. ’ V L:- . V Lt . \.. V Vl vb ..luc. V A c . Humbug]?! “A V Eff; . . P2 "tit! V Vr' M . n ~ V r so . . V I . 53‘ \ « “Agni: AL, ED}. {Wynn}? hunt-[$93555 V V .. ~ V V fry-.1“. §&.t5\2§ .... . l.-If1(£f x: I . . V» .V\ .V. A V f KL ; . In. t . u . V. xéifik “ i.f.fl«a L V. V V n . V .trtmtmvf» # .. .V... terrier; Riuteé'g‘lfg ..L i . V V . ..V VV_ . L... fix .. .3. V 7 V K tit»? ~ g V. ‘4 Lu") ffuflifin . $3 in ..V. V V... V. L V Kb.l.:a.‘i~.llnfl§. gig. high-u «Hp-g'rhhra‘ V . .HP-bllrnlkut. 14.59.. . . in . ISLE. V V vi V . V. 1 b q . Hit}... 9% V V V . .. 1 V V V . V V V. 1&‘zf6£f\l\ . V. V r i .a VVV. if? I. V 4?! . 1 V . t V V V . V a .. V V \ .34 AV... If. . x V V» r! V . t x i V. a . . . M. V..» . isthFAVVk a . . fl . I . l. I \- V y 0 .9- \1 . I. \x i I; n O9 I .1 V. . . IV if V V V .c 1 e .33“. aw!“ Van]? .1 . 5 A V Vi 0 .. y); ayV . {£{}W£HILI\V;{E V . . - Vrh'. V \l \ .x .. 1.!)H9otiitl, I. .fiFLVrM‘fis: .V. 55‘ II". I?!" 1 V \v. u. \ V V . V 1 5! ~ . .9. fit... V r Offlhflfflflrldu.hul(fvr.$§éd.ffmflst§‘ofi V ..V x V . V . V . . .1. 31 4' . f lerhh it. : Marti? .rMHrhIWQHU «fir-Err...” Vlr......|l.V£VthflV ._ e V . 9...: ..... L v V V WNUWV’VHUWILWIIOXWW fp..ffz.”mrunua§v kw ‘Vrr‘akls . .. I \ \hV . V.. 41}. 'lilllrali V)! .th V .lr ‘53! at? fret. V0.3. ..bt.....V.nnn-.. V»:- .V lfi‘rrhnvhflnfl. 1!! V :3 . . V. rial-(LP)! V 13:31.7 V. 1 V V.. V.. V 1 t V. I . . .ng . . V . a. V V. . L V V V" . )bemeumq V V .x g. . : .V .V \l‘ n . . . V V V » \.\ . V, . V V fi’.‘ V . Vighfwfhnfilttnmwzjffé n V . V: V . A... x a dwwsfiuivvs . . MOVE” .[ISsuoatV at”? 35%. V. V {Furl}. .ith t}, .5. V V V x . V V 1. V V . 5.. r3 . Vlalcihg1§5iorig§hnifklt§§fti Vr. V V x V \3)! “a... .00.. ii 3...... V .- 5 £3.11 n t (”P’flténfiEJVleufEt-£t\ V V V . . 15‘14‘13‘ I V LN” L? 3% ”Sufi? .t wit-)1}: If“. V V Vndqcizf. fttli V Ufrfit..ftr§.VVnnVrfi.VV .. . V . ~74:ng V «V V 33.53:! V V. :fifinfiizfgfbflfriflfVVi k V V 1:54 V». V (4 nVurC. Vistafulilretff} . ..i ‘IIVV..V;}J£&V‘I€VE V V 1" V . - g «“V .31! k. Iv... .erzcihfinmsmf V: x . ~ $3541.13 V :1 V \\ V . V : L V . . h l ‘ 2V 3 V V V 31.2. V V n «343%., $5,; VV . .. ggzél‘fixah) goggfivfi a x V V “2'31“ fli‘lflg’gg \ V». V. V 3.- .)tz‘ \- “I c n . «V k \‘L. r V ‘ 51“.: 33)! V 3‘“. yr: V . V. VI. ‘73)}? V V 531.: \ a #Rutziuflfaectéts albumin . V g3 V! V VV to“ .V . :iifizmu tgghxxfihfiuL fax; gfiVRVHVIfj . ..V . .rA zfisghfist .. t L .11.)...4... ILL! i£ \plf V. £3§§§if V. L... \ A. 3.0: $51!. {sthtltli.:F. sly... plk.$V..Ll.V~Lr ‘tanuinrl'fn. l’rrrerQIL . . . V V. . VV.V .VoVVVVOV, PHIL: H\ V . V git-154V gt; 3 . V. V Vuflnwcsfr: ighundwhil9 ,rIVV he.“ Ir (Vigiirzkftfcr. is»??? sift ...t..;...;. 3...... 5., mun... J. IVirVII . \V V IAVUIIIVLCEV 1).. krfillVfixLi}.llVrfllrI.V III-V V in; (\VIJ AVLa._vup’|47UIFI. 9.1””! 3V {HIP V.A|thi n V . V V fag. 1 t . . 31:17 V V V t . V . art :. VVV. . x . V g V VV \ .I ( . V, V V p x 4 V V: .V . VLW .. n q I V I . l V.. V will. 3 V5 3‘, V 9 . x 1'3 V V at: fig, V. ..3 V V . . mm” . _ V V . g; V V. V . at. 7ft§§ V t|§l‘nvl r l V f. V tritkunkmng . . (It; s V w an... frlflzv“ . V II. . .(‘i V1 ..V! s ‘ ~41 .1 ‘1. V I811}?! in.£1‘~ I; V 2 . V . ~2§ (tips! .tgflvgqnffv.f~tf 731 \ \ V . halitfin in V.. V. . 91‘? V .l .. I. \v... . . .A . V N V .r 1v. VV .V V .VHVVVVVHGEIVVVV ”WWW; 41:1ng figflzkxfksiiz. 131.3333 V V V. 811%}. V. ..T: .SL... Vlyffvro V Lyéifigz! . . . 1 VP. V . . :4 V}: V 3’ fgflflwf}ttvff 57‘91115. V17! «44.11 . .V 4.x?! V ZEVanuVVdu‘rrifitVfi.firfiu {ELHnHrlirf . mtg? V V.. V a thinggzghxiflxfiiv . i. E V . V L.) . lqu «flit-71:15.6 xV . Vmo V L . .V ll...- 3‘. 15.1.5315! I! a. . . V. V r V Alt)?! V fut... .2502), 1 £43523? 1% V V.. I .V. . .. V f‘gl‘ff>! . 3.317)}. is?# V. V?!- Pr » - - . 4 § I fi‘lvtggl. 3‘,§‘5“z’ EL V V 4. VwV I IV. V V. .Yurifffvtfv . 3:41.? . N. V.., . VVVVV C V . .‘lntfafgfig >. 0V . g . a V. . . lV i V O. Ian:- I \ b: T V.- IIIII II I IIIIII II IIIIIIIII III . 1293010119216 LIBRARY Michigan State University 1 3A {3321; This is to certify that the dissertation entitled AUTOMATIC CONTROL OF COMMERCIAL CROSSFLOW GRAIN DRYERS presented by ABBAS YOUSIF ELTIGANI has been accepted towards fulfillment of the requirements for Ph.D degree in Agricultural Engineering fl/MXw/iw Major professor Date 8/10/87 .MSU is an Affirmative Action/Equal Opportunity [run'lution 0-12771 MSU LIBRARIES RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. '7' fin; x 6’7;— ADTOHAIIC CONTROL.OF COMMERCIAL CROSSFLOW GRAIN DRYERS BY Abbas Yousif Eltigani A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY IN Agricultural Engineering Department of Agricultural Engineering 1987 ABSTRACT ' AUTOMATIC CONTROL OF COMMERCIAL CROSSFLOW GRAIN DRYERS By Abbas Yous if Eltigani Automatic control of continuous-flow grain dryers has been commercially available for a decade. These feedback control systems are temperature-activated and relatively inexpensive, but are not able to control the outlet moisture content of the grain in dryers adequately when the inlet moisture varies by more than two percentage points. The goal of this study was to develop a feedforward grain-moisture activated controller, and test the system commercially on crossflow grain dryers. The design objective was to control the exit moisture content to within i1.0% from the set point at inlet moisture variations up to 10 percent. First, an unsteady state model of crossflow grain drying was developed consisting of four differential equations. Solution Of the model requires excessive computer power and time, and thus several empirical models were tested as the process model in the feedforward dryer control system. The simplified empirical drying models predict the exit grain moisture from a crossflow dryer well compared to that predicted by the unsteady-state dryer simulation model and to that measured experimentally. Subsequently, the control system consisting of an empirical drying m0del, an on-line moisture meter, a tachometer, a data acquisition Software. a microcomputer, and the feedforward/feedback control software was implemented, and tested during two drying seasons, on two commercial crossflow maize(corn) dryers. The inlet grain moisture content varied between 16.1% to 34.3% (w.b.). The new control system controlled the outlet grain moisture content to i0.6% of the set point. The automatic control system can be adopted to different dryer types and different cereal types. Advantages of the feedforward control system include: (1) improved dryer control, (2) improved grain quality, (3) improved energy efficiency, (4) improved drying records, and (5) improved dryer economics. Approved Approved Dedicated To the Memory of My Mother, Mariam, My Sister, Madeena, and To Mahasin and Hiba iv ACKNOWLEDGMENT S My sincere thanks to many individuals and Institutions for aid, ideas and support during my stay at Michigan State University. I especially wish to express my deepest appreciation and gratitude to Dr. F.W. Bakker-Arkema for his moral and technical support. He has been an advisor, a friend, a colleague, and above all a true teacher. His example has influenced and will continue to influence all aspects of my life. I look forward to our continued association. Iam indebted to the members of my guidance commitee: Dr. James V. Beck, Department of Mechanical Engineering, Dr. Kris A. Berglund, Departments of Agricultural Engineering and Chemical Engineering and James F. Steffe, Departments of Agricultural Engineering and Food Science. Their advice and encouragement were very helpful. Special thanks goes to John C. Anderson for his friendship, help, and moral support during the last two and half years. Special thanks also goes to Paul Maag and Sons, In. (Eagle, MI) and Anderson Grand River Grain Terminal (Webberville, MI) for allowing me to use their drying facilities. My special appreciation goes to Shivvers, In. (P.O. Box 467, Corydon, Iowa 50060) for the financial support, the moisture meter and help that made this study possible. I would like to thank the graduate students of the Agricultural Engineering Department for their moral support and especially those in room 6. My thanks also goes to Dan Tyrrell for his help in the final preparation of the manuscript. My deepest appreciation goes to the University of Khartoum, Khartoum, Sudan and the Sudanse people, without their financial and moral support this study could not have materialized. Loving thanks is offered to my parents, my sisters, and my brother for all their sacrifices and support. And of course, I thank my wife Mahasin and my daughter Hiba for their help, understanding, and patience during my graduate study. Above all, my special praise and thanks be to Allah "God“, Cherisher and Sustainer of the worlds, Most Gracious, Most Mercifull, for his innumerable bounties. vi LIST OF TABLES TABLE OF CONTENTS 0000000000000000000000000000000000000000000000 LIST OF FIGURES ............................................. LIST OF ' SYMBOLS ............................................. 1. INTRODUCTION_.‘ ............................................ 2. OBJECTIVES 00000000000000000000000000000000000000000000000 3. Literature Review ........................................ 3.1 Types of Dryers ..................................... 3.1.1 Crossflow Dryers .............................. 3.1.2 Concurrent-Flow Dryers ........................ 3.2 3.3 WWW wwwoww:w 2. .2 on 3. 3. 3. unac» axuup o1.3 Mixed-Flow Dryers ............................. odeling of Continuous-Flow Dryers .................. l 2 Single-Kernel Drying Models ................... Deep Bed Drying Models ........................ trol of Continuous-Flow Dryers ................... 1 2 3 Basic Background in Control Theory ............ Definitions ................................... Classical Control Theory ...................... 3.3.3.1 Stability of Classical Control Systems. 3.3.3.2 Design of Classical Control System Feedforward Control Systems ................... Optimal Control Theory ........................ Adaptive Control Systems ...................... 3.3.6.1 Adaptive Controllers .................. Control of Grain Dryers ....................... 3.3.7.1 Control of In-Bin Low Temperature Grain ' Dryers ................................ 3.3.7.2 Control of Continuous-Flow High- Temperature Grain Dryers .............. 4. THEORY .................................................... 4.1 Modeling of Crossflow Dryers ....................... 4.1.1 Introduction ................................. 4.1.2 Development of Unsteady State Grain Drying .l. s 2. .2 .2 2. 3 1D 2D 3 4 Equations ..................................... 4.1.2.1 Energy Balance-Air .................... 4.1.2.2 Energy Balance-Product ................ 4.1.2.3 Mass Balance-Air ...................... 4.1.2.4 Mass Balance-Product ................. Computation Procedure ......................... e ign of the Crossflow Dryer Control System ........ Dryer (Process) Model ..... . .................... Dynamic Compensation ......................... Feedback Correction ........................... Crossflow Dryer Control Algor-thm, ............. vii 5 . EXPERIMENTAL INVESTIGATION ............................... 70 5.1 Equipment ........................................... 70 5.2 Instrumentation and Control System Implementation ... 75 5.3 Procedure ........................................... 81 6. RESULTS AND DISCUSSION ................................... 82 6.1 Simulation .......................................... 82 6.1.1 Unsteady-State Model Verification ............. 83 6.1.2 Empirical Model Verifications ................. 90 6.1.2.1 Exponential ........................... 91 6.1.2.2 Linear ................................ 91 6.1.3 Controller Stability Tests .................... 99 6.2 Experimental Results ................................ 109 6.2.1 Meyer-Morton Dryer ............................ 111 6.2.2 Zimmerman Dryer ............................... 129 6.3 Performance Index ................................... 