"Applied Science, Faculty of"@en . "Chemical and Biological Engineering, Department of"@en . "DSpace"@en . "UBCV"@en . "Dhanjal, Sanjiv K."@en . "2009-07-27T19:20:09Z"@en . "2001"@en . "Master of Applied Science - MASc"@en . "University of British Columbia"@en . "Rotary kilns are some of the largest pieces of equipment found in industry. Their resilience\r\nand versatility makes them suitable for a wide variety of applications, such as the\r\ncalcination of lime and alumina, reduction of iron ore, cement clinkering and solid waste\r\nincineration. Common to all of their uses is the need to thermally process a material,\r\ntypically particulate, by directly firing a fossil fuel. A kiln consists of a refractory lined\r\ncylindrical furnace with its axis at a slight angle to the horizontal. Raw material is fed at\r\nthe upper end and, by virtue of the kiln's rotation, it travels along its length. The burner\r\nis located at the lower end. As such, the kiln is a counter current heat exchanger, with\r\nhot combustion gases flowing upwards over a bed of particulate material traveling down.\r\nTo ensure efficient and quality processing, all of the material in a kiln needs to be\r\nheated up to temperature, without any of it overheating. This requires efficient heat\r\ntransfer from the combustion gases to the bed and efficient heat transfer within the bed,\r\nto ensure all of the material is evenly heated. However, this is not always achieved. For\r\nexample, it has been found in industry that fine material often remains unreacted. The\r\nreason has been attributed to the phenomenon of segregation, whereby the fines collect\r\ntowards the center of the bed and so are shielded from the hot combustion gases.\r\nExperiments on laboratory kilns have confirmed that segregation in the cross section\r\nof a kiln occurs rapidly. Within a few revolutions, the fines collect in the middle of the\r\nbed forming a core, often referred to as a \"kidney\". Since a kiln typically rotates at 1-2\r\nrpm and material residence time is 1-4 hours, a bed that contains fines will be segregated\r\nfor almost all of its time in the kiln. The fines will not be exposed to the hot gases\r\nand so may not react. On the other hand, a bed that contains mono-sized particles has\r\nbeen found to intimately mix. In such a bed all of the particles have an equal chance\r\nof being exposed to the hot gases. Since heat transfer within the bed is dominated by\r\nmaterial advection, the bed is likely to be isothermal. From this it has been presumed\r\nthat segregation is the cause of temperature non-uniformity in the bed cross-section.\r\nWhile there has been no experimental data published confirming this, modeling work\r\n\r\nhas concurred with the premise that segregation is the cause of temperature non-uniformity\r\nin the bed cross section. The aim of this study is to investigate this through experimentation\r\nand to determine what level of fines cause temperature non-uniformity in the cross\r\nsection of the bed.\r\nTo meet the study objectives a batch rotary kiln was designed. The furnace was lined\r\nwith a dense refractory that, when hot, was used as the heat source for the bed. This\r\nwould allow the heat rate to be calculated using radiation heat transfer theory. The\r\nfurnace was first heated using a gas flame. When it was hot, material was inserted and\r\n,the furnace rotated. The temperatures of the walls and in the cross section of the bed\r\nwere measured using thermocouples and a data logger.\r\nSand was selected as the bed material since it is inert. A number of sand mixes\r\nwith various levels of fines were made by using a sieve shaker. To quantify the level of\r\nfines the particle size standard deviation was calculated. This allowed the results to be\r\nextrapolated for other applications such as in industry where wide ranges of particle size\r\ndistributions are common.\r\nThe design of the furnace was carried out by the use of models written specifically\r\nfor this project. A one-dimensional model of the heat transfer between combustion gases\r\nand the furnace walls was written and used to ensure that the furnace could be heated\r\nusing a gas flame. It was then modified to ensure the walls could in turn heat up the bed\r\nmaterial. A two-dimensional model of the bed cross section was also written and used\r\noutput data from the one-dimensional model to determine whether the temperature profile\r\nin the bed cross section could be measured practically. The models were used to design\r\nthe pertinent furnace specifications. These were furnace I.D. = 400mm, sand bed depth\r\n= 100mm, lining = 70mm thick castable and insulation = 20mm thick blanket. These\r\nspecifications formed the basis of the furnace mechanical design and were not altered once\r\nthe experiments were started. Therefore, the models fulfilled their purpose.\r\nThe furnace was heated until the refractory hot face was 1100\u00B0C before the flame was\r\nextinguished. Sand was then inserted into it and the furnace rotated. Each test lasted for\r\nabout 20 minutes and the bed was heated up to about 800\u00B0C. A temperature profile was\r\nsubsequently obtained at select locations in the bed cross section. The number of bed\r\n\r\nthermocouples was limited to four, since they tended to agitate the bed, affecting heat\r\ntransfer. Since the locations were the same for each test, their range of temperatures was\r\nused to quantify the temperature non-uniformity in the bed. Using these temperatures,\r\ntogether with temperatures of the walls, the heat rate to the bed and corresponding heat\r\ntransfer coefficient were calculated for each run.\r\nIn all runs it was found that the spread of the temperatures in the bed was between\r\n100-200\u00B0C initially. The range dropped to less than 10\u00B0C after 20 minutes. The temperature\r\ndata was used to calculate heat transfer coefficients and these were in the range\r\n85-175 W / m 2K. A further interesting finding was that the rate of heat transfer across the\r\ncovered bed surface was 3-6 times the rate across the exposed bed surface. This highlights\r\nthe importance of conduction between the covered wall and particles adjacent to\r\nit. These data represent a significant contribution to the literature since very few studies\r\nare available at realistic kiln temperatures and for kilns with diameters above 0.25m.\r\nTo determine the effect of size distribution, the average range of temperature, heat\r\nrate to the bed and heat transfer coefficient were calculated for one minute time periods\r\nduring each test. The values were then plotted with the standard deviation of particle\r\nsize. Much scatter was observed in these plots and it appeared that size distribution had\r\nlittle effect on temperature uniformity in the bed and no effect on heat rate and heat\r\ntransfer coefficient.\r\nThe results were somewhat surprising, since segregation was evident during the some\r\nof the runs. Finer particles could be seen underneath the layer of larger particles, below\r\nthe bed surface. Therefore, for the size ranges studied segregation did not hinder heat\r\ntransfer across the bed. Perhaps heat transfer is a stronger function of overall material\r\nflow and effective thermal diffusivity, as opposed to segregation.\r\nThis study provides a substantial amount of data at realistic kiln temperatures, something\r\nthat is rare in the general literature. The data can be used for validating models\r\nof the heat transfer in the cross section of the bed, and so further understanding in this\r\narea."@en . "https://circle.library.ubc.ca/rest/handle/2429/11280?expand=metadata"@en . "10977337 bytes"@en . "application/pdf"@en . "E X P E R I M E N T A L S T U D Y OF T H E E F F E C T OF P A R T I C L E SIZE D I S T R I B U T I O N O N H E A T T R A N S F E R W I T H I N T H E B E D OF A R O T A R Y K I L N By Sanjiv K . Dhanjal B . Sc. (Chemical Engineering) University of Surrey, U K A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF CHEMICAL AND BIOLOGICAL ENGINEERING We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA November 2000 \u00C2\u00A9 Sanjiv K . Dhanjal, 2000 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of British Columbia Vancouver, Canada Department Date Qe-c, ^ DE-6 (2/88) Abstract Rotary kilns are some of the largest pieces of equipment found in industry. Their re-silience and versatility makes them suitable for a wide variety of applications, such as the calcination of lime and alumina, reduction of iron ore, cement clinkering and solid waste incineration. Common to all of their uses is the need to thermally process a material, typically particulate, by directly firing a fossil fuel. A kiln consists of a refractory lined cylindrical furnace with its axis at a slight angle to the horizontal. Raw material is fed at the upper end and, by virtue of the kiln's rotation, it travels along its length. The burner is located at the lower end. As such, the kiln is a counter current heat exchanger, with hot combustion gases flowing upwards over a bed of particulate material traveling down. To ensure efficient and quality processing, all of the material in a kiln needs to be heated up to temperature, without any of it overheating. This requires efficient heat transfer from the combustion gases to the bed and efficient heat transfer within the bed, to ensure all of the material is evenly heated. However, this is not always achieved. For example, it has been found in industry that fine material often remains unreacted. The reason has been attributed to the phenomenon of segregation, whereby the fines collect towards the center of the bed and so are shielded from the hot combustion gases. Experiments on laboratory kilns have confirmed that segregation in the cross section of a kiln occurs rapidly. Within a few revolutions, the fines collect in the middle of the bed forming a core, often referred to as a \"kidney\". Since a kiln typically rotates at 1-2 rpm and material residence time is 1-4 hours, a bed that contains fines will be segregated for almost all of its time in the kiln. The fines will not be exposed to the hot gases and so may not react. On the other hand, a bed that contains mono-sized particles has been found to intimately mix. In such a bed all of the particles have an equal chance of being exposed to the hot gases. Since heat transfer within the bed is dominated by material advection, the bed is likely to be isothermal. From this it has been presumed that segregation is the cause of temperature non-uniformity in the bed cross-section. While there has been no experimental data published confirming this, modeling work ii has concurred with the premise that segregation is the cause of temperature non-uniformity in the bed cross section. The aim of this study is to investigate this through experimen-tation and to determine what level of fines cause temperature non-uniformity in the cross section of the bed. To meet the study objectives a batch rotary kiln was designed. The furnace was lined with a dense refractory that, when hot, was used as the heat source for the bed. This would allow the heat rate to be calculated using radiation heat transfer theory. The furnace was first heated using a gas flame. When it was hot, material was inserted and ,the furnace rotated. The temperatures of the walls and in the cross section of the bed were measured using thermocouples and a data logger. Sand was selected as the bed material since it is inert. A number of sand mixes with various levels of fines were made by using a sieve shaker. To quantify the level of fines the particle size standard deviation was calculated. This allowed the results to be extrapolated for other applications such as in industry where wide ranges of particle size distributions are common. The design of the furnace was carried out by the use of models written specifically for this project. A one-dimensional model of the heat transfer between combustion gases and the furnace walls was written and used to ensure that the furnace could be heated using a gas flame. It was then modified to ensure the walls could in turn heat up the bed material. A two-dimensional model of the bed cross section was also written and used output data from the one-dimensional model to determine whether the temperature profile in the bed cross section could be measured practically. The models were used to design the pertinent furnace specifications. These were furnace I.D. = 400mm, sand bed depth = 100mm, lining = 70mm thick castable and insulation = 20mm thick blanket. These specifications formed the basis of the furnace mechanical design and were not altered once the experiments were started. Therefore, the models fulfilled their purpose. The furnace was heated until the refractory hot face was 1100\u00C2\u00B0C before the flame was extinguished. Sand was then inserted into it and the furnace rotated. Each test lasted for about 20 minutes and the bed was heated up to about 800\u00C2\u00B0C. A temperature profile was subsequently obtained at select locations in the bed cross section. The number of bed iii thermocouples was limited to four, since they tended to agitate the bed, affecting heat transfer. Since the locations were the same for each test, their range of temperatures was used to quantify the temperature non-uniformity in the bed. Using these temperatures, together with temperatures of the walls, the heat rate to the bed and corresponding heat transfer coefficient were calculated for each run. In all runs it was found that the spread of the temperatures in the bed was between 100-200\u00C2\u00B0C initially. The range dropped to less than 10\u00C2\u00B0C after 20 minutes. The temper-ature data was used to calculate heat transfer coefficients and these were in the range 85-175 W / m 2 K . A further interesting finding was that the rate of heat transfer across the covered bed surface was 3-6 times the rate across the exposed bed surface. This high-lights the importance of conduction between the covered wall and particles adjacent to it. These data represent a significant contribution to the literature since very few studies are available at realistic kiln temperatures and for kilns with diameters above 0.25m. To determine the effect of size distribution, the average range of temperature, heat rate to the bed and heat transfer coefficient were calculated for one minute time periods during each test. The values were then plotted with the standard deviation of particle size. Much scatter was observed in these plots and it appeared that size distribution had little effect on temperature uniformity in the bed and no effect on heat rate and heat transfer coefficient. The results were somewhat surprising, since segregation was evident during the some of the runs. Finer particles could be seen underneath the layer of larger particles, below the bed surface. Therefore, for the size ranges studied segregation did not hinder heat transfer across the bed. Perhaps heat transfer is a stronger function of overall material flow and effective thermal diffusivity, as opposed to segregation. This study provides a substantial amount of data at realistic kiln temperatures, some-thing that is rare in the general literature. The data can be used for validating models of the heat transfer in the cross section of the bed, and so further understanding in this area. iv Table of Contents A b s t r a c t ii List of Tables vi i i List of Figures x Acknowledgment xx i 1 Introduct ion 1 2 L i terature Rev iew 5 2.1 Material Flow 5 2.2 Segregation 7 2.2.1 Segregation Experiments 9 2.2.2 Segregation Modeling 10 2.3 Heat Transfer 12 2.3.1 Heat Transfer in the Free-Board 13 2.3.2 Heat Transfer in the Bed 15 3 Objectives and Scope of S tudy 22 4 Test Furnace Design 25 4.1 Bed Depth 26 4.1.1 Two-dimensional Bed Cross-Section Code 26 4.1.2 Boundary Conditions 30 4.1.3 Velocity Profile 30 v 4.1.4 Code Testing: 31 4.2 Furnace Refractory Design 33 4.2.1 Hot Gas Mass Flow and Temperature 35 4.2.2 Heat Up of Refractory by Gases 35 4.2.3 Heat Up of Bed by the Refractory Walls 36 4.2.4 Heat Transfer Within the Refractory Walls 37 4.2.5 Heat Transfer to Atmosphere 38 4.2.6 Implementing Boundary Conditions 38 4.2.7 Code Testing 39 4.3 Bed Depth and Furnace Sizing 39 5 Description of Experimented Set-Up 52 5.1 Furnace 52 5.2 Instrumentation 53 5.2.1 Refractory Temperatures 53 5.2.2 Bed Temperatures 54 5.3 Calibration 55 5.4 Preparation of Bed Material 55 5.5 Experimental Procedure 59 6 Results and Discussion 74 6.1 Recorded Data 74 6.1.1 Bed Temperatures 75 6.1.2 Wall Temperatures 77 6.1.3 Other Observations 80 6.2 Temperature Non-Uniformity 80 vi 6.3 Heat Transfer Rates and Coefficients 84 6.3.1 Calculation of Heat Transfer Rates and Coefficients 84 6.3.2 Results Obtained 87 6.3.3 Effect of Particle Size Distribution 89 6.4 Overall Comments 90 7 Conclusions and Recommendations 133 7.1 Conclusions 133 7.2 Recommendations For Future Work 136 A Refractory and Insulation Details 142 B Engineering Drawings of Furnace and Support 147 C Specifications of Instrumentation 151 D Sand Details 158 E Size Distr ibution Information For Sand Mixes 165 F Results: 184 vii List of Tables 4.1 2D Code Input Data for Test Case 49 4.2 ID Code Input Data for Test Case 49 4.3 Graphical Method Input Data for Test Case 49 4.4 ID Code. Test Case Results 50 4.5 Input Data - Heat Up of Furnace Using Flame Gases 50 4.6 Input Data - Heat Up of Bed Using Hot Refractory Walls 51 5.1 Calibration of Bed Data Logger 72 5.2 Details of Prepared Sand Mixes 73 6.1 Details of Test Runs 129 6.2 Measurements of Refractory Thickness During Installation of Thermocouplesl30 6.3 Sensitivity of Hot Face and Interface Temperatures to Changes in Refrac-tory and Blanket Thickness. Temperatures calculated using ID model after 120 minutes of Heat at lOOkW 131 6.4 Average Values for Temperature Range, Overall Heat Transfer Coefficient and Heat Rate 132 F . l Bed Temperatures at Select Times 260 F.2 Run 3 Wall Temperatures 268 F.3 Run 4 Wall Temperatures 268 F.4 Run 5 Wall Temperatures 269 F.5 Rim 8 Wall Temperatures 269 F.6 Run 9 Wall Temperatures 270 viii F.7 Run 10 Wall Temperatures 270 F.8 Run 11 Wall Temperatures 2 7 0 F.9 Run 12 Wall Temperatures 2 7 1 F.10 Run 13 Wall Temperatures 2 7 1 F . l l Run 14 Wall Temperatures 2 7 2 F.12 Run 15 Wall Temperatures 2 7 2 F.13 Run 16 Wall Temperatures 2 7 3 F.14 Run 17 Wall Temperatures 2 7 3 F.15 Run 18 Wall Temperatures 2 7 4 F.16 Run 19 Wall Temperatures 2 7 4 i x List of Figures 1.1 Schematic of Rotary Kiln 3 1.2 Rolling Bed Motion Depicting a Top Active Layer and Bottom Plug Flow Region 4 2.1 Modes of Bed Behavior in a Rotary Kiln 18 2.2 Rolling Bed Motion Depicting a Top Active Layer and Bottom Plug Flow Region 19 2.3 Heat Transfer Paths Inside a Rotary Kiln 20 2.4 Heat Transfer Paths Within a Packed Bed 21 4.1 Flux Entering and Leaving Finite Control Volume (Area in 2D case) . . . . 42 4.2 Depiction of Ghost Cell at i = / boundary 42 4.3 Velocity Profile in Two-Dimensional Grid 43 4.4 Boundary Conditions For Test of Transient Conditions: T ( z , y, 0) = x s'm(ny) 43 4.5 Boundary Conditions For Test of Convective Flux at Boundary 44 4.6 Simulation of Results 44 4.7 Comparison with Experiment .- 45 4.8 Refractory Model Structure 45 4.9 Cross Section of Test Furnace 46 4.10 Heat Up of Furnace - Slice 6 46 4.11 Heat Up of Bed - Slice 6 47 4.12 Heat Up of Bed Using 2D code 47 4.13 Temperature Profile in Center of Bed after 20 minutes 48 x 5.1 Test Furnace General Assembly 62 5.2 Experimental Furnace and Support 63 5.3 Location of Refractory Thermocouples 64 5.4 Instrumentation Set Up 65 5.5 Bed Probe and Data Logging Equipment 66 5.6 Dimensional Details of Bed Probe 67 5.7 Location of Probe Thermocouples in Bed Cross-Section 68 5.8 Temperature Indicated by Data Logger when Constant Millivolt Signal Applied 69 5.9 Coarse Sand Sample. 1/16-1/8 inch Filter Sand and Gravel 70 5.10 Fine Sand Sample. 10-20 Filter Sand 70 5.11 Relationship Between Rosin-Rammler Index, URR, and Normalized Stan-dard Deviation of Particle Size, a* 71 5.12 Furnace Filling Tube 71 6.1 Location of Thermocouples in Bed 93 6.2 Run 5 Bed Temperature Historj'. Mix 2 at 1 rpm (dp = 1.9mm, a* = 0.26, NRR = 7.3) 94 6.3 Run 8 Bed Temperature History. Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 95 6.4 Run 15 Bed Temperature History. Mix 5 at 0.43 rpm (dp = 1.2mm. a* = 0.22, nRR = 7.7) 96 6.5 Run 9 Bed Temperature History. Mix 5 at 3 rpm (dp = 1.2mm, a* = 0.22, TIRR = 7.7) 97 6.6 Run 11 Bed Temperature History. Mix 4 at 1 rpm (dp = 3.0mm, a* = 0.09. nRH= 15.5) 98 x i 6.7 Run 5 Bed Temperature Profiles. Mix 2 at 1 rpm (dp \u00E2\u0080\u0094 1.9mm, a* = 0.26, nRR = 7.3) 99 6.8 Run 8 Bed Temperature Profiles. Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, nRR = 7.7) 100 6.9 Location of Thermocouples in Refractory 101 6.10 Hot Face Refractory Temperature History of Furnace For Run 5 (Mix 2 at 1 rpm. dp = 1.9mm, a* = 0.26, nRR = 7.3) 102 6.11 Interface Refractory Temperature History of Furnace For Run 5 (Mix 2 at 1 rpm. dp = 1.9mm, a* = 0.26. nRR = 7.3) 103 6.12 Refractory Temperature History at Center of Furnace For Run 5 (Mix 2 at 1 rpm. dp = 1.9mm, a* = 0.26. nRR = 7.3) 104 6.13 Hot Face Refractory Temperature History of Furnace For Run 14 (Mix 5 at 1 rpm, dp \u00E2\u0080\u0094 1.2mm, a* \u00E2\u0080\u0094 0.22, nRR = 7.7) 105 6.14 Hot Face and Interface Refractory Temperature History at Center of Fur-nace For Run 5. As used in Heat Transfer Analysis. (Mix 2 at 1 rpm, dp = 1.9mm, a* = 0.26, nRR = 7.3) 106 6.15 Temperature Range of Bed Thermocouples B1-B4 for Runs 5, 10 and 11 . 