.0 n: '0 72!»? .. l If.A.-0< vr...)!.. I? ’0 ,al .. .1}... In. 3‘ i. u .l..4 an. "Eran. . .. ‘ . . 9 . u r.. ).'. H. .v!. u ._ 1.3. A... .. ‘ b) .1; I... .130: as .c :1... 3 z?» , . n.3,? . 53. .. i. i. e :5 . 9 .v y 7!... . . .1 .1 7.5.5! .6"!!! Eu ....\ I L: .. .a I F v+qfiu 29aW€73 I IIII IIZII IIIIIIIIIIIIIIIIIIIIIIIIII II IIIIIIIIII 380007 809274 II LIBRARY Michigan State University This is to certify that the thesis entitled MICROWAVE - CONVECTION DRYING OF READY -TO-EAT CEREALS presented by James C. Breslin has been accepted towards fulfillment of the requirements for M.S. degree in Agricultural Englneerlng Major professor Maj/MEL 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution X PLACE lN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE IfiI‘T‘E MSU is An Affirmative Action/Equal Opportunity Institution cmma-pd MICROWAVE - CONVECTION DRYING OF READY-TO-EAT CEREALS BY James C. Breslin A THESIS Submitted to Michigan State University in partial fulfillment of the requirements of the degree of MASTER OF AGRICULTURAL ENGINEERING Department of Agricultural Engineering 1988 ABSTRACT MICROWAVE - CONVECTION DRYING OP READY—TO-EAT—CEREALS BY JIH BRESLIN The performance of a convection - ndcmowave dryer is compared to a conventional conventional dryer for ready-to- eat cereals. A pilot scale microwave - convection dryer was built. The dryer has a working length of 4.25 m, with a maximum power of 20 KW. The residence time used in this study is approximately 120 seconds. Measurements were made to quantify the gross average environment within the pilot scale microwave — convection dryer.. Several assumptions were made to approximate the rate of microwave heating. The environmental measurements were used as boundary conditions to approximate the heat and. moisture transfer occurring within the microwave - convection dryer. Experiments were made to estimate the thermal and moisture diffusivity of the product. A mathematical model of the heat and moisture transfer was developed. The model approximates the performance of the pilot scale unit over a range of operating conditions. Several new designs are evaluated with the model. The model is used to estimate the thermal and moisture history of a cooked cereal product being dried in the production of a Ready—To-Eat cereal product in a conventional convection dryer. The time required to dry a cooked cereal product in a microwave - convection dryer is predicted by the model. The heat generated by microwave heating is based on the assumptions used to develop the model. Best results are obtained by limiting the microwave energy to 75 v/m in the first zone of the dryer. No microwave energy is applied in the second or third zones of the microwave - convection dryer. The drying time required to reduce the moisture content from 28 to 21% wet basis is reduced by 450% compared to the conventional convection system. The reduction in residence time can be used to decrease the length of the dryer from 10.1 to 3.1 m, or to increase the capacity of the production unit by 235%. 4 2.12.4 MaIior Professor////y/J77 A.E. Dept Chairman 1.0 2.0 3.1.1 3.1.2 3.1.3 3.2.1 3.2.2 TABLE OF CONTENTS PAGE NUMBER INTRODUCTION OBJECTIVES LITERATURE REVIEW MICROWAVE ENERGY MICROWAVES MICROWAVE HEATING DIELECTRIC PROPERTIES DRYING DRYING ANALYSIS MICROWAVE ASSISTED DRYING EXPERIMENTAL PROCEDURES DESCRIPTION OF THE MICROWAVE SYSTEM MICROWAVE FIELD STRENGTH TESTS PRODUCT DIFFUSIVITY MEASUREMENTS DATA COLLECTION THEORETICAL HEAT TRANSFER MASS TRANSFER NUMERICAL METHODS 1 10 10 12 13 17 23 28 33 35 35 42 47 SO 53 53 55 S7 7.0 7.1 8.0 9.0 10. 10.1 10.2 RESULTS AND DISCUSSION MICROWAVE FIELD STRENGTH MEASUREMENTS PRODUCT THERMAL AND MASS DIFFUSIVITY MEASUREMENTS OPERATIONAL DATA COLLECTION COMPARISON BETWEEN EXPERIMENTAL AND SIMULATED DATA USE OF THE MODEL ANALYSIS OF CURRENT DRYER DESIGN OF THE MICROWAVE - CONVECTION DRYER ADDITION OF MICROWAVE ENERGY IN THE EALLING RATE PERIOD ADDITION 0! MORE CONVECTION HEATING ADDITION OF MICROWAVE HEATING TO ALL ZONES ADDITION OF HIGH LEVELS OF MICROWAVE HEATING IN ZONE 1 POST SCRIPT CONCLUSIONS RECOMMENDATIONS FOR FUTURE WORK MICROWAVE YIELD STRENGTH PRODUCT DIPPUSIVITY 62 62 67 72 76 85 85 89 89 90 94 97 101 102 105 105 105 10.3 SURFACE EQUILIBRIUM 106 10.4 PRODUCT GEOMETRY 106 11. APPENDICES A. MICROWAVE FIELD STRENGTH MEASUREMENTS B. MICROWAVE - CONVECTION DRYER TEST CONDITIONS C. PROGRAM OVEN LIST OF TABLES TABLE 1.1 1 DOMESTIC SALES OF MAJOR READY-TO-EAT CEREALS BY U.S. FIRMS TABLE 3.1.3 DIELECTRIC pROpERTIES or SELECTED FOOD PRODUCTS TABLE 3.2.1 CONVENTIONAL DRYER SPECIFICATIONS TABLE 3.2.2 CONVENTIONAL DRYER OPERATIONAL CONDITIONS TABLE 4.4.1 pROPERTIES or RELLETS AT INLET or DRYER TABLE 6.1.1 NICRONAVE rIELD STRENGTH NEASURENENTS TABLE 6.1.2 NICRONAVE TIELD STRENGTH CALCULATIONS TABLE 6.1.3 CONBINED NICRONAVE rIELD STRENGTB BY POSITION TABLE 6.2.1 TRANSPORT MEASUREMENT BLOCK TEMPERATURE GRADIENT DATA TABLE 6.2.2 THERMAL DIFFUSIVITY AND CONDUCTIVITY CALCULATIONS TABLE 6.2.3 TRANSPORT NEASURENENT BLOCK MOISTURE GRADIENT DATA TABLE 6.2.4 22 24 26 52 63 65 66 67 70 71 72 TRANSPORT MEASUREMENT BLOCK MASS DIFFUSIVITY CALCULATIONS TABLE 6.3.1 73 TUNNEL OPERATING CONDITIONS RUN i 9 TABLE 6.3.2 74 PRODUCT TEMPERATURE AND MOISTURE MEASUREMENTS RUN 0 9 TABLE 6.4.1 79 TUNNEL OPERATING CONDITIONS RUN 0 23 TABLE 6.4.2 81 PRODUCT TEMPERATURE AND MOISTURE MEASUREMENT RUN 0 23 TABLE 7.1 89 DRYER OPERATING CONDITIONS MICROWAVE HEATING IN THIRD ZONE TABLE 7.2 92 DRYER OPERATING CONDITIONS MICROWAVE HIGH CONVECTION HEATING TABLE 7.3 94 DRYER OPERATING CONDITIONS MICROWAVE HEATING IN ALL ZONES TABLE 7.4 97 DRYER OPERATING CONDITIONS HIGH MICROWAVE IN FIRST ZONE LIST OF FIGURES FIGURE 1.1 CEREAL MANUFACTURING PROCESS FIGURE 1.2 FOOD DEHYDRATION FIGURE 3.1.2.1 PROPAGATION OF A PLANE WAVE IN A LOSSY MEDIUM FIGURE 3.1.3.1 DIPOLE ROTATION FIGURE 3.1.3.2 ELECTRIC DIPOLE MOMENT FIGURE 3.2.1 BELT DRYER FIGURE 3.2.2 PELLET DRYER AIR FLOW FIGURE 4.1.1 SCAN PRO MICROWAVE TUNNEL FIGURE 4.1.2 SCAN PRO MICROWAVE TUNNEL WITH CONVECTION HEATING FIGURE 4.1.3 MICROWAVE CONTROL PANEL FIGURE 4.2.1 MICROWAVE FIELD STRENGTH MEASUREMENT FIGURE 5.3.1 PROGRAM FLOW FIGURE 6.2.1 16 19 19 25 27 36 39 41 43 58 68 TRANSPORT MEASUREMENT BLOCK FIGURE 6.4.1 77 EXPERIMENTAL AND SIMULATED PRODUCT CONDITIONS RUN 0 9 FIGURE 6.4.2 80 EXPERIMENTAL AND SIMULATED PRODUCT CONDITIONS RUN 4 23 FIGURE 6.4.3 83 EXPERIMENTAL AND SIMULATED PRODUCT CONDITIONS RUN C 13 FIGURE 6.4.4 84 EXPERIMENTAL AND SIMULATED PRODUCT CONDITIONS RUN 0 17 FIGURE 6.6.1 86 ANALYSIS OF CONVENTIONAL CONVECTION DRYER FIGURE 7.1 91 SIMULATED DRYING CURVES MICROWAVE HEATING IN THIRD ZONE FIGURE 7.2 93 SIMULATED DRYING CURVES HIGH LEVELS OF CONVECTION HEATING FIGURE 7.3 96 SIMULATED DRYING CURVES MICROWAVE HEATING IN ALL ZONES FIGURE 7.4 99 SIMULATED DRYING CURVES INTENSE MICROWAVE HEATING IN ZONE 1 V“ CHAPTER 1 INTRODUCTION The ready-to—eat cereal industry has grown into one of the major forces in the food industry. At the current time the average American consumes 9.6 pounds of cereal per year. This adds up to a volume of 2.4 billion pounds of cereal produced each year in the United States (Level 1988). (1987) sales in the domestic Nielsen reported that the cereal industry exceeded 5.2 billion dollars in 1987. Table 1.1 details the market share of major products in the ready- tO-eat cereal industry. Corn Flakes and Frosted Flakes are the major products consumed in the United States. TABLE 1.1 DOMESTIC SALES OF MAJOR READY-TO-EAT CEREALS BY U.S. FIRMS 19 6 1986 URRENT 10 I SHARE 10 8 SHARE SHIPPED % SHIPPED 8 KELLOGG CORN FLAKES 135494 5.96 142890 6.11 KELLOGG FROSTED FLAKES 128761 5.66 128500 5.50 GENERAL MILLS CHEERIOS 128761 4.48 108500 4.40 KELLOGG RAISIN BRAN 104534 4.60 97362 4.16 KELLOGG RICE KRISPIES 83107 3.65 85123 3.64 GENERAL MILLS CHEERIOS HSNUT 60626 2.67 64587 2.76 GENERAL FOODS POST RAISIN BRAN 60311 2.62 57532 2.46 GENERAL FOODS POST GRAPE NUTS 58327 2.56 59641 2.55 KELLOGG FRUIT LOOPS 50751 2.23 52542 2.25 KELLOGG MINI WHEATS 49974 2.20 51259 2.19 GENERAL MILLS TOTAL 41329 1.82 43654 1.87 GENERAL MILLS WHEATIES 40417 1.76 39355 1.68 GENERAL MILLS LUCKY CHARMS 36927 1.62 37699 1.61 KELLOGG SPECIAL K 35460 1.56 38052 1.63 KELLOGG BRAN FLAKES 38216 1.68 37136 1.59 SOURCE: NIELSEN (1987) All of the products listed in Table 1.1 except Cheerios, Grape Nuts, and Fruit Loops, share a central processing theme. The raw material and processing conditions change for each product, but the unit Operations required to produce the products are the same. The standard unit Operations are Shown in Figure 1.1, they are: cooking, drying, milling, and toasting. There are two main Objectives in the cooking process. The first is to gelatinize the starch granules; Englyst 1986 reported that the gelatinization step improves the pallitabilty and nutrient value of the starch product. The second Objective of the cooking process is to penetrate the starch molecule with a flavor solution. This step improves the taste, and color of the final product (Hoseney 1986). According to Hirzel (1982) the main objective of drying in cereal production is to change the particle texture in order to meet the requirements of the milling process. In general, the texture required for milling is a hard exterior and a pliable center. The removal of excess moisture from a cooked cereal product often yields the acceptable texture. The milling process causes a physical change in the particle shape. The Change in shape prevents the gelatinization reaction from reversing and undergoing CEREAL MANUFACTURING BASIC PROCESS OPERATIONS RR COOKING DRYING \ MILLING —————————— TOASTING FIGURE 1.1 retrogration (Finley 1985). The net effect of the milling step is an increase in the particle surface area. Toasting has two main Objectives (Finley 1985). The first is to remove the moisture above the acceptable storage requirement of 2-3% wet basis; this step is necessary to insure a long Shelf—life of the product. The second Objective is to provide a suitable environment for the Mailard reactions which control the color and taste of the finished product. This thesis is an investigation of the drying process. From the perspective of the food processor, the most important characteristics of a dryer are the product quality, the dryer capacity, and the Operating and fixed drying costs. The quality Of a dried cereal product depends on the particle texture and uniformity. The texture is affected by the average moisture content and by the moisture gradient between the center and outer surface of each particle. The moisture uniformity of the dried product is critical. The drying operation needs to limit the variation in particle moisture. The capacity of a dryer is a function of the maximum bed depth which can be Obtained inside the dryer. As the bed depth increases, the humidity of the air exiting the bed rises. When the condition of the air leaving the product bed no longer changes with bed depth, the product has reached the maximum bed depth. Using the maximum bed depth and product flow rate, the maximum capacity Of the dryer can be estimated. The maximum capacity of a dryer is directly related to the size of the dryer. For a given size dryer the maximum capacity' is dependent on the rate of moisture removal. The rate of moisture removal is dependent on the rate of heat transfer. Figure 1.2 illustrates the energy and moisture transfer which occur inside a conventional dryer. The cost of the dryer can be broken into two parts: (1) the Operational cost per pound of product, and (2) the fixed cost of the equipment. The unit Operating costs include the energy costs required for drying one pound of product. The floor space required is a significant Characteristic of a dryer. The cost of the floor space is critical when a capacity increase is needed. The cost to construct additional floor space for a larger dryer is Often higher in an existing plant than the fixed cost of the dryer. mooo om1wmo._ ¢ ZH m>¢3 wz¢.._n_ ¢ n5 ZOHF¢o¢momm 16 Von Hipple (1954) developed an expression for the penetration depth of microwave energy into a thick food product. A thick product is defined as a product over 2.5 cm in thickness, or as a packed bed of small particles over 2.5 cm in thickness. The penetration depth is the distance which a plane wave travels before it has lost l/e Of its original power. The penetration depth is expressed as: A0 ( e')1/2 Dp - I 3.1.2.4 2 n s where, Dp - Penetration depth of material (m). 