135 6.4 CONCLUSIONS ......................................... 141 7. SUMMARY .................................................. 144 8. SUGGESTIONS FOR FUTURE STUDY ............................. 145 9. REFERENCES ............................................... 146 10. APPENDICES .............................................. 150 APPENDIX A Experimental Results ......................... 151 APPENDIX B Simulation Results ........................... 184 APPENDIX C Specifications for Data Acquisition Components 201 0.1 A/D Converter Specifications ............. 20 C.2 D/A Converter Specifications ............. 201 6.3 Digital I/O Specifications ............... 202 C.4 Real Time Clock and Counter/Timer ........ Specifications ........................... 203 viii TABLE 3.1 LIST OF TABLES PAGE Calculated energy requirements for the four dryer types operating under conditions which maintain grain quality and allow a grainflow rate of 48.5 kg of grain per hour per meter square of dryer area. .............................................. 9 MSU steady state crossflow drying model ............ 22 At, Ax, Ay, and physical properties of air and corn used in the unsteady state simulation model ... 61 Simulated outlet moisture contents and residence times for different inlet moisture contents; model equations 4.1-4.4. ................................. 62 Dryer specifications for the Meyer-Morton 850 crossflow dryer. ................................... 72 Dryer specifications for the Zimmerman ATP 5000 crossflow dryer. ................................... 74 CPU time for different amount of moisture removed and set points ..................................... 83 Testing the hypothesis with the Student t-test that the mean of the experimental and simulated grain outlet moisture contents for test #1 are equal .... 86 Testing the hypothesis with the Student t-test that the mean of the experimental and simulated grain outlet moisture contents for test #7 are equal .... 92 Parameter estimates for the exponential model(eqn. 4.16) data simulated by the unsteady state model (see Table 4.2) ................................... 93 Parameters estimates for the linear model(eqn. 4.17) using data simulated by the unsteady state model(see Table 4.2) .............................. 94 Inlet M.C. sets used as inputs in the simulation of crossflow grain dryer control system ........... 100 Summary of the results obtained from the different control tests (see Tables A.l-A.10) with the Meyer- Morton 850 dryer .................................. 128 Summary of the results obtained from the different control tests (see Tables A.11-A.14) with the Zimmerman ATP 5000 dryer .......................... 137 Performance indices for the different control tests in Table 6.7 ..................................... 142 0‘ .10 Ranking of different control tests according to different PIs ..................................... 142 6.11a Performance indices for tests 2,8,10, and 11 .... 143 6.1lb Ranking of tests 2,8,10, and 11. ................ 143 A.l Experimental results of an automatic control test on a crossflow dryer. Test Number 1 ............... 151 A.2 Experimental results of an automatic control test on a crossflow dryer. Test Number 2 ............... 153 A.3 Experimental results of an automatic control test on a crossflow dryer. Test Number 3 ............... 15S A.4 Experimental results of an automatic control test on a crossflow dryer. Test Number 4 ............... 157 A.5 Experimental results of an automatic control test ‘ on a crossflow dryer. Test Number 5 ............... 159 A.6 Experimental results of an automatic control test on a crossflow dryer. Test Number 6 ............... 161 A.7 Experimental results of an automatic control test on a crossflow dryer. Test Number 7 ............... 163 A.8 Experimental results of an automatic control test on a crossflow dryer. Test Number 8 ............... 165 . A.9 Experimental results of an automatic control test I on a crossflow dryer. Test Number 9 ................ 167 I A.10 Experimental results of an automatic control test I on a crossflow dryer. Test Number 10 ............. 169 A.11 Experimental results of an automatic control test on a crossflow dryer. Test Number 11 ............. 172 l A.12 Experimental results of an automatic control test on a crossflow dryer. Test Number 12 ............. 175 A.13 Experimental results of an automatic control test on a crossflow dryer. Test Number 13 ............. 178 A.14 Experimental results of an automatic control test on a crossflow dryer. Test Number 14 ............. 181 3.1 Simulation results of an automatic control test i on a crossflow grain dryer. Set Number 1 .......... 184 I 8.2 Simulation results of an automatic control test : on a crossflow grain dryer. Set Number 2 .......... 186 i 3.3 Simulation results of an automatic control test on a crossflow grain dryer. Set Number 3 .......... 188 B. 9 Ln 0" \l on Simulation results of an automatic control test on a crossflow grain dryer. Set Number 4 .......... 191 Simulation results of an automatic control test on a crossflow grain dryer. Set Number 5 .......... 193 Simulation results of an automatic control test on a crossflow grain dryer. Set Number 6 .......... 195 Simulation results of an automatic control test on a crossflow grain dryer. Set Number 7 .......... 197 Simulation results of an automatic control test on a crossflow grain dryer. Set Number 8 .......... 199 xi LIST OF FIGURES FIGURE PAGE 3.1 Schematic of Four Types of Convective Grain Dryers ........................................... 5 3.2 Schematic of a Conventional Continuous-Flow Crossflow Dryer .................................. 7 3.3 Block Diagram of a Single-Stage Concurrent-Flow Dryer with a Counterflow Cooler (Brooker et a1., 1974) ............................................ 11 3.4 Closed-Loop Control System ....................... 27 3.5 Block Diagram of a Control-Loop .................. 28 3.6 Feedback Control System .......................... 31 3.7 Components of a Feedforward Control System ....... 37 3.8 Block Diagram Representation of an Adaptive Control System ................................... 40 4.1 Mass and Energy Balances on a Control Volume Within a Crossflow Grain Dryer .................... 51 4.2 Flow Diagram of the Unsteady-State Computer Program ........................................... 59 4.3 Control Algorithm for Crossflow Grain Dryers ..... 69 5.1 Schematic of the Meyer-Morton 850 Dryer .......... 71 5.2 Schematic of the Zimmerman ATP 5000 Dryer ........ 73 5.3 Schematic of the Control System for a Crossflow Dryer ............................................ 76 5.4 Corn Outlet Moisture Content Comparison (Oven, COMP-U—DRY, Motomco) ............................. 78 5.5 Corn Inlet Moisture Content Comparison (Oven, COMP-U-DRY, Motomco) ............................. 79 6.1 Simulation vs Experimental for Test #1 with the Meyer-Morton Dryer; differential-equation model used ............. ................................ 84 6.2 The Difference between Experimental and Simulated Grain Outlet Moisture Content vs Time for Test #1 ................................. ~ .............. 87 6.3 Simulation vs Experimental for Test #7 with the MeyeréMorton Dryer; differential-equation model used ............................................. 88 xii 0" b U1 0‘ \l on \D .1 .1 .1 .1 .1 .1 .1 .1 .1 .1 .2 The Difference between Experimental and Simulated Grain Outlet Moisture Content vs Time for Test #7 ............................................... The Exponential Model vs the Unsteady-State Differential-Equation Model for the Drying of corn in the 850 Meyer-Morton Dryer .................... Exponential Model Outlet Moisture Content vs Unsteady-State Model Outlet Moisture Content in Drying Shelled Corn in the 850 Meyer-Morton Dryer The Linear Model vs the Unsteady-State Differential-Equation Model for the Drying of Corn in the 850 Meyer-Morton Dryer .................... The Linear Model Outlet Moisture Content vs the Unsteady-State Model Outlet moisture content in Drying Corn in the Meyer—Morton 850 Dryer ..... Simulation of the Automatic Control of the Meyer- Morton 850 Dryer(set #1) ......................... 0 Simulation of the Automatic Control of the Meyer- Morton 850 Dryer(set #2) ........................ H Simulation of the Automatic Control of the Zimmerman ATP 5000 Dryer(set #3) ................ N Simulation of the Automatic Control of the Meyer- Morton 850 Dryer(set #4) ........................ 0.) Simulation of the Automatic Control of the Meyer- Morton 850 Dryer(set #5) ........................ 9 Simulation of the Automatic Control of the Meyers Morton 850 Dryer(set #6) ........................ U'I Simulation of the Automatic Control of the Meyer- Morton 850 Dryer(set #7) ........................ 0‘ Simulation of the Automatic Control of the Meyer- Morton 850 Dryer(set #8) ........................ 7 Inlet and Outlet Moisture Contents, and RPM vs Time During Manual Control of the Meyer-Morton 850 Dryer (1984) ................................ 8 Inlet and Outlet Moisture Contents, and RPM vs Time During Manual Control of the Meyer-Morton 850 Dryer (1985) ................................ 9 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer- Morton 850 Dryer in Test #1 ..................... 0 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer- xiii 89 95 96 97 98 102 103 104 105 106 107 108 115 .21 .22 .23 .24 .25 .26 .27 .28 .29 .30 .31 .32 .33 Morton 850 Dryer in Test #2 ..................... 116 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer- Morton 850 Dryer in Test #3 ..................... 117 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer- Morton 850 Dryer in Test #4 ..... ................ 119 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer- Morton 850 Dryer in Test #5 ..................... 120 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer- Morton 850 Dryer in Test #6 ..................... 121 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer- Morton 850 Dryer in Test #7 ..................... 123 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer- Morton 850 Dryer in Test #8 ..................... 124 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer— Morton 850 Dryer in Test #9 ..................... 126 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control (linear model) of the Meyer-Morton 850 Dryer in Test #10 ....... 127 Inlet and Outlet Moisture Contents, and RPM vs Time During Manual Control of the Zimmmerman ATP 5000 Dryer .................................. 130 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Zimmmerman ATP 5000 Dryer in Test #11 ...................... 131 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Zimmerman ATP 5000 Dryer in Test #12 ...................... 133 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Zimmmerman ATP 5000 Dryer in Test #13 ...................... 134 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Zimmerman ATP 5000 Dryer in Test #14 ...................... 