107 6.16 Temperature Standard Deviation of Bed Thermocouples B1-B4 for Runs 5, 10 and 11 108 6.17 Temperature Range Divided by Temperature Standard Deviation of Bed Thermocouples B1-B4 for Runs 5, 10 and 11 109 6.18 Effect of Particle Size Distribution on Degree of Temperature Non-Uniformity at lrpm 110 6.19 Effect of Particle Size Distribution on Degree of Temperature Non-Uniformity at 2rpm I l l 6.20 Effect of Particle Size Distribution on Degree of Temperature Non-Uniformity. A l l runs 112 xii 6.21 Effect of Furnace Rotation on Temperature Non-Uniformity for Runs with Mix 5. ( dp = 1.2mm, a* = 0.22, nRR = 7.7) 113 6.22 Path of Particles in Cross Section 114 6.23 Heat Rate to Bed and Temperature Driving Force For Run 5. (Mix 2, (dp = 1.9mm, a* = 0.24, nRR = 7.3) 115 6.24 Heat Rate to Bed and Temperature Driving Force For Run 4. (Mix 1, (dp = 1.4mm, a*p = 0.29, nRR = 5.0) 116 6.25 Overall Heat Transfer Coefficient For Run 5. (Mix 2, (dp = 1.2mm, a* = 0.22, nRR = 7.7) 117 6.26 Overall Heat Transfer Coefficient For Run 4. (Mix 1, (dp = 1.4mm, a* = 0.29, nRR = 5.0) 118 6.27 Ratio of Heat Flow Across Covered Bed Surface and Exposed Bed Surface For Run 5 (dp = 1.2mm, a* = 0.22, nRR = 7.7) 119 6.28 Exposed Wall and Covered Wall Heat Transfer Coefficients For Run 5. (Mix 2, (dp = 1.9mm, a*p = 0.24, nRR = 7.3) 120 6.29 Sensitivity of Wall and Bed emissivity on Covered Wall to Covered Bed Heat Transfer Ratio. Run 5 Mix 2 at 1 rpm (dp = 1.9mm, a* = 0.24, nRR = 7.3) 121 6.30 Sensitivity of Bed Depth on Overall Heat Transfer Coefficient and Heat Rate to Bed. Run 5 Mix 2 at 1 rpm (dp = 1.9mm, a*p = 0.24, nRR = 7.3) . 122 6.31 Effect of Particle Size Distribution on Overall Heat Transfer Coefficient. Al l Runs, A l l Furnace Rotational Speeds 123 6.32 Effect of Particle Size Distribution on Heat Rate to Bed. Al l Runs, Al l Furnace Rotational Speeds 124 6.33 Effect of Particle Size Distribution on Final Bed Temperature after 20 minutes. Al l Runs 125 6.34 Relationship Between Temperature Range and Heat Rate to Bed. A l l Runs 126 xiii 6.35 Effect of Furnace Rotational Speed on Overall Heat Transfer Coefficient for Runs with Mix 5. {dp = 1.2mm, a* = 0.22, nRR = 7.7) 127 6.36 Effect of Furnace Rotational Speed on Heat Rate to Bed for Runs with Mix 5. (d p = 1.2mm, a* = 0.22, nRR = 7.7) 128 E . l Rosin-Rammler Distribution for Mix 1, Test 1 {nRR = 4.1, C = 0.95) . . . 166 E.2 Rosin-Rammler Distribution for Mix 1, Test 2 {nRR = 5.8, C = 0.95) . . . 167 E.3 Rosin-Rammler Distribution for Mix 2, Test 1 ( n ^ = 8.1, C = 1.00) . . . 168 E.4 Rosin-Rammler Distribution for Mix 2, Test 2 (nRR = 7.0, C = 0.98) . . . 169 E.5 Rosin-Rammler Distribution for Mix 3 {nRR = 11.7, C = 0.97) 170 E.6 Rosin-Rammler Distribution for Mix 4 {nRR = 15.6, C = 0.99) 171 E.7 Rosin-Rammler Distribution for Mix 5, Test 1 {nRR = 6.6, C = 0.99) . . . 172 E.8 Rosin-Rammler Distribution for Mix 5, Test 2 {nRR = 7.0, C = 0.99) . . . 173 E.9 Rosin-Rammler Distribution for Mix 6 {nRR = 11.8, C = 0.99) 174 E.10 Size Distribution for Mix 1, Test 1 {dp = 1.4, a* = 0.33) 175 E . l l Size Distribution for Mix 1, Test 2 {dp = 1.4, a* = 0.25) 176 E.12 Size Distribution for Mix 2, Test 1 {dp = 1.7, a* = 0.22) 177 E.13 Size Distribution for Mix 2, Test 2 {dp = 2.0, a* = 0.26) 178 E.14 Size Distribution for Mix 3 {dp = 2.4, a* = 0.17) 179 E.15 Size Distribution for Mix 4 {dp = 3.0, a*p = 0.09) 180 E.16 Size Distribution for Mix 5, Test 1 {dp = 1.2, a*p = 0.22) 181 E.17 Size Distribution for Mix 5, Test 2 {dp = 1.2, tr* = 0.21) 182 E . 18 Size Distribution for Mix 6 {dp = 1.3, a* = 0.15) 183 F . l Run 3 Bed Temperature History. Mix 2 at 1 rpm {dp = 1.9mm, a* = 0.26. nRR = 7.3) 185 xiv F.2 Run 3 Wall Temperature History. Mix 2 at 1 rpm (dp = 1.9mm, a* = 0.26, = 7.3) 186 F.3 Run 3 Exposed Wall to Exposed Bed, hew^eb, and Covered Wall to Covered Bed, hcw-+cb, Heat Transfer Coefficients. Mix 2 at 1 rpm (dp = 1.9mm, a*p = 0.26, n R R = 7.3) 187 F.4 Run 3 Overall Wall to Bed Heat Transfer Coefficient, h0, and Heat Flow, Qw.+b. Mix 2 at 1 rpm (dp = 1.9mm, a* = 0.26, nRR = 7.3) 188 F.5 Run 3 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 2 at 1 rpm (dp = 1.9mm, a* = 0.26, nRR = 7.3) 189 F.6 Run 4 Bed Temperature History. Mix 1 at 1 rpm (dp \u00E2\u0080\u0094 1.4mm, a* \u00E2\u0080\u0094 0.29, nRR = 5.0) 190 F.7 Run 4 Wall Temperature History. Mix 1 at 1 rpm (dp - 1.4mm, a* = 0.29, n R R = 5.0) 191 F.S Run 4 Exposed Wall to Exposed Bed, hew-+eb, and Covered Wall to Covered Bed, h cw->cbi Heat Transfer Coefficients. Mix 1 at 1 rpm (dp \u00E2\u0080\u0094 1.4mm, a*p = 0.29, n R R = 5.0) 192 F.9 Run 4 Overall Wall to Bed Heat Transfer Coefficient, ha, and Heat Flow, Qw->b- Mix 1 at 1 rpm (dp - 1.4mm, a* = 0.29, n R R = 5.0) 193 F.10 Run 4 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 1 at 1 rpm (dp = 1.4mm, a* = 0.29, nRR = 5.0) 194 F . l l Run 5 Bed Temperature History. Mix 2 at 1 rpm (dp = 1.9mm, a* = 0.26, nRR = 7.3) 195 F.12 Run 5 Wall Temperature History. Mix 2 at 1 rpm (dp = 1.9mm, a* = 0.26, n R R = 7.3) 196 F.13 Run 5 Exposed Wall to Exposed Bed, hew^eb: and Covered Wall to Covered Bed, h cw-^cb^ Heat Transfer Coefficients. Mix 2 at 1 rpm (dp \u00E2\u0080\u0094 1.9mm, a*p = 0.26, nRR = 7.3) 197 xv F.14 Run 5 Overall Wall to Bed Heat Transfer Coefficient, ha, and Heat Flow, Qw^b- Mix 2 at 1 rpm (dp = 1.9mm, a* = 0.26, nRR = 7.3) 198 F.15 Run 5 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 2 at 1 rpm (dp = 1.9mm, a* = 0.26, nRR = 7.3) 199 F.16 Run 8 Bed Temperature History. Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 200 F.17 Run 8 Wall Temperature History. Mix 5 at 1 rpm (dp = 1.2mm, a* - 0.22, n R R = 7.7) 201 F.18 Run 8 Exposed Wall to Exposed Bed, hew^eb, and Covered Wall to Covered Bed, h cw^ycb. Heat Transfer Coefficients. Mix 5 at 1 rpm (dp \u00E2\u0080\u0094 1.2mm, a*p = 0.22, n R R = 7.7) 202 F.19 Run 8 Overall Wall to Bed Heat Transfer Coefficient, h0, and Heat Flow, Qw->b- Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, nRR = 7.7) 203 F.20 Run 8 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, nRR = 7.7) 204 F.21 Run 9 Bed Temperature History. Mix 5 at 3 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 205 F.22 Run 9 Wall Temperature History. Mix 5 at 3 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 206 F.23 Run 9 Exposed Wall to Exposed Bed, hew-+eb, and Covered Wall to Covered Bed, hcw^.cb, Heat Transfer Coefficients. Mix 5 at 3 rpm (dp = 1.2mm, a; = 0.22, n R R = 7.7) 207 F.24 Run 9 Overall Wall to Bed Heat Transfer Coefficient, hQ, and Heat Flow, Qw^b- Mix 5 at 3 rpm (dp = 1.2mm, a* = 0.22, nRR = 7.7) 208 F.25 Run 9 Ratio Between Heat Flow to Covered Wall and Exposed Wall, \f>. Mix 5 at 3 rpm (dp = 1.2mm, a* = 0.22, nRR = 7.7) 209 F.26 Run 10 Bed Temperature History. Mix 3 at 1 rpm (dp = 2.4mm, a* = 0.17, n R R = 11.7) 210 xvi F.27 Run 10 Wall Temperature History. Mix 3 at 1 rpm (dp = 2.4mm, a* = 0.17, nRR = 11.7) 211 F.28 Run 10 Exposed Wall to Exposed Bed, hew->eb, and Covered Wall to Cov-ered Bed, hcw-+cb, Heat Transfer Coefficients. Mix 3 at 1 rpm (dp = 2.4mm, a*p = 0.17, n R R = 11.7) 212 F.29 Run 10 Overall Wall to Bed Heat Transfer Coefficient, h0, and Heat Flow, Qw-+b- Mix 3 at 1 rpm (dp = 2.4mm, a* = 0.17, nRR = 11.7) 213 F.30 Run 10 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 3 at 1 rpm (dp = 2.4mm, a* = 0.17, nRR = 11.7) 214 F.31 Run 11 Bed Temperature History. Mix 4 at 1 rpm (dp = 3.0mm, a* - 0.09, n R R = 15.5) 215 F.32 Run 11 Wall Temperature History.' Mix 4 at 1 rpm (dp = 3.0mm, a* = 0.09, n R R = 15.5) 216 F.33 Run 11 Exposed Wall to Exposed Bed, hew^eb, and Covered Wall to Cov-ered Bed, hcw^cb. Heat Transfer Coefficients. Mix 4 at 1 rpm (dp = 3.0mm, a*p = 0.09, n R R = 15.5) 217 F.34 Run 11 Overall Wall to Bed Heat Transfer Coefficient, hQ, and Heat Flow, Qw-*. Mix 4 at 1 rpm (dp = 3.0mm, a* = 0.09, nRR = 15.5) 218 F.35 Run 11 Ratio Between Heat Flow to Covered Wall and Exposed Wall, \I>. Mix 4 at 1 rpm (dp = 3.0mm, a* = 0.09, nRR = 15.5) 219 F.36 Run 12 Bed Temperature History. Mix 6 at 1 rpm (dp = 1.3mm, a* = 0.15, n R R = 11.8) 220 F.37 Run 12 Wall Temperature History. Mix 6 at 1 rpm (dp = 1.3mm, a* = 0.15, n R R = 11.8) 221 F.38 Run 12 Exposed Wall to Exposed Bed, hew^eb, and Covered Wall to Cov-ered Bed, hcw^.cb, Heat Transfer Coefficients. Mix 6 at 1 rpm (dp = 1.3mm, a; - 0.15, n R R = 11.8) 222 xvii F.39 Run 12 Overall Wall to Bed Heat Transfer Coefficient, hD, and Heat Flow, Qw.+b. Mix 6 at 1 rpm (dp = 1.3mm, a* = 0.15, n R R = 11.8) 223 F.40 Run 12 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 6 at 1 rpm (dp = 1.3mm, a* = 0.15, n R R = 11.8) 224 F.41 Run 13 Bed Temperature History. Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 225 F.42 Run 13 Wall Temperature History. Mix 5 at 1 rpm (dp = 1.2mm, a* - 0.22, n R R = 7.7) 226 F.43 Run 13 Exposed Wall to Exposed Bed, hew^eb, and Covered Wall to Cov-ered Bed, /icu,_^cfc, Heat Transfer Coefficients. Mix 5 at 1 rpm (dp = 1.2mm, a*p = 0.22, n R R = 7.7) 227 F.44 Run 13 Overall Wall to Bed Heat Transfer Coefficient, ha, and Heat Flow, Qw->b- Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 228 F.45 Run 13 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 229 F.46 Run 14 Bed Temperature History. Mix 5 at 1 rpm (dp \u00E2\u0080\u0094 1.2mm, a* = 0.22, nRR = 7.7) 230 F.47 Run 14 Wall Temperature History. Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 231 F.48 Run 14 Exposed Wall to Exposed Bed, hew^eb, and Covered Wall to Cov-ered Bed, hcw^cb, Heat Transfer Coefficients. Mix 5 at 1 rpm (dp = 1.2mm, a*p = 0.22, n R R = 7.7) 232 F.49 Run 14 Overall Wall to Bed Heat Transfer Coefficient, h0, and Heat Flow, Qw->b- Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 233 F.50 Run 14 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 5 at 1 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 234 F.51 Run 15 Bed Temperature History. Mix 5 at 0.43 rpm (dp = 1.2mm, a* = 0.22, n R R = 7.7) 235 xviii F.52 Run 15 Wall Temperature History. Mix 5 at 0.43 rpm (dp - 1.2mm, a; = 0.22, nRR = 7.7) 236 F.53 Run 15 Exposed Wall to Exposed Bed, hew^eb, and Covered Wall to Cov-ered Bed, hcu^cb, Heat Transfer Coefficients. Mix 5 at 0.43 rpm (dp = 1.2mm, cr*p = 0.22, nRR = 7.7) 237 F.54 Run 15 Overall Wall to Bed Heat Transfer Coefficient, ha, and Heat Flow, Qw-*b- Mix 5 at 0.43 rpm (dp = 1.2mm, a*p = 0.22, nRR = 7.7) 238 F.55 Run 15 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 5 at 0.43 rpm (dp = 1.2mm, a* = 0.22, nRR = 7.7) 239 F.56 Run 16 Bed Temperature History. Mix 4 at 0.43 rpm (dp = 3.0mm, a* = 0.09, n R R = 15.5) 240 F.57 Run 16 Wall Temperature History. Mix 4 at 0.43 rpm (dp - 3.0mm, a*p = 0.09, nRR = 15.5) 241 F.58 Run 16 Exposed Wall to Exposed Bed, hew^eb, and Covered Wall to Cov-ered Bed, hcw^cb, Heat Transfer Coefficients. Mix 4 at 0.43 rpm (dp = 3.0mm, a* = 0.09, nRR = 15.5) 242 F.59 Run 16 Overall Wall to Bed Heat Transfer Coefficient, /z0, and Heat Flow, Qw-*b- Mix 4 at 0.43 rpm (dp = 3.0mm, a* = 0.09, nRR = 15.5) 243 F.60 Run 16 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 4 at 0.43 rpm (dp = 3.0mm, a* = 0.09, nRR = 15.5) 244 F.61 Run 17 Bed Temperature History. Mix 5 at 2 rpm (dp = 1.2mm, a* = 0.22, nRR = 7.7) 245 F.62 Run 17 Wall Temperature History. Mix 5 at 2 rpm (dp = 1.2mm, a* = 0.22, nRR = 7.7) 246 F.63 Run 17 Exposed Wall to Exposed Bed, hew^.eb, and Covered Wall to Cov-ered Bed, hcw-yCb, Heat Transfer Coefficients. Mix 5 at 2 rpm (dp = 1.2mm, a*p = 0.22, nRR = 7.7) 247 x i x F.64 Run 17 Overall Wall to Bed Heat Transfer Coefficient, ha, and Heat Flow, Qw^b. Mix 5 at 2 rpm (dp = 1.2mm, a* = 0.22, nRR = 7.7) 248 F.65 Run 17 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 5 at 2 rpm (dp = 1.2mm, a* = 0.22, nRR = 7.7) 249 F.66 Run 18 Bed Temperature History. Mix 6 at 2 rpm (dp = 1.3mm, a* \u00E2\u0080\u0094 0.15, nRR = 11.8) 250 F.67 Run 18 Wall Temperature History. Mix 6 at 2 rpm (dp = 1.3mm, a* = 0.15, nRR = 11.8) 251 F.68 Run 18 Exposed Wall to Exposed Bed, hew^eb, and Covered Wall to Cov-ered Bed, hcw-+cb, Heat Transfer Coefficients. Mix 6 at 2 rpm (dp = 1.3mm, a*p = 0.15, nRR = 11.8) 252 F.69 Run 18 Overall Wall to Bed Heat Transfer Coefficient, h0, and Heat Flow, Qw->b- Mix 6 at 2 rpm (dp = 1.3mm, a* = 0.15, nRR = 11.8) 253 F.70 Run 18 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 6 at 2 rpm (dp = 1.3mm, a* = 0.15, nRR = 11.8) 254 F.71 Run 19 Bed Temperature History. Mix 4 at 2 rpm (dp = 3.0mm, a* = 0.09, nRR = 15.5) 255 F.72 Run 19 Wall Temperature History. Mix 4 at 2 rpm (dp = 3.0mm, a* = 0.09, TIRR = 15.5) 256 F.73 Run 19 Exposed Wall to Exposed Bed, hew^eb, and Covered Wall to Cov-ered Bed, hcw^cb, Heat Transfer Coefficients. Mix 4 at 2 rpm (dp = 3.0mm, a*p = 0.09, nRR = 15.5) 257 F.74 Run 19 Overall Wall to Bed Heat Transfer Coefficient, h0, and Heat Flow, Qw-+b- Mix 4 at 2 rpm (dp - 3.0mm, a* = 0.09, nRR = 15.5) 258 F.75 Run 19 Ratio Between Heat Flow to Covered Wall and Exposed Wall, Mix 4 at 2 rpm (dp = 3.0mm, a* = 0.09, nRR = 15.5) 259 xx Acknowledgment I am sincerely grateful to my supervisors Dr. A . P. Watkinson and Dr. P. V . Barr. Their guidance and support not only helped to develop this thesis but also my technical and personal skills. I thank my friends and colleagues, for they provided a forum for discussing and resolv-ing minor and major issues encountered in the course of this study. In particular, I am indebted to (in alphabetical order) Abba, Alissa, Diana, Gordon, Lee, Murugan, Nanda, Neeraj, Pierre, Robin (from University of Birmingham, U K ) , and Shawn for helping me take data during the high temperature trials and to Petar and Michael for providing computer support. Finally, I thank my dear wife, Susan, who always stood by me. Her unconditional love, affection and encouragement carried me through this episode in my life. Financial support for this work was provided by Dr. Watkinson's and Dr. Barr's N S E R C grants. x x i Nomenclature Symbol Description units A area m 2 Ax, Bx, Cx Factors used to simplify presentation Ay, By, Cy of discretized Equation 4.8 A variable in velocity profile in rectangular grid B depth of active layer m in rectangular grid b parameter in Rosin-Rammler distribution c variable in velocity profile C correlation coefficient cp specific heat capacity J /kg K d diameter m d average diameter m dsp Sauter mean diameter m di diameter of sand in m sample i Eajb notation used to write out discretized equations F flux T View Factor g acceleration due to gravity m/s 2 h heat transfer coefficient W / m 2 K [/] identity matrix xxn k effective thermal conductivity W / m K / linear distance m M mass of sand sample kg m variable used for clarification rrii mass of size fraction of sand kg in sample i N total number of particles n variable used for clarification rn number of particles in sand sample nRR parameter in Rosin-Rammler distribution indicating width of particle size range n unit vector Q heat transfer rate W . k W q heat flux W / m 2 R resistance to radiative heat transfer Res residual through sieve kg r radius m S sum of squares of residuals in in linear interpolation St sum of squares of differences between average and point values s standard deviation T temperature K T average temperature K xxi n f intermediate variable in approximate factorization t time s, min t* dimensionless time u velocity in horizontal direction m/s V velocity in vertical direction m/s X horizontal distance m y vertical distance m Greek a thermal diffusivity m 2 / s absorptivity of gas P bed angle, as described in Figure 4.9 e emissivity e angle in cylindrical coordinates radians e* dimensionless temperature V- viscosity Pa s p density k g / m 3 \u00C2\u00B0~SB Stefan Boltzmann coefficient, 5.67 X 1 0- 8 W / m 2 K o-p population standard deviation m of particle size normalized population standard deviation of particle size (^) ax population standard deviation K of temperatures B l to B4 <]> time advance variable = 1.0 for fully implicit = 0.5 for trapezoidal X X I V ^ ratio of heat transferred between covered wall and covered bed and exposed wall and exposed bed Q cu> \u00E2\u0080\u0094\ c b Q e w - t e b Q, temperature range \u00C2\u00B0C M a x [ r B l . . . TB4] - Min[Tj3i . . . TB4] u> rotational speed rad/s subscripts a ambient CV two dimensional control volume b bed bs bed surface exposed to wall eb exposed bed ew exposed wall 9 gas I max. horizontal cell number e.g. for 20 point mesh, I = 20 i horizontal cell number in initial j vertical cell number o overall p particle rad radiative s rectangular bed surface w wall x horizontal direction y vertical direction xxv superscripts time step normalized average xxvi Chapter 1 Introduction Industrial experience with rotary kilns has indicated that fine particles can be difficult to process. Isothermal experiments with particulate materials in laboratory kilns have have identified that when there are differences in the size of particles, segregation of finer particles to the center of the bed occurs. This has lead to the premise that segregation shields finer particles from the kiln heat source, leading to a temperature non-uniformity within the bed. To date, however, no experimental data has been found that verifies this. This thesis attempts to provide some. In particular this thesis will investigate the role of particle size distribution on heat transfer within the bed of a rotary kiln furnace. A rotary kiln is a refractory lined cylindrical furnace that is used to thermally process particulate material. It is typically 2-4m in diameter, up to 100m long and inclined a few degrees to the horizontal, see Figure 1.1. Raw material is fed into the upper end and it travels along the kiln length as the kiln rotates. At the lower end a fossil fuel is fired, using a suitably designed burner, to provide the energy for the process. A n induced air fan draws ambient air through the kiln providing oxygen for the fuel. Therefore, within the kiln there are two distinct regions; a free-board region, where the fuel burns and combustion gases flow, and a bed region, where the particulate material flows. A n efficient kiln is one where much of the energy from the fuel is transferred to the material being processed. This requires optimum fuel combustion, optimum heat transfer between the free-board and bed surfaces and optimum heat transfer through the bed to ensure all of the bed material receives the energy required to process it. In designing for such a system a thorough understanding of all of the energy transfer processes is necessary. This study attempts to further the understanding of one of these processes, heat transfer within the bed. Heat is transferred through the bed of a rotary kiln primarily through advection, or material flow. This material flow is complex and studies have identified a number of 1 CHAPTER 1. INTRODUCTION 2 different modes of bed flow behavior. The most desirable of them is the rolling mode, see Figure 1.2, because it provides good mixing with a steady angle of repose. The latter ensures continuous surface renewal, which in turn facilitates efficient particle agitation. Heat transfer is maximized, allowing reactions to proceed, without causing attrition and and excessive dust generation. Therefore, much of this study will focus on the rolling mode. In the rolling mode there are two distinct regions, see Figure 1.2. Nearer to the surface, the material cascades downwards in a continuously shearing region (active region or layer), while close to the walls a plug flow region is established, where material moves in a rigid body motion, | | = 0. The rolling bed maximizes heat transfer to and throughout the bed, helping to uni-formly process all of the material. However, as particles cascade down in the active layer, the particle ensemble dilates. Differences in the size of the particles then leads to seg-regation, as smaller particles pass through the voids created between the larger ones. Segregation, therefore, causes de-mixing and, it is believed, hinders heat transfer. CHAPTER 1. INTRODUCTION 4 Figure 1.2: Rolling Bed Motion Depicting a Top Active Layer and Bottom Plug Flow Region Chapter 2 Literature Review Relative to the free-board, there is little literature pertaining to the bed of a rotary kiln and it tends to concentrate on material flow and mixing. Few workers have extended the studies to include heat transfer in the bed. This section will first review work carried out in the area of material flow and mixing in the cross-section of a rotary kiln. Following this studies more specific to material segregation, which is a major cause of de-mixing, will be described. Finally heat transfer will be reviewed. 2.1 Material Flow Material flow in a rotary kiln may be resolved in the axial and transverse direction. However, it is transverse flow that determines most of the transport processes. Transverse flow determines the degree of mixing (or segregation) as well as the axial motion because material only advances axially when it is turned over in the cross-section. Therefore, one can appreciate the importance of transverse motion in heat transfer. Flow studies associated with the cross section of a kiln have indicated that bed motion changes with rotational speed (Henein 1980). As the rotational speed is increased, the bed motion changes from slipping to slumping followed by rolling, cascading, cataracting and finally centrifuging, see Figure 2.1. When a kiln with a particulate bed is rotated slowly, the bed behaves as a bulk and rotates with the kiln. If the gravitational force is greater than the centrifugal force and factional force between the bed particles and wall surface, then the bed will \"slip\". At higher rotational speeds and/or with rougher walls there comes a point when the gravitational force is overcome by the centrifugal and frictional forces, and so the bed will continue to rotate with the kiln. This occurs until the bed's radial inclination exceeds 5 CHAPTER 2. LITERATURE REVIEW 6 the static angle of repose of the particles. As the kiln is further rotated, the weight of the particles associated with the region above the static angle of repose will eventually overcome the shear stresses within the particulate body, and these particles will avalanche down the surface of the bed. The angle at which the particles begin to avalanche is sometimes called the maximum angle of stability (McCarthy, Shinbrot, Metcalfe. Wolf, and Ottino 1996) and once the avalanche is complete, the bed inclination adopts the static angle of repose. As the kiln continues to rotate, so will the bed until the maximum angle of stability is once again reached. Then the whole cycle is repeated. This type of bed behavior is referred to as \" slumping\". If the rotational speed of the kiln is high enough such that the bed rotates to its maximum angle of stability before the avalanching particles come to a halt, the bed is said to be \"rolling\". The avalanching particles produce a continuously shearing region, which is referred to as the \"active region\", \"active layer\" or \"shear layer\". This layer takes on a lens type shape. At the bottom of the avalanche the particles leave the active layer and enter a \"plug flow\" region where they rotate with the kiln in rigid body motion. In the rolling mode particles continuously enter and leave the active and plug flow regions. A further increase in kiln rotation causes the top part of the bed, called the apex, to rise above the bed surface before gravitational forces bring the particles down to \"cascade\" over the bed surface. Even further increases in rotational speed lead to \"cataracting\" and finally \"centrifuging\". In the centrifuging mode, the particles centrifugal force over-comes gravity, and hence a rotational Froude number (!