NUMERICAL ANALYSIS OF THE AIR FLOW IN A COOL STORE BY MEANS OF COMPUTATIONAL FLUID DYNAMICS

The postharvest quality of fresh fruits an vegetables is affected mainly by their cooling rate which depends to a large extent on the cooling air temperature and velocity. Whereas the optimal storage conditions is known for most products, in practice it is difficult to guarantee a uniform distribution of cooling air temperature and air velocity. This results in a variable cooling time, and hence end quality, of products which had equal quality attributes at harvest. For cool room design and operation purposes, it is necessary that the distribution of the processing variables (cooling temperature and velocity) is established. When the physical parameters of the materials involved (air, produce and processing equipment) and the geometry of the problem are known, the processing variables can be obtained from the conservation laws of mass, momentum and energy. Unfortunately these equations have inherited the complexity of the transfer processes they describe, and they are coupled and highly non-linear. Analytical solutions exist only in a small number of simplified cases, the most common approach is to use simplified equations, which result from a combination of approximations and dimensional analysis and require considerable experimental input. This empirical approach is very efficient when global parameters, such as heat transfer coefficients, are to be estimated for a specific geometry under specific conditions. Clearly, with the objective of design optimisation in mind, details of the distribution of the processing variables are necessary. At this scale of detail, experiments become very time-consuming and costly.

Alternatively, with the increasing computational power and memory capacity of computers and the development of efficient numerical algorithms, the numerical solution of the governing fluid flow, mass and energy equations has become possible. The availability of user-friendly software-codes (CFX, AEA Technology, Harwell, UK; Phoenix, Flowsolve Ltd, London, UK; Fluent/FIDAP, Fluid Dynamics International, Evanston, USA; STAR-CD, Computational Dynamics Ltd, London, UK) has brought the method within reach of the engineer. This approach of using of computers for solving fluid flow problems is known as Computational Fluid Dynamics (CFD). Being used for many years in high-technology engineering (aeronautic, aerospace and nuclear industries), the method has become increasingly popular in other fields such as food engineering (Datta and Teixeira, 1987; Mirade et al., 1995; Verboven et al., 1997). An important limitation of CFD is certainly the investment cost. For a typical design study cycle, one needs a few man-months of a highly-skilled CFD expert, a fast computer (workstations are mostly used, although PC_s are becoming increasingly important) with a high memory capacity (minimum 128Mb RAM) and a commercial software code according to the personal needs and preferences. This may add up to a considerable amount of money, which probably is still a major drawback for many food related companies. Nevertheless, the commercial benefits can be tremendous. Rhodes (1995) states a two-thirds reduction in the energy-cost of steel-making and a fuel cost saving of £500,000 per annum for a power station steam condenser, both through improvements resulting from the application of CFD. From a