136 xiv LIST of SYMBOLS Ao constant A1 constant Ai constant B constant 0 Bl constant C specific heat, kJ/kg-oc D diffusion coefficient, mz/hr 2 G flow rate, kg/hr-m h convective heat transfer coefficient, kJ/hr-mZ-oc fg latent heat Of vaporization for water in the grain, kJ/kg M moisture content, decimal, d.b. T air temperature, 0C t time, second, minute or hour V velocity, m/hr W absolute humidity of air, 1b water/lb dry air x coOrdinate direction along the width of the dryer, m y coordinate direction along the length of the dryer, m subscripts a air p product v water vapor w water liquid Greek Symbols 0 e p grain temperature, 0o bed porosity density, kg/m3 XV 1 INTRODUCTION Grain dryers normally operate by forcing hot air through a static or moving layer of grain. The drying process is an energy intensive process when weather conditions do not allow for low-temperature drying systems. In such cases, a high-temperature drying system is a suitable alternative to a low—temperature drying system. A high-temperature drying system is energy efficient when the dryer operates at a high inlet air temperature and a low airflow rate. Grain quality is one of the limiting factors in the use of high inlet air temperatures. The main objective of a drying process is to decrease the grain moisture content from one level of moisture content to a desired lower level of moisture content(set point). As the inlet moisture content of the grain entering the dryer changes, the above objective becomes difficult to achieve. Underdrying and overdrying usually take place due to the moisture variation in the grain entering the dryer. Overdrying and underdrying are not only caused by the variation in the inlet moisture content of the grain, but may also be a result of changing weather conditions and internal factors related to the dryer or grain. Underdrying is most serious, since wet spots and spoilage of the grain may occur. Overdrying is expensive due to the unnecessary costs in fuel, labor, maintenance, and investment. With a wide range of inlet moisture contents, grain dryer operators have a difficult task to control the grain outlet moisture content. The 2 usual approach taken to ensure that all grain is dried to or below a set point, is to overdry. Thus, the necessity exists for a system with the capability to control a dryer automatically. It should be noted that even an experienced dryer operator is not able to adjust the dryer parameters (such as the inlet air temperature or grain flow rate) properly to obtain exactly the desired average outlet moisture content. The unavailability of an inexpensive and yet accurate on—line moisture meter has had a negative effect on the development of an automatic dryer controller. This situation has lead some researchers to correlate the outlet grain temperature with the grain outlet moisture content, since grain temperature can be measured relatively easily. The use of such controllers for commercial grain dryers has not been successful, since many factors in addition to a change in the inlet grain MC can change the air exhaust temperature. The recent development of an on-line moisture meter, and the need for a dryer control system, constitute the inspiration for this work on a MC-based grain dryer control system. CHAPTER 2 mums The objectives of this dissertation are: 1. To develop a mathematical unsteady state grain-drying model for use in an automatic dryer control algorithm for crossflow grain dryers. 2. To develop a control algorithm for the control of commercial crossflow grain dryers. 3. To combine an on-line moisture meter, the grain-drying model, the control algorithm, and a motor controller into an automatic control system for crossflow grain dryers. 4. To test the control system on several commercial crossflow grain dryers. 5. To evaluate the newly—developed automatic dryer control system, and recommend alternative dryer models and control algorithms. CHAPTER 3 3 .W Wye—rs Various types of dryers are used in drying grains to the desired moisture content. The most common types of grain dryers make use of passing air through the grain. Heat and moisture are transferred between the passing air and the grain kernels by convection, and thus such grain dryers are named convective grain dryers. I Grain dryers fall into two categories, namely batch dryers and continuous-flow dryers. Batch dryers are characterised by the fact that grain is dried either with heated air or with near-ambient air in stationary bed depths up to several meters. In near-ambient, or low- temperature drying, the drying process takes place over many hours, days, or even months. Batch dryers will not be discussed here, a detailed discussion of the subject is given by Brooker et al.(1974). Continuous-flow dryers are classified by the relative direction of air and grain movement through the dryer. Several types are shown in Figure 3.1. In crossflow dryers, the flow of air is perpendicular to the flow of grain. The air and the grain move in the same direction in concurrent-flow dryers; in counter-flow dryers, the air and grain flow in opposite directions.1n mixed-flow dryers, the air flows partially in the grain direction and partially opposite to the grain direction. 3.1.1 Crossflow Dryers Crossflow dryers are simple in construction. They generally have a lower initial cost than other continuous-flow dryer types. Commercial crossflow dryers are usually non mixing type dryers. Groin leecl 390' .___,__. Crossflow .._>___ Air Alp 6'”an Grain __._. Concurrent ._,___ __,_ Counter ._,_ Flow Flow Alr - Alr Figure 3.1 Schematic of Four Types of Convective Grain Dryers. The drying process in a crossflow dryer is achieved by allowing wet grain to flow from the holding bin down the drying columns. Hot air is forced across the columns to heat the grain and to remove the evaporated moisture from the grain. The dried grain is cooled and unloaded.at the bottom of the dryer (Fig 3.2). The grainflow rate is regulated by the grain discharge augers at the bottom of the dryer columns. One of the major disadvantages of a crossflow dryer is the moisture gradient which develops across the drying column as the grain flows through the columns. Over-heating, over-drying, and over-cooling are characteristics of kernels at the air inlet side, whereas under-drying and under-cooling of kernels occur at the air outlet (Gygax et a1., 1974). Gustafson and Morey(l981) investigated experimentally the moisture gradient and grain quality across the drying column of a crossflow dryer. They found large differences in moisture content, grain temperature, breakage susceptibility, and germination across the drying column. Raising the drying temperature, and/or removing more moisture, reduced the overall quality of the grain. Grain turning midway through the drying section or reversing the airflow are the methods used to reduce the temperature and moisture gradients occurring across the drying columns. Air-recycling results in an improvement of the energy efficiency of crossflow dryers. The energy consumption of conventional crossflow dryers without air recycling is 7000-9000 kJ/kg(3017-3878 Btu/1b) of water removed (Nellist, 1982). Pierce and Thompson (1981) investigated the influence of various dryer operating parameters on the performance of severalcnxmsflow 7 FILLING AUGER <_ —>- FAN «t- AND .-.-_.~;.' :zg H E AT E R 45:: 3:1- «4— _,-;.-;_}‘.; -— — — 121-3:: 0211' o?” '32; 5.: HEATED AIR PLENUM ’ COOUNG AIR PLENUM -:,,/GRA|N METER UNLOADING AUGER Figure 3.2 Schematic of a Conventional Continuous-Flow Crossflow Grain Dryer (Brooker et a1., 1974). 8 grain dryers: (a) a conventional crossflow dryer, (b) a reversed air- flow model, (c) a reversed airflow dryer with air recirculation of the cooling air and 50% of the heating air (Hart-Carter), and (d) a recirculating model which re-uses the cooling air and the drying air from the second stage. The comparison of these units, drying corn at the same capacity from 25% to 15% moisture content (w.b.) under ambient conditions of 10 degrees C and 50% relative humidity, is shown in Table 3.1. It is clear that modification of the conventional crossflow dryer can decrease the energy requirements and improve the grain quality without affecting the dryer capacity. Differential grain-speed and tempering are two recent features added to the basic crossflow dryer design. Differential grain-speed refers to the movement of grain close to the air inlet side at a faster speed through the column than grain at the air outlet side (Bakker- Arkema et a1., 1982). The variation in grain speed is accomplished through dual discharge rolls rotating at different speeds. The optimum speed ratio depends on the grain type and the initial moisture content. Differential grain-speed improves grain quality, increases dryer energy efficiency and dryer capacity (Bakker-Arkema et a1., 1982). Tempering between subsequent drying passes or stages in multi-stage drying systems is practiced with rice. During tempering the temperature and moisture gradients within the individual rice kernels diminish (Steffe et al.; 1979, Ezeike and Otten; 1981), resulting in less subsequent fissuring and breakage. Bakker-Arkema et al. (1982) tested a commercial crossflow corn dryer which with differential grain-speed, tempering and air recycling features. The energy efficiency of the dryer was found to be 3700 kJ/kg (1600 Btu/lb) of water removed. Table 3.1; Calculated energy requirements for dryer types operating under conditions which maintain grain quality and allow a grainflow rate of 48.5 kg of grain per hour per meter square of dryer area. Dryer Total Energy Drying Airflow Maximum Moisture Type kJ/kg H20 Air Tem. Rate, Grain Differential (°c) (m3/mi m2) Tem. (°c) (a, w.b.) Conven. Crossflow 6940 68 42 60 5.0 Reversed Crossflow 7020 68 41 60 1.9 Hart- Carter 4890 65 58 60 1.3 Recicul. Air Dryer 4380 66 51 60 1.1 Source: Pierce and Thompson(l98l) 10 3.1.2 Concurrent-Flow Dryers A concurrent-flow dryer consists of one or more concurrent-flow drying beds coupled to a counter-flow cooling bed (see Fig 3.3). In multi-stage units, a tempering zone separates two adjacent drying beds. The air and grain flow in the same direction with the hottest air encountering the wettest grain. Concurrent-flow dryers have only recently become available commercially. The drying temperature in concurrent-flow dryers is not limited by the type or moisture content of the product since grain velocity is the governing factor. Air temperatures up to 500 0C are used in drying corn without affecting product quality (Hall and Anderson, 1980). The high rate of evaporation cools the air rapidly and prevents excessive grain- kernel temperatures. As the grain moves downward, its temperature increases rapidly, and then decreases slowly along with the drying air temperature. The high drying-air temperatures result in a high energy efficiency and a low airflow requirement for the concurrent-flow dryer. Moisture and temperature gradients among the dried kernels are small in concurrent-flow dryers since each kernel undergoes the same drying, tempering, and cooling treatment, in contrast to grain dried in crossflow dryers. The grain temperature is better controlled in concurrent-flow dryers and the maximum air temperature is maintained for a much shorter period of time in the drying section than in other dryers types. In the counterflow cooling section, hot grain is cooled gently due to the small difference (5-10 °C) in temperature between the warm grain kernels and the cooling air. Due to the beneficial effects in the ll GRAIN IN AIR IN 11 I HEATER II CONCURRENTFLOW DRYING In . ) DRYING AIR OUT ' A //”A ) I COOLING AIR OUT' A r COUNTERFLOW COOLING GRAIN OUT AIR IN Figure 3.3 Block Diagram of a Single-Stage Concurrent-Flow Dryer with a Counterflow Cooler (Brooker et a1., 1974). 12 drying and cooling sections, concurrent-flow dryers produce a higher quality grain than other dryer types(Bakker-Arkema et a1., 1981; Fontana et a1., 1982). The energy efficiency of concurrent-flow dryers with and without air- recirculation ranges from 3000 to 3800 kJ/kg (1293 to 1637 Btu/lb) of moisture removed(Nellist, 1982; Bakker-Arkema et a1., 1982). Thus, concurrent-flow dryers are energy efficient in comparison to crossflow dryers. The design of multi-stage concurrent-flow dryers allows the use of high grain velocities and high inlet-air temperatures. Increased dryer capacity, improved grain quality, dryer controllability, and improved thermal efficiency are the advantages of multi-stage concurrent-flow dryers compared to single-stage units. 