3 . 1 . 3 DIELECTRIC PROPERTIES Goldblith (1967) described the heat generated within an elemental volume of a particle as a function of the element field strength and the dielectric property of the particle. The dielectric property governs the amount Of microwave energy which is converted into heat. The energy results from the polarization of atoms by the electric field and the physical interaction between the electric field and the atoms. Heat is generated by creating and destroying the forces used to polarize atoms, and the inability of the polarization to follow the changes in the direction of the applied microwave field. 17 Detailed studies of polarization have been conducted by Debye (1929) , Daniel (1967), and Hasted (1973). The interaction of an electric field with a dielectric material has its origin in the response of charged ions to the applied electric field. The movement from the equilibrium position gives rise to induced dipoles which are created by polarization. The dipoles magnify the reaction to the electric field. This is termed the interfacial polarization. Some materials contain permanent dipoles due to an asymmetric Charge distribution. The dipoles tend to reorient under the influence of the electric field through so called orientation polarization described in Figure 3.1.3.1. The average dipole consists of two charged and slightly separated particles. The configuration results in a dipole moment when the charged particles attempt to reorient to the electric field as shown in Figure 3.1.3.2. The average dipole moment is given by Metaxes (1979) as: n ' q X 3.1.3.1 where, u - Average dipole moment q - The Charge of the dipole x - The distance separating the charged particles. 18 DIPOLE ROTATION (6‘ so) 1 @ EOE.“ a) f‘ (90 fig *9 a .Q .9 FIGURE 3.1.3.1 ALTERNATING ELECTRIC FIELD ELECTRIC DIPOLE MOMENT H———-—-—X1 U FIGURE 3.1.3.2 19 If the dipole moments within a given volume of N dipoles are summed, a measure Of the charge density of the volume is Obtained. The Charge density is termed the polarization vector P, and is described by Metexas (1979) as: P - - 3.1.3.2 The polarization vector is a measure of the energy stored within a product with N' individual dipole moments. Thus, P - p N' 3.1.3.3 The electrical potential of a system is expressed in the electric potential vector D. The difference between the vectors P and D is equal to the electric potential not due to polarization. Thus, D - so E + 1» 3.1.3.4 and since, D - so 3' E 3.1.3.5 rearrangement yields, P-so(s'-1)E 3.1.3.6 in which s' is the relative dielectric constant of the material. 20 If an electric field E' is applied to a single dipole the resulting dipole moment can be expressed as: p - a E' 3.1.3.7 in which a is the polarisability of the material and contains both the interfacial and orientation polarization affects. Combining equations (3.1.3.5),(3.1.3.6),and (3.1. 3.7) yields: so ( 1- s') - a N' E' 3.1.3.8 Equation 3.1.3.8 links the macroscopic quantities e' and E to the molecular properties N' and E'. In order to describe the dielectric constant of a material which is changing in temperature and moisture content, additional information about the dielectric property is required. The dielectric constant is divided into two parts. The complex form of the dielectric constant is given by Metexas (1979) as: c - e' - j s' 3.1.3.9 where, s' is the relative dielectric constant; and c' is the imaginary part which is called the dielectric loss factor. Physically, the two parts of the complex dielectric constant each describe a different process. The relative dielectric constant measures the ability of the material to store energy, and the dielectric loss factor measures the energy used by the reorientation Of polarized molecules within the 21 applied field. the loss factor to the Hipple 1954): tan 6 - ————— The dielectric properties Of some food materials are 8' shown in Table 3.3.1. TABLE 3.3.1 relative dielectric constant The loss tangent is defined as the ratio of 3.1.3.10 DIELECTRIC PROPERTIES OF SELECTED FOOD PRODUCTS RELATIVE DIELECTRIC DIELECTRIC LOSS TEMPERATURE CONSTANT FACTOR (°C) (unitless) (unitless) BEEF STEAK BOTTOM ROUND 25 40.0 12.0 FROZEN LEAN 0 3.9 0.3 BACON FAT 25 2.5 0.13 POTATO RAW 25 53.7 15.7 TURKEY COOKED 25 40.0 14.0 BUTTER O 4.05 0.39 BUTTER 35 4.15 0.44 WATER ICE -12 3.2 0.003 DISTILLED 25 76.0 12.0 MILK POWDER 30 2.29 0.048 WHEY POWDER 30 2.04 0.025 NEEPAWA WHEAT 21 3.0 0.90 SOURCE: VON HIPPLE (1954)/NEILSON (1974) 22 (Von 3 . 2 DRYING .A prerequisite for improving dryer performance is an understanding Of the conventional drying process. The Objective of the drying process in a cereal plant is to modify the textural Characteristics of the product. Two texture characteristics are modified by the drying process. The first is the outer surface of the particle; as the temperature of the outer surface rises, the surface becomes hard. The second Characteristic is the texture of the interior; as the average product moisture changes the moisture content and the texture of the interior changes little. This leaves a product with a hard, dry outer surface and a moist, pliable interior. The conventional cereal production process requires a milling step to modify the shape of the product. The product texture requirements for milling are a hard surface, and a pliable interior. Thus, the correct drying process supplies the product textures required for the milling process. The conventional convection dryer used in cereal production is the humidity - controlled belt dryer. A physical description of the dryer type is shown in Figure 3.2.1. Typical Operating specifications for the FEC belt are listed in Table 3.2.1. 23 TABLE 3.2.1 CONVENTIONAL BELT DRYER SPECIFICATIONS LENGTH SPEED DEPTH DRY BULB RELATIVE FLOW TEMPERATURE HUMIDITY OLUME ZONE (m) (m/min) (m) ( °C ) ( % ) (m /min) 1 10.1 4.0 0.05 61 65 600 2 9.7 2.25 0.06 61 65 600 3 10.5 1.25 0.12 61 65 600 The belt dryer uses air at constant dry bulb and wet bulb temperatures. The dry bulb temperature of the air is controlled by the use of a steam heat exchanger. The wet bulb temperature is automatically controlled by measuring the relative humidity and Changing the ratio between the volume of air removed as bleed, and the volume of air brought in the dryer as makeup. The dryer is divided into three zones or flights. The operating conditions of a belt dryer are controlled to remove the same amount of moisture in each flight. The moisture removal from a flight is adjusted by Changing the residence time of the product in the zone. As the product moisture content drops, additional time is required to remove the same amount of moisture. 24 H.N.m mmsoHu IQILV ualshN "omlof 9 WZOHFUUW 0 25 bun—2H Husoomm ozHJn—zcw m2 wmomm 3ND mw>mo ._..._mm Thus, the bed depth is largest in the third zone and smallest in the first zone. Air is supplied between the zones and is designed to move through and over the product bed, and then return to the air supply system (see Figure 3.2.2). Typical product temperatures and moisture contents at the inlet and outlet of each zone of the dryer are shown in Table 3.2.2. TABLE 3.2.2 CONVENTIONAL BELT DRYER OPERATING CONDITIONS TEMPERATURE RESIDENCE BED PRODUCT PRODUCT DRY BULB R.H. TIME DEPTB TEMP NOISTURE ( °C ) ( % ) (see) (cm) ( °C ) (t vb) zONE 1 61 65 151 S 38* 26.3* 2 61 65 258 7 46* 23.2* 3 61 65 504 10 52* 21.5* * MEASURED AT OUTLET OF FLIGHT The measurements shown in Table 3.2.2 do not imply that the dryer is operating at Optimum conditions. 26 N.N.m mason... Esggg .\ \ t/ / # I.. x \ / «.mzou / 3. Or... \ \ \.\. \w / as .\ .\ 5.. N 2...... s. p \\D\\.\ /, w/ PMS/ / a ed» 3&5“... PUG? \ II I M MD. I I I I A Go Ba K4835 Eugcdguk 304....— CE gs .5..me 27 3 . 2 . 1 DRYING ANALYSIS In order to analyze the drying process, it is necessary to understand the drying stages which the product undergoes inside the dryer. The cooked cereal product enters the dryer with a moisture content of 28% wet basis. At this moisture content the outer surface of the cooked cereal is saturated. AS the temperature of the outer surface increases, the moisture evaporation rate from the particle surface increases. This stage of the drying process is called the "constant rate period". The rate of drying is constant, and is dependent on the rate of evaporation at the outer surface. In this stage of the drying process, the resistance to moisture transport at the surface is much larger than the internal resistance. The time during which a product is in the constant rate drying period depends on the rate of moisture flow from the interior of the particle to the outer surface. When the moisture concentration of the outer surface drops below saturation, the constant rate period ends. The end of the constant rate period is characterized by a decrease in the rate of evaporation. This stage Of the drying process is called the "falling rate period”. In this stage the internal resistance to moisture transport is greater than the surface resistance. This causes a moisture gradient to occur within the particle, and the temperature 28 rises above the wet bulb temperature Of the air (Brooker 1978). The decrease in the evaporation rate is due to the increased resistance to moisture transport. The limiting factor is the diffusion rate of moisture from the interior of the product to the outer surface. The point at which the moisture content of the outer surface drops below saturation is defined as the critical moisture content. This point defines the Change from the constant rate period to the falling rate period. Fortes and. Okos (1981) stated that. the moisture migration in the constant rate period is in the form of liquid flow. As the falling rate period begins, moisture migration is split between liquid and vapor flow. .An evaporation front moves into the particle once the outer surface has reached the equilibrium moisture content. Geankoplis (1986) described the movement of moisture from the interior of the particle to the outer surface through the capillaries of a porous product. As the product dries, moisture is removed from the capillaries, and air moves in to fill the voids. AS these capillaries fill with air, they block liquid flow to the surface. At this point vapor movement becomes the dominant flow mechanism. When air pockets form in the interior of the product, the heat conduction from the outer surface to the center of 29 the product decreases. This is due to the change in heat transfer mechanism from conduction to convection as the heat moves through the air pockets (Geankoplis 1986). The net effect is a layer of insulation in the outer surface of the product. The thickness of the insulation layer is dependent on the depth of the evaporation zone. The insulation layer slows the conduction Of heat to the interior of the product and adds to the resistance of mass flow to the outer surface. Analysis of the internal liquid and vapor transport processes