~^) of 1.0 indicates the onset of centrifuging. Of the various types of bed behavior described above, rolling is preferred in industrial rotary kilns. Rolling provides good mixing at a steady angle of repose. Surface renewal is continuous which allows for efficient heat transfer. Cascading, in fact, appears to provide better particle mixing. Workers have operated test kilns in the cascading mode to ensure intimate mixing prior to segregation studies (Nityanand, Manley, and Henein 1986; Pollard and Henein 1989). This would imply that cascading provides better heat transfer also. However, cascading can lead to attrition and excessive generation of dust. Furthermore, kiln operation in a cascading mode would require extra mechanical power. CHAPTER 2. LITERATURE REVIEW 7 In the rolling mode, there axe two distinct regions, see Figure 2.2. Nearer to the surface of the bed the material travels downward in a continuously shearing region (the active region) while closer to the wall a plug flow region is established where material moves in rigid body motion. The particles enter the shearing region from the plug flow region once they reach their static angle of repose and re-enter the plug flow region towards the lower part of the active layer. In his extensive study of the transverse motion in the bed of a rotary cylinder, Henein (1980) was able to develop bed behavior diagrams that delineated the various types of bed behavior. He was able to relate bed behavior to rotational speed, bed depth, percentage fill and rotational Froude number. This allows one to predict the mode in which a bed behaves. Another important finding of Henein (1980) was that axial motion, to the extent that would occur in a rotary kiln, does not significantly alter transverse behavior. Therefore, he was able to utilize batch experiments to investigate bed behavior in a continuous kiln. In the case of particles with similar size, shape and density, experiments indicate that intimate mixing of particles eventually occurs with a rolling bed (Woodle and Munro 1993). The reason is that particle mixing within the active region is intense and so par-ticles entering the active region have an equal chance of re-entering the passive region at any radial location. The rate at which mixing occurs, however, is dependent on kiln load-ing and particle/wall and particle/particle friction (Woodle and Munro 1993; Nityanand, Manley, and Henein 1986). 2.2 Segregation In industrial rotary kilns mixing is not always adequate and it has been observed that, for example, smaller particles of lime take longer to calcine than larger ones (Sunnergren 1979; Von Wedel 1973). This implies that smaller particles are not exposed to the hot gases for as long as the larger particles, for if they were, then the time required for them to calcine would be less. The reason for this is attributed to particle segregation, which in the examples cited is due to size. Numerous workers have shown that smaller particles CHAPTER 2. LITERATURE REVIEW 8 tend to congregate in the middle of the bed rather than travel to the region of the bed that is exposed to the free-board. Segregation due to density differences has also been observed . One example is in the direct reduction of iron ore (Henein 1980), where denser iron ore particles segregate to the middle of the bed in the same way as the smaller limestone particles described above. The iron ore particles are surrounded by lighter coal particles which form an annulus around them. A review of segregation by Williams (1976) indicates that segregation may occur due to differences in particle size, density, shape and resilience, although he stated that segre-gation due to particle size is by far the most common in industry. The three segregation mechanisms that he described are Trajectory Segregation The distance that a particle is projected is equal to \" ' \" 1 ^ p . Therefore, a particle with larger mass and/or diameter will be projected further, if all other parameters are kept the same. Percolation In mixtures of different sized particles, the size of the voids between larger particles may become large enough for smaller particles to pass through them. This will result in smaller particles percolating through the voids of the larger ones leading to segregation. If the particles are in motion, then the voids will be larger, due to dilation of the particle ensemble. This allows percolation of particles of a size that would not pass through the voids formed when the particulate mass is stationary. Therefore, in a bed under motion, percolation can occur even when there is a only a small difference between the sizes of the larger and smaller particles. Vibrat ion In the case of a vibrating bed, large particles will rise to the surface while smaller particles will segregate to the bottom. Since there is little, if any, vibration in rotary kilns, this type of segregation will not be discussed further. Dolgunin, Kudy, and Ukolov (1998) conducted segregation experiments on mixtures of particles avalanching down an inclined surface. They found that larger and denser par-ticles tended to concentrate towards the center of the particulate mass. They attributed CHAPTER 2. LITERATURE REVIEW 9 this to particle collisions. When particles collide the exchange of momenta means that larger and denser particles end up with lower velocities than smaller and lighter particles. As particles travel down a chute, a high solids concentration occurs in the center which is favorable to particles with a smaller velocity fluctuation amplitude. Therefore, larger and denser particles migrate towards the center. This segregation mechanism would tend to negate the effects of trajectory segregation, described above, in granular flows. 2.2.1 Segregation Exper iments Segregation may occur axially or radially in a rotary kiln. Roseman and Donald (1962) investigated segregation in horizontal cylinders and concluded that radial segregation always occurs whereas axial segregation may or may not occur. The rate of axial segregation is relatively slow. Literature reports that axial seg-regation may not fully develop for 500-10,000 revolutions (Roseman and Donald 1962; Carley-Macauly and Donald 1962) . This equates to between 4 and 80 hours for a kiln that is rotating at 2 rpm. Since overall residence times in industry are less, axial seg-regation is unlikely to fully develop. However, a degree of axial segregation affects the residence time of individual particles. Although this may not be significant with respect to heat transfer, the extent of chemical or physical reaction may be affected, possibly to an extent that impacts quality. Radial segregation is a much more rapid process and takes 2-10 revolutions. Nityanand, Manley, and Henein (1986) and Pollard and Henein (1989) have conducted tests on the kinetics of radial segregation and concluded that the rate followed first zero kinetics, that is, the rate of segregation is a constant. These workers quoted, however, that other studies had reported the kinetics to be first order, that is rate is proportional to the degree of segregation. Henein (1980) took quantitative measurements of the segregation phenomenon. He used binary mixtures of sand and limestone, and investigated mixes with about 5-25% fines concentration. The size ratios were 2.2:1 for the sand and 8:1 for the limestone. He made the following significant findings CHAPTER 2. LITERATURE REVIEW 10 \u00E2\u0080\u00A2 Bed behavior is not affected by fines. For example, in the case of a rolling bed, the dynamic angle of repose and active layer thickness do not change and, in a slumping bed, the static angle of repose, shear angle and slumping frequency do not change when the quantity of fines is varied. \u00E2\u0080\u00A2 Since the flow characteristics of the bed are not significantly altered by the presence of fines, segregation by flow must be negligible. Therefore, the primary cause of segregation in a rotary kiln is percolation. \u00E2\u0080\u00A2 The segregated core is composed of all of the fines packed together with a quantity of the coarser particles such that the packing density is maximum for that particular mixture. The segregated core is surrounded by an annulus of coarse particles. \u00E2\u0080\u00A2 From the above, the size of the segregated core may be deduced. Its shape was found to be similar to the shape of the bed. \u00E2\u0080\u00A2 A second segregated region close to the apex is also formed. This consists of very fine material that is able to percolate through the shear layer and through the plug flow region. The size of these fines are less than about 15% of the size of the larger particles (Boateng and Barr 1996a). 2.2.2 Segregation M o d e l i n g Boateng (1993) developed a model of the transverse flow of particles in a rotary kiln . He employed the continuum assumption and developed a mean velocity profile by solving constitutive equations that are normally associated with granular flows. In such models analogy is made with kinetic theory. In granular flow individual particles have velocities that are different to the mean velocity associated with all of the particles within a finite volume, in a manner similar to molecules in a gas. The property that is analogous to thermal temperature of a gas is called granular temperature (units m2s~2) and this is a measure of the particle kinetic energy. A balance of granular temperature forms one of the constitutive equations and it needs to be solved together with the mass and momentum balance, see for example Lun, Savage, Jeffrey, and Chepurnity (1984). CHAPTER 2. LITERATURE REVIEW 11 Boateng (1993) applied granular flow theory to the active layer where it is appropriate for such analysis. In the plug flow region this is not required since plug flow exists. The resulting velocity profile was then used in a segregation model that was applied to the active region, implying that the fines are not small enough to percolate through the plug flow region. The segregation model considered a binary mixture of particles and used published modeling techniques for the phenomenon. Boateng's model, which is also described in Boateng and Barr (1996a), used the terms jetsam, to describe sinking particles and flotsam to describe floating particles. Essentially, a net percolation velocity is assigned to the bed and used to solve for jetsam particles. Determining the net percolation is the challenge and it is a function of the voids. The net percolation velocity was determined using a model developed by Savage and Lun (1988). The model results compared well with the experimental results of Henein (1980). Dolgunin, Kudy, and Ukolov (1998) developed a segregation model that is similar to the one used by Boateng. They added a migration term that is related to the mean distance between particles. With the addition of the new term, they found improved agreement with experimental results, especially for the case of segregation due to density differences. In addition to the continuum approach, several workers have developed discrete par-ticle models to investigate segregation. Meakin and Jullien (1992) describe two- and three-dimensional models of particles of different sizes being deposited onto a heap of particles. In these models the path of a single particle is tracked as it falls down the heap taking the steepest path. Once the particle arrives at a location where further movement down is prevented by other particles, the particle stops and its location is then fixed. Only one particle moves at any one time and its size is selected randomly from a specified range. Their results demonstrate segregation due to size. Ristow (1994) has used \"molecular dynamics\" to model particles of the same size but different density and particles of the same density but different size in a rotating drum in two dimensions. In his article he states that he modeled 600 particles, twelve of which were tracer particles. He tracked the tracer particles to determine how long they took to CHAPTER 2. LITERATURE REVIEW 12 segregate. Baumann, Janosi, and Wolf (1994) investigated segregation of particles due to size in a rotating drum using a tracking model that seems to be similar, in principle, to that of Meakin and Jullien (1992) above. They referred to their technique as \"bottom-to-top restructuring\", or \" B T R \" . They used a \"molecular-dynamics simulation\" type model and excluded inertial and elasticity effects. They started their simulation with the drum full of particles (which are in fact discs because the model is two-dimensional) that are well mixed. Then the drum is allowed to rotate by a small angle 1\u00C2\u00B0. The lowest particle is allowed to fall along the path of steepest descent until it arrives at a local minimum. After this the next particle above falls in the same manner. By doing this segregation occurs because smaller particles get stuck in niches along an incline that larger particles would pass over. Once all of the particles have moved, the drum is rotated further by the same small angle. Their model results show segregation, indicating that geometric affects alone can lead to it. They do recognize, however, that in the case of irregular shaped particles, friction, inertial effects and collective organization also play a role. 2.3 Heat Transfer Studies of the free-board have identified the heat transfer paths in a rotary kiln as those shown in Figure 2.3 (Barr, Brimacombe, and Watkinson 1989a). They are 1. Free-board gas to exposed bed (radiation and convection) 2. Free-board gas to exposed wall (radiation and convection) 3. Exposed wall to exposed bed (radiation) 4. Exposed wall to exposed wall (radiation) 5. Covered wall to covered bed (radiation and conduction) 6. Loss to surroundings (conduction through refractory and convection to atmosphere) CHAPTER 2. LITERATURE REVIEW 13 In addition heat is also transferred within the bed in a manner similar to a packed bed, but with an added advective component. The heat transfer paths within a packed bed are shown in Figure 2.4. These paths are 1. Within particle conduction 2. Conduction across the particle-particle contact area 3. Particle to particle radiation 4. Interstitial gas convection In this section the free-board studies will be reviewed first, followed by studies that concentrate on the bed. 2.3.1 Heat Transfer in the Free-Board Due to the high temperatures that occur in a kiln, radiation is the dominant mode of heat transfer. Analysis of such heat transfer is complex because of the arduous radiation paths. For example, a portion of the energy from the free-board that radiates to the exposed wall is reflected and re-absorbed back into the free-board. Another portion is reflected onto another part of the exposed wall only to be partly absorbed and partly reflected back to the bed, etc. Another complicating factor is that the walls play a regenerative role in the heat transfer process (Barr, Brimacombe, and Watkinson 1989a). When the wall is exposed to the free-board it gains a net amount of energy from the free-board. As the kiln rotates, the part of the wall that was previously exposed becomes covered by the bed and, since the bed is cooler (or warmer in some cases), energy transfer occurs. Studies by Barr, Brimacombe, and Watkinson (1989a) have shown that this regenerative effect produces a cyclical temperature in a region of the inner refractory wall, a region that they called the \"active\" refractory region. Barr, Brimacombe, and Watkinson (1989a) investigated the various heat transfer paths in a kiln by way of a two-dimensional cross sectional model that was marched along the CHAPTER 2. LITERATURE REVIEW 14 length of the kiln. The cross sectional model accounted for all of the heat transfer paths described above, except for the heat transfer within the bed, implying that the bed was well mixed. The results of the model agreed well with experimental data and suitable values for heat transfer coefficients for the paths shown in Figure 2.3 were determined. The fact that a cross sectional model that is marched along the length of the kiln can adequately describe the heat transfer process indicates, as Barr, Brimacombe, and Watkinson (1989a) mentioned, that heat transfer in a kiln is a local phenomenon, dom-inated by parameters at that particular axial location. This was further confirmed in a study by Jenkins (1998), who compared the results of a three-dimensional radiative model of a kiln with the same model but in only one dimension. It was discovered that the overall heat flux to the walls within the flame region only differed by a few percent, although the local heat fluxes differed more. In order to predict the heat transfer characteristics of the free-board one must appre-ciate a number of phenomena. First, combustion gases that have an electric (e.g. dipole or quadrapole) moment, such as CO2, CO, H 2 0 , N H 3 and hydrocarbons, absorb and emit radiation only within certain wavelengths. Their radiative characteristics need to be determined. Second, the velocity and concentration profiles of the gas need to be known. Since, these are a function of the combustion of the fuel and system aerodynamics, de-termining them can be a challenge. Only with knowledge of the radiative characteristics and the velocity and concentration profile of the free-board gas, can the heat transfer be calculated. One must also understand that all these phenomena are coupled. The radiative characteristics of the gas may be determined using the weighted sum-mation method (Smith, Shen, and Friedman 1982). Jenkins and Moles (1981) and Barr, Brimacombe, and Watkinson (1989b) have used this method in their models. The veloc-ity profile requires a flow model or measurements, if possible. Jenkins and Moles (1981) developed a relatively simple but effective aerodynamic flow model by calculating how much combustion air would be entrained by the flame jet. The velocity profile is then used as input to a combustion model to calculate a concentration profile. Finally, the Hottel zone method can be used to calculate heat transfer (Hottel and Sarofim 1967). Establishing the velocity profile inside a rotary kiln can be quite a daunting task, es-CHAPTER 2. LITERATURE REVIEW 15 pecially if the air entering a kiln is awkward and causes uneven aerodynamics. The aero-dynamics have a dominating effect on combustion, and hence, heat transfer. Therefore, this is an area of extensive study and it is making use of computational fluid dynamics, (Alyaser, Barr, and Brimacombe 1997; Bui , Simard, Charette, Kocaefe, and Perron 1995; Mastorakos, Massias, Tsakiroglou, Goussis, Burganos, and Payatakes 1998). 