A NEURAL NETWORK APPROACH FOR EVALUATION of SURFACE HEAT TRANSFER COEFFICIENT

An artificial neural network (ANN) approach for tackling the inverse heat conduction problems was explored - specifically for the determination of surface heat transfer coefficient at the liquid-solid interface using the temperature profile information within the solid. Although the concept is quite generic, the specific cases considered have a particular relevance to food process engineering applications. the concept was tested with two geometric shapes: a sphere and a finite cylinder, the former representing the simplest geometry and the latter representing a cross product of an infinite cylinder and an infinite plate. In developing the ANN model, two approaches were used. In the first one, the ANN model was trained to predict the surface convective heat transfer function, Biot number (Bi) from the slope coefficient (m) of temperature ratio curve under varying boundary conditions. the associated mean relative prediction errors were as high as 5.5% with a standard deviation of 8%. In the second ANN approach, m was related to tan-¹ (Bi) which significantly improved the model's predictive performance. the second ANN model could be used with Biot numbers up to 100 with a mean error less than 1.5% for either of the two geometries. Heat transfer coefficients evaluated using the developed ANN model were in agreement (<3% error) with those calculated using conventional numerical/analytical techniques under a range of experimental conditions.     

MODELLING of the MEDIA-SIDE HEAT TRANSFER IN SCRAPED SURFACE HEAT EXCHANGERS

The heat transfer on the media-side and in the tube wall of scraped surface heat exchangers was investigated when the product was heated with steam or cooled with water. The heat transfer coefficients found experimentally were 15% higher than predicted with the Nusselt theory for steam condensation, and 30% lower than predicted with a traditionally recommended model for cooling with water. New, much better models for the media-side heat transfer were developed. the choice of model for the media-side affects a subsequent modelling of the heat transfer on the product-side. When the product-flow was laminar in the scraped surface heat exchangers, the heat transfer was controlled mainly by the resistance on the product-side; while the resistance on the media-side was very small. Vortical flow decreased the resistance on the product-side considerably and made the choice of material in the tube wall important from a heat transfer point of view.     

VISCOSITY and SURFACE TENSION of CORN OIL AT FRYING TEMPERATURES

Surface active properties and viscosity changes of corn oil were determined under controlled frying conditions. Viscosity of corn oil was measured at a temperature range of 100–200C. Surface-active properties of thermally degraded corn oil were determined by measuring the stationary contact angle () of oil droplets placed on a low energy surface (polystyrene). the viscosity of oil decreased as temperature increased. At 200C, the kinematic viscosity had decreased 69% when compared with its initial value of 7.50 mm²/s at 100C. Oil was degraded by frying 400 g of potatoes for 5 min each hour for 36 h (low load frying). the experiment was repeated with 1200g of potatoes (high load frying). the viscosity of the oil increased 9.4% and 28.6% for low and high load fryings, respectively. the surface tension of the oil and its degradation time were inversely proportional (at high load frying) lowering the surface tension of the oil by 2.0%. A high correlation (r²= 0.90) between the viscosity and surface tension of oil was obtained.     

VISCOSITY and HEAT TRANSFER COEFFICIENTS FOR CANOLA, CORN, PALM, and SOYBEAN OIL

To understand the influence of frying oil's physical properties on heat transfer, heat transfer coefficient and oil viscosity were measured for combinations of oil type, temperature, and condition. the lumped capacity method for heat transfer in a high thermally conductive metal gave convective heat transfer coefficients. A capillary viscometer in a convective air heater provided viscosity data at frying temperatures. Frying time and oil temperature significantly affected viscosity. Oil viscosities were not statistically different between fresh and 12 h frying oil or 12 and 24 h frying oil, while between the remaining frying times the oil viscosities were statistically different. Corn oil viscosity showed the greatest increase over 36 h and the highest correlation between viscosity and heat transfer coefficient (−0.959).     

COLLECTION of ACCURATE EXPERIMENTAL DATA FOR TESTING the PERFORMANCE of SIMPLE METHODS FOR FOOD FREEZING TIME PREDICTION

Objective testing of the accuracy of food freezing time prediction formulae requires experimental data to be of high quality. It is shown that there are shortcomings in commonly used experimental methods, particularly relating to minimizing heat transfer in all but the required dimensions in experiments with the slab and infinite cylinder shapes, in maintaining uniformity of surface heat transfer coefficients across a sample, and in measurement of surface heat transfer coefficients. Broad guidelines to ensure that the errors introduced by experimental techniques are negligibly small are proposed. Major published experimental data sets are compared to these guidelines and comments made on their likely accuracy.     