which take place inside the product and control the rate of drying is beyond the scope Of this work. The equations which describe the mass, heat, and total pressure transport in a porous media are given by Luikov (1966), and Perkins (1979): SM --_- n a. V2 M + V2 k12 T + V2 R13 P 3.2.1.1 at u q __-_ - v2 k21 n + at v2 T + v2 k23 P + —-—— 3.2.1.2 8t pcp 8P _-_- - V2 R31 M + 72 R32 T + up V2 P 3.2.1.3 at where, M - Product moisture content dry basis (kg HZO/kg food) T - Product Temperature (°C) P - Total pressure of system (Pa) 30 Microwave heat source (watt) .0 I an - Mass diffusivity of product (mZ/sec) at - Thermal diffusivity of product (mz/sec) up - Pressure diffusivity Of product (mz/sec) kxx' Coupling coefficients. The moisture content at any point within the product is the sum Of the liquid mass and the vapor mass: A - M1 + Mv 3.2.1.4 Equations 3.2.1.1, 3.2.1.2, and 3.2.1.3 are valid for a solid which does not exhibit expansion or shrinkage, and has no internal convection. Brooker (1978) called the thermal diffusivity, mass diffusivity, and pressure diffusivity in equations 3.2.1.1, 3.2.1.2, and 3.2.1.3, the phenomenological coefficients. The terms R12 and k21 represent the coupling coefficients known as the Dufour and Soret affects (Bird, Stewart, Lightfoot 1960). The thermal diffusivity is not assumed to be constant. It is a combination Of the thermal conductivity, specific heat, and density Of the product: at - ------ 3.2.1.5 31 The specific heat of the product is dependent on the composition and changes throughout the drying process (Heldman 1979). Fortes and Okos (1980) assumed the thermal conductivity of the product to be constant over the range of product temperature and moisture content encountered in the drying of food products. Thus, K at - 3.2.1.6 vicpw lab + (1 - Heb) car) where, at - Thermal diffusivity Of product (mz/sec) K - Thermal conductivity of product (watt/m- oC) Cpx' Specific heat of component (J/kg—OC) “db' Moisture content of product (% dry basis). The mass diffusivity in equation 3.2.1.1 is a combination of the terms which influence the mass transport. The diffusivity is a function Of the product temperature (Fortes 1980): “I - C1 exp(Dl Tp) 3.2.1.7 where, Mass diffusivity Of product (mz/sec) a? C1 - Empirical coefficient (unitless) D1 - Empirical coefficient (unitless) Local Product temperature (°C). 32 The pressure diffusivity' and therefore, equation 3.2.1.3, can be eliminated from the analysis of conventional drying systems (Brooker 1978). 3 . 2 . 2 MICROWAVE ASSISTED DRYING Application Of microwave heating to the drying process offers advantages when combined with a conventional drying system. The application of heat from a microwave source does not depend on transfer of heat through the surface. The penetration Of ndcrowave energy delivers heat directly to the interior of a thick product. This Offers a unique opportunity to drying of thick products. The application of heat to the interior of a particle has the potential Of extending the constant rate drying period (Oda 1987). This can be found by analyzing the coupling between the product temperature and mass diffusivity. If energy can be delivered to the center Of the product and the resistance to moisture movement is inversely proportional to temperature, the net effect is faster movement of moisture from the interior of the particle to the outer surface. The result is an extension of the constant rate drying period. One of the potential advantages of a microwave power source is the local pressure which may be caused by a phase 33 Change of the moisture from liquid to vapor. When microwave energy is applied, the temperature of the interior is not limited to the boiling point temperature of the liquid in the product. If enough power is absorbed within a local area, the temperature of the liquid. may rise above the boiling point (Fortes and Okos 1980). This will result in an increase in the contribution of the total pressure term in. equation 3.2.3.1. As the local pressure rises, the driving force for moisture movement between the local area and the outer surface increases. Therefore the local rate of moisture transfer increases. This concept is valid only if the temperature of the liquid moisture trapped in the interior Of the product rises above the boiling point. If the local pressure rises above the pressure of the atmosphere, a change in the transport mechanism occurring may cause the product structure to become unstable. A mathematical model of the drying process includes equations 3.2.1.1, 3.2.1.2, 3.2.1.4, 3.2.1.6, and 3.2.1.7. A mathematical model Of the heating caused by microwave radiation includes equations 3.1.2.1, 3.1.2.2, 3.1.2.3, and 3.1.2.4. The combination of these equations forms the basis of the microwave - convection drying of ready-tO-eat cereals. 34 CHAPTER 4 EXPERIMENTAL PROCEDURES The experimental part Of this study has three objectives. The first is to Obtain measurements Of the environment within the microwave tunnel. This includes the air temperature, humidity, and velocity, and the microwave power to which the particles are subjected. The information is needed to describe the boundary conditions for the heat and mass transfer equations. The second step is to determine the thermal and mass diffusivity of the product. The diffusivity information is used to calculate the heat and mass transport occurring within the microwave tunnel. The third objective is to provide basic data to validate the mathematical simulation. Validation is required to check the assumptions made in developing the model. 4.1 DESCRIPTION OF THE MICROWAVE SYSTEM The drying experiments were conducted in a Scan Pro Microwave tunnel (Skandainviska Processinsturment AB Model 40 KW, Bromma, Sweden) which has been designed for continuous flow (see Figure 4.1.1). The Scan Pro tunnel has an operating length Of 14 feet. Of this, 6 feet consist of the chokes at the inlet and outlet Of the tunnel. 35 mODZ pmo ZHnmo€Dzoo :53 ($222. m>¢30mUHZ 0mm zoom HmJZH m? .29 The air temperature is adjusted by using a standard natural gas burner, and is measured with a RTD type thermocouple. A PID controller maintains the temperature set point within a range Of 4 oC. The supply air temperature is measured in the inlet duct at the tunnel face. The outlet air temperature is measured at the outlet duct. The air temperature inside the tunnel is not measured due to the electric field. A linear decrease in the air temperature is assumed. This assumption is based on the contact time between the product and the air stream. The residence time of the air in the tunnel is eight seconds based on the air velocity used in this study. The tunnel is divided into seven sections. The air temperature and humidity are calculated for each Of the seven sections. The field strength is adjusted by changing the electrical power to each of the magnetrons. The magnetrons have two power levels: 1.6 and 2.5 Kw (Figure 4.1.3). The frequency of the microwave radiation is fixed by the magnetron. The supply air humidity is adjusted by using saturated steam at 30 psi. The humidity is measured in the supply duct at the tunnel face with a Rotronix probe (Rotronix Instrument Corporation, Huntington N.Y.). The instrument has a range of 0 to 150 °C. 40 MICROWAVE CONTROL PANEL FIGURE 4.1.3 41 The outlet air humidity is measured in the outlet duct. The electric field prevents humidity measurement inside the tunnel. A.linear relationship between the humidity change and the length of the tunnel is assumed. 4.2 MICROWAVE FIELD STRENGTH TESTS The electric field strength determines the energy dissipated within the product. The electric field strength is a combination of the electric and magnetic fields generated by the magnetron. It is not possible to theoretical predict the magnituide of the microwave field strength in a controlled setting with amount of technology currently available. Therefore, the electric field strength within the microwave tunnel was measured experimentally. The procedure used to measure the field strength is based on several assumptions which are used to approximate the field strength. The experimental method consists of loading the tunnel with several containers filled with a known volume of water (see Figure 4.2.1). The temperature rise of the water was measured in discrete time segments. 42 ENG mmDoHu >._.H>¢o w>¢30muHZ mhzm2mm3w¢m2 Ihozmmhm 04w: m>¢30mo~2 43 The average strength Of the electric field as calculated by calorimetry, is given by Metexas (1979) as: Pcp(T"T0) Eavgz - 4.2.1 (som2nfs')t where, Eavg - Average field strength (v/m3). Equation 4.2.1 is valid if the dielectric properties Of the material are not a function of temperature. Between 20 and 30 C the dielectric properties do make a small change. The dielectric loss factor changes from 22 to 18, the dielectric constant changes from 82 to 78 (Von nipple 1954). This information supports the assumption Of constant dielectric properties Of water. Several approximations are made in order to use equation 4.2.1 for a 120 second experiment. First, it is assumed that the microwave field strength is uniform and constant within each Of the dryer zones. This approximation is theoretically incorrect. The electric field is time dependent and Changes direction at the rate of 2450 million times per second. With current technology there is no method to measure the temperature change at this high a rate. Therefore, the only pratical approch is to consider the electric field to be constant with respect to time. 44 Also, the variation of the field within the volume is a function of the source of the radiation and the geometry of the cavity. When more than one source is used, it is not possible to theoretically determine the electric field- strength at a given point. Therefore, the electric field strength is assumed to be constant in each dryer zone. The Change in field strength with location is not as critical in a continuous drying tunnel in which the product is moving, than in a batch system in which the product is stationary. In the continuous system, the product is subject to a different rate of heating at each point. The product's movements prevent hot spots from occurring due to the microwave field strength variation.- 2n: a batch system the product maybe subjected to hot spots which may occur due to the stationary position of the microwave field. The third assumption is that all the energy of the electric field enters the material. Under the prevous assumption Of a constant microwave field in each dryer zone, it is not possible to determine the direction Of incident radiation. Therefore, the ratio of energy reflected and transmitted cannot be determined. Since the main Objective of this study is to find the relationship between the energy transmitted to the product and the drying rate, the assumption that all of the energy generated enters the product is justified. A quantification of the reflected energy is required to size new magnetrons for the dryer, but 45 the absolute strength Of the field and the amount of energy reflected are not critical to this study. Seven rectangular containers of a microwave transparent material were filled with water and placed inside the microwave tunnel. The container size is 29 x 59 cm with a depth of 8 cm. The containers nearly filled the tunnel area. A Closed cylinder filled with water was placed inside each container. The cylinders acted as a control volume tO determine whather a phase change has occurred within the containers. The temperature of the containers and the cylinders are measured at set time periods. Several problems were prevented by using the large containers within the tunnel. They acted as a load in the microwave tunnel. Without a load inside the tunnel the magnetrons can be damaged. The containers averaged the variation in the electric field strength over the volume of the containers, and yielded an average value (see Figure 4.2.1). Thus, the tunnel length is separated into seven sections each Of which has a different field strength. The size of the containers, and the agitation received due to the moving belt, allowed a large amount of mixing to occur thereby eliminating hot spots. The small cylinders acted as a control volume. The cylinders were closed to prevent mass evaporation. The 46 cylinders were totally filled, thereby preventing evaporation. If a phase change occurred, the pressure inside the cylinder would have caused a rupture of the cylinder walls. This Check was used to determine if energy was lost by evaporation. The assumptions and approximations used to make a estimation of the microwave field strength available are Open to question. The results of the field strength tests must be evaluated on a order of magnitude approximation. 