2.3.2 Heat Transfer in the Bed Many of the earlier studies of kilns paid little attention to heat transfer within the bed. The bed was typically considered isothermal (Barr 1986; Gorog, Brimacombe, and Adams 1981) or not considered at all (Jenkins and Moles 1981). In the latter citation the bed and wall were considered as a single heat sink. These studies, nevertheless, provided useful insights into the operation and performance of rotary kilns. One reason is that bed studies have indicated that when a bed is well mixed the thermal resistance within it is small (Imber and Paschkis 1962). However, in the case of a segregated bed, this may not be the case and so heat transfer within such beds needs to be investigated. As mentioned above, the heat transfer paths within the bed of a rotary kiln are similar to those in a packed bed. However, the bed in a rotary kiln has an additional advective component due to material transport. The bed of a rotary kiln may, therefore, be analyzed in a manner similar to that of a packed bed as long as advection is accounted for. One method used to investigate packed beds is to consider the particles and the in-terstitial fluid as a continuum. By doing so, a single temperature represents a packet of particles and gas and so one equation can be used to represent the heat transfer. The drawback with this method is that particle surface temperatures, which determine ra-diative properties, and local particle dynamics, that can be used to describe a particle reaction, are not considered. Nevertheless, a continuum model has been developed for the bed and this will be described next. Following on from the flow and segregation model, described in Section 2.2.2, Boateng (1993) continued with the continuum assumption to develop a heat transfer model of the cross section of the bed. He solved the energy equation using an effective conductivity term (Schotte 1960) to describe the diffusive heat transfer component within the bed. CHAPTER 2. LITERATURE REVIEW 16 This effective conductivity term was enhanced in the active region by a factor relating to the mixing rate, as determined in the flow model. The boundary conditions for the cross-sectional model, essentially heat fluxes, were calculated independently using a one dimensional model of the kiln along the axial direction which was similar to the model developed by Barr, Brimacombe, and Watkinson (1989b). The validation of the above model was carried out largely by converting the two-dimensional model results into one-dimensional average values at axial locations and com-paring them with experimental data. Some two-dimensional validation was carried out but it appears that no measurements were taken in the bed, only in the free-board and walls. Even so, agreement with experimental data was good lending \"some confidence\" to the model (Boateng and Barr 1996a). Boateng (1993) reported that non-uniformity in the cross-sectional temperature is only significant when there is radial segregation. When uniformly sized particles were used the bed was well mixed and the resulting temperature profile was isothermal. This latter point would explain why studies employing the well mixed bed assumption have been successful. It is worth mentioning some other studies where heat transfer within the bed of a rotary kiln has been considered. Bui , Simard, Charette, Kocaefe, and Perron (1995) developed a three-dimensional model of a coke calcining kiln using a computational fluid dynamics package that is comprised of a number of sub models. The bed is simulated as a pseudo fluid with a higher viscosity in the plug flow region and a lower viscosity in the active region. A third value for viscosity is applied in the axial direction to ensure correct axial mass flow. The overall model results were validated using pilot and full size data and agreement was reasonable. However, the authors did recognize that analysis of the bed in such a manner limits the model's capabilities. Yang and Farouk (1997) developed a two-dimensional cross-sectional model and a one-dimensional axial model of the kiln. The two models are used together similar to the model by Boateng and Barr (1996a). However, these authors used the equations developed for granular flow theories throughout the whole of the bed cross-section. This is not strictly applicable because granular flow theories compare the motion of particles to CHAPTER 2. LITERATURE REVIEW 17 the motion of molecules in a gas. This is appropriate for granular flow down an inclined chute or in the active layer because such particles are in a state of random motion whereas material in the plug flow region of the bed behaves as a rigid body. Their model results for bulk temperatures along the kiln axis, however, compared well with experimental data. CHAPTER 2. LITERATURE REVIEW Slipping Rolling Slumping Cascading Cataracting Centrifuging Figure 2.1: Modes of Bed Behavior in a Rotary Ki ln Figure 2.2: Rolling Bed Motion Depicting a Top Active Layer and Bottom Plug Flow Region CHAPTER 2. LITERATURE REVIEW Figure 2.3: Heat Transfer Paths Inside a Rotary K i l CHAPTER 2. LITERATURE REVIEW Figure 2.4: Heat Transfer Paths Within a Packed Bed Chapter 3 Objectives and Scope of Study Studies of the material flow in rotary kilns have provided the following useful information. \u00E2\u0080\u00A2 Transverse flow plays the dominant role in the particle mixing rate and, therefore, heat transfer and reaction rate. Axial flow and, hence, overall residence time, is also determined by the transverse flow because the material only advances axially when it is turned over in the cross-section. \u00E2\u0080\u00A2 The are a number of modes of transverse bed behavior depending on material and operational properties, see Section 2.1. Of the various modes, rolling is preferred because it allows for maximum particle mixing and heat transfer without excessive particle agitation and dust generation. \u00E2\u0080\u00A2 When the size distribution of the particulate burden in the kiln is such that a quantity of fines below a certain size range exist, segregation occurs. Segregation may be axial or radial, see Section 2.2.2. \u00E2\u0080\u00A2 Axial segregation affects particle residence time. While this may not affect heat transfer significantly, it may affect the degree of chemical reaction. Therefore, axial segregation may play an important role in determining product quality. \u00E2\u0080\u00A2 When radial segregation occurs, the fines create a primary core at the center of the bed and, in cases of very small diameter fines, less than 15% of the diameter of larger particles, a secondary core at the kiln walls close to the apex. The overall effect of segregation is one of de-mixing and is, therefore, detrimental to the heat transfer process. \u00E2\u0080\u00A2 The fines do not alter the overall transverse bed behavior. \u00E2\u0080\u00A2 The mechanism by which segregation occurs is predominantly percolation. 22 CHAPTER 3. OBJECTIVES AND SCOPE OF STUDY 23 \u00E2\u0080\u00A2 The primary core consists of all the fines (excluding those in the secondary core) plus a quantity of coarser materials, such that the bulk density of the core is maximum for that particular mixture. Therefore, one can estimate the size of the core, see Henein (1980). The shape of the core is similar to the shape of the bed. \u00E2\u0080\u00A2 The rate of segregation is rapid, occurring between 2-10 revolutions of the kiln (Pollard and Henein 1989; Nityanand, Manley, and Henein 1986). Therefore, unless the bed is continuously agitated, the bed remains segregated throughout the length of the kiln. Radial segregation is suspected to be the primary cause of temperature non-uniformity in the cross section of a bed (Boateng 1993). However, most investigations of the segre-gated core are focused on flow and, hence, experiments are conducted for the isothermal case. While these studies have provided interesting and informative insights, the mineral processing rotary kiln industry requires to know the effect on heat transfer within the bed, which ultimately determines product quality. It is appreciated that axial segregation may also impact product quality, however, it will not be the focus of this study and will remain outside the scope. Boateng (1993) provided advancement in this area by his study of the transverse flow and heat transfer in a rotary kiln, see Sections 2.2.2 and 2.3.2. In his study he developed a model of the flow and heat transfer in the transverse section of the kiln. Included was a segregation model for a binary mixture of particles, which was validated independently of the heat transfer, using the cold segregation experimental data of Henein (1980). His hot model results were validated using experimental results from Barr (1986) plus his own experiments. However, the hot runs were largely for a well mixed bed with a narrow size range of particles. Giving that radial segregation is perhaps the dominant cause of temperature non-uniformity in the transverse section of a bed, which may lead to product inconsistencies, what appears to be lacking in the literature is a thorough study of heat transfer in a segregated bed. Therefore, the objective of this work is to initiate such a study. The first step is to provide hot data for a segregated bed that can be used to develop a CHAPTER 3. OBJECTIVES AND SCOPE OF STUDY 24 better understanding of segregation and its effect on heat transfer. With this information predictive models may be developed. A second objective is to establish what level of fines causes significant temperature non-uniformities. To quantify the level of fines, particle size distribution is to be used. This is a more practical method than using binary mixtures with fine and coarse particles. The results can be applied in industry to determine whether a particular kiln feed material is suitable or not for processing in a kiln. Finally, an inert bed material, sand, is to be used to avoid complexities associated with chemical reactions. In view of the above the following constitutes a scope of study that will meet the study objectives: 1. Design and build apparatus suitable for the study of heat transfer in the cross section of the bed. 2. Make a range of sand mixes with varying size distributions. For each mix make qualitative measurements of the flow characteristics. Apply a heat source to each mix and measure the temperature in the transverse section. 3. Determine how uniform the temperature is and calculate heat transfer data that can be used for modeling purposes. Chapter 4 Test Furnace Design Having established the scope of the research program, the next stage was to develop a test furnace in which suitable experiments could be performed. Since axial temperature gradients are negligible with respect to heat transfer, and axial motion has a negligible effect on radial motion, a batch furnace would suffice for this exercise. The batch furnace was developed by adopting a stepwise procedure. First a bed depth in which measurable temperature gradients occur was determined. The bed depth also needed to be small enough for laboratory convenience. Understanding that a practical rotary test furnace can be operated up to 20% full, this bed depth would also establish the furnace inside diameter. A two-dimensional model of the cross sectional heat and material transfer in a rotary furnace was developed as the tool to investigate bed depth. The second step was to design a rotating test furnace that provided a controlled and measurable amount of energy to the bed. This posed a challenge because, while the heat energy required to raise the bed temperature was relatively small, the energy required to raise the temperature of an amount of insulating refractory was large. A gas flame would have been able to provide this energy but its heat transfer characteristics are not fully predictable and, hence, a degree of uncertainty would be introduced if one were to be used. A heated element provides a more predictable heat source but it would not provide sufficient energy. Therefore, it was decided, if feasible, to use the hot refractory walls as the heat source. The heat flux to the bed could then be calculated using standard radiation theory. The furnace would be first heated. Once hot, the furnace is charged with sand and rotated. In designing such a furnace it was necessary to ensure that the refractory could be heated with a practical flame and then, once hot, it could provide sufficient heat for the bed material. A one-dimensional model of the heat transfer between the hot gas and walls 25 CHAPTER 4. TEST FURNACE DESIGN 26 and walls and bed material was developed in order to carry out the design. Once the furnace process design was complete, a mechanical and instrumentation design was carried out and finally the equipment was procured. 4.1 B e d D e p t h The objective here was to determine a bed depth in which one could measure a non-uniform temperature and yet is small enough to be convenient in a laboratory. To meet the objective, a model of the energy transfer in the transverse section of a bed was developed. Since the purpose of the model was to obtain approximate information, some simplifying assumptions were made. They were: \u00E2\u0080\u00A2 A rectangle was used to approximate the bed cross-section. This allowed for a simple grid structure. \u00E2\u0080\u00A2 The continuum assumption was made to represent the particle bulk. \u00E2\u0080\u00A2 An effective conductivity was utilized to represent diffusive heat transfer. \u00E2\u0080\u00A2 A velocity profile for the particles was approximated using bulk particle velocity data from Boateng (1993). \u00E2\u0080\u00A2 The particle mixing caused by shear in the active layer was modeled by enhancing the effective conductivity. 4.1.1 Two-dimensional B e d Cross-Section Code The system has advective and diffusive transport of energy and so can be represented as duT J_d_ f dT + dvT = 0 dx pcpdy V dy dy (4.1) CHAPTER 4. TEST FURNACE DESIGN 27 This equation was then solved in finite volume form as follows. First it was integrated over a two dimensional control volume1, see Figure 4.1. dF*-iA = 0 -dA+ At 4>At 4>At At At 4>At 2Ax U : 1 \u00E2\u0080\u009E\u00E2\u0080\u00A2 Ax2 a i - l - j + - L - 1 \u00C2\u00B1 + I A O 1 2Ax A x 2 2Ax A x 2 2Ax A x 2 l<3 2 ' 2Ay Ay2 H ir^r- + r 2Ay ' A y 2 ' 2Ay ' A y 2 V 2Ay A y 2 ) ( (4.8) and (p \u00E2\u0080\u0094 0-5 for a trapezoidal time advance and 4> = 1.0 for a fully implicit time advance. Note, that a \u00E2\u0080\u0094 \u00E2\u0080\u0094 is thermal diffusivity. At the interface of two cells the interfacial conductivity is best represented by the harmonic mean of the conductivity of the two adjacent cells (Patankar 1980) k : , l (4.9) CHAPTER 4. TEST FURNACE DESIGN 29 This allows for a better representation of the heat flux through the boundary when there is an abrupt change in the conductivity between cells, such as the one that occurs at the boundary between the active and plug flow regions. In such an event, the interfacial conductivity is dependent on the lower conductivity, and one can see that equation 4.9 allows for this. In saying this, however, in this case the plug flow region has a conductivity much lower than the active region, and so the effective conductivity at the boundary was taken to be that of the plug flow region. By using the notation Ea n N 2 + C 0 2 + 2 H 2 0 mg 64g 16g mg 44g 36g Since ambient air contains 21% (vol/vol) oxygen with the remainder nitrogen, n = 79/21 and m = 210.7. In the case of excess air, as is necessary to ensure complete combustion, additional N 2 and 0 2 appears on both sides of the above equation. The heat of combustion is 50.04 M J / k g CH4. Therefore, by fixing a heat input and excess air level the mass flows can be established. In order to calculate the temperature of the hot gases a mean value for specific heat capacity, c P i 9 , was used. With 25% excess air and at HOOK, c P i 9 = 1.35 kJ/kg K . The assumption of constant specific heat capacity is in accordance with all of the other sim-plifying assumptions made for this calculation. 4.2.2 Heat Up of Refractory by Gases The average heat flux from an isothermal gas volume at Tg to its entire bounding gas surface at, for example, Tw is = ewaSB(egT; - agT*) (4.14) where q = heat flux in W / m 2 ew \u00E2\u0080\u0094 emissivity of the refractory wall tg \u00E2\u0080\u0094 emissivity of the gas ag \u00E2\u0080\u0094 absorbtivity of the gas aSB = Stefan-Boltzman coefficient, 5.67 x 1 0 ~ 8 W / m 2 K 4 CHAPTER 4. TEST FURNACE DESIGN 36 Over a short axial length Tg and Tw are constant and so eg and ew can be considered constant. Therefore, equation 4.14 was used to determine the radiative exchange between the hot gas and refractory wall in one slice of the model. eg and ag were calculated using a weighted sum of gray gas model obtained from Barr (1999). A value of 0.87 was used 4.2.3 Heat Up of Bed by the Refractory Walls Figure 4.9 depicts the cross section of the test furnace when it is loaded with charge. Since the exposed surface of the charge \"sees\" only the refractory wall, the rate of heat transfer can be described by the following expression: n a SB (T*w - T*b) Qew-+eb \u00E2\u0080\u0094 \u00E2\u0080\u009E (4-15) where R = + -^w-A-ew Aeb*Peb\u00E2\u0080\u0094tew ^-b-^-eb Tew = exposed wall temperature in degrees K Teb = exposed bed temperature in degrees K Aew = exposed wall area per unit length in m 2 / m Aeb = exposed bed area per unit length in m 2 / m F = view factor In this calculation a uniform bed temperature is assumed, as would be in the case of perfect mixing. Also, the wall is assumed to be of uniform temperature in one axial slice. Hence, Teb \u00E2\u0080\u0094 Tb and Tew = Tw. From Figure 4.9 it can seen that Aeb = chord length XY and Aew \u00E2\u0080\u0094 2nr ( ^ ^ ) - Another assumption made here is that heat transfer between the walls and bed occurs dominantly through radiative exchange between the exposed wall and exposed bed. CHAPTER 4. TEST FURNACE DESIGN 37 4.2.4 Heat Transfer Within the Refractory Walls Heat transfer within the refractory can be described by the transient one-dimensional heat equation expressed in cylindrical coordinates: dT 1 d (, dT\ pCp-at-7d?[krd^)=0 (4.16) Re-arranging, integrating over a control volume and applying Gauss's rule gives: dT lev P P dt faicv) . ( ^ dr) n dA = 0 (4.17) In the one-dimensional case dV = A r and dA = 1. By carrying out the same analysis as for the two-dimensional bed cross section code, Equation 4.17 can be expressed in discretized form as At ripiCp^Ar2 J 2 2 STt r - fc . - - ir i -_ i (2T-^- i ) ] (4.18) Equation 4.18 can now be solved using either a tri-diagonal matrix inversion solver or guass seidel. The code allows for two different types of refractory insulation. Their interfacial con-ductivity is represented by the harmonic mean of the conductivity of the two insulations (Patankar 1980) * k + ki+i (4.19) CHAPTER 4. TEST FURNACE DESIGN 38 4.2.5 Heat Transfer to Atmosphere Heat transfer to atmosphere was calculated using a constant external heat transfer coef-ficient. A value of 7.5 W / m 2 K was used. 4.2.6 Implementing Boundary Conditions At the boundary the rate of heat exchange may be expressed in terms of a radiative heat transfer coefficient. For example, in the case of heat up of the bed, (refer to section 4.2.3) Qew-^eb \u00E2\u0080\u0094 hra(lyew-tebAe\){Tew Teb) (4.20) Equations 4.15 and 4.20 may be combined to give o-SB(Tw + T?)(Tw + Tb) \"rati.etu-feb \u00E2\u0080\u0094 RA > This radiative heat transfer coefficient can be applied at the boundary using the concept of the ghost cell, as in the case of the two-dimensional bed cross section code. For example at the i = I boundary, the heat flux at the surface is, F = h(TI+i - Ta) = \u00C2\u00B1-(Tl-TI+l) = ^ ( T i - T I H ) (4.22) 2 where F is the heat flux normal to the surface and Ta is the ambient temperature, in this case either the temperature of the gas or the temperature of the bed material. On re-arranging, - 4 f c / + . Tj - 2/zAr-r a CHAPTER 4. TEST FURNACE DESIGN 39 4.2.7 Code Testing The problem of transient heat conduction has been solved for some simple geometries. Incopera and Dewitt (1981) presents the solution for the infinite slab, infinite cylinder and sphere in graphical form. Use has been made of the solution for the infinite slab in checking the current code. The one-dimensional grid was modified to represent a thin cylindrical shell of a large inside diameter. The system may then be approximated as a infinite slab for heat transfer analysis thereby allowing the solution to be checked. Tables 4.2 and 4.3 give input data for the code and the graphical calculation described in Incopera and Dewitt (1981). Table 4.4 gives the solutions. It can be seen that the solu-tions are similar although not exactly the same. The difference is due to approximations in reading graphical data. However, this provides confidence that the code is correct. 4.3 Bed Depth and Furnace Sizing The two-dimensional bed cross section code was first run with approximate data. It was found that with a bed depth of 100mm the difference between the surface temperature and temperature inside the bed was about 25\u00C2\u00B0C, which is large enough to be reliably measured. Understanding that a rotary furnace that is up to 20% full may be used for the experimental study, this implied that a test furnace with an inside diameter of 400mm would serve our needs. To confirm this and size the refractory, the two codes described above were utilized. The one-dimensional code was used to size the refractory, ensuring that it could be heated up with a flame and that it would be hot enough and hold sufficient energy to heat the bed. A number of different refractory designs were investigated. They all consisted of a high density, highly conductive inner layer and an insulating outer layer. The final design only is presented here. The final design consisted of a 70mm inner refractory layer of Thor 60 castable (approx. 60% SiC, 23% AI2O3 and 14% SiO^) surrounded by a 20mm layer of Durablanket, see Appendix A . The exact inner and outer diameters were chosen for practical reasons. Table 4.5 gives the design data and computer input data when heat up of the furnace was CHAPTER 4. TEST FURNACE DESIGN 40 investigated. The results for a central slice of the furnace are given in figure 4.10. The refractory temperature profile was then used as input to investigate how a cold bed, with an initial temperature of 30\u00C2\u00B0C, heats up in the furnace. A l l other input data are for this run are given in Table 4.6. The results are presented in Figure 4.11 for a central slice in the furnace. The above results indicate that a furnace of this design may be heated up with a practical flame and, after the walls are hot, there is sufficient energy stored to heat the bed. The temperature distribution within the bed was then investigated using the two-dimensional bed cross section code. This was done as follows. The initial temperature of the bed material was again 30\u00C2\u00B0 C. The inside wall temperature history as the bed is heated was taken from the one-dimensional code and used as the boundary condition in the two-dimensional code. The wall to bed heat transfer coefficients were calculated in the same way as in the one-dimensional code. Also, the rectangular bed perimeter and depth to width ratio were made identical to the real bed, in an attempt to maintain the same resistance to heat transfer between the refractory wall and bed surface. This meant that the cross sectional area and, hence, bed volume were different in the two cases. To ensure that the bed's thermal capacity remained constant the specific heat capacity was modified. The results of the computer run are presented in Figure 4.12 and the temperature profile down the center of the bed is given in Figure 4.13. If one compares Figures 4.11 and 4.12, it can be seen that the bed is heated sufficiently for this experimental program. The two-dimensional results, Figure 4.12, show a slightly slower heating rate. After 20 minutes the average bed temperature is approximately 800\u00C2\u00B0 C with the two-dimensional code but almost 900\u00C2\u00B0 C with the one-dimensional code. This is due to mixing within the bed not being intimate and so the bed surface is hotter than the center of the bed leading to a reduced temperature driving force for heat transfer to the bed. After 30 minutes the bed temperatures are similar, see Figures 4.11 and 4.12. This provides some confidence in the trends of the results. Figure 4.13 shows that after 20 minutes a temperature difference of about 30-40\u00C2\u00B0 C between the center of the bed and bed surface should develop, which is indeed measurable. CHAPTER 4. TEST FURNACE DESIGN 41 From the results obtained one can conclude that this design is suitable for the current experimental program. The furnace can be heated using a gas flame. Then, once hot, the refractory walls will heat up the bed. Finally, during heat up, significant temperature gradients will form in the cross-section. The resulting temperature non-uniformity would be large enough to be measured. CHAPTER 4. TEST FURNACE DESIGN 42 Fy; ij+1/2 Ax x; i-i/2,j o 3> F \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 fy;ij-i/2 x; i+i/2j Figure 4.1: Flux Entering and Leaving Finite Control Volume (Area in 2D case) i = I boundary r o 1 o 0 i- ij ! IJ I+lj i \u00E2\u0080\u0094 ~ l Figure 4.2: Depiction of Ghost Cell at i = / boundary CHAPTER 4. TEST FURNACE DESIGN 43 Figure 4.3: Velocity Profile in Two-Dimensional Grid (0,1) T = 0 T = 0 T = sin(Tty) (0,0) (1,0) Figure 4.4: Boundary Conditions For Test of Transient Conditions: T(x,y,0) = xs'm(iry) CHAPTER 4. TEST FURNACE DESIGN 44 (0,1) T= 1 (0,0) dy dy = 0 (1,1) d r Ac = 2/3 T (1,0) Figure 4.5: Boundary Conditions For Test of Convective Flux at Boundary 1200 1000 800 g & i5 2 600 CD Q . E CD I -400 200 \u00E2\u0080\u00A2 mean bed exposed bed max. bed temp diff. 20 40 60 80 Time (min.) 100 120 140 Figure 4.6: Simulation of Results CHAPTER 4. TEST FURNACE DESIGN 45 1200 1100 1000 g- 900 2 800 700 600 500 400 20 40 60 80 Time (min.) calc. gas o meas. gas calc. bed x meas. bed 100 120 140 Figure 4.7: Comparison with Experiment axial slice refractory direction of calculation Figure 4.8: Refractory Model Structure CHAPTER 4. TEST FURNACE DESIGN 46 Figure 4.9: Cross Section of Test Furnace 1600 0 1 2 3 4 Time (hrs) Figure 4.10: Heat Up of Furnace - Slice 6 CHAPTER 4. TEST FURNACE DESIGN Figure 4.11: Heat Up of Bed - Slice 6 Figure 4.12: Heat Up of Bed Using 2D code CHAPTER 4. TEST FURNACE DESIGN 860 870 880 890 9O0 910 920 930 Temperature (\"C) Figure 4.13: Temperature Profile in Center of Bed after 20 minutes CHAPTER 4. TEST FURNACE DESIGN 49 Table 4.1: 2D Code Input Data for Test Case Bed Width (m) 0.32 Bed Depth (m) 0.08 Active Layer Size (%) 30 Mesh Size 20 x 20 Time increment (min.) 0.05 Free-board to Bed H T C (W/m2 K) 50 Wall to Bed H T C (W/m2 K) 10 Effective Conductivity (W/m K) 1.5 Specific Heat Capacity (kJ/kg K) 1100 Bulk Density (kg/m 3) 1460 Initial Free-board Temp (K) 800 Final Free-board Temp (K) 1150 v T (m/s) 0.0022 Table 4.2: ID Code Input Data for Test Case Inner Radius (m) 5 Outer Radius (m) 5.2 Ambient Temperature (K) 1100 Initial Temperature (K) 100 Heat Transfer Coefficient ( W / m 2 K ) 10 Refractory Conductivity (W/m K) 1 Specific Heat Capacity (kJ/kg K) 1000 Density (kg/m 3) 1000 Table 4.3: Graphical Method Input Data for Test Case Slab Thickness (m) 0.1 Ambient Temperature (K) 1100 Initial Temperature (K) 100 Heat Transfer Coefficient ( W / m 2 K ) 10 Refractory Conductivity (W/m K ) 1 Specific Heat Capacity (kJ/kg K) 1000 Density (kg/m 3) 1000 CHAPTER 4. TEST FURNACE DESIGN Table 4.4: ID Code. Test Case Results Time t* 0* Center Temp. Center Temp, (sec.) (f\u00C2\u00A3) (f^f-) ID code calculated (K) (K) 104 1.0 0.55 550 566 2 x 104 2 0.25 850 845 5 x 104 5 0.024 1076 1072 Table 4.5: Input Data - Heat Up of Furnace Using Flame G Furnace ID (mm) 416 Furnace Length (m) 1.0 Thor 60 Castable (mm) 70 k (W/m \u00C2\u00B0C) 5.9 p (kg/m 3) 2600 c p (J/kg \u00C2\u00B0C) 1100 Durablanket (mm) 20 k (W/m \u00C2\u00B0C) 0.08 P (kg/m 3) 128 c p (J/kg \u00C2\u00B0C) 1130 Fuel (Gas) input (kW) 50 Fuel heat content (MJ/kg) 50.037 Fuel air/fuel ratio (wt/wt) 17.17 Excess air (% of stoich.) 