3.1.3 Mixed-Flow Dryers In mixed-flow dryers, grain is dried by crossflow, concurrent-flow, and counter-flow. Grain flows over rows of alternate inlet and exhaust air ducts. Due to the combined effect of different drying mechanisms, mixed-flow dryers can be modeled as series of crossflow, concurrent- flow, and counter-flow submodels (Parry, 1985). The inlet air temperature in mixed—flow dryers can be higher than those used in crossflow dryers, since grain is not subject to the high temperature for long period of time. 13 3,2 Modeling of Continuous-Flow Dryers WM Some biological products when dried as single particles under constant external conditions, exhibit a constant-rate drying during the initial drying period, followed by a falling-rate drying period. A critical moisture content separates the two drying periods. During the constant-rate drying period, the material remains at the wet bulb temperature of the air. The rate of surface evaporation is determined by the rate of diffusion of water vapor through the film of air surrounding the product; thus, the drying rate is proportional to the difference between the partial pressure of the water vapor of the material and that of the drying air. The mechanism of moisture removal is equivalent to evaporation from a body of water and is essentially independent of the nature of the solid. The magnitude of the constant-rate drying depends upon three factors: (1) the heat or mass transfer coefficient; (2) the area exposed to the drying medium; and (3) the difference in the vapor pressure between the gas stream and the boundary layer surrounding the wet surface of the solid. The three factors are external; thus, the internal mechanism of liquid flow does not affect the constant—rate drying period. For individual grain kernels, the constant-rate drying only occurs when the moisture content is sufficiently high to maintain a surface layer of free water(Parry,l985). For corn, this only happens at moisture contents over 50%. Thus, harvested grain kernels dry entirely within the falling—rate drying periods. Theoretical, semi-theoretical, and empirical models have been developed to describe the transport of moisture from the interior to the surface of a grain kernel during the falling-rate drying period. l4 Luikov(l966) proposed a number of physical mechanisms to describe the transfer of moisture in capillary-porous products such as grains: (1) liquid movement due to surface forces(capillary flow); (2) liquid movement due to a moisture concentration differenCe (liquid diffusion); (3) liquid movement due to diffusion of moisture on the pore surfaces(surface diffusion); (4) vapor movement due to a moisture concentration difference (vapor diffusion); (5) vapor movement due to a temperature difference (thermal diffusion); and (6) water and vapor movement due to a total pressure difference (hydrodynamic-flow). Based on the above mechanisms, Luikov(l966) developed a mathematical model for describing the drying of capillary porous products. The model equations are a system of partial differential equations ' 6M 2 2 2 __ - V KllM + V K126 + V K13P (3.1) at 66 2 2 2 _ = V KZlM + V K229 + V K23P (3.2) at 6P 2 2 2 __ - V K31M + V K326 + V K33P (3.3) at where K 11, K22, and K33 are the phenomenological coeffic1ents while the other K—values represent the coupling coefficients. The coupling 15 results from. the combined effects of the moisture, temperature, and total pressure gradients on the moisture, energy, and mass transfer. Although, a modified form of Luikov's model was used in analyzing drying of rough rice (Husain et a1. , 1973) , lack of knowledge of the phenomenological coefficients hindered the application of Luikov's model to cereal grains. The. liquid/vapor diffusion theory has been used extensively in grain drying studies by different researchers, with the grain kernel shape assumed as a sphere. The following partial differential equations describe the moisture diffusion in spherical and rectangular coordinates: spherical 3E - _£_ 3_ (r2 D 3% ) (3.4) at 2 8r 8r r rectangular if - 1w 3”.“ > + a_ (D i“ > + 1(1) 35) (3.5) at 6x' 6x 8y By 62 az Bakker-Arkema and Hall(l965), Becker and Sallans(l955), Chittenden and Hustrulid(l966), Chu and Hustrulid(l968) , Hamdy and Barre(l969), I'Ienderson and Pabis(l96l,l962), Rowe and Gunkel(l972), Steffe and Singh(l980), Watson and Bhargava(l974) , and Young and Whitaker(l97l) , used the diffusion theory to analyze drying of different grain types. irhe majOr assumptions made by these researchers are: 1. the grain kernels are homogeneous and isotropic; 2. the diffusion coefficient is constant, or varies with temperature and/Or moisture content; 3.the mass transfer coefficient at the kernel surface is infinite, finite, or varies with time; 4.the initial moisture content distribution is uniform; and 5.the temperature gradient in the kernels during drying is negligible. Results have shown that the estimate of the diffusion coefficient(D) depends on the grain and the co-ordinate system of the diffusion equation. However, the general solution to the diffusion equation.has the form of a series of negative exponential terms, regardless of the particle geometry or the boundary conditions (Moon and Spencer, 1961) M - Me -B t MR - - 121 Aie 1 (3.6) M-M 0 e where, M - average moisture content of grain at time t (d.b) Ai = constant, characteristic of the material being dried, dimensionless B1 = constant, characteristic of the material being dried, hr.l 1H0 - initial moisture content of the material,(d.b) Die _ equilibrium moisture content,(d.b). fix moisture relationship analogous to Newton's law of cooling is often uSed in single-kernel drying analysis (Brooker et. a1, 1974). Thus, the t.att.e of moisture loss of aigrain kernel is proportional to the difference between the kernel moisture and its equilibrium moisture c olatent : __ - -k (M - Me) (3 7) dt where -l k- drying constant, hr . When the drying air is at constant temperature and relative humidity, Me is constant. Thus, solving equation (3.7) gives: M - M MR - e - e'kt (3.8) M - M o e Several purely empirical drying equations have been developed for cereal grains. Thompson (1968) proposed the following thin layer equation for shelled corn over the temperature range from 140 to 300 degrees F : c = A ln MR +3 ln (MR)2 (3.9) A = -1 862+.00488 T B - 427.4 exp(-.O33 T) Where t is the drying time in hrs, MR is the moisture ratio, T is corn temperature in 0F, and A and B are empirical coefficients that are flulctions of temperature. Flood et al.(1972) proposed the following empirical drying equation f0r‘shelled corn over the range 36 to 70 degrees F: MR - exp(-k c'66“ ) (3.10) where k = exp( -x ty ) (3.11) x,y are nonlinear functions of relative humidity and temperature: 1/2 1/2 x - (6.0142+.0001R2) -O.OlT(3.353+.001R2) 5 5RI - 5.8*10' T y - 0.1245-.0022R+2.3*10- 7’ "IV‘ 18 R - relative humidity, decimal T = temperature, F Henderson and Henderson (1968), Nellist and O'Callaghan (1971), Rowe and Gunkel(l972), Henderson(1974) and Nellist(l976) have fitted a two term series of negative exponentials to experimental thin-layer drying data for rice, rye grass seeds, alfalfa hay, shelled corn, and rye grass seeds, respectively. The time response equation has the general form : MR = A0 e-Bot + Al e'BIt (3.12) Sharaf-Eldeen et a1. (1980) found the two-term exponential model accurate over the whole range of drying for fully exposed ear corn. The model predicted the drying behavior of ear corn to within 1% moisture content of the experimental values. 3.2.2 Deep Bed Drying Models Deep bed drying of cereal grain has received major attention from researchers during the past 20 years. The moving bed is characteristic of continuous-flow dryers whereas the stationary bed is characteristic of batch dryers. Deep-bed drying models are generally divided into three types, logarithmic, heat and mass balance, and partial differential equation InOclels. The three types have some common features which suggest the division is arbitrary. Hukill(l954) made a simplified analysis of deep-bed drying and derived the one equation model 8T 6M G c _ - p h __ (3.13) a a 6x P fg 6t — ‘,:.—.— . - 19 where Ga - mass flow rate of moist air,kg/ m2 sec C a specific heat of moist air , J/kg-degree c x depth, m t - time, sec density of grain, kg/m3 1: I Z I moisture content of product, d.b. hfg- latent heat of vaporization of water, J/kg Using exponential temperature and moisture boundary conditions, .Hukill developed the following solution to eqn. (3.13): MR - (3.14) where x and t are dimensionless depth and time variables, respectively. Hukill's model underestimates the time required to dry grain to a Specified moisture content. Hukill suggested that this is due to inaccuracy in the boundary condition used for Me. Young and Dickens (1975) used Hukill's model to estimate the costs c>fgrain drying in fixed bed and crossflow systems. Baughman et al.(1971) proposed a relationship between the t:emperature and moisture gradients in a stationary bed of grain: ET 311 caca_ =--Qh __ (3.15) 8x fg 6x “fliere Q is the rate of advance of the drying zone. Using equations 3.13 aIKi3.15 they obtained a simplified drying equation: 6(MR) _ -l 6(MR) (3 16) 8t l-MR(0,T) 6X 20 h kp (M -M )x where r-kt, X- fg p o e GC(T -T) aa 0 e , and k is a drying constant. Barre et al.(1971) solved equation 3.16 assuming initial and boundary condition of the form MR (0,1) = exp(-1') (3.17a) MR (X,O) - l (3.17b) to model a crossflow dryer. They found the model to be fairly reliable in predicting the deep-bed drying in a crossflow dryer. The model was also used to compare the relative influence of parameters such as temperature, airflow, moisture content, and air humidity on the efficiency and capacity of a crossflow drying system. Sabbah et al. (1979) employed the log model to simulate the solar drying of grain. Thompson et a1. (1968) developed a series of deep-bed drying models based on heat and mass balances of a series of thin grain layers. Steady state crossflow, concurrent-flow, and counter-flow drying were Simulated with good accuracy. Boyce(1966), and Henderson and Henderson (1968) used similar simulation procedures to simulate the drying of Stationary deep beds of grains. A more fundamental approach, based on the laws of simultaneous heat and mass transfer and resulting in a series of coupled partial differential equations, was developed by Bakker-Arkema et al.(1974) at Fiichigan State University (MSU). Separate sets of three partial differential equations (PDE), plus an appropriate thin-layer rate e’CII-!.ation, were employed to model various stationary and continuous flow drYing systems. The MSU steady-state crossflow drying model is shown in Table 3.2. 21 Equations 3.18-3.22 can be solved by numerical integration employing an explicit finite difference technique. The PDE for crossflow, concurrent flow, and counter flow dryers are similar in form to the fixed-bed drying model. Laws and Parry (1983) presented the MSU PDE models in a general form. The PDE models have a sound thermo-mechanical basis in contrast to the other types of deep bed drying models(Parry, 1985). 22 Table 3.2 MSU steady state crossflow drying model 6T 67 80 8y 8W '3;— am 3y— an F Boundary T(0.y) - 9(x,0) - W(0,y) - M(x,0) = h L (H) V p C a a am h a V p C C V P P Pm Pm P pm P an appropriate thin-layer equation Conditions: T(inlet) 0(initial) W(inlet) M(initial) (3.18) (3.19) (3.20) (3.21) (3.22) 23 3.3 Control of Continuous-Flow Dryers WW Automatic control has played a vital role in the advancement of engineering and science. In addition to its extreme importance in space-vehicle, missle-guidance, and aircraft-piloting systems, automatic control has become an integral part of modern industrial manufacturing. For example, automatic control is essential in controlling pressure, temperature, humidity, viscosity, and flow in the food processing industry. 