A Model of Synovial Fluid with a Hyaluronic Acid Source: A Numerical Challenge

Initially motivated by the analysis of the flow dynamics of the synovial fluid, taken as non-Newtonian, this paper also reports on a numerical challenge which occurred unexpectedly while solving the momentum equation of the model. The configuration consists of two infinitely long horizontal parallel...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Fluids (Basel) 2021-04, Vol.6 (4), p.152
Hauptverfasser: Ozan, S. Canberk, Labrosse, Gérard, Uguz, A. Kerem
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Initially motivated by the analysis of the flow dynamics of the synovial fluid, taken as non-Newtonian, this paper also reports on a numerical challenge which occurred unexpectedly while solving the momentum equation of the model. The configuration consists of two infinitely long horizontal parallel flat plates where the top plate is sheared at constant speed and the bottom plate is fixed. The synovial fluid shows a shear-thinning rheology, and furthermore it thickens with the hyaluronic acid (HA) concentration, i.e., it is also chemically-thickening. Accordingly, a modified Cross model is employed to express the shear rate and concentration-dependent viscosity, whose parameter values are determined from experimental data. Another significance of the study is the investigation of the effect of an external stimulus on the flow dynamics via a HA source term. The resulting flow exhibits peculiar features resulting from extremely large and small, but positive, numerical quantities, such as the viscosity and the shear rates. This requires constructing a parametrized zero-machine level solver, up to 300 accurate digits or so, for capturing the correct length scales of the flow physics. As a conclusion, the physical model, although simple, but original, leads to interesting results whose numerical determination turns out to be successful only once the real cause of the numerical trap is identified.
ISSN:2311-5521
2311-5521
DOI:10.3390/fluids6040152