4.3 PRODUCT DIFFUSIVITY MEASUREMENTS The second Objective Of the experiments was to analyze the resistance of the product to heat and mass transfer. The terms which need to be determined experimentally are the thermal and the mass diffusivity. The heat flow inside a solid with constant thermal properties and without internal convection, heat generation, or shrinkage is described by Geankoplis (1983) as: 3T --—- - at 02 T 4.3.1 at in which, K “t - ---- 4.3.2 PCP 47 where, “t Thermal diffusivity of product (mz/Sec) Thermal conductivity of product (watt/m °C) 8 I cp - Specific heat (J/kg oC) Density of product (kg/m3) o I The thermal diffusivity is a measure of the product's resistance to the flow of energy. It can be estimated by measuring the temperature change within the solid over time. Holman (1976) described the one dimensional case using a backward difference method: (Vxlz (t+atTn - tT“) at - -—-- . 4.3.3 Vt (tTn—l - tTn) - (tTn - tTn+1l where, at - Thermal diffusivity of product (mz/sec) tTm - Temperature at node m in time step t (C0) The specific heat of a product is a measure of the ability to store thermal energy. Heldman (1974) proposed the following relationship between the effective specific heat and the product moisture content as: CD - cp' Mdb + cpf ( 1 - Mdb ) 4.3.4 cp - Specific heat of product (J/kg Co) cpw - Specific heat of water (J/kg CO) Cpf - Specific heat of dry product (J/kg/C°) 48 "db - Moisture content of product (% dry basis). Using equations 4.3.2 and the measurements of product temperature, the thermal conductivity Of the product can be calculated. The moisture diffusion inside a bed of product is described by Geankoplis (1983) as: an -——— - a. v2 M 4.3.5 at where, am - Mass diffusivity of product (mz/sec) M - Moisture content of product (% dry basis) The mass diffusivity is a measure of the resistance to moisture movement within a solid. The effective mass diffusivity is an average of the resistance to mass movement in both the liquid and vapor phases. In order to use equation 4.3.5, information describing the effective mass diffusivity’ Iq- of the product must be determined. It is Obtained using a one dimensional finite difference analysis of the moisture profile within the bed at known points at discrete time steps. 49 Geankoplis (1983) described the one dimensional case BS: (V*)z (t+at"n - tun) a. - ———- . 4.3.6 Vt (tHn—I - tnn) ‘ (tfln - tfln+11 where, tum - Moisture content at node m in time step t (% dry basis) An Arrhenius relationship based on product temperature is used to describe the mass diffusivity during the drying process as a function of product temperature (Liou 1982)(see equation 3.2.1.6). 4.4 DATA COLLECTION The third Objective is to validate the assumptions used in the model. In order to accomplish this objective, several sets of product temperature and moisture content were recorded under different experimental conditions. The data collection procedure consists of two parts. The first part provides information on the conditions inside the tunnel; the second part gives information on the change in temperature and moisture content of the product inside the tunnel. 50 The parameters controlled in the tunnel are the product residence time and the microwave power level. The parameters which constitute the environment inside the tunnel are a function Of the product conditions. The accuracy with which the initial and boundary conditions were measured are: 1. Supply air temperature 1 2.0 ( °C ) 2. Exhaust air temperature t 2.0 ( °C ) 3. Supply air humidity i 1.0 (% RM) 4. Exhaust air humidity i 1.0 (% RH) 5. Product mass flow H- 0.5 (lb/min) The second part Of the data collection centered around the determination of the rate Of energy absorption and the rate of moisture loss. Removing a sample of product from the tunnel without disrupting the system proved to be difficult. The method selected consisted of allowing the process to come to steady state, stopping the system, and removing the samples. This was the only practical method of taking samples without the risk of high levels of microwave energy escaping from the tunnel. There is an error associated. with the sampling technique. It is caused by the time required to take the measurements. The time required to make a set Of 51 measurements was about 20 seconds. In order to minimize the time, a number of people were required. The error associated with product temperature loss is larger than the moisture content loss, so three temperature measurements are made at each point. As soon as the conveyor stopped, the tunnel was Opened and the product samples were removed. Measurement of the product temperature were made using an Omega (Omega Instrument Corporation, Stanford, CT.) infrared radiation detector. The average product moisture content was measured using the AOAC (Association Of Analytical Chemists, 1975) two stage moisture determination method. The system used to supply the product to the microwave dryer is outside the scope Of this study and will only be described in broad terms. The product enters the microwave tunnel at the conditions shown in Table 4.4.1. TABLE 4.4.1 PROPERTIES OF PELLETS AT INLET OF DRYER INITIAL TEMPERATURE 98.0 (r0 ) INITIAL NOISTURE CONTENT 28.2 ( 3 db INITIAL DENSITY 44.9 (lb/ft ) SPECIFIC HEAT DRY FOOD 0.4 (BTU/lb-F) CRITICAL MOISTURE CONTENT 18.0 ( 3 db) SURFACE AREA 4.2 (ftz/lb) PROTEIN 0.0012 (lb/in3) CARBOHYDRATE 0.0093 (lb/in3) SODIUM 8.2 E-6 (lb/in3) POTASSIUM 6.9 E-6 (lb/in3) DIETARY FIBER 2.0 E-3 (lb/in3) 52 CHAPTER 5 THEORETICAL Describing the drying process with a mathematical model provides a method of analyzing the basic relationships Of the process. In this section, the mathematical model is described for the drying of a two dimensional food product. The heat flow in a two-dimensional solid bed without internal convection, expansion or shrinkage is described by Geankoplis (1983) as: ———— I at ( ----- '0' ----- ) ‘8' ----- 5.1.]. where, “t - Thermal diffusivity Of product (mz/sec) T - Product temperature (°C) q - Microwave heat generation (watt) Cp - Specific heat of product (J/kg oC). The boundary condition for the heat transfer at the outer surface is given by Fortes and Okos (1980) as: an at where, Jq - Heat flux at the outer surface ( watt/m2) ht - Convective heat transfer coefficient (watt/mz- C0) 53 Ts - Temperature of the outer surface (C0) Ta - Dry bulb temperature of the air (C0). Lv - Latent heat of vaporization (J/kg H20) The convective heat transfer coefficient is defined by Holman (1976) as: Nu Ka ht - 5.1.3 X where, Nu - Nusselt Number (dimensionless) Ka - Thermal conductivity of the air (watt/m- °C) x - Distance air has moved over bed (m). The Nusselt number is a dimensionless number and relates the convective heat transfer coefficient ht to the thermal conductivity of the air. The Nusselt number is a combination of two dimensionless numbers, the Prandtl number and the Reynolds number. Holman (1976) gives the following relationship for a flow over a flat plate: Nu - 0.332 Prl/3 Rel/2 5.1.4 where, Pr - Prandtl number (dimensionless) Re - Reynolds number (dimensionless). The Prandtl number relates the thickness of the hydrodynamic and thermal boundary layers. The Reynolds number expresses the dimensionless flow velocity. 54 The initial conditions used in equation 5.1.1 are: Tn'o - To 5.1.5 where, TO - Temperature of the product as it enters the tunnel (Co)- 5.2 MASS TRANSFER The moisture transfer inside a two—dimensional solid food product is described by Geankoplis (1983) as: ---- - a. ( ————— + ----- ) 5.2.1 where, am - Mass diffusivity of product (mz/sec) M - Moisture content Of product (% dry basis). The boundary conditions at the outer surface of the product bed are: where, Jm - Moisture flux at outer surface (kg/m2). The convective mass transfer coefficient hm is defined by Geankoplis (1983) as: 55 n. - ———————— 5.2.3 where, 5h - Sherwood number (dimensionless) ca - Mass diffusivity of moisture in air (m/sec). The Sherwood number is a dimensionless relationship between the convective mass transfer coefficient and the product diffusivity. The Sherwood number is given by Holman (1976) 85: sh - 0.664 Rel/2 Sc 1/3 5.2.4 where, Sc - Schmidt number (dimensionless). The Schmidt number relates the thickness of the hydrodynamic and mass transfer boundary layers. The Schmidt number is defined as: Sc - ------- 5.2.5 9 “a where, u - Viscosity of the air (kg/m sec) p - Density of the air (kg/m3) “a - Diffusivity of the air (m/sec). The initial conditions used in the solution of equation 5.2.1 are stated as: 56 where, Mo - Moisture content of the product entering the tunnel (% dry basis). A steady state condition inside the tunnel is assumed. The product temperature rise and moisture decrease are calculated at each point in the tunnel. This procedure has the effect of tracking the change in product temperature and moisture of the food product in the tunnel. 5 . 3 NUMERICAL METHODS The differential equations for the heat and. mass transfer are solved using the Finite Element Method (FEM) (Segerlind 1978). The FEM is a widely used numerical procedure for solving differential equations. The method has become the standard technique to solve problems in structural mechanics, heat transfer, and fluid flow. Program Oven, a generalized FEM oven - simulation program was employed. Figure 5.3.1 gives a description of the basic steps used by the program to estimate the Change in product temperature and moisture content. 