beverly and Tanner [24], Bullough et al [46] and Burgos et al [51].…”

Section: Numerical Regularisationsupporting

confidence: 77%

“…The general rules of conduct for appropriate regularisation were not explored in this work. As for most works [ 21 , 25 , 28 , 29 ], the regularisation itself is ad hoc and applied for a specific application or problem [ 25 ]. Restrained to the context of this work, some general rules of thumbs may be defined, for which less accurate numerical results can be expected.…”

Section: Discussionmentioning

confidence: 99%

“…This means that, when the yield stress contribution becomes more than

(

) of the pressure loss or flow regime, less accurate simulations are expected. Numerical inaccuracy or even instability occur for quiescent flow conditions (i.e., low or zero shear zones) [ 25 , 28 , 29 ]. When the unyielded region becomes relatively large, less accurate results are obtained due to the (un)regularised viscoplastic problem [ 25 , 28 , 29 ].…”

Section: Reliability Irrespective Of Rheological Inputmentioning

confidence: 99%

“…Numerical inaccuracy or even instability occur for quiescent flow conditions (i.e., low or zero shear zones) [ 25 , 28 , 29 ]. When the unyielded region becomes relatively large, less accurate results are obtained due to the (un)regularised viscoplastic problem [ 25 , 28 , 29 ]. A pressure loss number of below

(

) means a plug radius of at least

of the pipe radius (i.e.,

) in accordance with Figure 14 or

in terms of plug area (i.e.,

).…”

Section: Reliability Irrespective Of Rheological Inputmentioning

confidence: 99%

“…Irrespective of the rheological input, some works assessing the numerical accuracy for Bingham flows exist [ 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 ]. However, these studies are in terms of error in numerical residuals or numerical stability criteria with regard to numerical regularisation techniques.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Numerical Reliability Study Based on Rheological Input for Bingham Paste Pumping Using a Finite Volume Approach in OpenFOAM

et al. 2021

View full text Add to dashboard Cite

Rheological quantification is important in many industries, the concrete industry in particular, e.g., pumping, form filling, etc. Instead of performing expensive and time-consuming experiments, numerical simulations are a powerful means in view of rheological assessment. However, due to the unclear numerical reliability and the uncertainty of rheological input data, it is important for the construction industry to assess the numerical outcome. To reduce the numerical domain of cementitious suspensions, we assessed the numerical finite volume simulations of Bingham paste pumping flows in OpenFOAM. We analysed the numerical reliability, first, irrespective of its rheological input by comparison with the literature and theory, and second, dependent on a certain rheological quantification by comparison with pumping experiments. Irrespective of the rheological input, the numerical results were significantly accurate. Dependent on the rheological input, a numerical mismatch, however, existed. Errors below 1% can be expected for proposed numerical rules of thumb: a bi-viscous regularisation, with pressure numbers higher than 5/4. To improve bias due to uncertain rheology, a rheological configuration close to the engineer’s aimed application should be used. However, important phenomena should not be overlooked. Further assessment for lubrication flows, in, e.g., concrete pumping, is still necessary to address concerns of reliability and stability.

show abstract

Section: Numerical Regularisationsupporting

confidence: 77%

Section: Discussionmentioning

confidence: 99%

“…This means that, when the yield stress contribution becomes more than

(

Section: Reliability Irrespective Of Rheological Inputmentioning

confidence: 99%

(

) means a plug radius of at least

of the pipe radius (i.e.,

) in accordance with Figure 14 or

in terms of plug area (i.e.,

).…”

Section: Reliability Irrespective Of Rheological Inputmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Numerical Reliability Study Based on Rheological Input for Bingham Paste Pumping Using a Finite Volume Approach in OpenFOAM

et al. 2021

View full text Add to dashboard Cite

show abstract

Representative velocity scale of Rayleigh‐Bénard convection with shear‐thinning fluids

Masuda,

Iyota,

Ohta

2023

Can J Chem Eng

View full text Add to dashboard Cite

Although natural convection is frequently encountered in various chemical processes, Rayleigh number (Ra) cannot be defined fully in shear‐thinning fluid systems. In particular, the velocity scale, which is necessary to estimate the effective viscosity of the system, should be discussed carefully. Thus, in this study, the representative velocity scale of Rayleigh‐Bénard (RB) convection, which is a typical example of natural convection, with shear‐thinning fluids was investigated based on the velocity fields obtained using computational fluid dynamics. Numerical simulations revealed that the critical temperature difference at which RB convection starts to fully develop decreases with an increase in the shear‐thinning property. The shear‐thinning property also induced subcritical bifurcation. In addition, the velocity scale of convection increases with an increase in the shear‐thinning property. Thus, the shear‐thinning property is considered to accelerate convection. Compared with several types of velocity scales used by other researchers, significant deviations from the actual scale were observed. Therefore, a new type of velocity scale, including the buoyant to viscous force ratio, arbitrary parameter, and thermal diffusivity, was proposed. The proposed velocity scale allowed an approximate estimation of the actual velocity scale. Although further investigation of the validity is necessary with varying geometries and rheological parameters, this velocity scale will be useful for controlling RB convection with Newtonian/shear‐thinning fluids.

show abstract

Development and Validation of an Open-Source Finite-Volume Method Solver for Viscoplastic Flows

Mirkov

Ouyahia

Lahlou

et al. 2021

Current Problems in Experimental and Computational Engineering

View full text Add to dashboard Cite

In the present paper, we discuss implementation details of a free and opensource numerical solver based on the finite volume method for numerical simulation of viscoplastic non-Newtonian fluids. In addition to the fact that they are involved in many industrial applications, both their physical properties and their rheological behavior make them challenging for numerical simulation. Viscoplastic fluids are known to behave as solid unless the shear stress reaches a critical level, known as yield-stress, beyond which they behave as liquid. In most cases, both yielded and unyielded regions coexist in the fluid domain. In mathematical model of viscoplastic fluid, the constitutive equation is a non-differentiable function. This is often overcome by using the approximate constitutive equation that has a regularized form, e.g. the Papanastasiou regularization model. Using the same approach, we assess the influence of regularization parameters on simulation convergence and results accuracy. In this study, we give implementation details of viscoplastic fluid models in freeCappuccino open-source Computational Fluid Dynamics code. Moreover, we perform validation on several well known benchmark cases and compare proposed approach with those existing in published literature. We also perform a parametric analysis and show the effect of Reynolds and Bingham numbers on the extent of the yielded regions. Conclusions of the study have relevance in practical application of computational fluid dynamics to viscoplastic fluids in particular and to non-Newtonian fluids in general.

show abstract

A quantitative evaluation of viscosity regularization in predicting transient flows of viscoplastic fluids

Cited by 7 publications

References 19 publications

Numerical Reliability Study Based on Rheological Input for Bingham Paste Pumping Using a Finite Volume Approach in OpenFOAM

Numerical Reliability Study Based on Rheological Input for Bingham Paste Pumping Using a Finite Volume Approach in OpenFOAM

Representative velocity scale of Rayleigh‐Bénard convection with shear‐thinning fluids

Development and Validation of an Open-Source Finite-Volume Method Solver for Viscoplastic Flows

Contact Info

Product

Resources

About