3 3.2 Definitions The terminology used in describing control systems includes the following terms (Ogata, 1970; Baumeister et a1., 1978): (1) Plant: A plant is a piece of equipment performing a particular operation. (2) Process: A process is an operation or development marked by a series of gradual changes which succeed one another in a relatively fixed way and lead toward a particular result or end. (3) System: A system is a combination of components which act together and perform a certain objective. (4) Disturbance: A disturbance is a signal which tends to adversely affect a system. (5) Feedback Control: A feedback control is an operation which, in the presence of a disturbing influence, tends to reduce the difference between the output of a system and the reference input. (5) F ”L ' Control System: A feedback control system is one which maintains a prescribed relationship between the output and the reference input by using the difference as a means of control. (7) Servo-mechanism: A servo-mechanism is a feedback control system in which the output is a valve position, velocity, or acceleration. 24 (8) Automatic Regulating System: An automatic regulating system is a feedback control system in which the reference input or the desired output is either constant or slowly varying with time, and in which the primary task is to maintain the output at a desired value in the presence of a disturbance. (9) Process Control System: A process control system is an automatic regulating system in which the output is a variable such as temperature, pressure, flow, liquid level, or moisture content. (10) Closed-Loop Control System: A closed-loop control system is a system in which the output signal has a direct effect upon the control action. -(11) Open-Loop Control Systems: An open-loop control system is a system in which the output has no effect upon the control action. (12) Adaptive Control System, : An adaptive control system is a system which has the ability to self-adjust or self-modify under unpredictable changes in input or environmental conditions. (13) Controlled Variable: A controlled variable is the variable of the controlled system which is directly measured or controlled. (14) Response Time: The response time is the time required for the controlled variable to reach a specified value after the application of a disturbance. (15) Peak Time: The peak is the time required for the controlled variable to reach a maximum following the application of a stepwise disturbance. (16) gise Time: The rise time is the time required for the controlled variable to increase from 10 to 90%, 5 to 95%, or 0 to 100% of its final value, following the application of a stepwise disturbance. OJ) SggglgyLlimg: The settling time is the time required for the absolute value of the difference between the controlled variable and 25 its final value to become (and remain) less than a specified value, following the application of a step disturbance. (18) Transfer Function: The transfer function, 6(5), of a linear system is the ratio of the output transform, Y(s), to the input transform, U(s), given the initial system conditions are zero. 3,3,3 Classical Control Theory Classical control theory deals with single input-single output (SISO) linear systems, and utilizes the block diagram approach for system representation. Figures 3.4 and 3.5 show simple and detailed block diagrams of a closed-loop feedback control system, respectively. The system components are described by the transfer functions of each component. The closed-loop transfer function of the control system is used for analysis, design and synthesis of the control system. The closed-loop transfer functions of the two control systems shown in Figures 3.4 and 3.5 result in the following two equations: G (s) G (s) 32 = c xuanummu mmmuorm Lomcmm I 959". Amy! mm 33 A35 38 a .lwl + w + 3m u + _+ «bosom .m. cmzohhoo 9%.“qu 3.5.80 is». W}... nmvwou W > A a a I Lavage sou Amun m a mUcdnczva ill")! ll\\\s- 29 to produce multiple jw axis poles and are bounded otherwise. Thus, a linear system with poles in the left half plane and multiple poles on the jw axis is clearly unstable since the multiple jw axis poles give ,rise to a time response of the form tp_l, p > 1. The key to determine whether a linear system is stable, unstable or marginally stable is to locate the system poles in the S-plane by solving the system characteristic polynomial explicitly for the system poles. An alternative method which does not require solution of the characteristic equation has been proposed by Routh(1877). The Routh's Stability criterion is based on the value and sign of the elements of _the first column of the Routh array. If the elements of the first column are positive and non zero, the system is stable. If any of the elements is negative, the system is unstable. Thus, the Routh stability criterion eliminates solving the characteristics equation, but requires the system to have a polynomial characteristic equation. 3.3.3.2 Design of Classical Control System Figure 3.6 shows a block diagram of a feedback control system which will be used in the discussion of classical control system design. Gp(s) is the given transfer function of the process being controlled with C(s) as the control variable. R(s) is the the desired (reference) value for C(s), E(s) is the error signal and U(s) is the controllable input to the process represented by Gp(s). Transfer functions GC(s) and R(s) are transfer functions which can be specified by the designer to achieve a desired behavior for the controlled variable, C(s). The closed-loop transfer function for the system in Figure 3.6 with feedback control is: C“) — GC(S) GP(S) (3.27) R(s) 1+H(s) GC(s) Gp(s) 30 The three control (design) objectives are: 1) system stability under all system operating conditions; 2) “good" steady-state error performance; 3) "good" system dynamic or transient performance. The Routh criterion is helpful in determining the GC(s) and R(s) values which result in the desired stability. However, design stability is often considered along with the design of dynamic performance. Design of steady-state error performance starts with the application of the final value theorem which requires that the limit of the error exists as time goes to infinity. The steady state error performance tends to worsen as the number of poles at 8-0 of the input increases; it tends to improve as the number of poles at S=O of Gp(s) increases. Since Gp(s) is usually fixed, poles at S=0 are added to the controller function GC(s) by the so-called "proportional plus integral control". In proportional plus integral control, the input u(t) is computed as a function of the error and the integral of the error: t u(t) = K e(t) + K I e(¢) d1 (3.28) p I 0 or KI E(s) U(s) - K E(s) + (3.29) P s where u(t) - input from the controller l R( ) + E(s) l s GC(s) 31 U(s) Gp C(s) H(s) Figure 3.6 Feedback Control System. 32 Kp — proportional parameter KI - integral parameter e(t) (E(s)) = error between the set point and the actual output. The drawback of adding integral control to a system with proportional control is the tendency of integral control to reduce the range of parameter values (KP’KI) for which the system is stable (Manetsch and Park, 1982). Design of dynamic performance is usually an objective for systems which have to adjust quickly to input changes. There are several dynamic performance criteria which should be measured when a step change occurs in the system input: 1. rise time (see section 3.3.2); 2. settling time: the time required for the output to reach and remain within a given percentage i a % of the input; and 3. maximum overshoot: the maximum overshoot of the output as a proportion of the input value. To achieve the above three dynamic performance measures, the Root Locus design technique can be used to choose GC(s) (and perhaps E(s)) so that the resulting pole locations will result in the desired values for the dynamic performance measures (Manetsch and Park, 1982). A basic technique for improving the dynamic performance of a system is the use of derivative control along with proportional control (Manetsch and Park, 1982). Proportional plus derivative control is represented by the following equation: u(t) = er(c) + Kr i“; (3.30) dt Where u(t) = input from the controller to the plant 33 Kp = proportional parameter K . . r - derivative parameter e(t) = error between set point and the plant output. In control problems requiring improvement in both dynamic performance and steady state error, it is common to use the so-called PID control (Proportional-Integral-Derivative Control) (Manetsch and Park, 1982). Integral control is used for steady state error improvement while the derivative control operates to improve dynamic performance: C de u(t) - K e(t) + K f e(1)dr + K _ (3.31) p I 0 r dt U(s) then becomes: KIE(s) U(s) - K E(s) + + S K E(s) (3.32) P S r The transfer function GC(s) (of equation 3.32) is: 2 K (S +S K /K + K / K ) GC(S) _ U(s) _ r p r I r (3.33) E(s) S The main effect of the PID control is to introduce one pole (at S=O) and two zeros into the S-plane. By properly choosing KP, K and KI, the r control engineer has the option of locating two zeros in the S-plane. 3.3.4 Feedforward Control Systems In all processes the point at which the material enters the process and the point at which it leaves the process are not the same. The longer it takes for a material to move from the entrance to the exit of 34 a process, the more difficult it is to control the process. Tfiuns, the lenger the process dead-time, the more difficult it is to maintain the controlled variable at the desired set point. This is especially true when the load variables of a process change frequently, and the rate of ichange is large. To control a long dead-time process, it is desirable to account for a variation in the load at the time the variation takes place. This is done in so-called feedforward control systems. The elements needed in l implementing a feedforward control are shown in Figure 3.7; they linclude a process model, a dynamic compensator, and a feedback corrector. The feedforward process model is developed by using material and energy balances, and several empirical relationships. The manipulated variable is computed as a function of the measured variable and the set point. Changes in the load are corrected by the feedforward controller. If the load variables are measured correctly and the relationships between the manipulated variables are exactly known, perfect control can be achieved. Major load variables are identified according to tfiuazfrequency of change and the magnitude of the change. The major load variables are always measured; the minor load variables are not because they cause only small disturbances in the process. When the load and the manipulated variables enter the process at different locations, a dynamic imbalance may take place, and dynamic compensation in the form of lag, lead/lag and/or dead-time is required to minimize the effect of the dynamic imbalance. Dynamic compensation greatly improves the performance of a feedforward control system (Badavas, 1984). 35 A feedforward control system can provide excellent control if the process can be modeled accurately. Inadequate feedforward control results from: (1) inadequate modeling of the process; (2) inaccuracy in the load- variable measurements; and (3) computational errors. The cumulative effect of errors in feedforward control computations results in an offset of the controlled variable from the set point. To eliminate the offset, a feedback controller must be added to the control system. The feedforward controller corrects the variations in the major load variables while the feedback controller corrects errors due to the minor load variables. The feedback controller has a smaller cOrrective action than the feedforward part, and is referred to as the feedback "trim". The feedback trim can provide an adjustment to a model coefficient, and thus can result in a major change to the controlled variable. 