57 H.m.m meon zoou zomoomm Subroutine Airchr calculates the air characteristics, at the start of each time step. The air characteristics include the temperature, absolute humidity, dewpoint temperature, thermal diffusivity, and mass diffusivity. The free surfaces are identified. and the ZReynOlds number, Nusselt number and PTandtl number, Schmidt number and Sherwood number are calculated for each element with a free surface by Subroutine Number. Next the product heat and moisture transfer Characteristics are calculated by subroutine Kscalc. This includes the product density, specific heat, thermal conductivity, and moisture diffusivity. The convection and evaporation terms are calculated for each free edge in subroutine Conv. The pointers are calculated to determine how the data vector is broken down into its component parts. The size and breakdown of the data vector is needed for subroutine Iterat. Iterat is the controlling subroutine in a general (FEM) solution technique. Subroutine setstp calculates the maximum time step which can be used from the stiffness and capacitance matrices for 59 the smallest element. The procedure is executed once for heat and once for moisture transfer and the minimum value is taken as the time step. This step eliminates the possibility of numerical instability. In subroutine Di Sp the attenuation of the microwave field strength is calculated for each element based on the minimum distance from a free surface, and the microwave attenuation a. The heat generated at the center of each element due to the microwave radiation is calculated in subroutine Femsor. The product moisture content is calculated by Iterat. The evaporate cooling at the surface and the energy transported into and out of each element by the moisture movement are calculated by the subroutine couple. The energy transferred into and out of each element is added to the source term and the product temperature is calculated by Iterat. The model consists of two coupled differential equations (ie 5.1.1, 5.2.1) with a microwave heat source and the associated initial and boundary conditions. A two- dimensional grid is used to describe the food product bed to 60 be dried in both the conventional convection dryer and the microwave - convection dryer. A listing Of the computer program is contained in Appendix C. The solution technique is based on a finite difference technique in time using a central difference method. The method is inherently stable, due to dynamic adjustment in the time step based on the method described by Segerlind (1976) A product residence time Of 128 seconds requires 322 seconds of computer time on a Digital Vax 8700. 61 CHAPTER 6 RESULTS AND DISCUSSION In this chapter the experimental and simulation results are presented and discussed. The experimental results are divided into three parts: 1. Tests to determine the microwave field strength. 2. Experiments to determine the product thermal and mass diffusivity. 3. Comparison of the experimental and simulated data to evaluate the simulation model. 6.1 MICROWAVE FIELD STRENGTH MEASUREMENTS The technique used to measure the microwave field strength is indirect in nature. The amount of energy absorbed by' a material Of known thermal and dielectric properties is used to calculate the intensity of the microwave field. Currently, there are In: reliable instruments on the market which measure the field strength. Table 6.1.1 presents the change in temperature of water in the containers and cylinders for zones A through G of the microwave tunnel (see Figure 4.2.1). 62 TABLE 6.1.1 MICROWAVE FIELD STRENGTH MEASUREMENTS CONTAINER A B C D E F G INITIAL CONTAINER TEMPERATURE 0C 19.1 19.5 19.1 19.3 19.3 19.6 19.4 INITIAL CYLINDER TEHPERATURE °C 19.1 19.5 19.1 19.3 19.3 19.5 19.4 FINAL CONTAINER TEMPERATURE °C 19.3 20.4 32.4 22.3 30.2 31.2 20.5 PINAL CYLINDER TEMPERATURE °C 19.3 20.5 31.8 22.2 30.4 31.2 20.5 TENPERATURE RISE CONTAINER °C/sec 0.2 0.9 13.3 3.0 10.9 1.6 1.1 TEMPERATURE RISE CYLINDER °C/sec 0.2 1.0 12.7 2.8 11.3 1.7 1.0 Water has a penetration depth of 1 - 2 cm, therefore all of the energy gained by the water is absorbed at the outer surface of the. container. The microwave field strength calculated will be assumed to be available at the outer surface. The temperature Of each container is measured at three locations with a RTD thermocouple after 120 seconds of exposure to the microwave field. Approximately 30 seconds after the completion. of the test, the three temperature values were averaged to approximate the temperature rise Of the container. 63 The basic assumptions made are, that the field strength is constant over the container volume and that all of the microwave energy generated enters the water bath. These approximations are necessary to estimate the field strength inside the tunnel. The data in Table 6.1.1 shows that zones 3 and 5 exhibit an increases in water temperature of more than 10 °C. Zones 2, 4, and 6 experience an increase in water temperature between 1.0 and 2.9 °C. The two levels Of water temperature indicate that there are two levels of microwave energy within the tunnel. The increase in water temperature is used to calculate the microwave field strength. The containers function as a load inside the microwave cavity. The load eliminates the variation in field strength due to non-uniform loading. The cylinders function as the control volume to measure the rise in water temperature. The difference between the water temperature in the cylinders and the containers is due to one Of three possible reasons. The first is the variation in the microwave field. The assumption was made that the field strength is constant inside each container. The second possible source of error is that evaporation occurred only in the large containers. The third source of error is non-uniform mixing occurring inside the tunnel; the temperature inside any 64 container may not have been uniform through out the container. In most cases, the differences are small enough to be assigned to experimental error. TABLE 6.1.2 MICROWAVE FIELD STRENGTH CALCULATIONS CONTAINER A B C D E F G CYLINDER ENERGY GAIN Cal 0.2 0.8 10.2 2.3 8.8 1.4 0.9 POWER DISSIPATED CONTAINER Watt 2.3 11.4 145.2 33.2 126.9 19.5 12.6 MEASURED FIELD STRENGTH v/I 9.9 23.4 83.4 39.8 77.9 30.5 24.5 Table 6.1.2 presents the energy calculations used to determine the field strength. The data in Table 6.1.2 shows that zones 3 and 5 have a microwave field strength of between 83.38 to 77.95 v/m. Zones 2, 4, and 6 have a microwave field strength between 23.4 to 39.84 v/m. The range in microwave field strength and the variation in the heating seen in the zones are due to the location of the magnetrons. The magnetrons are located in zones 3 and 5 (see Figure 4.1.1). The heating in zones 2, 4, and 6 is due to microwave energy escaping from zones 3 and 5. The value of microwave field strength is calculated for each container, and constitutes an average over the volume occupied by the container. The field strength values are to 65 be used as a guide in determining the energy input to the product. The field strength information is used to calculate the amount of heat being added by the microwave field; therefore, the accuracy required in using the field strength data is critical. The accuracy of the data is based on the assumptions made in designing the experiments. Appendix A contains the measured data and calculated values of the microwave field strength for each of the five test runs. The data verifies how repeatable the measurements are. TABLE 6.1.3 COHHINED MICROWAVE FIELD STRENGTH BY POSITION SECTION A B C D E F G AVERAGE FIELD STRENGTH 10.2 33.5 87.1 61.7 86.1 37.1 14.7 HINIHUH FIELD STRENGTH 0.0 2.4 5.9 4.5 3.1 8.4 12.0 NAXIHUH‘ FIELD STRENGTH 21.6 37.4 93.5 67.5 89.7 52.9 35.8 STANDARD DEVIATION FIELD STRENGTH 6.9 2.4 5.9 4.6 3.1 8.4 4.4 Table 6.1.3 represents a tabulation of the calculated field strength for each of the five experiments. The 66 average, the range and the standard deviation values are given. The data show that the measured field strength can be replicated to within 15% of the full range value. 6.2 PRODUCT THERNAL AND HASS DIFFUSIVITY HEASUREHENTS The second goal of the experimental investigations is to obtain the thermal and mass diffusivity of the product being dried. Table 6.2.1 contains the data of the product temperature measurements ‘which occur inside a (controlled convection dryer at discrete points inside a block of cooked cereal shown in Figure 6.2.1 TABLE 6.2.1 TRANSPORT NEASURENENT BLOCK PRODUCT TEHPERATURE GRADIENT DATA BLOCK SECTION TIEE A B C D E F (sec) (deg F) (deg r) (deg r) (deg r) (deg r) (deg I) 600. 106. 87. 86. 86. 86. 86. 1200. 113. 89. 86. 86. 86. 86. 1800. 117. 92. 87. 86. 86. 86. 2400. 120. 95. 87. 86. 86. 86. 3000. 122. 98. 88. 86. 86. 86. 3600. 123. 100. 89. 87. 86. 86. 4200. 125. 102. 91. 87. 86. 86. 4800. 126. 104. 92. 87. 86. 86. 5400. 127. 106. 93. 88. 87. 86. 6000. 127. 107. 94. 88. 87. 87. 6600. 128. 108. 95. 89. 87. 87. 7200. 128. 109. 96. 90. 87. 87. 7800. 129. 111. 98. 90. 88. 87. 8400. 129. 111. 99. 91. 88. 88. 9000. 130. 112. 99. 92. 89. 88. 67 fiNd mmbo: \ \ \ \ \ \ \ \ \ \ 1nnlr| 1\| r ._\\ L1....x..x..|+1|;\|\..|e \ \ \ T \ \ \ _ \ _ \ _ \ . \ . \ \ \ ‘ ‘ llllll 1|llT|I1lllTll1lllTll1TllflliIllflli. 5 RE v3.0.5 hzwzwmamcmx hmommz¢mk 6Q The uniform block of product in figure 6.2.1 has one free surface exposed to the controlled air stream. The product temperatures are measured and are used to calculate the product thermal diffusivity and thermal conductivity. Table 6.2.2 presents the thermal diffusivity and thermal conductivity of the product calculated at each point at each time step. The thermal conductivity of the cereal product is calculated from the thermal diffusivity, density and specific heat values. The thermal conductivity of the product is not constant. Table 6.2.2 shows that a change occurs in the thermal conductivity from .356 to .789 BTU/sec-ft-Fo. This pattern is repeated for each of the tests and cannot be attributed to experimental error. The range in thermal conductivity is attributed to an evaporation front moving from the outer surface. The change in thermal conductivity cannot be separated from a density change as detailed data is not available. The product block used to determine the thermal diffusivity and thermal conductivity is used to calculate 69 the mass diffusivity if the product moisture profile through the block is measured at discrete points in time. Table 6.2.3 presents the moisture profile measured within the product block. TABLE 6.2.2 TRANSPORT MEASUREMENT BLOCK THERMAL DIFFUSIVITY AND CONDUCTIVITY CALCULATIONS TIME minutes 50 60 70 80 90 100 110 120 T BLOCK B deg r “t BL CK B ft /sec K BLOCK B B/s-ft—r T BLOCK C deg r 4! BL CK C ft /sec K BLOCK C B/s-ft-r T BLOCK D deg r “t BL CK D ft /sec K BLOCK D B/s-ft-r 98 .743 .356 .676 .361 86 .655 .356 100 .743 .353 89 .687 .362 86 .663 .360 102 .752 .351 91 .697 .362 87 .674 .363 104 .759 .350 93 .706 .362 .684 .367 106 .767 .349 94 .716 .363 88 .695 .371 107 .774 .349 95 .726 .363 89 .706 .374 108 .781 .349 96 .736 .363 90 .717 .377 109 .789 .350 98 .748 .365 90 .727 .381 70 TABLE 6.2.3 TRANSPORT MEASUREMENT BLOCK PRODUCT MOISTURE GRADIENT DATA BLOCK SECTION TIME A B C D E F (sec) (% Vb) (% vb) (% vb) (% vb) (4 vb) (% vb) 600. 18. 28. 28. 28. 28. 28. 1200. 14. 26. 28. 28. 28. 28. 1800. 12. 23. 28. 28. 28. 28. 2400. 11. 21. 27. 28. 28. 28. 3000. 10. 19. 27. 28. 28. 28. 3600. 10. 17. 25. 28. 28. 28. 4200. 9. 16. 24. 28. 28. 28. 4800. 9. 15. 23. 27. 28. 28. 5400. 9. 14. 22. 27. 28. 28. 6000. 9. 13. 21. 26. 28. 28. 6600. 8. 13. 20. 26. 28. 27. 7200. 8. 12. 19. 25. 27. 27. 7800. 8. 12. 18. 24. 27. 27. 8400. 8. 12. 18. 24. 27. 27. 9000. 8. 11. 17. 23. 26. 27. Table 6.2.4 presents the mass diffusivity of the product in each section of the product bed at each time step. The mass diffusivity of the cereal product is calculated from the product moisture data and equation 4.3.5 (Figure 6.2.1). The mass diffusivity is calculated and then used to develop a overall mass diffusivity which is a function of product temperature. The mass diffusivity is modified by the local product temperature in accordance with equation 3.1.6. Coefficients A and B are developed from a least square fit of the product moisture measurement and the raw mass diffusivity. 