TIME-VARIABLE RETORT TEMPERATURE PROFILES FOR CYLINDRICAL CANS: BATCH PROCESS TIME,ENERGY CONSUMPTION,and QUALITY RETENTION MODEL1

The batch retort model developed uses a heat transfer equation for heat conduction in cylindrical cans, first order kinetics for microbial inactivation, first order kinetics for quality losses and a transient energy balance to estimate steam consumption. For a given retort, lethality process and quality retention, the transient energy balance equation in the model allowed the identification of feasible time-temperature profiles reducing energy consumption, total process time or both. In the examples analyzed and depending upon product specifications, time-variable retort temperatures reduced process time by 18–55 min. These examples suggested that a change from constant to time-variable retort temperatures could increase canning capacity by 20–50%.     

EFFECT of VOLUME CHANGE IN FOODS ON the TEMPERATURE and MOISTURE CONTENT PREDICTIONS of SIMULTANEOUS HEAT and MOISTURE TRANSFER MODELS

Temperature and moisture content predictions of two models of simultaneous heat and moisture transfer (SHMT) in foods with and without the assumption of volume change were compared. In the first model, thermophysical and transport parameters were kept constant. In the second model, these were taken as variable functions of temperature and moisture content. Comparison of the results of both models with and without shrinkage assumption showed that there can be significant differences in predicted moisture and temperature gradients, and average moisture contents and temperatures. Biot numbers for heat and mass transfer, and the extent of overall shrinkage with moisture loss are important factors affecting these differences. the predictions of the second model were compared to experimental drying data with shrinkage. the variable-grid finite difference method can be used successfully to solve model equations involving volume changes.     

A MODEL FOR DIFFUSION of SODIUM IN GREEN OLIVES AT DIFFERENT TEMPERATURES and LYE CONCENTRATIONS

A diffusion model for a composite flat plate, consisting of a thin skin of thickness o, and a thicker flesh of thickness L, with constant flesh and skin effective diffusion coefficients (D_F and D_S) was successfully adjusted to the experimental data obtained during the debittering of green olives variety Arauco (also known as Criolla) at 15C, 20C and 25C and lye concentrations of 1.50%, 2.25% and 3.00% of NaOH. the D_F values ranged between 1.64 × 10^?10 and 9.20 × 10^?10 m²/s, and increased with increasing temperature and lye concentration. the D_S values ranged between 1.27 × 10^?12 and 5.63 × 10^?12 and also increased with increasing lye concentration and temperature. Activation Energies (Ea) of diffusion in the flesh and the skin were evaluated with an Arrhenius‐type relationship. the Ea values for the flesh were 93.2, 72.7 and 64.8 kJ/mol, and 92.7, 91.46 and 33.67 for the skin for lye concentrations of 1.50%, 2.25% and 3.00%, respectively.     

ESTIMATION of CONVECTIVE HEAT TRANSFER BETWEEN FLUID and PARTICLE IN CONTINUOUS FLOW USING A REMOTE TEMPERATURE SENSOR1

The convective heat transfer coefficient (h_fp) between fluid and particle in continuous tube flow was estimated using a temperature pill, a remote electronic temperature sensor which uses a quartz crystal as the temperature sensing element. the temperature history of a particle (with temperature pill mounted within) was monitored as it moved through a test section. A finite element algorithm was then used to back calculate h_fp from the time-temperature data. the value of h_fp ranged from 134 W/m²K to 669 W/m²K (Nu=3.6 to 17.3) over a fluid generalized Reynolds number range (restricted by experimental constraints) from 19 to 196.8. As expected, h_fp decreased with increasing carrier medium viscosity and increased with increasing flow rates. Even with several conservative factors at play, the lowest Nusselt number was greater than the 2.0 expected for a spherical particle in a stagnant fluid. Potential applications of the technique include noninvasive measurement of the temperature (and h_fp values) of the particles moving within a fluid stream in otherwise inaccessible locations.     

SIMULTANEOUS HEAT and MASS TRANSFER DURING the DEEP FAT FRYING of TORTILLA CHIPS

The heating and water loss of a single tortilla chip during deep fat frying was investigated experimentally and the data were analyzed by writing two heat and mass transfer balances. the equations were solved using explicit finite difference technique. Oil uptake as a function of frying time was described by a first order exponential equation. the empirical and theoretical heat and mass transfer agreed well.
The effect of oil temperature on the moisture loss and oil uptake as a function of frying time was analyzed. Moisture loss rate increased as temperature increased. the effects of temperature on oil uptake were not significant during the first 15 s of frying, however, the final oil content was higher for the tortilla chip fried at 190C than at 150C for 60 s.     