3,3,5 Optimal Control Theory In conventional(classic)h control theory, the analysis and design of a control system is carried out with transfer functions and graphical techniques. A major disadvantage of the classical control theory is the fact that it is limited to linear time-invariant systems with a single input and single output. Thus, conventional control is powerless for time-varying systems, non-linear systems, and multiple-input-multiple- output (MIMO) systems. Due to the complex nature of many engineering systems, a new approach has been developed to analyze and design control. systems for such systems. The approach is based on the state variable concept (smallest set of variables which determine the state of a system). It is applicable to MIMO linear, nonlinear, time-invariant or time-variant MIMO systems . 36 Application of optimal control requires the selection of a performance index and a design procedure which can yield an optimum 1;;thin the limits imposed by the physical constraints. The performance index results in a number which indicates the "goodness" of performance. It is optimal if the values of the parameters are chosen so that the selected performance index has reached a minimum or maximum. A quadratic performance index is frequently used in optimal control systems. The performance index determines the optimal system configuration. It must be pointed out that an optimal control system operating under a given performance index is not optimal under other performance indexes. Thus, in practical systems, it is more sensible to seek optimal control which is not rigidly tied to a single performance index. Analysis of a given optimal control strategy is important since it aids the designer in determining whether a performance index is realistic for a given system and set of constraints. Controllability and observability are the two most important questions regarding the existence of an optimal control point. A system is said to be controllable at time to if it is possible to transfer the system from an initial state x(to) to another state in a finite interval of time. A system is said to be observable at time to if it is possible to determine the state of the system by observing its output over a finite time interval. The concepts of controllability and observability are important in the optimal control of multivariable systems. The solution of an optimal control problem may not exist if the system is not controllable. Although most physical systems are controllable and observable, corresponding mathematical models may not possess the 37 .8393 Houucoo pumaowpomm m mo mucoconrnoo 5m oudwfim mm3d_td> zoos L05: 0 63.2.3) _omzo.._u.coo , . mmiaic> i mmmuocd mde_L0fidz y cofiamcmdeoo : .mid_Ld> 0.233 _omvgzdidz .fi Lmzotficoo Il‘lY. V mfiEom xuanpmmm vmw , cofidvsdcoo .Ochou pLdztowpmmu 38 property of controllability and observability. Therefore, it is necessary to analyze the conditions under which a system is controllable and observable. 3,3.6 Adaptive Control Systems The interest in adaptive control systems has increased rapidly. The term adaptive system has a variety of meanings, but usually implies that the. system is capable of accommodating changes, whether these changes arise within or external to the system. Adaptive control has a great advantage to the system designer since it tolerates moderate design errors or uncertainties. In most feedback control systems, small deviations of a parameter value from the design value do not cause problems in the normal operation of the system, provided the parameter is inside the loop. If a parameter varies widely with environmental changes, the control system may respond satisfactorily to one environmental condition but may be unstable under other conditions. If a model parameter can be estimated continuously, variations in modeling can be compensated by adjusting the controller parameters so that satisfactory system performance is achieved under various environmental conditions. Such an adaptive approach is useful for solving a problem in which the plant parameters change from time to time. Different definitions of adaptive control systems can be found in the literature. The vagueness surrounding the definitions and classification of adaptive systems is due to the large variety of mechanisms by which adaptation can be achieved. The various definitions arise because of the different classifications and definitions which. divide control systems into adaptive and non-adaptive systems. 39 An adaptive control system can be defined as a system which measures continuously and automatically the dynamic characteristic of the plant. The difference between the measured and the desired dynamic characteristics is used to generate an actuating signal so that optimal performance is maintained regardless of an environmental change. Also, such a system may continuously measure its own performance according to a given performance index and modify its own parameters (Ogata, 1970). 3,3,6,l Adaptive Controllers An adaptive controller has the following three functions: (1) the estimation of the dynamic characteristics of the process; (2) the decision-making based on the estimated parameters of the process; and (3) the modification or actuation based on the decision. If the process model is not well known due to random time-varying parameters or the effect of an environmental change on the plant dynamic characteristics, identification, decision, and modification procedures must be carried out continuously, or at intervals of time based on the rate of change of the plant parameters. A block diagram representation of an adaptive control system is shown in Figure 3.8. In this system, the process is identified and the performance index measured continuously or periodically. The performance is compared with the optimum, and the decision is made based on the actuating signal needed to achieve the optimum. The dynamic characteristic of the process must be measured and estimated continuously, or at least frequently. Estimation of the process parameters may be made from normal operating data of the process or by use of test signals. 40 .Eoummm Houuaoo m>Hudmp< cm mo sowumquwmudmm Emuwmwa xooam w.m muswam W?U0¢¢U HGVCOECOLt’r—U a ”.3330 mmmuocd ill V cofiduificmg fil cmzorvcoo llllY :29qu co$au¢=ooz 41 Parameter estimation must be rapid to account for any variation in the process parameters. Estimation time should be short compared to the environmental changes. Once the process has been estimated, it is compared with the desired characteristic. Subsequently, the decision is made how to vary the adjustable parameters in order to obtain the desired performance. The control signals are modified according to the results of the estimation and decision. In most schemes, the decision and modification are conceptually a single operation with the modification consisting of a means of mechanizing the transformation of a decision output signal into a control signal(the input to the process). The control or input signal to the process can be modified in two ways. The first approach is to adjust the controller parameter in order to compensate for changes in the process dynamics. This is called controller parameter modification. The second approach is to synthesize the optimal control signal based on the process transfer function, the performance index, and the desired transient response. This is called control signal synthesis. The choice between controller parameter modification and control signal synthesis is primarily a hardware decision since the two methods are conceptually equivalent. In cases in which reliability is important, the use of parameter change adaptation is favored over the use of control signal synthesis (Ogata, 1970). . In conclusion, most control systems which require precise performance over a wide range of operating conditions are adaptive to some extent. When high adaptability is required, an estimation- decision-modification system is needed with either sequential or continuous modification, depending on the rate of change of the varying parameters. g f— 42 3.3,7 Control of Grain Dryers The optimum operation of grain dryers is accomplished by obtaining / 51:) the desired grain moisture content at minimum energy use and grain I--H. . deterioration and af/maXimum capacity. A considerable amount of extra energy- is consumed during incorrect operation, such as overdrying. In addition to a waste of energy, overdrying impairs grain quality, and increases fuel cost, labor, and maintenance. The control of grain dryers is usually achieved manually. In manually-controlled dryers, the dryer operator adjusts the grain flow rate and/or thedrying-air temperature so that the desired moisture content is reached. A skillful operator is required for adequately controlling a grain dryer. Automatic control of grain dryers has recently become a popular research topic. The literature on automatic grain dryer control can be divided into two catogeries: 1. control of in-bin low-temperature grain dryers; and 2. control of continuous-flow high-temperature grain dryers. 3.3.7.1 Control of In-Bin Low—Temperature Grain Dryers Kranzler (1976) developed a control scheme for low—temperature ..Md-JII-Hy-h ... .'-‘ drying _of shelled corn using long-term weather data and simulation of several control modes. The control schemes were wired into an array of integrated circuit elements. The operator can input an'anticipated combination of harvest conditions. The control system then determines the humidistat and fan control strategy at the optimum operating points. Morey et al.(1978) simulated for the Corn Belt region of the US several different fan-management strategies for ambient drying systems by using a low-temperature drying model and the appropriate weather data. They concluded that continuous fan operation proved to be more 43 energy efficient than fan control based on relative humidity, temperature or time. Simonton et a1. (1981) investigated a microprocessor-based grain drying control system. Their objective was to predict the performance of a low-temperature drying system using a simulation based on the logrithmic drying model. They also developed a method for controlling the output moisture of a continuous-flow dryer; grainflow rate was used in controlling the dryer output. The control algorithm was implemented using a microprocessor, a digital-to-analog (D/A) converter, interfacing circuitry, an analog-to-digital (A/D) converter, and a motor controller. The grain flow rate was controlled by varying the motor speed of the unload auger. The Simonton control system is yet to be implemented on a continuous flow grain dryer. Derret and Allison (1981) reported experimental results of a microprocessor-based control for an in-bin grain drying system. Bin radius, grain depth, air flow rate, initial grain moisture content, desired grain final moisture content, and allowable drying time were input variables used in the drying algorithm. The control algorithm of Simonton et al.