71 TABLE 6.2.4 TRANSPORT MEASUREMENT BLOCK PRODUCT MASS DIFFUSIVITY CALCULATIONS TIME minutes 50 60 70 80 90 100 110 120 M BLOCK B 3 vb 18.7 17.1 15.9 14.8 14.0 13.3 12.7 12.3 “- BL K B ft /896 .812 .779 .262 .159 .116 .091 .077 .067 BASE B .502 .428 .130 .071 .048 .035 .028 .023 M BLOCK C 3 vb 26.5 25.4 24.3 23.2 22.1 21.1 20.1 19.2 “- BLOCK C ftz/sec .074 .088 .102 .120 .144 .179 .237 .246 BASE C .079 .085 .092 .102 .115 .135 .168 .232 M BLOCK D 3 vb 28.1 27.9 27.6 27.3 26.8 26.3 25.7 25.1 “n BL CK D ft /sec .042 .051 .055 .060 .064 .069 .075 .082 BASE D .053 .057 .061 .064 .067 .071 .074 .077 6.3 OPERATIONAL DATA COLLECTION The third objective of the experimental section was to collect operational data which can be used to verify the 72 assumptions made in the model. The operational data can be divided into two parts; the oven conditions and the product conditions. The oven conditions for run 9 are presented in Tables 6.3.1. TABLE 6.3.1 MICROWAVE CONVECTION DRYER TUNNEL TEST RUNS TUNNEL OPERATING CONDITIONS RUN 3 9 Residence Air Air Food Air Field ‘--—* Tine Temp Humidity Width Velocity Strength Zone (sec) (r) (lbW/le) (ft) (ft/s) (v/m) 1 17.7 115. 0.005 2.0 2.0 6. 2 17.7 120. 0.015 2.0 2.0 24. 3 17.7 126. 0.025 2.0 2.0 95. 4 17.7 132. 0.036 2.0 2.0 35. 5 17.7 138. 0.046 2.0 2.0 79. 6 17.7 144. 0.057 2.0 2.0 29. 7 17.7 150. 0.068 2.0 2.0 12. The operational conditions shown in Table 6.3.1 contain the heat transfer supplied to the product by the microwave tunnel. The dead zone velocity represents the average velocity of the air above the product. The microwave field strength represent an average of microwave field strengths measured for each zone. The product temperature and moisture measurements are presented in Table 6.3.2. The data shown in Table 6.3.2 gives a history of the product temperature and moisture changes occurring inside the tunnel for run 9. The product temperatures exhibit a 73 range which average 97 13 F0. is dependent on measurement was made. The temperature measurements made in zone TABLE 6.3.2 the 6 had an average value of 170 123 F0. point at which The initial temperatures for run 9 PRODUCT TEMPERATURE AND MOISTURE CONTENT MEASUREMENTS MICROWAVE DRYING TEST RUN 3 9; NO EXTERNAL CONVECTION POSITION TIME TEMP 1 TEMP 2 TEMP 3 MOIST 1 MOIST 2 ft sec deg r deg r deg r % vb 3 wb 0.00 0.0 99.0 96.0 96.0 26.3 26.3 2.00 17.7 108.0 96.0 99.0 25.5 26.3 4.00 35.4 111.0 105.0 119.0 25.4 26.5 6.00 53.1 117.0 119.0 119.0 26.1 25.3 8.00 70.8 138.0 144.0 158.0 25.2 24.0 10.00 88.5 170.0 175.0 164.0 23.1 21.2 12.00 106.2 155.0 178.0 177.0 20.2 19.8 14.00 123.9 165.0 178.0 182.0 18.1 17.2 The range in temperatures can be attributed to two sources of error. The first source is the non-uniformity in the microwave field strength. If the field strength is more intense on one side of the tunnel, a range in product temperatures will be measured. The second source of error is the measurement technique (see section 4.3). The method used had a non-uniform time lag due to the time required to open the tunnel. This lag allows the product to cool, thereby increasing the temperature variation. The product moisture content measurements for run 9 in Table 6.3.2 also show a range dependent on position. The 74 minimum moisture range occurs near the inlet, and increases to a maximum of 1.9% at point 5. The range is small, but the accuracy of the moisture content data is critical. The product moisture content is used to calculate the pounds of water evaporated in the tunnel. Error in the moisture content has an influence on the calculated effectiveness of the tunnel at higher loads. The range in moisture content data can be assumed to stem from three sources. The product bed may not have been at a uniform moisture content when the measurements were made. Also, the non-uniformity in the microwave field may have induce a non-uniformity in the product bed. The second possible error in the moisture contents may be due to a change in the structure of the product. Pie (1987) theorized that moisture bound to the structure of the product changes and becomes free water. The technique used for moisture analysis does not detect bound water, therefore a transition between bound and free water is not detected. If the change in the product binding structure does occur, it will cause a range in product moisture. The most likely cause of the range in product moisture is the range in product temperature. If a range in product temperature is caused by the non—uniformity of the microwave field, then the product moisture will also be non-uniform. 75 6.4 COMPARISON BETWEEN EXPERIMENTAL AND SIMULATED DATA The objective of this section is to present a comparison between the experimental data and the simulated results. Each of the experimental data sets were compiled, and divided into two parts. The information about the tunnel operating conditions from runs 7 - 23 was used to describe the boundary conditions inside the tunnel. The measured product temperature and moisture content values from test runs 7 -23 were used to evaluate the performance of the simulation. Figure 6.4.1. presents the simulated product temperature, and product moisture content as a function of residence time in the tunnel for run 9. Superimposed on the curves are the experimental temperature and moisture profiles. Figure 6.4.1 shows three errors. The first is the slope of the temperature curve in zones 3 and 5. In these sections, the model is predicting a faster rise in the product temperature than the experimental data indicates. This is due to the method used to calculate the microwave field strength. 76 M Ameqooomv maze. .mo.mawmmms ” .owu .oun .93.. .90 an I .rL. _:_. a; . . . .. . 1.1,.» .2710; 1..-»; 0 H c. w . O m. n 0 Wm . 0 0 S Ir M . e. an mm 1 “TH M w H m r. v n _ .mu 3 T A T. «a . A.” L Mm. . T 3 w H 1...... WM... r3 1 T” TC TO. M .. mu w. u «a W n... m .0 T8 r Tm .. . m m m M o r. Tu W“ {\\ .TL 9.. M 0 TM \1 .n _ _. T m . . . .5 w. a n o a 4 es" :2. mm on w mmwma ZoHsom>zoo I m>¢komoH2 « mo m Ju W“ aZmaZoo mmDBmHo: 924 mmbsdmmmzma . ,l\ . Bobaomm 92.4452; 92¢ adezgummmxm 77 (szsuq :94 %) annmsxow HTOILHVd The field strength measured by a container may not necessarily correspond to the zones set up in the simulation. This is due to the volume inside the microwave tunnel used as a partial choke between the magnetrons. A container may be located between the high intensity field near a magnetron and low intensity field inside the choke. Therefore, a higher than actual microwave field strength may have been measured in zone 3, and a lower than actual field strength may have been measured in zone 4 (see Figure 4.1.1). The second error occurs due to the equilibrium at the surface of the product bed. The model predicts an instantaneous drop in the moisture content of the outer node to match the vapor pressure of the air. This results in a decrease in the average moisture content of the bed. The decrease is due to oversized finite elements used at the product surface. The number of surface elements causes a large drop in the average moisture content of the product bed. The inaccuracy of the FEM grid at the surface must be balanced against the computational time of refining the model at the surface. The third error may be caused by the values used for the product mass diffusivity and thermal conductivity. Note that the relationship between the product temperature and 78 mass diffusivity influences the slope of the moisture content curve. The relationship between the experimental data and the simulated values for run 9 as shown in Figure 6.4.1, are within the accuracy limits needed to evaluate the possible advantages to adding microwave into the energy drying operation, in the opinion of the author. Table 6.4.1 presents the conditions in the tunnel for run 23. The product load for run 23 was more than twice that of run 9. Therefore the FEM grid for the bed depth was changed. The new grid has the same shape and number of nodes but the distance between the nodes has been increased and new initial to compensate for the larger bed depth, conditions. TABLE 6.4.1 PRODUCT TEMPERATURE AND MOISTURE CONTENT MEASUREMENTS MICROWAVE DRYING TEST RUN I 23: EXTERNAL CONVECTION Residence Air Air Food Air Field ——-' Time Temp Humidity Width Velocity Strength Zone (sec) (F) (lbW/le) (ft) (ft/s) (v/m) 1 17.7 170. 0.012 2.0 5.0 6. 2 17.7 166. 0.019 2.0 5.0 24. 3 17.7 161. 0.026 2.0 5.0 95. 4 17.7 156. 0.033 2.0 5.0 35. 5 17.7 151. 0.040 2.0 5.0 79. 6 17.7 146. 0.047 2.0 5.0 29. 7 17.7 141. 0.055 2.0 5.0 12. 79 m.w.o ohsufim Aueqooemv.mslec.mozmo1mmms .ova .ow" .oo— F . .- wit..» ... f .1. O; .b. _,._ 1.. F ”a 4 zen pane sawmo Zomeom>zoo I m>¢komuHa 4 2H azmazoo mmpammo: n24 amps4mmmams Bopnomm ame¢gp=_m 92¢ aqazmaHmmmxm '00! '08! 'OIJ '09! '09! ’07! ‘08! ‘08! ‘01! '00! '00° ( d 38p ) EHHLVHHdNHL 3073803 038 '88 '98 ‘88 '08 '83 '98 '88 '08 80 f‘l'f‘T'TTT‘T'TTr‘I-I‘F 1 1'r‘r‘r-r1'1'rr1-‘rrrr1‘r1‘r1‘I-r11'm m7 ( 91339 19‘ %) annrsxon ETOILHVJ Figure 6.4.2 shows the relationship between the experimental and the simulated temperature and moisture content values for run 23. The same errors noted in run 9 are found in run 23. Runs 9 and 23 represent two levels of loading. Run 9 had a load of 6.03 pounds of dry product per minute, run 23 had 14.03 pounds of dry product per minute. The simulation represents the experimental adequately, considering the assumptions of microwave field strength which are used to calculate the microwave heating. The ability of the model to calculate the effect of a change in bed depth is obvious. TABLE 6.4.2 PRODUCT TEMPERATURE AND MOISTURE CONTENT MEASUREMENTS MICROWAVE DRYING TEST RUN 3 23; EXTERNAL CONVECTION POSITION TIME TEMPl TEMP2 TEMP3 MOIST1 MOISTZ ft sec deg F deg F deg F 3 wb 3 vb 0.00 0.0 90.0 96.0 96.0 26.3 26.3 2.00 18.3 93.0 97.0 97.0 34.2 34.2 4.00 36.6 103.0 97.0 101.0 34.0 33.7 6.00 54.9 135.0 122.0 127.0 34.2 34.0 8.00 73.2 152.0 145.0 144.0 33.2 33.5 10.00 91.5 149.0 153.0 159.0 29.9 30.7 12.00 109.8 170.0 160.0 163.0 29.9 29.2 14.00 128.1 172.0 167.0 163.0 28.1 29.4 The test conditions for runs 7 - 23 are compiled in Appendix B. 81 Reviewing the relationship between the experimental data and the simulated values over the entire range of test runs provides an indication of the performance of the simulation. The deficiencies of the model appear to be: 1. The exact value of microwave field strength in each zone of the tunnel is unknown. 2. The accuracy of the FEM grid at the outer surface is limited. 3. The relationship between the product temperature and mass diffusivity is inaccurate. Runs 9 and 23 represent the best fits between the experimental data and the simulated values. The other test runs represent a range in loading and initial conditions. The model is general enough to accept the variations, and produces reasonable results, in the opinion of the author. Figures 6.4.3 and 6.4.4 represent the average performance of the model. The model is not required to perfectly describe the relationship between microwave and convection drying, but to give an indication of the trend. It can then be employed for the design the next generation of dryers. 82 EXPERIMENTAL AND SIMULATED PRODUCT 140. ...!. .’ .2. we: 1.3;“ RESIDENCE TIME (seconds) TI TIST RUN 4 18 .1" 00 ‘ I e TEMPERATURE AND MOISTURE CONTENT IN A MICROWAVE — CONVECTION DRYER T“r‘T¢fi —r' T rfi“n" - - ‘OBI '08! ‘OLI ‘091 '091 (J 39p) HHHLVHHdEEI HOVJHHS GEE ‘ T” wT“T""‘T"’ f “T“T_r T ‘ 'c '071 '081 '08! '01! 'OOI '06 'T'T'TTYT‘TTT'T 'TI'TT‘I‘T‘IT l' T'Y'V'I‘I’Y‘I‘TTI'TT‘F V" TTTTT1 '98 '88 '08 '83 '93 '88 03 '81 '9! (SISFQ 191 %) ENflLSION HTDILHVd 6.4.3 Figure R? I. «.¢.m oHsuHH Amvnooomv. HHEHBS omozmonmmS .owH .omH ooH . rvazr. H. H -H. .H H r H H H--H H. H H: 0H H. H- H ¢HIHLIHH. r mi: NHL EOVEHOS GHH 111(111TW—r '081 'OZI ‘01! '001 I 'OII (d 39p) aanmvaaa [ I '091 I IW 091 1 ‘OLI ' I I fiT 'OBI ‘09! HH H Hem Hn-H ”HESS 20:05:80 I 52:28:— 4 Ho Hzmazoo $558: 52 mmpimmnfime Hopaomm ganja—Hm 92¢ .ngmamumxu '83 '98 88 ‘08 '08 '93 '83 '03 04 r'T'fT'I' [T T TTT‘I' T T'T‘I’T‘I'FT’TT'T'Y "1'1"!