EFFECT of LOW TEMPERATURE BLANCHING, CYSTEINE-HCi, N-ACETYL-L-CYSTEINE, Na METABISULPHITE and DRYING TEMPERATURES ON the FIRMNESS and NUTRIENT CONTENT of DRIED CARROTS

Low temperature long time (LTLT) blanching (70C for 20 min) together with calcium treatment can be used to significantly improve the texture of rehydrated dried carrots when compared to high temperature short time (HTST) blanching (100C for 3 min). LTLT blanching allows pectin methyl esterase to deesterify pectin which can then react with calcium to form salt bridges. 0.3% L-cysteine-HCl was found to be most effective in preventing ascorbic acid loss and obtaining a product with the highest rehydration ability, compared to pretreatments with 0.3% N-acetyl-L-cysteine and 0.1% sodium metabisulphite. On the other hand 0.1% sodium metabisulphite was most effective in preserving the carotenoids content of dried carrots. Ascorbic acid and rehydration ability were more adversely affected by long drying time than high drying temperature, while carotenoids were more sensitive to high drying temperature than drying time. Hence, 60C drying temperature was good for ascorbic acid and rehydration ability, while 40C drying temperature was good for carotenoid and color of dried carrots.     

COMPARISON of TWO OPERATING METHODS of A PLATE HEAT EXCHANGER UNDER CONSTANT HEAT FLUX CONDITION and THEIR EFFECT ON the TEMPERATURE PROFILE DURING MILK FOULING

Two methods of operating a plate heat exchanger under constant heat flux were studied: Method A) control by increasing the flow rate of the heating medium; and Method B) control by increasing the inlet temperature of the heating medium. the cross-sectional temperature variation in the plate used in this study was smaller than that of an intermating-type plate. A change in the axial temperature profile was discovered as a result of the uneven formation of milk deposit along the channel. the temperature changes of both the milk and the solid-milk interface (stainless steel-milk or deposit-milk) during fouling were larger when the effects of fouling were overcome by increased flow rate of heating medium than when the effects of fouling were overcome by increased temperature of the heating medium. A quadratic empirical model underpredicted the temperatures in the lightly fouled region and overpredicted the temperatures in the heavily fouled region.     

QUALITY LOSS of DOUBLE CONCENTRATED TOMATO PASTE: EVOLUTION of the MICROBIAL FLORA and MAIN ANALYTICAL PARAMETERS DURING STORAGE AT DIFFERENT TEMPERATURES

This study deals with nonsterile canning of double concentrate tomatoes (30° Brix) stored for 210 days at three different temperatures (4, 10 and 25C) in 200 kg drums. the evolving analytical composition (soluble solids, total solids, glucose, fructose, pH, total acidity, volatile acidity, citric acid, malic acid, succinic acid, hydroxymethylfurfural (HMF) and color parameters) that the product underwent during storage was dependent both on storage temperature and on different aerobic levels within the drum (top and bottom sections). the microbial profile (yeast, lactic acid bacilli and molds) was correlated with many important metabolites (D- and L-lactic acids, ethanol, acetic acid and diacetyl). the results indicate that the increase of these substances is dependent both on storage temperature as well as the oxygen tension within the drums.
Taken all together, the analytical findings offer a great help in evaluating the quality of semifinished tomatoes. We also found that lactic bacteria grow rapidly at 25C and after 15 days their number from both sampling areas in the drums (i.e., 10 cm below the sample surface and 15 cm above the bottom of each drum) is already greater than 10⁵ cfu/ml. At 10C, 30 days were needed to reach such a cell concentration, and after 45 days the level reaches 10⁷-10⁸ cfu/ml. By contrast, at 4C there were differences between top and bottom sampling areas. In the top area, 10⁵ cfu/ml was reached after 60 days, while for the bottom area this was reached after 120 days. Regarding yeast at 25C.     

PARTICLE SIZE, SALT CONCENTRATION and TEMPERATURE EFFECTS ON NITROGEN EXTRACTION of CHICKPEA (Cicer arietinum) BY RESPONSE SURFACE METHODOLOGY

This study was conducted to evaluate the effect of time, pH, particle size, %NaCl and temperature on total nitrogen (TN) extraction of low commercial grade chickpea ( Cicer arietinum ) grain, in order to find the optimum extraction conditions for further processing. Extraction times used were 30, 60, 90 and 120 min and the nitrogen extraction curve was carried out at a pH range of 2.0 to 11.0. Response surface methodology was used to evaluate the effect of particle size (100, 150 and 200 mesh), salt concentration (0, 0.3 and 0.6% NaCl) and temperature (25, 35 and 45C). No significant time effect was found on nitrogen extractability. However, a great pH effect on nitrogen extraction was observed. Results from the RSM analysis showed that the most adequate conditions for total nitrogen extraction were particle size, 200 mesh; 0.0% NaCl; pH 8.0; temperature, 25C; and time, 30 min. By using the above mentioned parameters a yield over 90% of nitrogen was obtained.