Dimensionless correlations for heat transfer coefficients during reciprocating agitation thermal processing (RA-TP) of Newtonian liquid/particulate mixtures

• Dimensionless correlations were developed for reciprocation agitation thermal processing.
• Contribution from both natural and forced convection included.
• Developed correlations were statistically significant.
• Separate correlations developed for liquid only and liquid particulate mixtures.
• Applicability of correlations was assessed for non-Newtonian fluids and different particle shapes.

Reciprocating agitation thermal processing (RA-TP) is a recent addition to the container agitation mechanisms available to reduce quality loss during canning. Data on heat transfer coefficients associated with RA-TP of canned particulates in Newtonian fluid were used to develop dimensionless correlations for predictive modeling of RA-TP. Developed correlations were validated against data from processing of non-Newtonian liquid-particulate mixtures. Effect of container orientation was included in the correlations by considering the aspect ratio (AR) of can. Sums of amplitude with container's axial dimension (A + Daxis) and particle diameter (A + Dp) were found most suitable as characteristic length in correlations for overall (U) and fluid-to-particle (hfp) heat transfer coefficient, respectively. Gr/Re2 (0.01–4.46 and 0.02–6.86 for U and hfp, respectively) during RA-TP was lower than those during other modes of agitation, suggesting predominance of forced convection during RA-TP. However, mixed-convection model gave better R2 (0.93–0.95) than pure-forced-convection model, suggesting mixed convection scenario under mild agitation conditions. Developed correlations fitted well for data from non-Newtonian fluid mixtures on replacing the viscosity term with apparent viscosity of non-Newtonian fluid. It was also seen that Sauter mean diameter can be used to approximate the effect of varying particle shapes and sizes in the developed correlations.Industrial Significance: Reciprocating agitation of containers is receiving interest from thermal processing industry for enhancing product quality. The aim of this study was to develop dimensionless correlations for predictive modeling of heat transfer coefficients to be used in computational programs for process establishment and validation. Additionally, importance of natural convection in forced convection heat transfer correlations has been demonstrated by developing the U and hfp correlations using the mixed convection approach as the combination of natural and forced convection heat transfer. This study also analyzes the applicability of developed correlations for real foods which may be of non-Newtonian nature and of varying shapes and sizes. The developed correlations would be useful for modeling the time–temperature profiles of a canned particulate mixture and will be helpful in determining the contribution of natural and forced convection heat transfer. These dimensional numbers would give a better understanding of the physical phenomenon and can also be easily used for scale-up purposes.