“ 1' m TT‘I'T‘I’Y’I'TT'] '07 (Ssqu 19A %) HHDLSION 3101i3Vd 6.5 USE OF THE MODEL The goal of this project was to determine if an accelerated drying rate can be obtained with the addition of microwave energy. The model is adequate to give an indication of the effect of microwave energy on the drying rate. The next step in this study is to remove the conditions used in the test runs, and to employ the model to fully analyze the use of microwave energy in the design of a new dryer. 6.6 ANALYSIS OF A CONVENTIONAL CONVECTION DRYER The experimental and simulated temperature and moisture content values in the convection dryer without the addition of microwave energy is presented in Figure 6.6.1. The operating conditions are shown in Table 3.2.1. The belt dryer is designed to heat the product at the outer surface, and to dry the product from the outer surface. Figure 6.6.1 shows how the product temperature and moisture content values change through the convection dryer. There is good agreement between the experimental and simulated values. In order to correctly model a three flight dryer, the model was subdivided into three sections. 85 ’2' ANALYSIS OF EXPERIMENTAL AND SIMULATED PRODUCT 600. 000. 1000. 700. zoo. 400. 500. RESIDENCE TIME (seconds) TEMPERATURE AND MOISTURE CONTENT 300. IN A CONVENTIONAL CONVECTION DRYER 100. 'TTijrFFrITI1frlrlf1lT!)TthTt1IT‘T11111111111111 000 ‘ ‘TT‘I'W‘F? T‘T‘I‘1 ’I‘T’TW’TT 1 ‘l I rTT‘F‘T‘T'T r-t-r—Trt c. 'OQI 'OII '081 '02! 'OII '001 '06 '08 (a 39p) HHHLVHHdNHI aovaans 038 TI"! ‘1' 1 ’l’ I’TT T T'T‘l’ 1"! 1' ['1’1' F11” YT T'I‘TI‘I‘T’TTT Tm 'EETTGZ '98 '78 ‘22 '02 '01 '91 '71 ”Z! ‘0! ( szqu 19M %) HHDISION ETDILHVd 6 .(3. 1 Figure OC The product temperature and moisture content gradients which occur at the outlet of the first flight are averaged, and used as input to the second flight. The averaging of the gradients inside the product bed approximates the product blending which occurs between dryer flights. At the end of the first, zone the product drops from the belt in the first zone to the belt in the second zone. The drop between zones is treated as a mixing step which averages the temperature and moisture gradients. The mixing action results in a discontinuity of the heating and the drying curves between the flights. This procedure is repeated between the second and third flights. The temperature and moisture gradients which occur at the end of the second and third zones are tabulated in Appendix D. The fit between the production data and the model predictions are best for the first two flights. The predicted product temperature is low for the third flight. This may be due to a change in the product characteristics or in the bed depth. The temperature at the end of the third flight cannot be measured in production, and was estimated. The predicted product moisture content is low in the third flight of the dryer. This is due to the error in the product temperature. 87 The results shown in Figure 6.6.1 match the current pellet dryer performance within the accuracy needed to evaluate design changes. The results shown in Figure 6.6.1 will be used as a reference to compare new designs of a microwave convection dryer with the conventional convection dryer. 88 CHAPTER 7 DESIGN OF A MICROWAVE CONVECTION DRYER In this chapter, the model discussed in chapter 6 will be used to evaluate new operating conditions of the microwave - convection dryer. Different combinations of microwave heating and convection heating are tested. The assumptions made in developing the model have been justified. 7.1 ADDITION OF MICROWAVE ENERGY IN THE FALLING RATE PERIOD In the falling rate period of the drying process, extensive energy is required to remove water. Applying microwave energy in this section, increases the rate of drying. Table 7.1 contains the conditions used in a microwave convection dryer during the falling rate period. TABLE 7.1 MICROWAVE CONVECTION DRYER OPERATING CONDITIONS MICROWAVE HEATING IN THIRD ZONE Residence Ai r Ai r Food Ai r Field Time Temp Humidity Width Velocity Strength Zone (sec) (F) (lbW/le) (ft) (ft/s) (v/m) 1 151.1 141. 0.085 4.0 12.0 0.0 2 260.9 141. 0.085 4.0 12.0 0.0 3 458.5 141. 0.085 4.0 12.0 25.0 89 The microwave — convection dryer described in Table 7.1 has the same size and product flow rate as the conventional system analyzed in section 6.6. The addition of microwave energy occurs in zone # 3. .As in section 6.6, the dryer model used three simulations to approximate the mixing and the change in bed depth from 5 to 10 cm between the first and third zones. Figure 7.1 presents the results of a simulation based on the conditions shown in Table 7.1. The product temperatures and moisture contents in the microwave - convection dryer are identical to the production system until the product enters the third zone. There, the rate of heating increases due to the application of microwave energy. This causes an increase in the rate of moisture removal. The final product moisture is 17.53 wet basis compared to 21.5% in the conventional convection dryer. Thus, the addition of microwave heating in the third zone causes a decrease in the final moisture content of 4 percent. 7.2 ADDITIONAL CONVECTION HEATING Results shown in section 7.1 that the addition of microwave heating in the third zone of the dryer causes an increase in the drying rate. Heat can be applied from any SOUICO . 90 H.H .HsuHH Amvnooomv m2; mozmaHmmm .Jmm 1J4»; “mm WHH.-Wmme.meH mmmm._wH : ”HM :.HH.. .9. '071 '021 '001 '0 '091 '081 ( J 89p ) HHHLVHHdNHL HDVJHHS GHQ '00? 33:3 02:53 zgaomgoo 44229szsz mb. lacs nanny H: I>zou I m>mbo quMma QHB . 2:8 .54 HH 283-58 3:. a :3 :3: .2 £382.. n o 6558 22.55380 I EZEBSH: mm>mbo 02:23 ausdgazm '93 '73 '0: I 91999 19‘ %) EHDLSION ETOILHVd MTV“?! '7" TT’I'I“? 99 The change in residence time can be used to decrease the size of the dryer flights from 10.1 (meters) to 2.32 (meters) if the width and bed depth remain the same as the convection system. An conservative estimate of the minimum length of the dryer flights is 3.1 (m). The microwave field strength of 75 (volts/meter) is within the standard operation of the Scan Pro microwave tunnel. This level is possible in a larger dryer if the cavity is designed to maintain that level. 100 CHAPTER 8 POST SCRIPT The results shown in chapter 7 are based on a model developed in chapters 3 through 6. There are a number of assumptions made in developing this model. The most critical assumptions involve the rate at which the microwave energy is transferred to the product. The assumptions made in chapter 6 to approximate the microwave field strength within the Scan Pro microwave tunnel are: The microwave field strength within the tunnel is independent of time. The microwave field strength in the tunnel is constant over the volume measured. All of the microwave energy generated is transmitted to the water bath. These assumptions were made to estimate the field strength within the tunnel, and represent a macroscopic approach to the problem and not a microscopic view. 101 CHAPTER 9 CONCLUSIONS The first objective of this study was to build a microwave - convection test unit to develop the basic information needed to develop a mathematical model of the process. Figure 4.1.2 illustrates the test unit which was constructed to collect basic information required. Detailed experiments were undertaken to describe the amount of energy being transferred to the product by convective heating and microwave radiation. A mathematical model was developed to simulate the heat and mass transfer occurring as the product passes through the test unit. The simulation program OVEN was obtained and modified to incorporate microwave heating of the product. A verification step was undertaken to establish the ability of the model to predict the product temperature and moisture content within the test. A range of operational and initial conditions were used. Experimental measurements of the product temperature and moisture content were compared with the model predictions. The results show that the model is adequate agreement between the experimental and calculated data. 102 The model was used in Chapter 7 to test different operating conditions for a microwave - convection dryer and compare them to a conventional convection dryer. The model predicts reduction in the residence time for the microwave-convection dryer of between 50 and 4003 compared to a conventional convection system. The reduction in residence time is of the same magnitude as the reduction in residence time for pasta reported by Decareau (1986). The decrease in residence time is due to the addition of microwave heating in the first zone of the dryer. The range in residence time is due to the level of microwave heating applied in the first zone. The operational conditions shown in Table 7.4, a 450% reduction in residence time when a 75 v/m field strength is applied in zone one. The operational conditions in Table 7.3 show a 50% reduction in residence time when a 25 v/m field strength is applied in all zones. Additional levels of convection heating did not significantly affect the residence time required. The decrease in residence time can be used to increase the capacity of the current system, or to decrease the size of the dryer. A capacity increase of 95% can be expected if a high level of microwave heating is applied in zone one. The length of the flights can be reduced from 10.1 (meters) 103 to 3.1 (meters). A combination of change in bed depth and dryer length can be selected to obtain the desired capacity. The model has been used as a guide to develop engineering recommendations on the maximum capacity of a microwave-convection dryer. The accuracy of the model is within the limits needed to evaluate design changes to a drying system. The final results must be evaluated subject to the assumptions used to develop the model. 104 CHAPTER 10 RECOMMENDATIONS FOR FUTURE WORK 10.1 MICROWAVE FIELD STRENGTH The model developed assumes a uniform microwave field strength within each zone of the dryer. This assumption is in question as pointed out in section 4.2. Part of the error illustrated in Chapter 6 can be attributed to the non—uniformity of the microwave field. In the opinion of the author, the capacity increase shown in Figure 7.4 is not possible unless the microwave field strength is uniform inside the dryer. :I believe the manufactures of microwave equipment will need to develop methods to predict the field strength, and modify the design of the microwave cavity to match the product requirements. This process is currently ongoing, but outside of the microwave manufactures. 10.2 PRODUCT DIFFUSIVITY The technique used to measure the product mass diffusivity outlined in section 4.3 is acceptable when relatively low rates of heat and mass transfer are occur. Microwave convection drying is characterized by very high rates of both heat and mass transfer. Therefore, testing is required to verify the relationship between the product 105 temperature with the rate of moisture movement. In the opinion of the author, the thermal conductivity and the mass diffusivity are a function of the local product temperature and of local moisture contente In the model the mass diffusivity is assumed to be a function of the product temperature only. Thus, an expression for the moisture diffusivity as a function of product temperature and moisture content must be developed. 10.3 SURFACE EQUILIBRIUM The model does not accurately predict the conditions at the outer surface of the drying particle because of the coarseness of the finite element grid at the surface. This can be corrected with the addition of more elements. Several researchers ( Forets, and Okos 1980) who recommend a non equilibrium approach. Implementation of this method allows more accurate analysis of the movement of moisture to the outer surfaces of drying food products. Implementation of a non equilibrium approach is recommended. 