25 Flue gas cp (J/kg \u00C2\u00B0C) 1350 Mean beam length (mm) 416 Part, press. C 0 2 and H 2 0 (atm) 0.25 Inside refractory emissivity 0.87 Initial temp (\u00C2\u00B0C) 20 Ambient temp (\u00C2\u00B0C) 20 External wall to ambient H.T.C. ( W / m 2 \u00C2\u00B0C) 7.5 Radial cells in refractory 90 Axial cells 10 CHAPTER 4. TEST FURNACE DESIGN Table 4.6: Input Data - Heat Up of Bed Using Hot Refractory Walls F i l l (%) 20 Initial bed temp (\u00C2\u00B0C) 20 Bed bulk p (kg/m 3) 1460 Bed c p (J/kg \u00C2\u00B0C) 1100 Covered wall/bed H.T .C . ( W / m 2 \u00C2\u00B0C) h e w ^ e b Bed surface emissivity 0.9 Exp bed to exp wall view factor 1.0 Chapter 5 Description of Experimental Set-Up The dimensions and refractory selected in Chapter 4 formed the basis for the mechanical design of the test furnace that was manufactured for this work. A description of the mechanical design will be given in this chapter. A description of the instrumentation and the experimental procedure used in this study will also be presented here. Prior to the description, the principle of the experimental procedure will be recapit-ulated. In essence the furnace walls are heated using a gas flame. Once the walls are hot, the flame is extinguished and the furnace is charged with an inert granular material, which, for this study, is sand. The furnace is then rotated and, as it rotates, the walls heat the bed. The temperature at select locations in the transverse section is recorded to deter-mine the degree of temperature uniformity in that section. The refractory temperatures are also monitored so that the heat flux to the bed surface may be calculated. 5.1 Furnace Figure 5.1 shows the pertinent details of the furnace. A photograph of the furnace is contained in Figure 5.2. The body of the furnace consists of a 600mm (24\") N .B . , sch. 10, carbon steel pipe that is 1110mm long. The furnace is lined with an approximately 70mm thick layer of Thor 60 castable refractory which is backed with 25mm (1\") of Fiberfax Durablanket. These products were obtained from V R D Canada, in Langley, B C . Their details can be found in Appendix A . The Thor 60 was chosen for its high density (2600 kg/m 3) and high conductivity (5.9 W / m K at 800\u00C2\u00B0 C). Its high density allows it to hold sufficient thermal energy so that once hot it has sufficient energy to heat the bed as required. Its high conductivity ensures that 52 CHAPTER 5. DESCRIPTION OF EXPERIMENTAL SET-UP 53 the hot face temperature remains high enough to provide a thermal driving force between it and the bed. Furnace insulation is provided for by the Durablanket which has a low thermal con-ductivity (0.6 W / m K at 800\u00C2\u00B0 C). The ends of the furnace have 200mm diameter openings that are concentric with the furnace diameter. They are lined with 100mm of Plicast LWI 24R which is a light castable. This material provides thermal insulation. It was also purchased from V R D Canada and is described in Appendix A. The furnace rests on a support structure whose load bearing beams are manufactured from W4 x 13 \"I\" section, see Figure 5.2. The structure holds four rollers on which the furnace rests and is allowed to rotate. The furnace is coupled to a 5.5m long pilot rotary kiln in the high head room area of UBC's Advanced Materials and Process Engineering Laboratory ( A M P E L ) which is in The Brimacombe Building. This pilot kiln, which is described in Barr, Brimacombe, and Watkinson (1989a), drives the test furnace rotation. Further details of the mechanical design of the furnace and the support structure can be found in the engineering drawings contained in Appendix B . 5.2 Instrumentation Instrumentation for the furnace allows for temperature monitoring of the high density castable, which is used as the heat source for the bed, and also the temperatures within the bed. 5.2.1 Refractory Temperatures Inserted into the refractory at three axial and two radial locations are nine type K ther-mocouple probes. The exact locations are shown in Figure 5.3. At each axial location two probes are designed to indicate the refractory hot face temperature and the other one is designed to measure the refractory cold face temperature. Note that the refrac-tory cold face temperature is in fact the interface between the Thor 60 castable and the Durablanket. CHAPTER 5. DESCRIPTION OF EXPERIMENTAL SET-UP 54 The thermocouples were manufactured in-house using either 18 or 20 A . W . G . type K thermocouple wire and ceramic sheaths. The hot junction was formed by welding the wires and it was embedded into the refractory at the locations indicated on Figure 5.3. A l l nine of the thermocouples are connected to a 2 pole 10 position rotary selector switch, model OSW3-10, and then to a single input digital thermometer, model HH21. Cold junction compensation is included for electronically within the thermometer. Both of these instruments were supplied by Omega and details can be found in Appendix C. The instrumentation set-up is depicted in Figure 5.4. As one can deduce from the above description, the refractory temperatures are to be manually recorded. Refractory temperatures are indicated on the digital thermometer after selecting the desired thermocouple using the rotary switch. These temperatures are then manually recorded by the experimentalist. 5.2.2 Bed Temperatures Figure 5.5 shows the probe that was designed specifically to measure bed temperatures. Its dimensional details are given in Figure 5.6. The probe consists of four thermocouples, each one passing through a 3/16\" by 0.035\" wall stainless steel tube. The thermocouples are type K and were produced using duplex insulated thermocouple wire. This wire has an A . W . G . number of 20 and has nextel ceramic insulation. It was purchased from Omega and details can be found in Appendix C. At the hot junction of each thermocouple the insulation has been removed and the wire has been passed through a 25mm long ceramic sheath and then welded together to form an exposed hot junction. From Figures 5.5 and 5.6 one can see that the thermocouple ends are positioned at different radial locations. These locations were selected to obtain the temperature distri-bution in the cross-section of the bed. Figure 5.7 gives the location of each thermocouple in the bed cross section. For data logging, each thermocouple in the probe was connected to a signal con-ditioning board to convert the thermocouple signal to a i 5V signal. Cold junction CHAPTER 5. DESCRIPTION OF EXPERIMENTAL SET-UP 55 compensation was also carried out by this board. The signal conditioning board was connected to a data acquisition board in a 486 computer that ran at 66MHz with 24MB of R A M . Figure 5.5 shows a photograph of the set-up with the bed probe and Figure 5.4 depicts it diagrammatically. The signal conditioning board is model CIO-EXP-16 and the data acquisition board is model CIO-DAS08. Both of these items were purchased from Computer Boards in Vancouver and further details of them can be found in Appendix C. 5.3 Cal ibra t ion The portable digital temperature indicator used to measure the thermocouples in the refractory was purchased calibrated to within 0.1% or 1\u00C2\u00B0C accuracy. This instrument was used to check the calibration of the bed temperature data logging equipment. In order to check the calibration of the data logging equipment a micro-volt signal generator was used. A signal was provided across the terminals that the thermocouples were connected to on the data logging equipment. The temperatures displayed on the computer were subsequently compared to those displayed by the digital thermometer to determine accuracy. Table 5.1 shows the results of this test and it can be seen that the data logging equipment is no more than 4\u00C2\u00B0C inaccurate in approximately 680\u00C2\u00B0C, which is quite acceptable for this study. Stability of the data logging equipment was checked using the same signal generator and recording the temperature for one minute. Figure 5.8 shows the results. It can be seen that the signals were quite stable. The electrical technician responsible for the signal generator advised that the slight drift was most likely caused by drift in the signal generator. 5.4 Preparat ion of B e d Mate r i a l Sand was selected as the bed material for this study. It is inert, readily available and its properties are well known. Two types of sand were used and both were obtained from CHAPTER 5. DESCRIPTION OF EXPERIMENTAL SET-UP 56 Target Products Ltd. in Burnaby, B C . One was a coarse sand and referred to as \"1/16 to 1 /8 inch filter sand and gravel\". It was composed of 66% silica with the remainder largely alumina with some iron oxide and calcium oxide. The other sand was finer and referred to as \"10-20 filter sand\". The numbers indicated approximate mesh size range. This sand was over 93% silica. While it is appreciated that the two sands used were not of a consistent composition, the properties of interest in this study were relatively similar. These properties are specific heat capacity, thermal conductivity and density. Also the particle shapes were irregular and appeared to be similar and, as will be reported in the results, the dynamic angle of repose was constant. Although no measurements were carried out, it was expected that the coefficient of restitution for the sand particles was also similar. Further details, including composition and MSDS sheets, of the sand are provided in appendix D. Photographs of the sand are contained in Figures 5.9 and 5.10. The purpose of this study was to investigate the effect of particle size distribution on temperature gradients in the cross-section of the of the bed. Therefore, a number of sand mixes with varying size distributions were prepared using the two sands described above. A sieve shaker, model T S l manufactured by Gilson Screen Company, with appropriately sized sieves was used to prepare the sand mixes. Six mixes in total were prepared for this study. They ranged from a wide size distri-bution to a narrow one. Table 5.2 gives details of the sand mixes and the size distribution charts are given in Appendix E. Each mix has been characterized by an average particle size and standard deviation. The average particle size is based on volume and calculated as follows where mt- and d,- are the mass and characteristic linear dimension1 respectively for each size fraction and M is the total mass of the sample. For comparative purposes the Sauter 1The mean size of the aperatures of the two sieves used to create a particular size fraction have been used. For example, for the size fraction greater than 1.0mm and smaller than 1.4mm, d, is 1.2mm (5.1) CHAPTER 5. DESCRIPTION OF EXPERIMENTAL SET-UP 57 mean diameter, which is the diameter of a particle with the same volume to surface area ratio as the bulk, was also calculated. The Sauter mean diameter, dsp, is calculated as \u00C2\u00A3 n i d i where n,- is the number of particles in size fraction z, of characteristic linear dimension di. The standard deviation is calculated by Y>,( rt co co co co co o s Mrt o CO XT CM CM CO CO OO CM o f\u00E2\u0080\u0094I rt CO 1 3 OJ t H rt PH OJ rt OJ P CM i O I 1:1 .2 | r2 O .3 rt .3 O \u00C2\u00AB f ? <+H OJ 6 s * a a; rt P Pi 3 \u00C2\u00A7 e P4 CO o 0) 13 m rt CO p rt V f t rt ft 3 OJ rt 's! ? t d . rt OJ R 13 co o oq co co CM CO r\u00E2\u0080\u0094I I IO o co CO I o CO CM CM CM CM i co o CM i CM CM 05 Ol o o co o co 05 o CO co CO I o CM CM CM CM T\u00E2\u0080\u0094I I co CM 05 co co co rt .2 o. ..\u00E2\u0080\u0094i S h O to OJ p 13 13 P. rt T i rt rt 13 b, to the bed as dTb Qw^b = M f c C p t \u00E2\u0080\u0094 (6.4) where Mb = mass of bed (kg) cp,b = specific heat capacity of bed (kJ/kg K) = rate of temperature change of average of B l to B4 (degrees K per second) dfb dt ^ is calculated simply by calculating A ^ t from the discrete data points collected by the data-logger. Having calculated the heat taken up by the bed, Qw-+b, an overall heat transfer coef-ficient, ha, can be calculated using Ab(Tw \u00E2\u0080\u0094 Tbs) ^ ^ In Equation 6.5 Tw is the wall temperature. Section 6.1.2 described that the inside wall CHAPTER 6. RESULTS AND DISCUSSION 85 temperatures were measured by thermocouples R l , R2, R4, R5, R7 and R8, see Figure 6.9. The temperatures recorded by these thermocouples were generally similar except for R5, which was consistently higher. No apparent reason for this was discovered but it is unlikely that one region of the refractory surface is much hotter than the rest. As such it can be assumed that the average of R l , R2, R4, R7 and R8 represents the hot wall temperature, Tw. Ai, in Equation 6.5 is the area of the bed surfaces exposed to the wall. To calculate this refer to Figure 4.9. Ab is the straight distance X Y plus the arc distance X Y multiplied by the length of the furnace. Tbs is the temperature of the bed surface and may be assumed to be the temperature measured by T B 4 . T B 4 is closest to the covered bed area in contact with the wall. However, it also can be assumed to represent the exposed bed surface temperature. This latter assumption is valid if one considers the movement of particles in the cross section, as depicted in Figure 6.22. Particles entering the plug flow region maintain their radial position since motion in this region as a rigid body. Therefore, the temperature recorded by T B 4 is likely to be the temperature of particles closest to the exposed bed surface as well. The above arguments indicate that Tw = Tew = Tcw and Tbs \u00E2\u0080\u0094 Teb = Tcb- This can be used to calculate the rate of heat transfer from the exposed wall to the bed surface by radiation 2 using Qew^eb = \u00E2\u0080\u009E (6.6) 2 A t the temperatures of the experiment and since there was no gas flow over the bed, heat transfer by conduction and convection is negligible. CHAPTER 6. RESULTS AND DISCUSSION 86 where T e U; = wall temperature in degrees K (T^) Teb = exposed bed temperature in degrees K (TBA) Aew = exposed wall area per unit length in m 2 / m Aeb = exposed bed area per unit length in m 2 / m J- = view factor In this expression Kirchoff's law is assumed, i.e. a = e = 1 \u00E2\u0080\u0094 p. Also Aeb = chord length XY and Aew = 2rrr (^~), on Figure 4.9. Finally, the difference between Q w ^ b and Qew-teb must be the rate of heat flow between the covered wall and covered bed, Qcw-ycb- Qew-teb and Qcw-icb can then be used to calculate their respective heat transfer coefficients, h e w ^ e b and hcb^cw. > Qew-yeb ,\u00E2\u0080\u009E A-eb\J-w ~ J-eb) i Qcw-tcb ,\u00E2\u0080\u009E c^ h c ~ * * \" 4*(r\u00C2\u00AB - T*) ( 6 - 8 ) where A eb = straight distance X Y on Figure 4.9 mutiplied by furnace length (m2) A^ = arc distance X Y on Figure 4.9 mutiplied by furnace length (m2) TW = wall temperature in degrees K = exposed bed temperature in degrees K (TBA) Tcb = covered bed temperature in degrees K (TBA) Another quantity that has been calculated is the ratio of the heat transferred through CHAPTER 6. RESULTS AND DISCUSSION 87 the covered bed surface and exposed bed surface, where (6.9) 6.3.2 Results Obtained The wbove analysis has been carried out for all runs. Therefore, for every run a plot of the heat rate to the bed, Qw->b, the overall heat transfer coefficient, hQ, the exposed wall to exposed bed heat transfer coefficient, hew^eb, the covered wall to covered bed heat transfer coefficient, hcw^cb, and the ratio of heat transferred across the covered bed surface and exposed bed surface, \Ir, have been made. These plots are contained in Appendix F. To highlight the trends observed, Runs 4, 5 and 8 will be used as examples. Figure 6.23 shows the heat rate to the bed, Qw^bi a n d the temperature driving coef-ficient, Tw \u00E2\u0080\u0094 Tbs, for Run 5. It can be seen that, as expected, the heat rate to the bed is initially high but tapers away towards the end of the test as the temperature driving force (difference between Tw and 2],) diminishes. The heat rate to the bed and temperature driving force profile for Run 4 is similar, see Figure 6.24. This trend was found in all runs. The overall heat transfer coefficient, h0, profile, however, was not always similar. In Run 5, it was relatively constant, see Figure 6.25. In Run 4, Figure 6.26 it followed a wave-like trend, gradually rising, falling and rising again. Overall it was found that the trend for this curve was different from one run to the next. It should be noted that during the final minutes of the plot, while the bed temperature is closer to isothermal conditions, it also is similar to the wall temperature. Therefore, slight errors in the thermocouple readings manifest into large errors in the value of the heat transfer coefficients. As such, in some runs, such as Run 4, the overall heat transfer coefficient becomes very large and/or negative. The ratio between the heat flow across the covered bed surface and exposed bed surface, ^ for Run 5 is shown in Figure 6.27. While the profiles were slightly different for all runs, \&, was always more than 1.0 indicating that more heat is transferred across the covered bed surface. Table 6.4 gives the average values of $ between 5-6 minutes and 2-3 revolutions for all runs. The mechanism for heat transfer across the covered bed surface is CHAPTER 6. RESULTS AND DISCUSSION 88 radiation and conduction between the covered wall surface and the sand particles adjacent to it. Since the covered wall area is much less than the exposed wall area, its resistance to radiative heat exchange is greater. However, since the heat rate between the covered wall and covered bed is higher, then conduction heat transfer must be significant. As such, the covered wall to covered bed heat transfer coefficient must be a function of both radiation and conduction. The rate of heat transfer across the covered bed and exposed bed surfaces was used to calculate the respective heat transfer coefficients. As an example Figure 6.28 shows the profile of the covered wall to covered bed heat transfer coefficient and the exposed wall to exposed bed heat transfer coefficient. Prior to further analysis with heat transfer coefficients, it is worth determining the sensitivity of the results to input parameters. The bed emissivity was taken to be 0.9, as quoted for sand in Incopera and Dewitt (1981). However, since the bed is granular, the effective emissivity may be larger. This is because any radiation reflecting off the bed particles would most likely be directed towards another particle. Therefore, the bed absorptivity is closer to 1.0 and consequently, through Kirchoff's law, the emissivity would also be 1.0. Also, the wall emissivity was taken as 0.87, as quoted for silicon carbide (the major constituent of the refractory) in Incopera and Dewitt (1981). However, the emissivities for other refractories are closer to 0.75. Calculations with a bed emissivity of 1.0 and refractory emissivity of 0.75 were carried out using the data for Run 5. Figure 6.29 shows the effect on the covered bed to covered wall heat transfer ratio. It can be seen that the change is some 5-10% in either case. For all analyses bed depth was constant and the depth was measured to be 100mm. To investigate how sensitive the results are with bed depth the heat rate to the bed and overall heat transfer coefficient was calculated for a 110mm bed also. Data for Run 5 was again used and the results are