A mixed experimental-CFD-data science approach for rheological measurement of polishing fluids

• Published in: Mechanics of advanced materials and structures Vol. 27; no. 13; pp. 1167 - 1177
• Main Authors: Thomas, Naveen, Srinivasa, Arun R., Bukkapatnam, Satish T. S.
• Format: Journal Article
• Language: English
• Published: Abingdon Taylor & Francis 01.07.2020; Taylor & Francis Ltd
• Subjects: AIC; Computational fluid dynamics; Data reduction; Data science; Exact solutions; Experimental methods; material modeling; Mathematical models; Mixed CFD-experimental method; Parameter identification; particulate suspension; Polishing; rheological characterization; Rheological properties; Rheology; Slurries; Statistical inference; Wall effects; yield stress
• ISSN: 1537-6494; 1537-6532
• DOI: 10.1080/15376494.2020.1745967

We discuss a novel means for obtaining rheological properties of polishing slurries by combining statistical inference techniques (the Akaike Information Criterion), CFD and Torque vs Speed Data. The data was obtained by using a custom built torsional rheometer that subjects the polishing fluid to conditions that are similar to polishing. Our comparison indicates that side wall and inertial effects significantly affect the values of the parameters of any given model even under nominally slow rates of rotation. When these are considered, the Herschel-Bulkley model seems to be a significantly better fit compared to two other popular other models for the slurry. The results suggest that a systematic combination of computational-statistical-data science approach is necessary for identifying model parameters even for a slow flows as compared to currently used data reduction methods based on analytical solutions for torsional flow that ignore inertial and side-wall effects.