10.4 PRODUCT GEOMETRY The product geometry used in this model does not allow for internal convection 1J1 the product. bed. In order to include the internal convection, a method is needed to develop a pseudo conductivity - convection between the 106 particles which form the bed. A product bed with internal convection would move this model closer to representing the real world. Therefore internal convection should be added to the model of a microwave - convection drying of food particles in a belt dryer. 107 LI ST OF REFERENCES LIST OF REFERENCES A. O. A. C. 1975. .Association of Official Agricultural Chemists Handbook. AOAC, Inc., Washington, D.C. A. C. Nielsen Company 1986. Market Servey Ready-to—eat Cereal Consumption 1986 Summary. Volume 11, 1986 A. C. Nielsen Company 1987. Market Servey Ready—to-eat Cereal Consumption for the April - May Market Segment. Volume 4, 1987 Appleton, J. R., and Kwan, P. C. 1986 Proceedings of the Workshop on Microwave Applications in the Food and Beverage Industry. December 1986. Ontario Hydro Research Division, Toronto, Ontario, Canada. Bird, R. 8., Stewart, W. E, and Lightfoot, E. N. 1960. Transport Phenomena. Wiley and Sons, New York, N.Y. Brooker, D. 8., Bakker-Arkema, F. W., and Hall, C. W. 1981. Drying Cereal Grains. The AVI publishing Company Inc., Westport, CN. Copson, D. A. 1975. Microwave Heating. AVI Publishing Company, Westport, CN. 108 "1'1 .4. 1" k... Daniel, V. T. 1967. Dielectric Relaxation. Academic Press, New York, N.Y. Debye, P. W., 1929. Polar Molecules. Chemical Catalog, New York, N.Y. Decareau, R. B. and Peterson, R. A. 1986. Microwave Processing and Engineering. Ellis Horwood Ltd., Chelster, England. Englyst, H. N. 1988. New Concepts in Starch Digestion in Man. Proceedings of the Nutrition Society, V47, nl, p 64-75 Dunn Clinical Nutrition Center, Cambridge, MA. Finley, J. W. 1985. Chemical Changes in Food During Processing. AVI Publishing Company, Westport, CN. Fortes, M., and Okos, M. R. 1980. A Non- Equilibrium Thermodynamic approch to Transport Phenomena in a Capillary- Porous Media., Transactions of ASAE 1981. Geankoplis, C. J. 1983. Transport Properties and Unit Operations. Allyn and Bacon Inc. Newton, MA Goldblith, S. A. 1967 Basic Principles of Microwaves and Recent Developments. Ellis Howard Ltd., Chelster, England. 109 H“ r‘M "FISH; r“ Basted, J. B. 1973. Dielectric Properties and Molecular Behavior. Chapman and Hall, London, England Heldman, D. R. 1979. Food Process Engineering. AVI Publishing Company, Inc. Westport, CN Hirzel, R. W., 1982. Kellogg Company Advanced Technology Division, Personal Communications. Holman, J. P. 1976. Heat Transfer. McGraw - Hill Book Company, St. Louis, MO. Hoseney, R. C. 1986. Priniciples of Cereal Science and Technology; .American .Association of Cereal Chemists, St. Paul, MN. Level D. C., 1987. Kellogg Co. Communications Department Technical Services, Personal Communications. Liou, J. M., 1982. An .Approximate Method for Nonlinear Diffusion Applied to Enzyme Inactivation During Drying. .Agricultural University Wageningen, The Netherlands. LUIKOVA, A. V. 1966. Heat and Mass Transfer in Capillay- porous Bodies. Pergamon Press, London. 110 Metexas, A. C., 1974. Effect of Real and Imaginary Parts of the Dielectric Constant on the Performance of a Microwave Oven. J. Microwave Power 11(2),1OS-115. Metexas, A. C., 1979. Raido Frequancyf and Microwave Industrial Applications. 17th Universities Power Enginnering Conference, Research Center, Electrophysics Section, Manchaster, England Oda, S. J. 1986 Proceedings of the Workshop on Microwave Applications in the Food and Beverage Industry. December 1986. Ontario Hydro Research Division, Toronto, Ontario, Canada. PIE, T. J. 1986 Proceedings of the Workshop on Microwave Applications in the Food and Beverage Industry. December 1986. Ontario Hydro Research Division, Toronto, Ontario, Canada. Segerlind, L. J. 1973. Applied Finite Element Analysis. J. Wiley and Sons, New York, N.Y. Van Arsdel, W. E. 1984. Food Dyhdyration Volume 2, The Avi Publishing Company, Inc., Westport, CN. Von Ripple .A. R” 1954 Dielectric Materials and Applications. The Technology Press, Cambridge MA. 111 .‘T‘afllh‘m APPENDICES MICROWAVE APPENDIX A FIELD STRENGTH MEASUREMENTS MICROWAVE FIELD STRENGTH TESTS TEST # 1 INITIAL INITIAL FINAL FINAL TUNNEL TEMPERATURE TEMPERATURE TEMPERATURE TEMPERATURE SECTION CONTAINER CYLINDER CONTAINER CYLINDER A 19.10 19.10 19.50 19.50 B 19.50 19.50 20.40 20.50 C 19.10 19.10 25.90 26.10 D 19.30 19.30 23.00 23.20 E 19.30 19.30 26.20 26.20 F 19.60 19.50 21.80 21.90 G 19.40 19.40 20.50 20.50 CYLINDER ENERGY POWER FIELD DEL T / TIME GAINED DISIPATED STRENGTH TUNNEL UNIT / VOLUME WATTS SECTION cal V / M A 0.01 16.88 1.18 21.60 B 0.02 42.20 2.95 34.15 C 0.12 295.40 20.66 90.36 D 0.06 164.58 11.51 67.45 E 0.11 291.18 20.36 89.71 F 0.04 101.28 7.08 52.91 G 0.02 46.42 3.25 35.82 112 TEST * 2 INITIAL INITIAL FINAL FINAL TUNNEL TEMPERATURE TEMPERATURE TEMPERATURE TEMPERATURE SECTION CONTAINER CYLINDER CONTAINER CYLINDER A 19.50 19.50 19.80 19.50 B 20.70 20.50 22.90 22.90 C 26.20 26.10 40.20 41.10 D 23.40 23.30 29.40 30.20 E 26.40 26.30 40.10 40.00 F 21.80 21.90 23.70 23.90 G 20.50 20.50 20.60 20.60 CYLINDER ENERGY POWER FIELD DEL T / TIME GAINED DISIPATED STRENGTH TUNNEL UNIT / VOLUME WATTS SECTION cal V / M A 0.00 0.00 0.00 0.00 B 0.02 101.28 3.54 37.41 C 0.13 633.00 22.13 93.53 D 0.06 291.18 10.18 63.44 E 0.11 578.14 20.21 89.39 F 0.02 84.40 2.95 34.15 G 0.00 4.22 0.15 7.64 113 TEST # 3 INITIAL INITIAL FINAL FINAL TUNNEL TEMPERATURE TEMPERATURE TEMPERATURE TEMPERATURE SECTION CONTAINER CYLINDER CONTAINER CYLINDER A 19.70 19.50 19.80 19.70 B 22.40 22.40 24.10 24.10 C 37.90 37.40 51.10 51.80 D 29.20 30.10 35.90 36.20 E 37.80 40.00 51.60 52.10 F 24.20 24.00 26.20 26.20 G 21.10 20.80 21.30 21.30 CYLINDER ENERGY POWER FIELD DEL T / TIME GAINED DISIPATED STRENGTH TUNNEL UNIT / VOLUME WATTS SECTION cal V / M A 0.00 8.44 0.30 10.80 B 0.01 71.74 2.51 31.49 C 0.12 607.68 21.25 91.64 D 0.05 257.42 9.00 59.65 E 0.10 510.62 17.85 84.01 F 0.02 92.84 3.25 35.82 G 0.00 21.10 0.74 17.08 114 TEST # 4 INITIAL INITIAL FINAL FINAL TUNNEL TEMPERATURE TEMPERATURE TEMPERATURE TEMPERATURE SECTION CONTAINER CYLINDER CONTAINER CYLINDER A 20.80 20.60 21.00 20.70 B 24.10 24.20 26.00 26.20 C 45.40 45.60 55.80 56.60 D 34.90 34.40 41.20 41.40 E 47.90 47.50 60.20 60.20 F 26.40 26.80 27.90 28.90 G 21.80 21.80 22.10 22.10 CYLINDER ENERGY POWER FIELD DEL T / TIME GAINED DISIPATED STRENGTH TUNNEL UNIT / VOLUME WATTS SECTION cal V / M A 0.00 4.22 0.15 7.64 B 0.02 84.40 2.95 34.15 C 0.09 464.20 16.23 80.10 D 0.06 295.40 10.33 63.89 E 0.11 535.94 18.74 86.06 F 0.02 88.62 3.10 35.00 G 0.00 12.66 0.44 13.23 115 TEST # 5 INITIAL INITIAL FINAL FINAL TUNNEL TEMPERATURE TEMPERATURE TEMPERATURE TEMPERATURE SECTION CONTAINER CYLINDER CONTAINER CYLINDER A 20.50 20.50 20.90 20.70 B 25.70 25.90 27.50 27.50 C 50.60 51.20 61.30 62.10 D 40.40 40.60 45.40 45.60 E 56.60 56.80 68.10 68.20 F 27.40 27.80 28.90 29.10 G 22.00 22.30 22.40 22.30 CYLINDER ENERGY POWER FIELD DEL T / TIME GAINED DISIPATED STRENGTH TUNNEL UNIT / VOLUME WATTS SECTION cal V / M A 0.00 8.44 0.30 10.80 B 0.01 67.52 2.36 30.55 C 0.09 459.98 16.08 79.73 D 0.04 211.00 7.38 54.00 E 0.10 481.08 16.82 81.54 F 0.01 54.86 1.92 27.53 G 0.00 0.00 0.00 0.00 116 APPENDIX B MICROWAVE - CONVECTION DRYER TEST UNIT OPERATIONAL DATA COLLECTION AND MODEL VERIFICATION FEE-Its. 713-3.:bmonsufih -[J W Ananoommv male .mozmalmmma m .oWsHTF... 14.—NH _ “on.” .1.”er .:., ~11.-. #6”. .37...»- -— TH.“ m '03 'BI '92 '93 '08 (sysoq :94 %) HHHLSION HTOILHVd FT’T'T'TT'F T'TT“! 'FT’TTTT'T'T' f'rT‘! ’I’T‘I 'r'r Y 1 'I'F‘I’ TT'T'T'T'T‘I"! (J 39p) HHHLVHHdWHL HQVJHHS qaa 06 o m m. s A zpa puny um mmwmn zo~soa>zco I a>¢komoH= 4 mo rm m azmazoo ampemmo: 924 Hmaa4romoH2 d no 5209200 559202 024. 5:53.553... Bonn—om“. naa<fl§~m 024 AdazmfiHmmmNH (J 39p) HHHLVHHdNHL HOVJHDS 038 '88 '08 '88 '98 '88 ’08 'BI '91 f‘l’f‘TT’TTT'T'I'TT'l'T‘I’T T‘I’TT T‘TT’I'T‘I 'I 'T'I'T'I'TT‘I’T‘T‘T'TT‘H '98 (szqu 39M %) aanmsxom HTOILHVd ‘!'l4 0 ’0 D lbliilx'l...‘ ' 4.14- n‘lll!) IIIDIIL. i -_J 119 BETHLE- o rm.- -o mm... -..-....- -..-- :3.-- 3... - --1--..--..--- .msqooom. .:.. .mozma.mmm3 . 3.. .3. .93 . ” ”Er-L.-- . . .:.-1..., 3... . 0.... ..- .. 00... 3.. . . .-..-.-.,.-W-.r:..--r1fmu v T J ‘91 0 '8! '03 '82 EHHLVHHdNHL HOVJHDS 038 '92 '02 'ee (SISQQ 39A %) HHHLSIOH HTDILHVd o e. . zen pun. - $25 235528 I 5:28.: .q no m 5......on 3:922. oz... muss-Emma"... Sumo»... 3.3452; 95. gqszmaammxm |I_ Wh'flf'rITm‘rTTTEETT‘TTWrTW '98 (I 39p) 120 .mGBOOQ... .m2.. .mozma.mmmz .3; 6.3. .2: ;. .o FF1 3... :..tr,.. .3....»0 2,. .. 0.-.-. 9.1.27.0“ '91 '81 _111 {Ha-m: o..-H-H.Hu-H..m.---.-.----.--.:.-.-:-:--1.1:.--1.-.-.-.-...1.|-# . u M _ " w ” 8 d m H V “ nu m "H. . Tm . S .- 1... u n m 0 . no .r. n... . Thu r8 3 n V .10 n 3 h. N . Um. .- 0 m .13 I . m .” Mm . . d a . o m w 3 . V mm x) _ o . M .. % m 0 1 T . . O M O w Ah... m m“ M . B. t i. .0 \I/ .l ... p E q . H. . a... .:.... -.0 cm W”. M m 5...... 22.2.3.8 - 5:63... a .3 am 1. mm... m. . azmazOo amnammo: DZ... mmaafimmmzmb (\ . I.\ 905000.. QHBzco I H>4komoHS d ho azaazou ampsmmoa 02¢ HQDadMHmHHB subcomm nmadqpaum 02¢ AdazmflHmmmNm IDIIII-IIII v'l IOIIIVtOQIIOIlIiInID Iltlill‘l’I-Itlll’zi ll!’ I'll! '9: (51990. 191 %) HHHLSION 'IHOILHVd H 3 O m 0 S r. . fl mu Tn w o .. V mu 3.. O T T. H .r. .1: nm re Tm r. “.0 m fl 9 W mm .m o H h x H U rm V n _ .. L - _ 1T. w we 0. 1 Eu 8 o I H v.0 0 M; o o .m \\I w o F P mm o .w e . 3.28.5... . on W Tm. ( 'OBI 122 j}: -1m WM? @- M-nu NH-- 1:-1-1-1-1. :.....-1-..:.....-..--. 1.1-11 . Amvnoommv ESE... omoznmmmmmmom . .3.. ..2: 3.2 .8 .a 6 U... I 10 . VLF-#11. P .. -. . :1- .:.—.1» . L .% 0...- . r. w 3... r .- my. N1. .11. O 3 in: V . n . G n. H . _ m . L . 0 S 11 I _ .1. 0.. n1 3 m rm m n m . u . "V “W m hm O n. N . _ mu 3 r 0 fl «IT: 18 I . M w “1% L ”mu S _ m . . 3 n M m m m ”a H m . H n. Gnu . m Wm ”3 ) m u H- m u. z _ m w .M nu no. A m . . I nHu 1.0 a _ M TM 1. m . O . ) . w o M .11 P mm a. . . T8 9 . U u _ 2 . a... .:.: m 3 m S _ W was... 78.3528 1 “.5253; 4 z. W a. m m. . . 2530 $8.22. a: Enigma... . ( . (1 _ ' aobnomm nae<fl§~m 024 Adszgumummxm [— 123 hi3LEQI£ .’ Amvnooumv mama HUZMQHmmmg .ow. .om. .oo. .oo. PIP-11F.- Fl h11h15r11h-hl1:11—1 — .ee w — .e¢ .c u . . . .LeflF- 1. v I I '01! '00! (J Bap) HHHLVHHdNHL EOVJHDS G38 1 fii-lz - :1- 31-111.21.1w-wz-m- . .0-HQMw-m31-1-1: 1:1- -:.-:11--- . . . r l 1 1 I l '071 '081 '08! l '00! IV, 'OLI ‘09! I '09! #u ~ aha hunk . mahmn onpom>zoo 1 m>zou I HbdkomoHS 4 ho Wm H n . Bobaamm nfiadabfl—m 02¢ AdBZHHHMHmNH 1g . ‘ H u . r1111111111-11111111111131;-1-111111111111-11111111111111...--11111111--.-...-11111.11: 125 _1P1:3.1:3:3-.1.-1-.---3...11-bam 0.3:: .m ..1113113-_-11::11-! Amunooem. m2.. mozmn.mmm .3.“ Wow." .2: .3 .931 . .331 .3 .mu; rILtL: .3 .. . .1. r3 r3. . 1 3.: pa ‘ and Hans magma onsongzoo I m>¢komo2~ 4 .mo azuazoo HEP—.20: E: Huaadmfigaa sumac»; naszoo 1 m>¢komoHa 4 mo rm . azuazoo HmDam.o: nz¢ Hubsdmfimama . ponncmm QHBdAD:.m 024 Adazmammamxa [111111111111 1 11 1 (1 Bap) HHQLVHHaNHL HDVJHHS GHQ '93 '82 '02 ‘83 '38 (stvq 191 %) HHnLSIOH THOILHVd '93 '98 ['I'T'I'T‘I'Y'T T'TTT‘I‘FT‘I—T WYWWW '07 i ..J 128 Air---.--- cam-10.....-» Hum ..¢.00@.. .:.. .mozmaHmmmg .ov. .9». .oo. .o r1r1+1L-.1r1r-...1 1»- 1r .. ..3 r h .1 r. r .1.m11k1. .1.1171~1L1 F“7“f‘1"“T“1“_V'1"V‘T“”F‘T“ ‘09! '09: '0‘“ ‘08! '02! ‘OII '00! (J 36p) SHHLVHEdWHL HOVJHHS aaa ‘OLI “I“? 081 cm . zpm pun. mawma Zo.eom>zoo 1 m><30mo.: ¢ ho Bzmezoo mmpam.c= n24 mmaa¢mmm2ma summon. amyzoo I m>¢komuH2 « mo Bzmpzoo ampemHoE 92¢ mm394mmm2mb Bunaomm QHBzoo I m><¢omuHS 4 ho Bzmszoo HMDBmHoz mz< amba