given in Figure 6.30. A small change can be seen. The sensitivity analysis indicates that the results are only mildly affected by input data, other than temperatures. CHAPTER 6. RESULTS AND DISCUSSION 89 6.3.3 Effect of Particle Size Distribution As in the case of the analysis of temperature non-uniformity, Section 6.2, the heat transfer coefficients and heat rates in each run were compared by calculating the average values over a one minute period. Also, for similar reasons, the averages over the same time periods, 5-6 and 9-10 minutes and 2-3 and 6-7 revolutions, were used. Table 6.4 gives results for values at 5-6 minutes and 2-3 revolutions for all runs. Figure 6.31 shows the average overall heat transfer coefficient, ha, plotted against the normalized particle size standard deviation, cr*, for all runs. The average heat rate to the bed, Qu>-\u00C2\u00BB6, is plotted in Figure 6.32. The above results imply that particle size distribution has little effect on overall heat transfer and heat rate to the bed. To further investigate this the final bed temperature after each run was plotted with particle size standard deviation, see Figure 6.33. It would appear that particle size distribution does not influence the final bed temperature either, which agrees with the premise made in the first sentence of this paragraph. One question that can be asked is whether the temperature uniformity within the bed has an influence on the heat rate to the bed. It would be expected that a well mixed bed of uniform temperature would be a better heat sink. This is because its surface temperatures would be lower than those in a bed with cooler particles in the middle and warmer particles at the surfaces. This would lead to a greater temperature driving force for energy transfer between the heat source, refractory walls in the current case, and bed. Figure 6.34 shows a plot of averaged temperature range, Q T , and average heat rate to the bed, Qw-+b- The data points are scattered indicating that no relationship was evident. The effect of furnace rotational speed was investigated to see how much heat transfer is influenced by it. Figure 6.35 shows the relationship between overall heat transfer coefficient, h0, and rotational speed. Figure 6.36 show the shows the relationship between average heat rate to the bed, Qw^bi and rotational speed. For the earlier data, between 5 and 6 minutes, increasing rotational speed can be seen to increase heat transfer coefficient and heat rate to the bed. Between 9 and 10 minutes the same trend is not clear. Increasing rotational speed increases the rate of material mixing within the bed and CHAPTER 6. RESULTS AND DISCUSSION 90 consequently the adjective component of heat transfer is higher. As such the data for 5-6 minutes is understandable, increasing rotational speed increases overall heat transfer coefficient and heat rate to the bed. This provides some credibility to the experimental results. Between 9-10 minutes one would expect a similar trend but the reason for not finding it are likely related to the lower temperature driving force, Tw \u00E2\u0080\u0094 Tb, at that time. Slight errors in the temperature measurements may lead to this. The error bars on Figures 6.35 and 6.36 were calculated based on the range of data for the runs at 1 rpm, which were the same and so demonstrate repeatability. The error ranges further highlight the wide data scatter. 6.4 Overal l Comments The results provided little evidence of a relationship between particle size distribution and temperature uniformity in the cross-section. Since particle segregation was evident, heat transfer within the bed, therefore, appears to be a function of bed effective ther-mal diffusivity (due to effective conductivity and advective transfer) rather than overall particle flow and re-arrangement. Prior to making such a statement, however, other reasons for not finding a relationship between particle size distribution and temperature uniformity should be considered. It is possible that during each test particle motion was in a transient mode and segregation had not yet established. This would mean that the bulk particle flow in all runs was similar, leading to similar heat transfer rates. However, pertinent data from each test was obtained within 20 minutes, which equates to 17 revolutions at 1 rpm, since the furnace was not rotated for the first 3 minutes. Work done on the kinetics of segregation indicates that transverse mixing is, in fact, a fast process. The work of Rogers and Clements (1972) indicated that segregation was well developed after 10 revolutions but Nityanand, Manley, and Henein (1986) found it to be faster, within 1 revolution. The former workers did not take data prior to 10 revolutions since their tests were conducted at higher rotational speeds. Since the current tests were conducted for 17 revolutions, it is likely that the bed was segregated for a significant part CHAPTER 6. RESULTS AND DISCUSSION 91 of the test and so it is unlikely that the kinetics of segregation account for not finding a relationship between particle size distribution and temperature uniformity. If the rate of particle mixing in the active layer is so high that the active layer is effectively isothermal, then all material entering the plug flow region is at the same tem-perature. Any temperature difference in the plug flow region would then only occur due to radial heat diffusion. This would make all of the results similar. However, since mate-rial in the plug flow region remains there for no more than 30 seconds before re-entering the active layer, the temperature of B l should be higher than the temperature of B4 30 seconds earlier. This was not found to be the case and so an isothermal active layer could not explain the results. If indeed particle size distribution does not affect temperature non-uniformity within the bed, then an explanation of phenomena that lead to the notion that it does is war-ranted. In industry it was observed that finer particles of lime remained uncalcined (Sun-nergren 1979; Von Wedel 1973). Material flow studies, e.g. (Henein 1980), identified that finer particles remained in the center of the bed, creating the so called \"kidney\". From this an inference was made that the segregation of the finer particles prevented them from traveling to the exposed bed surface and receiving heat from the walls and free-board. The findings of the current study imply that a core of segregated fine particles may not create a cold center, but rather, a cold core is a result of poor heat transfer through the bed. Naturally, in a lime kiln for example, the finer particles are in the middle and hence would remain uncalcined if the bed core was cold The heat transfer analysis, Section 6.3, also showed no relationship with particle size distribution. The lack of a relationship between the degree of temperature uniformity, O T 5 and average heat rate to the bed, Qb^w is particularly interesting. It implies that transverse heat transfer to the bed may be independent of heat transfer within the bed. Therefore, a one-dimensional model of a kiln that only considers overall heat transfer between the free-board and bed can provide useful information about kiln performance. Analysis of temperature gradients within the bed can be carried out independently which lends credibility to pseudo three-dimensional models such as Boateng and Barr (1996b). In experiments like this at such high temperatures, errors are inevitable. The experi-CHAPTER 6. RESULTS AND DISCUSSION 92 mental procedure attempted to minimize these but some of the errors were: \u00E2\u0080\u00A2 Differences in initial hot refractory temperature between each run. \u00E2\u0080\u00A2 Differences in the time the sand bed was stationary inside the furnace prior to rotation. \u00E2\u0080\u00A2 Disruption of particle flow by the bed probe, which could have enhanced particle mixing. CHAPTER 6. RESULTS AND DISCUSSION 93 CHAPTER 6. RESULTS AND DISCUSSION CHAPTER 6. RESULTS AND DISCUSSION CHAPTER 6. RESULTS AND DISCUSSION 6. RESULTS AND DISCUSSION FRONT VIEW 60 Dimensions in mm, U.O.S. Not To Scale Figure 6.9: Location of Thermocouples in Refractory CHAPTER 6. RESULTS AND DISCUSSION CHAPTER 6. RESULTS AND DISCUSSION CHAPTER 6. RESULTS AND DISCUSSION 106 SH H-l 13 OJ cn rt rt rt S-i OJ o rt rt (H <4H o fH OJ rt OJ O O J C O rt1 ^ H II SH OJ P H e 5 H ^ * ft 43 ^ OJ II rt i . f t HH 'IS 0 rt CH SH OJ O i-t \u00C2\u00A3 -s \u00C2\u00AB .a CO 03 rt 13 bfj JH CHAPTER 6. RESULTS AND DISCUSSION 111 CHAPTER 6. RESULTS AND DISCUSSION 112 CHAPTER 6. RESULTS AND DISCUSSION c E o I CD < c 'E CD LO O \u00E2\u0080\u00A2 r \u00E2\u0080\u00A2 1 -* \u00E2\u0080\u00A2 1 \u00E2\u0080\u0094 c \u00E2\u0080\u0094f K I |-\u00E2\u0080\u00A2 r\u00E2\u0080\u0094 0 \u00E2\u0080\u0094 | 1 o d o M PH QJ ^ Q \u00C2\u00AB? 13 ^ OJ II m QJ ^ d II taO. ft CHAPTER 6. RESULTS AND DISCUSSION 123 13 CO CO Cu \u00E2\u0080\u00A2\u00E2\u0080\u0094t rt o hp < CO CXI O o < < < OO o O < < J\u00E2\u0080\u0094<_> S J cu Q T3 i CO T3 C CO CO TD V N E CD N CO o o '\u00E2\u0080\u00A2c CO CL (>1 2LU/M) \u00C2\u00B0u luapjuaoo jajsueji 1B8H | |BJ9AO 3 A V o CO o d d d d CO m CO o O i-i CO d -<-= rc3 CO W ce CO > o d o d .2 '+=> d r2 '(H -(J CC \u00E2\u0080\u00A2 \u00E2\u0080\u00A2\u00E2\u0080\u0094l P co N in jo TJ PH o fct=l CO CD CU d CHAPTER 6. RESULTS AND DISCUSSION O < o o o a o CD < < O < u u oo <1 o < < o o o o o o o CD in TJ- co C\J ^ -(M>0 0 'P^ a oi ajey 1B9H e6ejaAv CHAPTER 6. RESULTS AND DISCUSSION 0 in oo o o to o m o LO \u00E2\u0080\u00A2* c\u00C2\u00BB (OB) sainujiu oz J8UV ejniBjeduuai psa CHAPTER 6. RESULTS AND DISCUSSION 126 CHAPTER 6. RESULTS AND DISCUSSION 127 CN CM c E o i CO < c E CD i CO O \u00E2\u0080\u00A2 -> fH O H-= O OJ OJ bO 9 O CO OJ fH rt H-= rt fH OJ r l \u00E2\u0084\u00A2 o o OJ cj MH fH OJ rt rt rt OJ W <+H o CO H-= MH O rt W t -H ^ OJ <+H o o >> g '> P co bO rt rt OJ cn CD a o MH aj to rt rt .rt Q \u00E2\u0080\u00A2a O rt o CM co co oj n r r t 13 I I O O PH a OJ O O o ,rt ^ MH fH OJ PH l\u00E2\u0080\u0094l ,0J F H OJ O (H HJ PH o a OJ ' ^ CO a s J2 rt Irt fH o v\u00E2\u0080\u0094' y co rt CD * a OJ ri .2 .rt FH CO ^ O 9 + X io - H C M ^ , + CM O Oi CO iO OO CI CD O N CM CM CM CO CM C) N 0 0 ^ O CO Oi CM CO t\u00E2\u0080\u0094 oo co oo oo oo co co io i\u00E2\u0080\u0094i ^ r- Oi i>- t~-O O O O O oo co oo i o 0 CM CM CM CM CM m o r\u00E2\u0080\u0094 c-co co IT- co co rt CQ CO < i\u00E2\u0080\u0094I CM CO ^ rt -P. ^ \u00C2\u00A3 rt H cj CHAPTER 6. RESULTS AND DISCUSSION 132 Table 6.4: Average Values for Temperature Range, Overall Heat Transfer Coefficient and Heat Rate Run f ir hQ 5-6 min 5-6 min \u00C2\u00B0C W / m 2 3 176.3 105.2 4 107.2 174.1 5 105.5 124.9 8 177.5 102.4 9 103.0 152.5 10 95.4 131.1 11 86.7 104.7 12 88.3 119.5 13 113.3 86.9 14 130.6 126.8 15 210.8 116.4 16 252.8 106.3 17 96.5 112.2 18 124.8 164.8 19 85.0 103.7 Q * ClT I-6 min 5-6 min 2-3 revs kW \u00C2\u00B0C 28.4 4.0 176.3 36.2 6.2 107.2 31.9 4.7 105.5 30.7 3.7 177.5 34.1 5.3 91.0 28.2 4.2 95.4 32.1 3.5 86.7 36.7 4.7 88.3 31.3 3.5 113.3 35.6 4.0 130.6 28.4 3.1 205.2 23.6 3.1 232.3 22.9 3.0 183.8 31.4 4.6 172.6 28.4 3.8 60.7 h0 Q * 2-3 revs 2-3 revs 2-3 revs W / m 2 K kW 105.2 28.4 4.0 174.1 36.2 6.2 124.9 31.9 4.7 102.4 30.7 3.7 132.7 57.5 2.6 131.1 28.2 4.2 104.7 32.1 3.5 119.5 36.7 4.7 86.9 31.3 3.5 126.8 35.6 4.0 114.3 27.0 3.0 102.9 22.8 3.0 110.4 27.7 3.1 149.7 45.8 5.3 91.5 33.6 3.8 Chapter 7 Conclusions and Recommendations 7.1 Conclusions In this study high temperature heat transfer in the cross-section of a particulate bed in a rotary furnace has been investigated through experimentation. A refractory lined batch furnace (I.D. = 400mm, Length = 900mm) was designed and manufactured. It was heated using a gas flame and, when the walls reached 1100\u00C2\u00B0C, the gas flame was extinguished and the furnace was charged with sand and rotated. The energy in the walls was used to heat the sand bed. The wall temperature was monitored using thermocouples embedded in it and the temperature rise of the sand bed was monitored using a probe with four thermocouples located at different radial locations. The duration of each test run was about 20 minutes. The wall and bed temperature values allowed the rate of heat transfer from the wall to the bed to be calculated. The four temperatures in the bed itself also allowed one to determine the heat uptake by the bed and to quantify the degree of temperature uniformity. Such experiments have not been reported in the literature and so the first objective of this study was successfully met. In lime kilns it has been reported that when the particulate feed has a wide size distribution, the fine particles may remain unreacted. It is suspected that since the fines segregate to the center of the bed, they are shielded from the hot free-board gases, that is, segregation hinders cross sectional heat transfer. To investigate this, six sand mixes with varying size distributions were tested in this study. The size distribution was quantified by measuring the particle size mean and standard deviation. The uniformity of the bed temperature was quantified by calculating the range of temperatures recorded by the four bed thermocouples. Most of the runs were carried out with a rolling bed. The rotational speed was 1 rpm or greater and the fill was approximately 20% of the cross-sectional area. The bed tem-perature profiles were similar in all cases. The uppermost thermocouple, which is closest 133 CHAPTER 7. CONCLUSIONS AND RECOMMENDATIONS 134 to the exposed surface of the bed, was always the coolest and the lowest thermocouple, closest to the refractory walls, was the warmest. This implied that the upper thermocou-ple was in the cold core of the bed, below the active layer. This thermocouple was only 15 mm from the bed surface. While it would have been useful to have the active layer temperature, an additional thermocouple would have over agitated the sand particles and so affected the results. The bed temperature profiles were generally smooth in nature. Two runs were carried out with a slumping bed, with a rotational speed of about 0.4 rpm. In these runs a step wise temperature rise was observed, which corresponds with the mixing nature of such bed motion. This result provides confidence that the probe and instrument set up was able to respond adequately to temperature changes within the bed. The refractory hot face temperature was found to be relatively constant for the du-ration of all of the tests. This indicated that the refractory was able to hold sufficient energy for heating the bed to the required temperatures. Also, its conductivity was high enough so that the rate of energy transfer to the hot face, and ultimately the bed, ensured that the required bed temperatures were achieved. Therefore, the furnace was meeting its design objectives. In all runs it was found that the spread of the temperatures in the bed was between 100-200\u00C2\u00B0 C initially. After about 20 minutes the bed was virtually isothermal. Since the four bed thermocouples were always located at the same position and the furnace fill was constant, the spread of these four temperatures was used to make comparisons between the runs. The average spread between 5 and 6 minutes was selected. A plot of this value with particle size distribution demonstrated a somewhat random relationship. This would imply that, for the ranges studied, the particle size range has minimal effect on temperature gradients. At 1 rpm the temperature spread within the bed increased as particle size distribution increased, but there was considerable data scatter and insufficient data points at all size distributions. Further tests at 1 rpm with the existing sand mixes and sand mixes with wider and narrower size ranges should to be carried out to confirm this finding. By making the assumption that the average of the four bed temperatures indicated CHAPTER 7. CONCLUSIONS AND RECOMMENDATIONS 135 average bed temperature the heat rate to the bed was calculated. Furthermore, by as-suming that the average of the hot face thermocouples represented the average hot face temperature, heat transfer coefficients between the wall and bed surface were calculated. The wall to bed overall heat transfer coefficients were in the range 85-175 W / m 2 K . The values tended to either remain relatively constant for the duration of the test runs, or their profiles were wave-like. These data represent a significant contribution to the literature since very few studies are available at realistic kiln temperatures and for kiln diameters above 0.25m. The average values of heat rate to the bed and heat transfer coefficient at 5-6 minutes were compared with particle size distribution and it was found that the latter did not affect the former. This was not surprising considering that temperature uniformity was also not affected by particle size distribution. It was interesting to note, however, that a bed with a wide temperature range (highly non-uniform) did not necessarily result in a lower heat up rate, implying that overall transverse heat transfer may be independent of heat transfer within the bed. Therefore, a one-dimensional model of a kiln that only considers overall heat transfer between the free-board and bed can provide useful information about kiln performance. Analysis of temperature gradients within the bed can be carried out independently which lends credibility to pseudo three-dimensional models such as Boateng and Barr (1996a). The mechanism of heat transfer from the exposed wall to the exposed bed is radiation. Conduction is negligible since particles at the exposed bed surface are not in contact with the walls. Also there was no gas flow and so convection is also negligible. The rate of heat transfer was, therefore, calculated using radiation theory. Since the overall rate of heat transfer to the bed was calculated also, the difference between these two values was the covered wall to covered bed heat fiowrate. The ratio of heat transferred from the covered wall to covered bed and from the exposed wall to exposed bed surface was, therefore, also determined. It was discovered that the rate of heat transferred across the covered bed surface was between 3 and 6 times as much as the rate across the exposed bed surface, especially during the initial stages of the test runs. This result highlights the importance of conduction between the covered wall and bed particles adjacent to it. CHAPTER 7. CONCLUSIONS AND RECOMMENDATIONS 136 In general it was found that particle size distribution has little or no impact on thermal gradients within the bed or overall heat transfer to the bed. One possible reason for this finding is that segregation had not established during the tests, so that the particle flow was the same in all runs. Another is that the active layer was so intensely mixed that it was effectively isothermal, and so radial temperature gradients were due only to thermal diffusion in the plug flow region. Both of these were negated for reasons described in the discussion. As such, for the size ranges tested, particle size range had little impact on the transverse heat transfer. This study has reported a substantial quantity of heat transfer data within the bed at temperatures close to those found in industrial rotary kilns. Such data has not been available for general use in the past. Use of this data can be made for the purposes of validating mathematical models. As such, this constitutes one of the recommendations. 7.2 Recommendations For Future Work Based on the findings of the current study the following work is recommended: \u00E2\u0080\u00A2 Further tests at 1 rpm with the existing sand mixes and sand mixes with a wider size distribution should be carried out. This will confirm whether there is a relationship between size distribution and temperature non-uniformity. \u00E2\u0080\u00A2 The data reported should be used to validate either existing two-dimensional models or to develop new ones. Such models give a clear picture of the temperature profile in the cross-section of the bed and, hence, can provide greater insight into the heat transfer characteristics. Once validated, such models can be modified and applied to industrial kilns \u00E2\u0080\u00A2 An appropriate bed model requires a knowledge of the material flow profile, in particular the size, shape and location of the active layer. A novel technique has been developed at the University of Birmingham, U K , and described in (Parker, Dijkstra, Martin, and Seville 1997). Such a technique, or indeed any suitable one, should be applied to cold tests using the same materials in the current study to obtain a velocity profile. CHAPTER 7. CONCLUSIONS AND RECOMMENDATIONS 137 \u00E2\u0080\u00A2 Another unknown in the bed model is the material effective conductivity in the active layer. The effective thermal conductivity in the plug flow region can be determined using models developed for packed beds, e.g. Schotte (1960). In the active layer the effective conductivity is enhanced due to particle mixing. The degree of particle mixing is, therefore, likely to be related to the effective conductivity. As such a profile of the mixing in the active layer, defined perhaps using the granular temperature (Boateng 1993), needs also to be measured. This can then be associated with the effective thermal conductivity enhancement term to develop the effective conductivity profile in the active layer. Bibliography Alyaser, A . H . , P. V . Barr, and J . K . Brimacombe (1997). Modelling of axi-symmetric flow and combustion in an industrial scale furnace. 