Milk-fouling phenomena of plate heat exchangers (PHEs) result in an increase of energy consumption, extra maintenance, higher labor costs, and a decrease of production capacity. In addition, they can cause the growth of unwanted microorganisms on the ...
Milk-fouling phenomena of plate heat exchangers (PHEs) result in an increase of energy consumption, extra maintenance, higher labor costs, and a decrease of production capacity. In addition, they can cause the growth of unwanted microorganisms on the corrugated surfaces of PHEs. It is important to understand fouling kinetics on PHE surfaces and the preventive strategy. This research was aimed to develop the transient 3-D model to describe fouling phenomena and estimate the amount of surface fouling using computational fluid dynamics (CFD) based on the hydrodynamic and thermodynamic performances of the PHEs and surface characteristics. Lower flow rates create lower shear forces acting on the fouling. It was found that there is a linear relationship between the amounts of fouling and Reynolds numbers. Therefore, the pre-exponential factors related to milk fouling kinetics were determined by mass flow rate of milk empirically associated with different coating materials (conventionally uncoated stainless steel (SS-316) and stainless steel surfaces with Lectrofluor-641 and graded Ni-P-Polytetrafluoroethylene (Ni-P-PTFE)). The constant heat flux delivered from hot water to skim milk was used as a boundary condition. It was the first time to simulate the fouling pattern and temperature distribution of the PHEs in which all the realistic corrugation profiles have been successfully imported to the CFD codes using AutoCAD software. The amount of fouling on the plate coated with Lectrofluor-641 showed an approximate 90% decrease compared to the control plate. However, there was no significant difference of milk deposits between Lectrofluor-641 and graded Ni-P-PTFE coated surfaces. Consistency was observed between the measurement and model prediction for fouling masses with a maximum prediction error of 7%.