(1981) was utilized with the drying air as the control variable. They obtained at the laboratory level acceptable agreement between the calculated and measured moisture contents. A low-temperature corn drying control system was investigated by Mittel and Otten (1983). Ambient air temperature and relative humidity were used as the drying parameters in the control algorithm; a microcomputer with a dual disc drive and 48k memory was employed. The relative humidity and temperature sensors were interfaced to the microcomputer through analog to digital converters and a timer-counter board. The authors utilized the the thin-layer drying and wetting equation of Mishra and Brooker (1979), the desorption equilibrium 44 moisture content equation of Gustafson and Hall (1974), and the sorption equilibrium moisture content of Thompson as quoted by Morey et a1. (1979). The Mittel-Otten control algorithm is based on five indices to be specified before drying is started: 1. the relative humidity to control the drying fan; 2. the relative humidity to start searching for alternatives other than continuous fan operation without supplemented heat; 3. the relative humidity to control the heater; 4. the initial time period for which continuous fan operation is acceptable; and 5. the moisture content in the upper 10% of the bin. The above indices are used in making the following decision and control steps: a) If the relative humidity of the air is less than the set relative humidity to control the drying fan or the total drying time is less than the set time at which the continuous fan operation is acceptable, the drying fan is on but the the heater and aeration fan remain off. b) If the relative humidity of the air is greater than the set relative humidity to control the drying fan and less than or equal to the set relative humidity at which alternatives other than operating the fan without supplemental heat is searched for, the heater is turned on to decrease the relative humidity. c) If the ambient air relative humidity is greater than the relative humidity at which alternatives other than the continuous operation of the drying fan without supplemental heat is searched for, the heater and the dryer fan are turned off and the aeration fan is started, or the fans and heater are turned off depending on the moisture content of the grain at the upper 10% of the bin. 45 The ’Mittel-Otten simulation results of the control algorithm show that 5 to 31% of the energy can be saved compared with high-temperature drying, and 10 to 19% compared with uncontrolled low-temperature drying, depending on the weather conditions. The control algorithm was not tested on an actual low—temperature drying system. 3 ontrol 0 Co t'nuous- ow Hi h-Tem erature Grain Dr ers The first significant paper on the automatic control of continuous- flow grain dryers was co-authored by Zachariah and Isaacs (1966). Classical control theory was applied to a crossflow dryer. Three control systems were tested -- a proportional-integral-derivative (PID) system, a feedforward system with feedback trim, and an on-off feedback system; the drying process was modeled by Hukill (1954) deep-bed drying equation. .Due to the unavailability of on-line computing and moisture measurement in the sixties, the Zachariah/Isaacs control system‘was simulated, but not implemented on commercial dryers. Holtman and Zachariah (1969a) compared the Hukill drying model with limited experimental data, and with an empirical model in which the moisture content in the continuous-flow dryer is assumed to vary linearly with time. The linear model was recommended for dryer-control applications on the basis of accuracy and simplicity. In a later study, Holtman and Zachariah designed an optimal control system for a crossflow grain dryer using quadratic programming in conjunction with the linear drying model. The Holtman-Zachariah optimal control system could not be implemented due to the excessive on-line calculation requirements. Borsum et a1. (1982) utilized microprocessor-based technology for the automatic control of a concurrent-flow grain dryer. An inferential proportional-integral feedback control algorithm, based on the outlet air and the outlet grain temperatures, was experimentally tested. 46 Although acceptable control-accuracy was reached, the authors recommended development of a continuous moisture-content meter to be used in conjunction with a feedforward controller for control of the varying dead-times and reaction rates in commercial-scale dryers. Schisler et a1. (1982) investigated the optimal dryer-control strategy for concurrent-flow drying assuming the inlet gradxirmoisture content and the outlet grain temperature are measured continuously. The control algorithm is based on the transient solution of the partial- differential-equation steady-state drying model. Lack of an inlet moisture-measuring device prevented implementation of the control system. Fbrbes et al. (1984) first employed a continuous-flow moisture meter for the control of a commercial grain dryer. They compared two exponential-decay model-based feedforward controllers with a PID feedback controller and a lead/lag feedforward controller, using simulation. The first of the model-based controllers employed the inlet grain moisture content as the load variable, while the second utilized for this quantity the average of the moisture content of the inlet grain and of the grain presently in the dryer; the second controller best controlled the outlet grain moisture, and was subsequently tested. successfully on a commercial scale. Adaptive control was investigated for continuous-flow grain drying, by Nybrant and Regner (1985) and by Nybrant (1986); they developed a microprocessor controller based on the dryer air-exhaust temperature. A linear-difference form of a time-discrete model constitutes the process model; it combines recursive least-square identification with minimum variance control law. The controller was implemented on a laboratory- scale crossflow wheat dryer. Nybrant and Regner suggested that a 47 controller based on direct moisture measurements might lead to an improvement of the adaptive dryer control. Marchant (1985) reviewed the state of continuous-flow dryer control, and concluded that proportional—integral (PI) controllers are unlikely to meet the. control requirements of grain dryers. He conducted a simulation study of a model-based control algorithm containing an exponential drying equation of similar form as utilized by Forbes et al.(l98'4). No experimental data was presented by Merchant; he suggested intermittent measurement of the moisture content every five minutes if a continuous moisture meter was developed. A partial-differential-equation steady-state simulation model of a grain dryer was adapted by Whitfield (1986) to predict the unsteady states resulting from varying inputs; the approach is similar to that of Schisler et a1. (1982). The simulated data formed the basis for the choice of the parameters in a feedback PI controller. The non- linearities in the drying process are not taken into account in this controller-type; therefore, the PI controller is unstable under certain operating conditions. In conclusion, it is clear that automatic control of grain dryers requires microcomputer process-control in conjunction with continuous or semi-continuous measurement of the controlled variable (i.e. grain moisture} content). Because of the long (1-3 hours) dead-times and the frequent and large load upsets, feedforward controllers have innate advantages for continuous— —flow grain dryers over proportional, PI and PID'controllers. Feed- forward controllers require a model for the (i.e. moisture content) which calculates the correct control signal for the present input-load condition and set point A number of drying models (i.e. linear, exponential, adaptive) have been proposed but none has 48 thus far proven to be superior for the control of continuous-flow grain dryers . 49 CHAPTER 4 4 . THEORY 'The theoretical part of this investigatiOn is divided into two section-s. In the first section, the modeling of the crossflow dryer during steady and unsteady state operation is discussed. In the second section, the design of the control system for crossflow dryers is considered. 4.1 Modeling of Crossflow Drvers 4,1,1 Introduction Drying of agricultural products such as grain depends on the cxnumct between the drying-air stream and the bed of grain kernels during which both heat and mass transfer take place. The heat transfers from the hot air to the cold grain, while the moisture is transferred from the grain to the air. Heat and mass balances are made to develop mathematical models to describe the drying process. The models are derived with certain assumptions to facilitate their development, solution and applications. Equations 3.18-3.22 represent the simulation model for crossflow grain drying obtained from energy and mass balances. The model is a steady-state model; it used extensively in analyzing and designing cmossflow grain dryers (Brooker et a1., 1974). However, due to the steady-state nature of the model, it is not suitable for use in automatic control of crossflow grain dryers. Thus, an unsteady-state model for crossflow dryers needs to be developed. The development of this model is presented in the following section. 50 4,1.2 Development of Unsteady-State Grain Drying Equations Unsteady-state energy and mass balances for air and grain are written on a differential volume located at an arbitrary position in the grain bed of a crossflow dryer. Figure 4.1 shows the control volume along with the air and grain as they enter and leave the control volume. In developing the unsteady-state crossflow drying equations the following assumptions are made: 1. no appreciable volume shrinkage occurs during the drying process; 2. no temperature gradients exist within the grain particles; 3. particle to particle conduction is negligible; 4. air and grain flowrates are plug type; 5. dryer walls are adiabatic with negligible heat capacity; 6. the heat capacities of moist air and grain are constant;and 7. Vp is constant during a dt time step. Assumption (1) is disputable since shrinkage occurs during drying. The shrinkage effect has been considered by Spencer(l972) in simulating wheat drying in a fixed—bed dryer; however, he did not indicate whether correction for shrinkage improves the simulation results. The other assumptions have been shown to be valid for continuous- flow dryers (Bakker-Arkema et a1., 1974). 51 ill GRAIN —> —> AIR I I —-—> —> Y X v c v M pp p pm pp p l p V C T v c H a a m —-D— —+. pa a m(T+a_ dx) l x w l paVaW . _ dy . paVa(W+§_ dx) . X | dx as am v c 9+ d pp p pm( a— y) ppr(M+5 dy) Figure 4.1 Mass and Energy Balances on a Control Volume Within a Crossflow Grain Dryer 52 4 l 2 Ener Balance-A‘r energy in - energy out - energy transferred = energy accumulated 6T pavacm T dydt - paVaCm (T + 5; dx) dydt - h a (T-6) dxdydt - epaCm 6T dx dydt at or, 6T 6T ep C - p V C - h a (T-0) a m 5;- a a m 3;- or, 6T Va aT h a ___ - - ___ .___ -_______(T-0) (4.1) at 6 6x ep C a am where C - C + W C m a v 4 a a - oduc energy in + energy transferred - energy out + energy to evaporate water + change in sensible heat of grain w.r.t. time + change in sensible heat of water vapor 60 8M ,V C 6d dt+ h T—6 d d dt - V C 6+ d d dt + h - pp p pm x a( ) x y pp p pm( —ay 3*) x fg( ppat 60 6M dxdydt) + p C dxdydt + (C (T-6))(- p dxdydt) P Pm V P 8t at or, 8M 69 86 h a (T-0) + (p C (T-fi) + p h ) -p V C - p C p v p f5 5;“ p p Pm3;- p pmat 53 c (T-0) h 60 = h a (T-9) + ( V + f3 ) 6M - v 33_ (4.2) 6y at p c c c at P p pm Pm Pm where C = C + C M pm P W 4.1.2.3 Mass Balance-Air water vapor in - water vapor out + change of water vapor in the air within the control volume - rate of water vapor evaporated from the grain p v w dydt - p v (w+aw dx)dydt + 6p aw dxdydt aa aa a 8x at - 6M dxdydt at V p 6W = a 8W + p 8M (4 3) at 6 6x spa 8t 4.1.2.4 Mass Balance-Product water in solids in - water in solids out = change of MC of the solids in the control volume w.r.t. time V M dxdt-p V ( M + 8M dy )dxdt= p dxdydt 8M P P P By P at or, 61 = - v .3”. (4.4) at P 8y 54 W ‘The finite difference technique is used to solve equations 4.1 - 4.4 along with the empirical thin-layer equation for corn proposed by Thompson (1968), the DeBoer empirical equation for the equilibrium Imoisture content (Bakker—Arkema et a1.,1974), and the SYCHART package for moist air properties given by Bakker-Arkema et a1. (1974). The following finite difference terms are substituted for the corresponding partial differential terms: 8T _ Tx+Ax,y,t+At ' Tx,y,t+At (4 5) 5E7 Ax T - T 8T = x+Ax,y,t+At x+Ax,y,t (4 6) at At 9 - 0 80 = x+1/2Ax,y+Ay,t x+l/2Ax,y,t (4 7) 6y Ay 0 - 0 60 _ x+1/2Ax,y,t+At x+1/2Ax,y,t (4 8) a? At M - M 6M _ x+1/2Ax,y,t+At x+1/2Ax,y,t (4 9) at At M — M 8M = x+l/2Ax,y+Ay,t x+l/2Ax,y,t (4 10) 8y Ay 8W = Wx+Ax,y,t+At - wx,y,t+At (4 11) 8x Ax 55 6W Wx+Ax,y,t+At - wx+Ax,y,t at At (4.12) Equations 4.5-4.12 are substituted into equations 4.1-4.4. Three equations are formed.for three of