5th Annual Conference of Com-putational Fluid Dynamics Society of Canada, 23-28. Barr, P. V . (1986). Heat Transfer Processes in Rotary Kilns. Ph. D. thesis, University of British Columbia, Vancouver, B C , Canada. Barr, P. V . (1999). Emmisabsor - computer subroutine. Barr, P. V . , J . K . Brimacombe, and A . P. Watkinson (1989a). A heat-transfer model for the rotary kiln: Part i , pilot kiln trials. Metallurgical Transactions B 20B, 391-402. Barr, P. V . , J . K . Brimacombe, and A . P. Watkinson (1989b). A heat-transfer model for the rotary kiln: Part i i . development of the cross-secton model. Metallurgical Transactions B 20B, 403-419. Baumann, G. , I. M . Janosi, and D. E. Wolf (1994). Particle trajectories and segregation in a two-dimensional rotating drum. Europhysics Letters 27(3), 203-208. Boateng, A . A . (1993). Rotary Kiln Transport Phenomena: Study of the Bed Motion and Heat Transfer. Ph. D. thesis, University of British Columbia, Vancouver, B C , Canada. Boateng, A . A . and P. V . Barr (1996a). Modelling of particle mixing and segregation in the transverse plane of a rotary kiln. Chemical Engineering Science 51(17), 4167-4181. Boateng, A . A . and P. V . Barr (1996b). A thermal model for the rotary kiln including heat transfer within the bed. Int. J. Heat Mass Transfer 39(10), 2131-2147. Bui , R. T., G. Simard, A . Charette, Y . Kocaefe, and J . Perron (1995). Mathemati-cal modeling of the rotary coke calcining kiln. The Canadian Journal of Chemical Engineering 73, 534-545. Carley-Macauly, K . W. and M . B . Donald (1962). Particle mass segregation in a two-dimensional rotating drum. Chemical Engineering Science 17, 493-506. 138 BIBLIOGRAPHY 139 Chapra, S. C. and R. P. Canale (1988). Numerical Methods For Engineers. McGraw-Hi l l , Inc. Dolgunin, V . N . , A . N . Kudy, and A . A . Ukolov (1998). Development of the model of segregation of particles undergoing flow down an inclined chute. Powder Technol-ogy 96, 221-218. Gorog, J . P., J. K . Brimacombe, and T. N . Adams (1981). Radiative heat transfer model in rotary kilns. Metallurgical Transactions B 12B, 55-70. Henein, H . (1980). Bed Behaviour in Rotary Cylinders with Applications to Rotary Kilns. Ph. D. thesis, University of British Columbia, Vancouver, B C , Canada. Hottel, H . C. and A. F . Sarofim (1967). Radiative Heat Transfer. New York: McGraw-Hil l Book Company Inc. Imber, M . and V . Paschkis (1962). Int J Heat Mass Transfer 5, 623. Incopera, F. P. and D. P. Dewitt (1981). Fundamentals of Heat Transfer. John Wiley and Sons, Inc. Jenkins, B . G. (1998). Modelling - it's plastic, mathematic, stochastic, elastic and fantastic, edited by dong-ke-zhang and nathan, g. Adelaide International Workshop on Thermal Energy Engineering and the Environment, Adelaide, Australia. Jenkins, B . G. and F. D. Moles (1981). Modelling of heat transfer from a large enclosed flame in a rotary kiln. Trans IChemE 59, 17-25. Lun, C. K . K . , S. B . Savage, D. J . Jeffrey, and N . Chepurnity (1984). Kinetic theories for granular flow: inelastic particles in couette flow and slightly inelastic particles in a general flowfield. J. Fluid Mech 140, 223-256. Mastorakos, E. , A . Massias, C. D. Tsakiroglou, D. A . Goussis, V . N . Burganos, and A . C. Payatakes (1998). Cfd predictions for cement kilns including flame modelling, heat transfer and clinker chemistry. Applied Mathematical Modelling 23, 55-76. McCarthy, J . J . , T. Shinbrot, G. Metcalfe, J . E. Wolf, and J . M . Ottino (1996). Mixing of granular materials in slowly rotated containers. AIChE Journal ^\u00C2\u00A3(12), 3351-3363. BIBLIOGRAPHY i w Meakin, P. and R. Jullien (1992). Simple models for two- and three-dimensional particle size segregation. Physica A 180, 1-18. Nityanand, N . , B . Manley, and H . Henein (1986). A n analysis of radial segregation for different sized spherical solids in rotary cylinders. Metallurgical Transactions B 17B, 247-257. Ollivier-Gooch, C. (1998). Computational Methods in Transport Phenomena I and II. Vancouver, B C , Canada: University of British Columbia. Parker, D. J . , A . E. Dijkstra, T. W. Martin, and J . P. K . Seville (1997). Positron emission particle tracking studies of spherical particle motion in rotating drums. Chemical Engineering Science 52(13), 2011-2022. Patankar, S. V . (1980). Numerical Heat Transfer and Fluid Flow. Hemisphere Publish-ing Inc., McGraw-Hill Book Company. Pollard, B . L. and H . Henein (1989). Kinetics of radial segregation of different sized irregular particles in rotary cylinders. Canadian Metallurgical Quarterly 17B(1), 29-40. Ristow, G. H . (1994). Particle mass segregation in a two-dimensional rotating drum. Europhysics Letters 28(2), 97-101. Rogers, A . R. and J . A . Clements (1971/1972). The examination of segregation of granular materials in a tumbling mixer. Powder Technology 5, 157-168. Roseman, B . and M . B . Donald (1962). Mixing and de-mixing of solid particles, i i -effect of varying the operating conditions of a horizontal drum mixer. Brit Chem Eng 7(11), 823-827. Rosin, P. and E. Rammler (1933). The laws governing the fineness of powdered coal. The Institute of Fuel, 29-36. Savage, S. B . and C. K . K . Lun (1988). Particle size segregation in inclined chute flow of dry cohensionless granular solids. J. Fluid Mech 189, 311-335. Schotte, W. (1960). Thermal conductivity of packed beds. AIChE Journal 6(1), 63-67. BIBLIOGRAPHY \u00C2\u00AB i Smith, T. F . , Z. F. Shen, and J . N . Friedman (1982). Evaluation of coefficients for the weighted sum of gray gases model. Transactions of the ASME 104, 602-608. Sunnergren, C. E . (1979, September). A new mixer for rotary lime kilns. National Lime Association Operators Meeting, Seattle WA. Von Wedel, K . (1973). Beurteilung des drehrohrofens zur erzieling eines kalkes definierter qualitat (an appreciation of the rotary kiln for producing lime of a spec-ified quality). Zement-Kalk-Gips 26(3), 110-113. Williams, J . C. (1976). The segregation of particulate materials, a review. Powder Technology 15, 245-251. Woodle, G. R. and J . M . Munro (1993). Particle motion and mixing in a rotary kiln. Powder Technology 76, 241-245. Yang, L. and B . Farouk (1997). Modeling of solid particle flow and heat transfer inrotary kiln calciners. Journal of the Air and Waste Management Association 4% 1189-1196. Appendix A Refractory and Insulation Details 142 APPENDIX A. REFRACTORY AND INSULATION DETAILS 143 12-17-99 10:48P p . 0 ii60-9080 - 196A Street. Langley. British Columbia. Canada. V1M3B4 Tal: (604) 888-8685 Fax; (60\u00C2\u00AB) 888-5211 B r a n d Class THOR 60 C A S T / L O W - M O I S T U R E P R E L I M I N A R Y L A B O R A T O R Y T E S T OATA Maximum R e c o m a e n d e d T o a p e r a t u r e , *F - *< M a t e r i a l R e q u i r e d , l b / f t 3 - k g / m 3 W a t e r R e q u i r e d t o C a s t , u t . % ASTM F l o w R a n g e , % A b r a s i o n L o s s , cm3 A f t a r 1 5 0 0 * F ( 8 1 6 * C ) A f t e r 2000*F ( 1 Q 9 3 \" C ) B e t h l e h e m C y c l i n g T e s t , % S t r e n g t h L o s s S a m p l e s F i r e d 8 2 0 O 0 * P ( l o 9 3 * C ) & C y c l e d 5 X \u00C2\u00AB 2 0 0 0 * F ( 1 0 9 3 * C ) Thermal Conductivity, BTU-in / f t 2 \u00C2\u00AB 3 9 2 ' F ( 2 0 0 * C ) * 7 5 2 F ( 4 0 0 * C ) 6 1 1 1 2 * F ( 6 0 0 * C ) \u00C2\u00AB 1 4 7 2 \" F ( 8 0 0 * C ) *P h r 2 7 0 0 - 1482 162 - 2 6 0 0 5 . 5 - 6 . 0 28 - 36 0 . 0 4 5 . 2 4 0 . 9 4 1 . 3 4 1 . 2 TOTAL MODULUS COLO TEST BULK L I N E A R OF HOT MOR C R U S H I N G T E M P . D E N S I T Y CHANGE RUPTURE 6 T E M P . S T R E N G T H ( \" F ) U b / f t 3 ) (%) ( l b / i n 2 ) ( I b / i n 2 ) ( l b / i n 2 ) 230 165 2400 \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 1500 162 - 0 . 2 4 9 0 0 1 5 0 0 0 2000 162 5 5 0 0 2500 161 200 CC) ( k g / m 3 ) (t) ( H P a ) ( M P a ) ( H P a ) 110 2640 17 816 2600 - 0 . 2 34 1 0 3 1093 2600 38 1371 2580 1 .4 C H E H I C A L A N A L Y S I S C A L C I N E D B A S I S , w t * S i C 5 9 . 0 S i 0 2 14 . 4 CaO 2 . 7 A1203 2 3 . 4 F e 2 0 3 0 . 1 N a 2 0 0 . 2 A P P A R E N T POROSITY (%) 13 I B 17 14 (t) 13 18 17 14 The data given above are based on averages o f the r e s u l t s oC a small number o f t e s t specimens made under c o n t r o l l e d c o n d i t i o n s i n the l a b o r a t o r y and are determined by standard ASTH procedures where a p p l i c a b l e . v a r i a t i o n from the above r e s u l t s nay occur i n i n d i v i d u a l t e s t s and i n a c t u a l p l a n t p r o d u c t i o n . These r e s u l t s cannot be taken as minima or maxima f o r s p e c i f i c a t i o n purposes. 03/95 APPENDIX A. REFRACTORY AND INSULATION DETAILS 144 NARCO CANADA INC. Brand Plicast LWI-24R Quality Number: IZ66 This is a medium weight insulating castable formulated with a decreased iron content. Typical uses include petrochemical heater walls and roofs; regenerator linings in FCCU processes; and walls and arches in heat treating furnaces. TYPICAL TEST DATA Maximum Recommended Temperature, \u00C2\u00B0F - \u00C2\u00B0C Pyrometric Cone Equivalent, P.C.E. Material Required, lb/ft3 -kg/mJ Water Required to Temper, wt.% Setting characteristics Minimum time before firing Average storage life Tested after firing at test temperature TOTAL MODULUS TEST BULK LINEAR OF TEMP. DENSITY CHANGE RUPTURE (\u00C2\u00B0F) (lb/rV) (%) (Ib/inJ) 230 92 -0.1 350 1000 87 -0.2 300 1500 85 -0.3 220 2000 85 -0.5 300 2500 88 -1.5 500 C O (g/cm3) (%) (MPa) 110 1.47 -0.1 2.4 538 1.39 -0.2 2.1 816 1.36 -0.3 1.5 1093 1.36 -0.5 2.1 1371 1.41 -1.5 3.4 Up to 2500 15-16 i&y- 1362 26.1 Hydraulic 24 hours 12 months 1371 COLD CRUSHING STRENGTH (lb/in1) 750 700 650 650 1800 (MPa) 5.2 4.8 4.5 4.5 12.4 Data at temperature THERMAL HOT M0R CONDUCT. @ TEMP. (BTO-in/WffVF) (lb/in2) 2.3 2.6 275 2.8 300 3.5 300 225 (W/m-K) (MPa) 0.33 0.37 1.9 0.40 2.1 0.50 2.1 1.6 CHEMICAL ANALYSIS, dried basis, wt.% Si03 47.56 A1A 37.45 FejO, 0.79 TiO, 0.62 CaO MgO NajO-KjO L.6.I. 9.32 0.16 2.12 0.19 The data given above are based on averages of test results on samples selected from routine plant production, by standard A.S.T.M. pnx^ ures^ where applicable. Variation from the above data may occur in individual tests. These results can^riBQaken as rninima or maxima for specification purposes. A Member of TV P I D l i H VeUsch fctfa Group L1797 \u00E2\u0080\u0094 19427 - 92 AVE^ktTlJRREY. B.C. CANADA V4N 4G6 TELEPHONE: (604) 888-868S FAX: (604]1888-5211_ APPENDIX A. REFRACTORY AND INSULATION DETAILS 145 Typical Physical Properties Duraback* OurablanKet\" S Ourablanket H P - S Ourablanket 2600 Color While White While White Temperature Grade\" 982^C (1800'F) 1260\u00C2\u00B0C (2300'F) 1260\"C (2300=F) 1430\u00C2\u00B0C (2600'F) Melting Point 1648X (3000\u00C2\u00B0F) 1760\u00C2\u00B0C (3200\u00C2\u00B0F) 1760\"C (320CTF) 1760\u00C2\u00B0C (3200\u00C2\u00B0F) Fiber Diameter 2-4 microns 2.5-3.5 microns 2.5-3.5 microns 3.5 microns (mean) (mean) (mean) (average) Specific Heat 1130 J/kg -C 1130 J/kg \u00C2\u00B0C 1130 J/kg \u00C2\u00B0C 1130 J/kg ' C \u00C2\u00AE 1093-C (2000'F) (0.27 Btu/lb 'F) (0.27 Btu/lb \u00C2\u00B0F) (0.27 Btu/lb \u00C2\u00B0F) (0.27 Btu/lb \u00C2\u00B0F) Specific Gravity 2.73 g/cm3 2.73 g/cm3 2.73 g/cm3 2.73 g/cm3 Average Tensile Strength \u00E2\u0080\u0094 5.5 lb/in2 <@ 4 PCF \u00E2\u0080\u0094 _ (ASTM 686-76) 9.9 lb/in' @ 6 PCF 12.5 lb/in2 Cowalion AH PigMs \u00C2\u00AB*servod PimbW n US* PMK 4 of a APPENDIX A. REFRACTORY AND INSULATION DETAILS 146 Typica l Phys ica l Proper t ies Flbermat\u00E2\u0084\u00A2* Blanket Color: Temperature Grade\": Fiber Diameter: Specific Gravity: Nominal Weight: Tensile Strength (ASTM 686-76): White 760'C (1400oF) 2.5-3.5 microns (mean) 2.73 g/cm3 Vi\" thickness = 3.7 oz/fv 1\" thickness = 7.3 oz/ft* 2\" thickness =14.7 oz/ft? 7-10 psi (typical) Typica l Mechan ica l Propert ies C o m p r e s s i o n Recovery Percent Compression 10 30 50 Percent Recovery 93 82 71 Typica l P h y s i c a l Proper t ies F i b e r m a x * Mat Color: Temperature Grade': Melting Point: Fiber Diameter: Specific Gravity: Specific Heat Capacity at 1093\u00C2\u00B0C (2000'F): Fiber Surface Area: White 1650'C OOOOT) 1870<,C(3400CF) 2-3.5 microns (mean) 3 g/cm3 1246 J/kg \u00C2\u00B0C (0.297 Btu/lb \u00C2\u00B0F) 7.65 mJ/g Ouraback* Durablanket* S Durablanket H P - S Durablanket 2600 Thermal Conduct iv i ty vs . M e a n Temperature (per A S T M C - 1 7 7 ) \" a o e o E .375 (2.6] .346 (2.4) .3t7(2.Z| .288 (2.0) .260(1.8) .231 (1.6) .202(1.4) .173(1.21 .144|1.0| .115(0.8) .08710.6) .058 (0.4) .029 (0.2| 0 64 kg/m1 (4 lb/ft3) y ** \u00E2\u0080\u00A2 \ \u00E2\u0080\u0094 * ^ s \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2> \u00E2\u0080\u00A2 * * 96 kg/m 3 (6 lb/ft3) 128 kg/mJ (8 lb/ft\") \u00E2\u0080\u00A2129 (-200) \u00E2\u0080\u00A218 (0| 93 12001 204 (4001 316 427 (600) (800) Mean Temperature 538 110001 \u00E2\u0080\u00A2 \u00C2\u00BBC (*F) 6 4 9 (1200) 760 (1400) 871 (1600) 982 (18001 'The temperature graoe of FiberTrair* insulation is determined by irreversible linear change criteria, not product melting point. For additional information about product performance or to identify the recommended product (or your application, pleaae contact the Unifrax Application Engineering Group at 716-278-3899. Data are average results of te3U> conducted undor standard proce-dures and are suoject to variation. Results sfiould not be used tor specification purposes. \u00E2\u0080\u00A2'All heat flow calculations are based on a surface emissivity (actor o' 0.9O. an ambient temperature of 27\"C (80 :F). and zero wind velocity, unless otherwise stated. All thermal conductivity values for Fibedrax materials have oeen measured in accordance wltn ASTM Teal Procedure C-t 77. when comparing similar data, it is advisable to check the validity of all thermal conductivity values and ensure the reaulting heat flow calculations are based on the same condition factors. Variations in any of these (actors will result in significant differences in me calculated data. Fofm C-U21 Etlocro* 9*25 C1999. uwfrai Conwat-on AH F-onls flese^o Printed in US* Page 5- ol S Appendix B Engineering Drawings of Furnace and Support 147 APPENDIX B. ENGINEERING DRAWINGS OF FURNACE AND SUPPORT 148 Appendix C Specifications of Instrumentation 1. Thermocouple cable 2. Rotary Selector Switch 3. Portable Temperature Indicator 4. Signal Conditioning Board 5. Data Acquisition Board 151 APPENDIX C. SPECIFICATIONS OF INSTRUMENTATION 152 Thermocouple Wire Duplex Insulated MADE IN USA ^5 K Duplex Insulated CHROMEGA*-ALOMEGA* Duplex ANSI Type K ANSI Color Code: Positive Wire, Yellow; Negative Wire, Red; Overall, Browrr^ OMEGA Engineering does not use reprocessed Teflon or PVC in manufacturing of thermocouple wii '^j-1 Insulation AWG No. Model Number Price 1000\" Type Wire Insulation C o n d u c t o r ^ s\"-jg O v e r a l l A , Max. Temp Nominal Size Wt.t f lb/10001 Ceramic 14 20 20 20 24 24 24 14 20 24 XOK-14 XC-K-20 XT-K-20 XL-K-20 XC-K-24 XT-K-24 XL-K-24 XS-K-14 XS-K-20 XS-K-24 $2695 1425 1275 1200 995 895 845 1875 JJQSO.. 825 Solid Solid Solid Solid Solid Solid Solid Solid Solid Solid Nextel Ceramic*? ^ Nextel CeramiQ$ss& Nextel C e r a m i c $ \u00C2\u00BB 'Nextel Ceramic I Nextel Ceramic- > Nextel.Geramic-..--: Nextel Cerarnia^fe Silfa Silica?. * Silfa Sihc-i , , 1 -. Silfa S i l i c a ^ - J i g t Nextel,CeramicpS\u00C2\u00A7fe Nextel CeTafnicjSi * Nextel Cejarjiic| w Nextel; Ceratrijej LNSxfelC^rarfjicr N^elCepnfic Ne%el:eeTamic.a5\u00E2\u0084\u00A2 Silfa'Slica*'-'- \u00C2\u00AB,';: Silfa^SihcateJ&M^ SilfCSilica^^S 2000 1800 1800 1800 1600 1600 1600 1900 1800 1600 1090 980 980 980 870 870 870 1038 980 870 38 16 15 14 12 11 10 35 12 10 High Temp. Glass\" 20 24 HH-K-20 HH-K-24 492 348 Solid Solid High Temp* Glassf High Temp Glassy 1300 1300 704 704 .060 x.105 .055 X .090 3* Glass 20 20 24 24 26 28 30 36 GG-K-20 GG-K-20S GG-K-24 GG-K-24S GG-K-26 GG-K-28 GG-K-30 GG-K-36 410 600 290 395 270 230 240 320 Solid 7x28 Solid 7x32 Solid Solid Solid Solid Glass Braid^S^Tj Glass B r a l d M * ^ Glass Braid Glass Braid Glass Braids Glass Wrai.\u00E2\u0080\u009E. Glass Wrap^ Glass B r a i d ^ e L ^ Glass Braid. Glass BraidV-Glass Braids Glass Braid-*p Glass Braid; Glass Braid,-;-:\u00C2\u00AB Glass Braid- w 900 900 900 900 900 900 900 900 482 482 482 482 482 482 482 482 .060 X .095 .060 X.100 .050 X .080 .050 x .085 .045 X .075 .045 X .070 .045 X .070 .045 X .070 Teflon* Glass 30 36 40 TG-K-30 TG-K-36 TG-K-40 530 590 650 Solid Solid Solid PFA J' * f\" \u00E2\u0080\u00A2tt*?Z. PF * PFAWft\u00C2\u00AB\u00C2\u00BBva1 Glass Braid Glass Braid ' ^ Glass B r a i d > r 500 500 500 260 260 260 .034 X .047 .028 X .038 .026 X .035 MS Teflon* Neoflon* PFA (High Performance) 20 20 22 24 24 30 36 40 TT-K-20 TT-K-20S TT-K-22S TT-K-24 TT-K-24S TT-K-30* TT-K-36* TT-K-40* 495 755 735 370 515 335 375 500 Solid 7x28 7x30 Solid 7x32 Solid Solid Solid PFA PFA PFA, PFA'' PFA PFA PFA PFA PFA--\" PFAr PFA PFA'\"5 PFA PFA, PFA^ PFA 500 500 500 500 500 500 500 500 260 260 260 260 260 260 260 260 .068X.116 .073 X . 126 .065 x.133 .056 x .093 .063 x .102 .022 x .038 .017 x .028 .015 X.024 11 11 9 6 6 2 2 2 Teflon* Neoflon* FEP 20 24 FF-K-20 FF-K-24 490 350 Solid Solid FEP FEP FEP FEP 392 392 200 200 .068X.116 .056 x .092 11 6 Polymers 24 20 24 PR-K-24 KK-K-20 KK-K-24 230 1055 735 Solid Solid Solid Polyvinyls Kapton Kapton (Rip,Cord) \u00E2\u0080\u00A2 Kapton Kapton _ 221 600 600 105 315 315 .050 x .086 .057 x. 103 .045 x .079 5 11 6 tWeight of spool and wire rounded up to the next highest lb., does not include packing material \"Overall color: clear \"Has color tracer Wire available with special limits of error, consult Sales for pricing and availability Spool Pricing 1. Determine price per 1000 ft. spool 2. Look up price category for the standard lengths available: 25 ft. = Price per 1000' x 2 . 5 1000 50 ft. = Price per 1000' x 2 1000 100 ft. = Price per 1000' x 1.75 1000 200 ft = Price per 1000' x 1.5 1000 500 ft = Price per 1000' x 1 1000 1000 ft = Price per 1000' (Net) To determine total cost, use established price, category (from step 2) times total length Round price to nearest dollar. Examples: * *s Discount Scheduje <=f ^ l66(fft sp6ois~6rTly 3-4... 5-9 10-19. 'K . .81 \u00E2\u0080\u00A2- -.. :;...i5 Wire Type: EXPP-J-14 @ 555/1000' * 50 ft. EXPP-J-14 @ $.555/ft x 2 = $1.11/ft. ($55.50 per 50 ft. spool)__ 100 ft. EXPP-J-14 @ .555/ft x 1.75 = .97/ft. ($97.00 per 100 ft. spool)_ 150 ft. EXPP-J-14 @ .555/ft x 1.75 = .97/ft. ($146.00 per 150 ft. lot) Note: Customer will receive one 100 ft. spool and one 50 ft spool _ 200 ft. EXPP-J-14 @ .555/ft x 1.5 = .83/ft. ($167.00 per 200 ft. spool)_ 300 ft. EXPP-J-14 @ .555/ft x 1.5 = .83/ft. ($250.00 per 300 ft. lot) Note: Customer will receive one 200 ft. spool and one 100 ft. spool __. 500 ft. EXPP-J-14 @ .555/ft x 1 = .555/ft. ($278.00 per 500 ft. spool)_ \"ief Wis H-19 APPENDIX C. SPECIFICATIONS OF INSTRUMENTATION 153 Standard Thermocouple and RTD Rotary Selector OSGW line has Gold Plated Contacts WVVUV/I I v O : 1 Op f f - - 1 fTfe ' Shown smaller than actual size 3 and 5\" Sizes 3 Pole --4 / / / / / - B --c-7b order iv/fh optional pistol grip, add suffix '-PG' to model no. Comes complete with mounting hardware, panel mounting template and instwction manual. Add $6.50 to price. Ordering Example: OSWG3-20-PG, 3\" (76 mm) switch, 20 positions, gold contacts, with optional pistol grip, $204 + 6.50 = $210.50. Insulation Resistance: 20 Mfi at 300 Vdc Contact Resistance: 0.004 Q. or less Also Available: Pistol grip handle retrofit kit; specify OSW-PG and switch size, $10. OSW Switch Dimensions Switch Size Dim. A mm (in) Dim. B* mm (in) Dim. C mm (In) Dim. D mm (in) Dim. E mm (in) Bezel Size mm (in) 76 mm 0\") 28.6 (ISO 54 {2%\") 62 (2KO 44.4 (iy.-) 66.7 {2%\") 82.55 (3V) 127 mrr (5\") 38.1 (UP) 54 (2IO 62 (2XO 58.7 (2X\u00C2\u00AB\") 92.07 (3*\") 133.35 {5vn To Order (Specify Model Number) | I No. of No. of Contact Silver Plated Contacts Gold Plated % Contacts % 1 Poles Positions Action Size Model No Price Model No. Price \ ftf 2 OSW3-2 $95 OSWG3-2 $142 m 3 OSW3-3 97 OSWG3-3 145 H 4 Break OSW3-4 99 0SWG3-4 148 jH 5 Before 3\" OSW3-5 102 OSWG3-5 153 m 6 Make OSW3-6 104 OSWG3-6 156 I Z 8 OSW3-8 108 OSWG3-8 162 9 OSW3-9 111 OSWG3-9 166 p 10 OSW3-10 113 OSWG3-10 169 m \u00E2\u0080\u00A2 12 OSW3-12 118 OSWG3-12 177 M 0 14 OSW3-14 122 OSWG3-14 183 m \u00E2\u0080\u00A2 16 3\" OSW3-16 127 OSWG3-16 190 m L 18 OSW3-18 131 OSWG3-18 196 1 E 20 Make OSW3-20 136 OSWG3-20 204 H 24 Before OSW5-24 145 OSWG5-24 217 m 28 Break OSW5-28 154 OSWG5-28 231 ft 30 5\" OSW5-30 158 OSWG5-30 237 \u00E2\u0080\u00A2 | 32 OSW5-32 162 OSWG5-32 243 % 36 OSW5-36 171 OSWG5-36 256 w 40 OSW5-40 185 OSWG5-40 277 SB 3 6 OSWT-6 111 OSWGT-6 166 10 OSWT-10 124 OSWGT-10 186 m 12 OSWT-12 131 OSWGT-12 196 P 18 OSWT-18 151 OSWGT-18 226 M 20 Make OSWT-20 158 OSWGT-20 237 0 24 Before 127 OSWT-24 168 OSWGT-24 252 \u00E2\u0080\u00A2 28 Break (5\") OSWT-28 181 OSWGT-28 271 L 30 OSWT-30 188 OSWGT-30 282' E 32 OSWT-32 195 OSWGT-32 292 36 OSWT-36 208 OSWGT-36 312 40 OSWT-40 222 OSWGT-40 333: BRI G-77 APPENDIX C. SPECIFICATIONS OF INSTRUMENTATION 154 ANALOG SIGNAL CONDITIONING & EXPANSION CIO-SSH 16 & CIO-SSH 16/DST Features \u00E2\u0080\u00A2 Adds simultaneous sample & hold front end to A/D board \u00E2\u0080\u00A2 4 channels standard, expandable to 16 channels \u00E2\u0080\u00A2 Compatible with CIO-DAS1600 -DAS 1400 and -DAS 16 families Analog Input Specifications Aperture time 175nS avg, 250nS max Aperture uncertainty \u00C2\u00B125 nS Acquisition time 4 uS Max sample rate 250 KHz Droop rate \u00C2\u00B1100uV/mS Input gains I, 10, 100, 200, 300, 500 600, 700 and 800 Gain selection Switch selectable. Channels may be at different gains. Accuracy 0.01%\u00C2\u00B1l-bit Channels 4 standard, up to 16 total. Additional channels may be added by purchasing/installing the CIO-SSH-AMP part. Each SSH-AMP adds one channel. Software Description Includes InstaCal\u00E2\u0084\u00A2, installation, calibration and test software. The CIO-SSH16 is supported by the UniversalLibrary\u00E2\u0084\u00A2 for use with the DAS-16, DAS-1600 and DAS-1400 families. UniversalLibrary\u00E2\u0084\u00A2 (provides programming language support for all DOS and Windows languages). Also supported by many third party, high-level data acq. programs. Ordering Information CIO-SSH 16 4 channel SSH board CIO-SSH 16/DST w/ detachable screw terms. CIO-SSH-AMP' 1 channel expansion (you may add up to 12 channels) Interconnect Cables Standard: C37FF-# series Shielded: C37FFS-5, C37FFS-10 CIO-EXP-32,CIO-EXP-32/DST, CIO-EXP-16 & CIO-EXP-16/DST Analog Input Specifications Channels 32, CIO-EXP-32, differential 16, CIO-EXP-16, differential Gain weights 1, 10, 100, 200, 300, 500 CJC sensor +24.4 mV/\u00C2\u00B0C (0V at 0\u00C2\u00B0C) User Field Configurable Options Input filters (7 Hz), Open Thermocouple detection, lOKohm ground reference resistor. Cascading with the EXP-16 & EXP-32 Features \u00E2\u0080\u00A2 A/D board channel expansion \u00E2\u0080\u00A2 32 channel (EXP-32) \u00E2\u0080\u00A2 16 channel (EXP-16) \u00E2\u0080\u00A2 User selectable gains \u00E2\u0080\u00A2 On-board CJC sensor Software Description Includes InstaCal\u00E2\u0084\u00A2, installation, calibration and test software. The CIO-EXP series is supported by the UniversalLibrary\u00E2\u0084\u00A2 for use with most A/D boards. /DST versions provide all field I/O wiring through detachable screw terminals. Operating Modes The active EXP-32/16 channel is selected by the host A/D board's digital output. An A/D board ana-log input must be dedicated to each EXP board. 128 CHANNELS TOTAL WtTH CIO-OAS08 256 CHANNELS WITH CIO-0AS16 FIRST CAOE m SERES F A D BOAROiS OASIS MUSTBEOMUXAOI6-10 FOR OAS08 AMY t: 1 37 PIN Interconnect Cables DAS-800/8 series DAS-1600/1400/16 series C37FF-# series C-EXP2DAS16-10 CIO-EXP-BRIDGE16 & CIO-EXP-BRIDGE16/DST Analog Input Specifications Channels 16, fully differential Gains 1, 10, 100, 1000 Field Installable Precision Resistors It is critical that you use high quality resistors. We offer these values that match many sensors CIO-EXP-RES-120 120Ohm,5ppm,0.1% CIO-EXP-RES-350 350 Ohm, 5ppm, 0.1% CIO-EXP-RES-1000 1000 Ohm, 5ppm, 0.1% Features \u00E2\u0080\u00A2 16 chan., A/D board expansion \u00E2\u0080\u00A2 Ideal strain gauge interface \u00E2\u0080\u00A2 On-board Wheatstone bridge circuitry \u00E2\u0080\u00A2 User selectable gains S a m p l e C h a n n e l S T R A I N G A U G E O R O T H E R B R I D G E S E N S O R EXCITATION VOLTAGE <+) S E N S E HIGH (+) Software Description Includes InstaCal\u00E2\u0084\u00A2, installation, calibration and test software. The CIO-EXP series is supported by the UniversalLibrary\u00E2\u0084\u00A2 for use with most A/D boards. /DST version provides all field I/O wiring through detachable screw terminals. Operating Modes The active EXP-BRIDGE16 channel is selected by the host A/D board's digital output. An A/D board analog input must be dedicated to each EXP board. S E N S E L O W ( . ) EXCITATION VOLTAGE (-) Interconnect Cables DAS-800/8 series DAS-1600/1400/16 series C37FF-# series C-EXP2DAS16-10 Premium Performance at Amazingly Low Prices, See the Table of Contents for Our Prices! page 59 APPENDIX C. SPECIFICATIONS OF INSTRUMENTATION 155 p nd TM i l l ed is. I s) 8 id rM ill \u00E2\u0080\u00A2A s. d M I I d s. ntfj>l !-H&>-bt? O O o CD w d CO CU & CD cq cc3 c o d d P S c o 2 ll APPENDIX F. RESULTS: 191 APPENDIX F. RESULTS: 192 APPENDIX F. RESULTS: 193 APPENDIX F. RESULTS: 195 APPENDIX F. RESULTS: 196 APPENDIX F. RESULTS: 197 APPENDIX F. RESULTS: 198 APPENDIX F. RESULTS: 201 APPENDIX F. RESULTS: 202 APPENDIX F. RESULTS: 203 APPENDIX F. RESULTS: 206 APPENDIX F. RESULTS: 208 APPENDIX F. RESULTS: 211 APPENDIX F. RESULTS: 212 APPENDIX F. RESULTS: 213 APPENDIX F. RESULTS: 217 APPENDIX F. RESULTS: 218 APPENDIX F. RESULTS: 221 APPENDIX F. RESULTS: 222 APPENDIX F. RESULTS: 223 APPENDIX F. RESULTS: 224 APPENDIX F. RESULTS: 226 APPENDIX F. RESULTS: 227 APPENDIX F. RESULTS: 232 APPENDIX F. RESULTS: 233 APPENDIX F. RESULTS: 237 APPENDIX F. RESULTS: 238 CM CM S-I co CO IC3 r 3 Or O CO CD K d \u00C2\u00A3 d CD o CP O O CQ d \u00C2\u00AB3 CD 1) PQ S-i CD > o d d ri XT' i O r ' CM CM CD O Ui 3 || bO E H b APPENDIX F. RESULTS: 241 APPENDIX F. RESULTS: 242 APPENDIX F. RESULTS: 243 o co a a, ui co cc3 Or\" fa CP K -a a o a CP \u00E2\u0080\u00A2 \u00E2\u0080\u00A2\u00E2\u0080\u0094I O m CP o O ui CO I cc3 CP W CP PQ ccj Ul CP > o co in ri fa Q5 Q5 c oT o CP o Ul 3 II bO fa b APPENDIX F. RESULTS: 244 APPENDIX F. RESULTS: 246 ) OS cs~ cs OH u CN O u s-i Pi CN fa u bp APPENDIX F. RESULTS: 247 APPENDIX F. RESULTS: 248 APPENDIX F. RESULTS: 251 co c T\u00E2\u0080\u0094I o U C S H J cd co X >> u O \u00E2\u0080\u00A2\u00C2\u00AB\u00E2\u0080\u0094t SH H J a; OH a oo i \u00E2\u0080\u0094 i a CO fa SH OJO fa \ APPENDIX F. RESULTS: 252 APPENDIX F. RESULTS: 253 APPENDIX F. RESULTS: 256 APPENDIX F. RESULTS: 257 APPENDIX F. RESULTS: 258 APPENDIX F. RESULTS: Table F . l : Bed Temperatures at Select Times Run Time B l B2 B3 B4 Ave Temp (min) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 3 4.98 221.5 282.9 345.4 399.8 312.4 5.00 222.7 283.9 347.1 402.2 314.0 5.02 223.6 285.3 348.9 405,5 315.8 7.48 407.2 450.5 510 537.6 476.3 7.50 408.7 451.2 511.1 539.4 477.6 7.52 410 452.2 511.7 540.6 478.6 9.99 546 552.4 592.2 611.4 575.5 10.00 546.8 553.1 592.9 612.3 576.3 10.02 547.5 553.4 593.6 612.9 576.9 12.49 616.9 623 654 667.5 640.4 12.50 617.3 623.1 654.5 667.5 640.6 12.52 617.7 623.4 654.8 667.9 641.0 14.99 675.3 670.9 695 705 686.6 15.01 675.7 671.5 695 705 686.8 15.02 675.9 671.9 695.4 705.3 687.1 17.49 708.1 709 726 732.3 718.9 17.51 708.2 709.3 726.2 732.6 719.1 17.52 708.3 709.2 726.2 732.6 719.1 19.99 734.9 732.5 744.9 749.8 740.5 20.01 734.9 732.7 744.8 749.9 740.6 20.02 735.1 732.9 744.9 749.8 740.7 APPENDIX F. RESULTS: 4 4.99 357 347.4 453.2 466.4 406.0 5.01 358.4 349.6 454.4 467.1 407.4 5.02 360.2 351 455.7 468.6 408.9 7.49 528.5 545.6 594.9 599 567.0 7.51 529.6 546 596 599.2 567.7 7.53 530.6 546.7 596.2 600 568.4 9.99 618.2 630.2 660 664 643.1 10.01 619.1 630.9 660.9 664 643.7 10.03 620 631.4 661 664.5 644.2 12.48 678.8 683.1 704.1 705.5 692.9 12.49 679.3 683.4 704.3 705.7 693.2 12.51 679.5 683.6 704.8 705.7 693.4 15.00 719.7 720.3 734.6 734.5 727.3 15.02 719.8 720.7 734.5 734.6 727.4 15.03 720.2 720.8 735 734.7 727.7 17.48 746.9 744.3 754.6 753.1 749.7 17.50 747 744.4 754.8 753.2 749.9 17.52 747.3 744.6 755.1 753.4 750.1 19.98 765.3 760.4 768 766.4 765.0 20.00 765.2 760.4 768 766.3 765.0 20.02 765.4 760.5 768.1 766.4 765.1 5.00 316.1 321.2 395.6 408.6 360.4 5.01 316.5 323.3 400 409.5 362.3 5.03 316.9 324.8 402.5 410.6 363.7 7.49 484.7 512.2 566.6 575.9 534.9 7.50 485.5 513.3 566.7 576.7 535,6 7.52 486.7 514.6 567.8 577.5 536.7 9.99 600.7 611.9 645 652.4 627.5 10.01 600.7 611.9 645 652.4 627.5 10.03 601.3 612.6 645.6 653.1 628.2 12.50 669.5 674.6 697.2 700.7 685.5 12.52 670 674.8 697.3 700.9 685.8 12.54 670.6 675.1 697.9 701.3 686.2 14.99 716.7 716.1 731.3 733.2 724.3 15.01 716.8 716.2 731.4 733.6 724.5 15.02 717 716.3 731.7 733.4 724.6 17.50 747.8 745 756.5 754.6 751.0 17.51 747.8 745.3 756.7 754.6 751.1 17.53 748 745.6 757 754.7 751.3 19.98 769.1 764.5 772.3 770.6 769.1 20,00 769.1 764.7 772.4 770.6 769.2 20.02 769.3 764.9 772.5 770.7 769.4 APPENDIX F. RESULTS: 8 4.99 173.3 5.01 175.1 5.03 176.5 7.47. 351.6 7.49 352.7 7.51 353.6 9.99 494.6 10.01 495.3 10.03 496 12.48 581.3 12.50 581.6 12.52 582.4 14.98 645.4 15.00 645.6 15.02 645.7 17.50 693.2 17.52 693.4 17.54 693.8 19.99 727.4 20.01 727.5 20.03 727.6 9 5.00 298.5 5.01 300.7 5.03 302.5 7.49 515.7 7.51 515 . 7.53 514.3 9.99 606.2 10.01 606.7 10.03 606.7 12.49 660.8 12.51 661.6 12.53 661.7 14.99 694.5 15.01 695.3 15.03 695.6 17.48 729.1 17.50 728.9 17.52 728.9 19.98 755.8 20.00 757.3 20.02 758.2 269 340.2 270.9 341.4 272.7 342.7 446.3 501.1 447.3 501.9 448.6 503 553.8 595 554.7 595.7 555.9 595.7 623.4 656.6 623.7 656.6 624.2 656.8 675.1 700.4 675.4 700.7 676.1 700.7 713.4 732.2 713.6 732.4 713.8 732.5 741.6 756.2 741.8 756.7 742.2 756.6 347.3 416.4 350.2 417.5 352.5 418 533.5 572.7 533.7 574.4 534.3 575.9 613.7 642.1 614.5 642.4 615.3 642.6 670.5 691.4 670.7 691.7 671.3 691.8 706.6 725.7 706.6 725.9 706.8 726.4 735.8 752.1 736.3 752.3 736.5 752.7 759.3 773 759.7 773.2 759.6 772.8 350.7 283.3 352.6 285.0 354 286.5 508.6 451.9 509 452.7 510.1 453.8 596.4 560.0 597.1 560.7 597.8 561.4 656.5 629.5 656.7 629.7 657 630.1 698.1 679.8 698.3 680.0 698.5 680.3 728.5 716.8 728.8 717.1 728.9 717.3 751 744.1 751.3 744.3 751.4 744.5 419.5 370.4 421.6 372.5 423.4 374.1 568.9 547.7 569.4 548.1 570 548.6 635.2 624.3 635.5 624.8 636 625.2 684.2 676.7 684.6 677.2 684.8 677.4 718 711.2 718.1 711.5 718.4 711.8 742.6 739.9 743 740.1 743 740.3 763.3 762.9 763.6 763.5 763.6 763.6 APPENDIX F. RESULTS: 5.00 5.02 5.03 7.49 7.51 7.53 9.99 10.01 10.03 12.49 12.51 12.53 14.99 15.01 15.03, 17.49 17.51 17.53 19.99 20.00 20.02 4.97 5.01 5.03 7.49 7.51 7.53 9.99 10.01 10.02 12.48 12.50 12.52 15.00 15.02 15.04 17.49 17.51 17.53 19.99 20.01 20.03 352.4 355.8 357.6 512.3 513.2 513.6 611 611.8 612.3 676.9 677.4 677.6 720.3 720.7 720.9 750.5 750.7 751.3 769.1 769.2 769.5 294.2 296.6 297.6 478.5 479.4 480.8 586.4 586.3 587 650.3 650.5 651 697.6 698 698.3 727.8 727.8 728.2 749.7 749.8 749.9 406 407.2 409.4 562.8 563.3 564.3 639.8 639.9 640.5 696.5 696.5 696.7 732.5 732.5 732.5 757.2 757.2 757.2 772.5 772.7 773.1 323.7 327 328 524.6 526.2 527.4 611 611.6 612.6 667 667.6 668.1 705.7 705.9 706.2 733.5 734 733.9 752.4 752.6 752.7 435.9 437.5 439.3 576.5 577.2 577.9 652.9 653 652.9 706 706.4 706.7 741.2 741.7 741.6 764.6 764.8 765.1 779.9 780.1 780 354.3 357.4 358.9 529.2 531 531.9 617.4 617.5 618 672.6 673 674 711.1 711.6 712.1 739.1 739.2 739.3 758.6 758.7 758.9 457.1 458.6 459.5 575.8 576.4 577.4 650.4 650.5 651.6 704.1 704.4 704.7 739.4 739.5 739.8 762.1 762.9 762.3 778.5 778.1 778.5 396.9 398.9 399.2 533.8 534.5 535.4 618.3 619 619.3 674.1 674.7 675 713.8 713.8 714 741.1 741.2 741.5 760.2 760.6 760.6 412.9 414.8 416.5 556.9 557.5 558.3 638.5 638.8 639.3 695.9 696.2 696.4 733.4 733.6 733.7 758.6 758.9 759.0 775.0 775.0 775.3 342.3 345.0 345.9 516.5 517.8 518.9 608.3 608.6 609.2 666.0 666.5 667.0 707.1 707.3 707.7 735.4 735.6 735.7 755.2 755.4 755.5 APPENDIX F. RESULTS: 12 5.00 268.7 5.02 269.2 5.03 270.9 7.49 461.8 7.51 462.5 7.53 464.3 9.99 579.9 10.01 580.7 10.03 580.9 12.49 653.9 12.51 653.9 12.53 655.1 14.99 705.1 15.01 705.2 15.03 705.4 17.49 737.3 17.51 736.9 . 17.53 737.8 19.99 758.8 20.00 758.5 20.02 759.3 13 5.00 252 5.02 252.9 5.04 253 7.50 416.1 7.51 417.3' 7.53 418.2 9.99 535.5 10.01 536.2 10.03 537.1 12.49 610 12.51 610.7 12.53 611.5 14.99 667.5 15.01 667.4 15.03 668.4 17.49 706.7 17.51 706.9 17.53 706.3 19.99 734.2 20.01 734 20.03 734.5 310.7 351.6 313.8 354.3 317.3 357.7 511.5 532.9 512.7 534.4 513.5 535.1 611.7 625.9 612.7 626.7 612.8 626.5 676.5 686.1 677.1 686.9 677.9 687.5 718.5 727.1 .718.8 727 718.5 726.9 746.1 753.1 745.5 753 746.3 753.6 764.6 771.4 764.2 771.8 764.7 772.4 284 332.9 285.5 334.9 288.2 337.1 474.3 501.4 475.4 502.5 476.8 503.4 574.6 588.9 574.9 589.5 575.8 590.1 639 653.8 638.8 653.5 639.8 654.3 684.7 697.9 684.7 698 685.3 697.9 719.8 730.6 720.5 729.6 720.3 730.7 743.2 752.2 742.8 752.8 743.3 753 351.2 320.6 353.3 322.7 355.1 325.3 542.8 512.3 543.4 513.3 544.2 514.3 634.2 612.9 634.8 613.7 635.3 613.9 692.2 677.2 693.1 677.8 692.8 678.3 731.5 720.6 731.9 720.7 733 721.0 758.1 748.7 757.4 748.2 757.4 748.8 775.4 767.6 776.8 767.8 776 768.1 351.5 305.1 354.1 306.9 355.7 308.5 522.5 478.6 523.3 479.6 525.1 480.9 605.9 576.2 606.1 576.7 607 577.5 664.8 641.9 666.4 642.4 665.8 642.9 706.9 689.3 706.8 689.2 708.4 690.0 738.8 724.0 738.1 723.8 738.5 724.0 759 747.2 759.1 747.2 759.7 747.6 APPENDIX F. RESULTS: 14 4.99 281.4 5.01 282.6 5.03 284.1 7.49 483.3 7.51 484.5 7.53 485.1 9.99 593.3 10.01 593.3 10.03 593.7 12.49 666 12.51 665.7 12.53 665.9 14.98 717.2 15.00 717.4 15.02 717.5 17.48 751.2 17.50 750.9 17.52 751.4 20.00 774.8 20.02 774.8 20.04 774.9 15 4.99 264.4 5.01 266.2 5.03 267.2 7.49 419.5 7.51 420.3 7.53 421.1 9.98 527.5 10.00 528.4 10.02 529.6 12.50 609.3 12.52 610.3 12.54 610.4 15.00 663.9 15.01 664.5 15.03 664.6 17.49 704.2 . 17.51 704.7 17.53 704.4 19.99 734.6 20.01 734.3 20.03 734.4 320.8 372.7 323.4 374.4 326.2 376.2 521.4 534.6 523.3 536.1 524.9 537.9 616.9 620.1 617.6 620.2 618.3 621.3 681.2 686.7 682 687.2 682.2 687.5 725.8 731.8 726.1 732.6 726.6 732.4 756.1 763.8 756.3 763.7 756.8 764 776.6 783.8 776.6 784.1 777 784.3 266.7 378.6 269.1 382.2 271.4 385.6 467.2 503.5 468.2 504.4 468.8 505.1 570.1 588.8 571.1 589.7 571.4 590.2 639.9 654 640.8 654.6, 641.1 655.4 686.5 696.2 687.2 696.3 687 696.5 719 728.7 719.3 729.3 719.9 729.5 744 752.6 744.2 752.7 744.2 752.8 416.1 347.8 417 349.4 419 351.4 579.1 529.6 580.1 531.0 581.2 532.3 653.5 621.0 654.8 621.5 654.2 621.9 708.5 685.6 708.1 685.8 708.8 686.1 746.6 730.4 747.6 730.9 746.7 730.8 775.7 761.7 776.3 761.8 776.2 762.1 794.4 782.4 794.5 782.5 794.5 782.7 492.6 350.6 492.7 352.6 492.2 354.1 557.4 486.9 557.9 487.7 557.9 488.2 627.2 578.4 627.9 579.3 629.7 580.2 688.3 647.9 689.2 648.7 690 649.2 724.6 692.8 725.2 693.3 726.1 693.6 749.9 725.5 751.1 726.1 750.2 726.0 769 750.1 769.2 750.1 768.9 750.1 APPENDIX F. RESULTS: 16 4.99 197.8 5.01 200.9 5.03 203.5 7.49 397.9 7.51 398.2 7.53. 398.7 9.99 511.4 10.01 512.1 10.03 512.9 12.49 590.6 12.50 590.9 12.52 591.3 14.98 643.2 15.00 643.8 15.02 644.3 17.50 687.5 17.52 687.8 17.54 688.2 20.00 718 20.02 717.8 20.03 717.8 17 4.99 389.1 5.01 390.7 5.03 392.9 7.49 538.3 7.51 540.2 7.53 541.2 9.99 632.5 10.01 632.6 10.03 633.4 12.49 699.9 12.51 700.4 12.52 701 14.98 743 15.00 743.4 15.02 743.6 17.48 769.1 17.50 769.3 17.52 769.3 20.00 785.7 20.02 785.9 20.04 785.9 269.1 ' 383.9 273 385.4 275.8 386.6 457.2 486.8 457.9 487.8 458.4 488.9 557.2 579.5 557.2 580.1 557.5 580.3 623.2 642.2 623.3 642.8 623.4 643.2 666.4 682.5 666.7 682.9 667.1 683.6 700 716.1 700.5 715.6 701 716.3 724.4 737.5 724.6 737.8 724.8 737.6 407 457.7 408.4 458.4 409.7 459.7 551.7 571 552.1 571.5 552.8 571.8 642.8 656 643.4 656.6 644.2 657 705.4 716.1 705.5 716.3 705.5 716.5 743.9 753.7 744.1 754.1 744.5 754.6 768.6 777.9 768.8 778.2 769.2 778.3 783.8 792.6 783.7 792.7 783.7 792.8 498.9 337.4 500.6 340.0 501.6 341.9 521.8 465.9 522.6 466.6 524 467.5 603.8 563.0 604.7 563.5 605.4 564.0. 660.6 629.2 661.4 629.6 660.8 629.7 695.4 671.9 695.5 672.2 695.7 672.7 726.1 707.4 725.6 707.4 725.6 707.8 745.8 731.4 745.7 731.5 745.8 731.5 517.7 442.9 517.8 443.8 517.8 445.0 598.1 564.8 599.4 565.8 600.2 566.5 670.9 650.6 ,671.6 651.1 672 651.7 727.3 712.2 727.6 712.5 727.9 712.7 762.3 750.7 762.4 751.0 762.9 751.4 785.7 775.3 786 775.6 786.1 775.7 800.2 790.6 800.3 790.7 800.3 790.7 APPENDIX F. RESULTS: 18 4.99 355.4 5.01 357.3 5.03 359.7 7.49 539.8 7.51 541 7.53 543 9.98 623.5 10.00 624.2 10.02 624.7 12.50 689.1 12.52 690 12.54 690 15.00 731.7 15.02 732 15.03 732.6 17.49 758.3 17.51 758.4^ 17.53 758.6 19.99 776.6 20.01 776.4 20.03 .776.7 19 5.00 346.8 5.01 349.7 5.03 350.5 7.49 519.2 7.51 519.5 7.53 520.5 9.99 623.7 10.01 624.6 10.03 625 12.49 688.7 12.51 688.9 12.53 689.4 14.99 731.2 15.01 731.7 15.02 731.8 17.48 760.6 17.50 760.7 17.52 761.1 19.98 779.6 20.00 780 20.02 779.4 381 427.6 383.8 429.1 387 429.7 557.4 567.8 558.4 569.2 558.8 570.1 643.8 653 643.9 653.7 644.4 654.1 701.7 712.6 701.9 713.2 702.6 713.2 738.2 747.9 738.4 748.3 738.8 748.5 761.8 770 762 769.8 762.1 770.7 777 786.9 777.6 787.2 777.6 786.8 401.2 421.4 402.6 423.4 404.4 423.9 551.3 570.2 552 571.6 551.9 573.2 641.7 656 642.5 656.7 643.3 657.3 699.3 711.2 700 711.9 700.2 712.1 737.1 747.8 737.7 747.4 737.7 747.4 762.6 771.3 762.7 772.1 763.3 771.5 780.3 788.3 780.7 788.4 -780.3 788.9 508.8 418.2 510.7 420.2 512.6 422.3 618.9 571.0 619.2 572.0 619.8 572.9 678.7 649.8 679.4 650.3 679.6 650.7 729.2 708.2 729.6 708.7 729.9 708.9 760 744.5 760.4 744:8 760.9 745.2 780.8 767.7 780.4 767.7 780.6 768.0 794.2 783.7 794.4 783.9 794.2 783.8 432.9 400.6 434.2 402.5 435.7 403.6 572.5 553.3 573.3 554.1 573.7 554.8 658 644.9 658.1 645.5 658.5 646.0 713.1 703.1 713.4 703.6 713.9 703.9 749.6 741.4 749.7 741.6 750 741.7 775.2 767.4 775.9 767.9 775.2 . 767.8 ' 791.8 785.0 791.9 785.3 792.7 785.3 APPENDIX F. RESULTS: Table F.2: Run 3 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 953 951 808 816 992 751 963 930 851 2.3 750 872 806 815 914 748 796 864 845 5.3 745 790 803 817 842 744 781 778 840 7.3 739 774 797 797 824 740 774 767 829 9.3 741 772 787 796 814 735 773 764 811 11.3 743 769 776 776 813 729 772 768 797 13.3 750 770 763 771 813 722' 775 769 781 15.3 754 770 751 762 811 714 776 771 765 20.3 766 773 721 739 812 693 779 770 731 Table F.3: Run 4 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 987 974 821 1008 1029 759 1013 930 2.0 741 885 822 752 936 755 830 864 -4.3 733 826 819 740 848 755 820 778 -6.7 V 719 790 815 741 822 752 807 767 8.7 724 779 807 748 820 748 802 764 -10.7 733 780 797 755 817 743 800 768 12.5 745 778 785 767 827 735 794 769 _ 14.5 747 778 773 770 826 729 798 771 20.3 767 780 736 789 828 703 795 770 APPENDIX F. RESULTS: Table F.4: Run 5 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 776 918 812 776 961 734 845 886 872 3.8 779 833 813 746 859 739 815 794 872 5.8 753 797 810 769 852 735 795 785 866 7.9 750 789 804 769 840 734 795 778 854 9.9 753 788 796 763 835 732 796 774 839 11.9 759 788 785 774 833 726 794 776 822 15.0 768 789 765 782 833 715 800 772 795 20.0 781 790 735 798 834 695 802 770 760 f Table F.5: Run 8 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 1028 989 842 1038 1057 749 1041 _ _ 3.7 791 832 843~ 808 908 748 852 - -5.5 762 818 841 788 866 746 842 - -7.6 763 816 836 791 860 745 824 - -9.5 768 815 827 790 854 743 826 - -11.5 773 813 816 792 850 738 825 - -13.6 778 810 804 796 848 732 825 - -16.4 781 790 735 798-' 834 695 802 - -20.4 784 808 785 802 847 721 823 - -24.4 791 805 759 807 846 704 822 - -28.4 795 800 738 812 843 687 819 _ APPENDIX F. RESULTS: Table F.6: Run 9 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0c \u00C2\u00B0C \u00C2\u00B0c \u00C2\u00B0C \u00C2\u00B0C 0.0 1011 836 1025 1047 747 1032 - 905 4.5 767 - 834 777 884 747 823 - 896 6.5 754 - 831 770 853 747 822 - 887 8.5 758 - 823 774 849 745 819 - 872 12.5 775 - 803 803 845 736 823 - 842 19.3 790 - 757 757 849 743 830 - 784 Table F.7: Run 10 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 1028 1065 832 1041 1059 767 1039 1046. 885 4.6 787 857 830 772 873 754 831 842 879 6.8 724 801 827 773 851 754 812 825 870 9.3 751 784 819 776 848 750 804 810 853 12.3 759 785 801 785 846 741 806 811 828 15.3 763 791 780 796 841 727 810 813 802 21.3 784 798 742 807 842 700 811 811 761 25.3 787 800 722 810 839 684 808 811 741 Table F.8: Run 11 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 1051 1077 833 1051 1063 733 1039 1044 869 4.3 773 845 830 792 887 729 823 863 864 5.8 720 794 828 766 853 728 804 817 859 8.9 732 769 820 763 833 705 788 797 842 11.9 742 767 803 771 829 699 784 793 818 16.5 759 778 774 778 829 655 788 790 781 20.3 769 785 747 789 828 626 790 793 753 24.3 776' 787 721 796 825 588 791 793 726 APPENDIX F. RESULTS: 271 Table F.9: Run 12 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 1030 1064 825 1045 1062 - 1042 1049 874 4.2 789 849 823 777 898 - 817 829 871 6.2 761 805 820 774 864 - 804 .810 864 8.8 759 792 814 777 851 - 802 807 852 11.3 765 789 799 781 837 - 808 815' 824 15.2 769 787 776 794 838 - 808 813 799 19.3 778 793 752 801 837 - 808 812 771 24.7 788 800 723 804 833 - 805 806 740 Table F.10: Run 13 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 1045 1078 830 1060 1076 720 1059 1065 890 4.7 752 834 827 795 882 726 854 867 884 6.7 756 803 824 790 864 727 832 841 875 8.7 749 790 817 790 850 728 825 832 863 11.7 756 793 802 792 843 727 820 824 840 15.7 768 792 777 799 840 714 818 822 808 20.7 778 794 747 806 839 694 815 819 772 25.7 781 795 725 809 836 679 812 815 750 30.3 782 794 699 808 830 643 806 806 722 APPENDIX F. RESULTS: 272 Table F . l l : Run 14 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 1043 1080 853 1065 770 751 1071 1078 918 4.2 807 868 849 799 780 748 852 862 907 6.2 781 828 844 796 779 744' 835 845 899 9.0 776 808 834 802 774 740 828 834 879 13.2 780 804 809 808 761 735 828 834 848 18.2 789 806 777 818 741 729 828 832 844 24.3 796 810 742 824 714 722 827 830 771 30.5 798 810 715 822 689 714 822 822 742 34.5 792 807 700 822 673 693 817 818 726 Table F.12: Run 15 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 1036 1066 833 1054 1074 715 1058 1062 893 5.7 803 845 827 810 892 709 844 873 885 9.0 778 808 816 812 862 725 843 848 868 12.3 790 809 801 809 856 696 825 844 845 15.8 779 802 782 813 845 717 830 835 821 19.1 789 804 763 809 845 669 822 830 798 25.3 786 796 733 814 839 677 818 818 764 33.5 785 795 699 802 827 623 804 806 726\" 40.7 772 786 675 797 814 613 795 791 698 47.7 768 777 656 783 803 584 780 781 680 APPENDIX F. RESULTS: 273 Table F.13: Run 16 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C. \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 1017 1051 834 1036 1062 709 1038 1045 889 4.2 761 833 825 790 846 728 848- 835 879 6.4 738 791 819 776 832 728 819 813 871 9.7 756 785 806 777 833 698 802 820 851 14.2 759 779 781 780 826 680 802 813 818 18.8 \"764 780 754 786 824 674 801 810 784 25.7 773 784 717 792 821 646 799 804 745 32.6 775 783 689 791 815 624 792 796 714 39.5 771 777 665 784 805 599 782 785 690 46.4 761 768 647 775 793 584 771 773 671 48.8 758 765 642 772 789 579 767 767 665 V Table F.14: Run 17 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 \u00E2\u0080\u00A2 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 1048 1074 839 1061 1081 700 1061 1066 901 ,3.8 786 877 885 782 893 697. 853 884 896 5.4 756 821 832 780 865 695 829 847 892 7.8 752 793 825 777 840 675 818 828 880 11.2 756 785 810 788 837 668 813 819 854 15.5 768 787 783 805 840 666 816 820 819 21.3 781 794 748 816 842 658 817 818 778 26.3 786 797 722 820 840 646 817 817 749 31.3 788 796 697 815 834 625 812 812 724 36.3 783 793 681 812 828 615 806 806 709 APPENDIX F. RESULTS: 274 Table F.15: Run 18 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 1042 1075 819 1066 1087 709 1066 1072 887 3.8 786 879 820 776 892 719 851 883 886 5.8 755 812 818 781 861 696 825 842 881 8.1 752 789 811 783 843 729 816 ' 827 871 12.0 762 787 794 792 839 726 817 823 843 15.4 773 790 774 803 839 714 819 823 816 20.3 784 798 745 813 840 703 822 825 781 25.6 791 801 717 817 838 678 821 821 752 30.5 792 801 697 815 834 661 816 816 731 35.3 790 797 680 810 826 643 809 809 713 40.4 785 791 665 804 818 637 800 800 698 Table F.16: Run 19 Wall Temperatures Time R l R2 R3 R4 R5 R6 R7 R8 R9 (min.) \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C \u00C2\u00B0C 0.0 . 1037 1070 833 1059 1082 732 1061 1068 897 4.2 763 845 830 771 873 752 830 854 893 5.6 743 800 828 756 845 753 813 829 889 8.5 736 776 819 763 825 756 803 814 874 11.9 749 776 802 778 826 746 807 815 847 15.9 768 786 777 796 833 724 816 821 813 20.7 785 798 747 812 838 717 821 824 779 25.6 794 805 722 818 840 695 822 824 752 30.5 795 805 700 819 837 672 818 819 732 35.3 794 802 684 814, 830 661 811 811 716 40.8 788 795 667 808 821 642 802 802 700 "@en . "Thesis/Dissertation"@en . "2001-05"@en . "10.14288/1.0058975"@en . "eng"@en . "Chemical and Biological Engineering"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en . "Graduate"@en . "Experimental study of the effect of particle size distribution on heat transfer within the bed of a rotary kiln"@en . "Text"@en . "http://hdl.handle.net/2429/11280"@en .