the four unknowns, namely 0, W, and T: oi,j,k+l - (1'A6)*91,j,k + A3*THT/B3 - A5*(C$THT+hfé*(Wi+1,j k - Wi’j’k)/B3 + A6* TP (4.13) wi+1,j,k+1 ‘ (wi+1,j,k ' Al*wi,j,k+l )/A8 ' A4*(Mi,j,k+1' Mi,j,k)/A8 (4.14) Ti+1,j,k+1 ‘ (Ti+l,j,k + A1*Ti,j,k + Bl*91,y,k+1 )/32 (“'15) Mi j k+1 is calculated using the thin-layer equation evaluated at the following temperature, specific humidity, and relative humidity values: ai,j,k + Ti,y,k+1 2 Temperature - Wi,j,k + wi+1,j,1< 2 Specific humidity - ( + W i,j,k+1 )/2 Relative humidity = RH (Temperature, Specific humidity) where, the subscripts i,j,k are equivalent to x+l/2Ax,y,t for M and 0, and to x,y,t for T'and W. Other subscripts should be interpreted accordingly. Also, _ * Al Ga At / (pa*e*Ax) 56 A2 = h * a * At / (pa* 6) A3 - h * a * At / pp A4 = pp / 6*pa A5 - pa* At / (Ax * pp) A6 = pp* At / (pp* Ay) A7 = p / p Bl - A2 / (Ca + Cv* (wi,j,k+l + wi+l,j,k)/2 ) BZ - 1 + Al + Bl BS - C + C * M. . p w i,j,k THT “(T1,j,k + Ti+l,j,k ) / 2 ' 0i,j,k TP ' (91,3,k + 91,j+1,k )/2 The following calculation scheme is followed after the first time step and during which Vp is assumed to be constant: 1. increment dryer depth; 2. increment time; 3. calculate 0i,j+l,k+l u51ng equation (4.13); 4. calculate Mi.J+1,k+1 uSing the thin layer equation, (3.9); 5. calculate W. 1+1 j+1 k+l using equation (4.14); 6. calculate Ti+l,j+l,k+l uSing equation (4.15); 57 7. increment x and repeat steps 3 through 6 until the air exit has been reached; 8. read the new value for Vp; and, 9. go back to step 1 unless the total length of the dryer or grain exit has been reached. The above scheme along with the equations for the four unknowns are implemented in a computer program written in Fortran. Figure 4.2 shows the flow diagram of the computer program. The program simulates the unsteady state drying of a crossflow dryer and acts as the basis for the simulated portion of the automatic control of the crossflow dryer. The values of Ax, At, and Ay along with the physical properties for air and corn are given in Table 4.1. The values for Ax and At are kept constant due to stability reasons. At At - .006 hrs (21.6 sees) the program is stable for all grain flow rates used during the simulation (i.e. 5.6 to 13.7 m/hr). The simulation program is not affected by a change in Ay (.034 to .082 m) due to changes in the grain flow rate. The heat transfer coefficient is assumed to vary with the airflow rate only. Since the air flow rate is constant, the heat transfer coefficient is also constant. This may introduce an error due to the lack of information on how the heat transfer coefficient varies with grain flow rate. The surface area of corn per unit volume of bed is assumed to be constant. In the development of the unsteady-state model one, of the assumptions is that no shrinkage occurs during drying. As discussed earlier, shrinkage does occur but is considered to be of minor influence. The remaining properties (i.e. specific heat, density, etc) for air and grain may vary with temperature. It is assumed that the variations 58 Reodlmtml condVUons :4 + Read IMet nc. - lid ' i Increment flme,depth ——»+ Calculate gram temp Qrmn mc. Mr spec.hum. and Mr temp V Increment mdth Is math Mcrements 59 @@ <9 4 l‘ + Reod new IMet mt. Is depth Increments done Figure 4.2 Flow Diagram of the Unsteady-State Computer Program. 60 are small and result in negligible errors in the simulation results. A sample output of the unsteady-state simulation model is shown in Table 4.2. The Table shows the relationship between the inlet moisture content, the outlet moisture content, and the residence time. 4,2 Design of The Crossflow Dryer Control System Commercial grain dryers are characterised by large dead-times and frequent inlet moisture content variations, especially at terminal grain elevators. This creates a difficulty in controlling dryers with regular feedback controllers, since dead-time represents an interval during which the control system has no information about the effect of a previously taken control action. A better control system will be one that corrects for the variation in the grain inlet moisture content by measuring the load variable at the dryer inlet. Such a control system is known in the literature as a feedforward control (Badavas, 1984). A feedforward control strategy is used in this study for the control system of commercial crossflow dryers. To design and implement a feedforward control system, three elements are needed; (1) a process model, (2) a dynamic compensation model, and (3) a feedback correction model (see Section 3.3.4). The three elements are investigated below with reference to the control of continuous-flow grain dryers. 4.2.1 Dryer (Process) Model The partial differential equation model developed in Section 4.1 to model the crossflow dryer is accurate, but needs main-frame capability 61 Table 4.1 At, Ax, Ay, and physical properties of air and corn used in the unsteady-state simulation model. Parameter Units Value At hr .006 Ax ft .01 Ay ft .11 to .27 a ft'1 239 ca Btu/lb 0F .242 cp Btu/lb 0F .268 cv Btu/lb 0F .45 cw Btu/lb °F 1.0 h Btu/ hr-ftz-OF .363*GA 59 for CA < 500 .69*Ga'49 for Ca 2 500 hfg Btu/lb 1000 for M 2 .17 (1094-576)[l+4.35exp(-28.25M)] for M < .17 3 pa lb/ft .075 3 pp 1b/ft 38.7 e dimensionless .45 62 Table 4.2 Simulated outlet moisture contents and residence times for different inlet moisture contents; model equations 4.1—4.4. Inlet M.C Outlet M.C. Residence Time (%,w.b.) (%,w-b.) (hr) 20.00 11.34 1.13 20.00 11.74 1.06 20.00 12.42 0.95 20.00 13.27 0.83 20.00 14.04 0.72 20.00 14.35 0.68 22.00 12.03 1.28 22.00 12.36 1.22 23.00 10.78 1.64 23.00 13.84 1.12 23.00 14.04 1.09 23.00 14.53 1.02 23.00 15.05 0.95 23.00 15.44 0.89 23.00 15.85 0.84 24.00 18.18 0.67 24.00 18.03 0.68 24.00 17.49 . 0.75 24.00 16.91 0.82 24.00 16.49 0.88 25.00 12.52 1.60 25.00 14.19 1.33 25.00 18.49 0.74 25.00 18.40 0.76 25.00 18.10 0.79 26.00 13.62 1.55 26.00 14.69 1.38 27.00 18.90 0.93 27.00 18.47 0.98 27.00 18.05 1.04 27.00 17.72 1.08 27.00 17.36 1.13 28.00 15.71 1.49 30.00 16.49 1.64 Notezcrain VeloCity 5.6 to 13.7 m/hr . Column Length 9.1 m (30.ft) Dryer Width .305 m“(1 ft) .Airflow Rate 24.4 m3/m2-min (80 CFM/ftz) Air Temperature 104.4 0C (220 oF) Air Specific Humidity .0032 Initial Grain Temp. 4.4 °C (40 0F) 63 to be used for onmline calculation. Thus, a simple dryer model, ___ _._~._ ..4.._ -..- ... accurate enough for control purposes, needs to be_developed if‘a feed- ...,p..- ...—...... ,. -.., w-..— forward dryercontrol system is to be used (Holtman and Zachariah, {Sggzgiw'fl‘ . Two empirical dryer models have been proposed in the literature for the design of control systems for grain dryers: an exponential model (Matthews, 1985) M(t) - exp(-filt) (4.16) M(O) and a linear model (Holtman and Zachariah, 1969a) _E£El_ - 62 + B3t (4.17) M(O) The two empirical models compare well with the partial differential equation model for the crossflow dryer (see Section 6.2.2). The parameters in equations (4.16) and (4.17) are computed every time the grain outlet moisture content, grain inlet moisture content, and the unloading auger rpm are measured. The value of '31 in equation (4.16) is calculated using the following equation: 6 - (1n M(O) )/t (4.18) 1 M(t) where, t is the residence time of the grain exiting the dryer, hours; li(t) is the outlet moisture content, decimal (w.b.) corresponding to the inlet moisture content M(O) for the given residence time t. The two parameters 62 and ,93 in equation (4.17) can not be estimated directly, since only one set of measurements is available eadh time the two parameters have to be estimated. To estimate the two 64 parameters with one set of data, a sequential least square estimation method is employed (Beck and Arnold, 1977): A1 = xlpll + x2912 (4.19) A2 - x1912 + x2922 (4.20) A = A x + A x + 02 (4 21) 2 2 1 1 A? p11 _ - + P11 (4 22) A A A 1 2 p12 — - + P12 (4.23) A A3 922 - - + p22 (4.24) A Y-X b -x b e 1 1 2 2 (4 25) A A 8 b1 - A1 + bl (4 26) A 9 b2 — A2 ___ + b2 (4.27) A where, X1 = 1 x2 - T Y M(t) M(O) b1 and b2 are estimates of 62 and 63, respectively. The initial conditions are b1 = b2 = 0. P11 = P22 = 100000. P = 0. 12 65 2 0 =1. Because of the inaccuracy in estimating b1 and b2 during the initial dryer start up, the parameters are not used in the dryer control decision until the estimates are converging. LAW Dynamic compensation is needed when dynamic imbalance exists in a system. Dynamic imbalance is the result of the different response of the controlled variable to changes in the manipulated variable compared to changes in the load variable. To improve the performance of the feedforward control syStem, a dynamic compensation is required. In grain dryers, the dynamic imbalance is the result of the large dead-time which varies with the grainflow rate. When the manipulated variable (grainflow rate) is adjusted due to a major load change (i.e. the inlet moisture content), the adjustment affects the grain already in the dryer to a different degree based on how long a layer of grain has been in the dryer. To reduce the dynamic effect, a pseudo inlet moisture content is defined. The pseudo inlet moisture content (Mps) is defined as a weighted average of the moisture content of the inlet grain and the grain currently in the dryer (Olesen, 1976; Forbes et a1., 1984). The weights are chosen such that the incoming grains and the grain at or near the top of the dryer, have a larger influence than the grain near or at the bottom of the dryer. The pseudo inlet moisture content is calculated from the following equation: MpS - b1M(1) + b2M(2) + ...... + bnM(n) (4.28) where, b + b + ...... + b = 1 0 66 where n - the number of samples used in the calculation of the pseudo inlet moisture content, the subscript l = the present inlet moisture content, and the subscript n - the moisture content of the grain near the outlet of the dryer. The value of n was chosen between 10-20 depending on the residence time, and thus the length of the drying column and flow rate of the grain. Different values for b1, b2, ..., bn were investigated in the calculation of the pseudo inlet moisture content; b1 - l/(lISZZ/i) and bi - (2/i)/(lf222/i) were found to give a value for Mps which results in excellent automatic control of a crossflow dryer. .4.2 3 Feedback Correction A feedforward control system controls the dryer perfectly if the drying process is modeled correctly, and accurate measurements and computations are made. However, errors do occur due to inaccurate assumptions in the drying model, inaccuracies in the moisture content measurements, changes in the minor loads variables (grain test weight, wind effects, BCFM, etc.), and due to computation errors. The feedback correction corrects for feedforward model inaccuracies. The feedback correction is achieved by incorporating the present value of the estimated parameters in the control decision for the next time interval. For the exponential model (see eqn. 4.16), an exponential smoothing is used to "correct" the parameter, 61 (Montgomery and Johnson, 1976): 316(11)- (l-A)*Bl + A*filc(n-1) (4.29) 67 where, filc(n) - the parameter value used in the present control decision fllc(n-l) - the parameter value used in the previous control decision 61 - the present estimated parameter value A - the smoothing constant ( O S A s 1). For the linear model, no filtering is necessary since sequential parameter estimation provides filtered estimates of the parameters (Beck and Arnold, 1977). 4,2,4 Crossflow Dryer Control Algorithm The control system algorithm is a feedforward model-based type with feedback correction and dynamic compensation. The control algorithm for the crossflow dryer is shown in Figure 4.3. The flow chart is drawn for the exponential model, and is equally valid for the linear model. The calculation scheme for the control algorithm (Figure 4.3) is as follow: 1. initial conditions (rpm, set point, inlet MC) are set; 2. Inlet and outlet moisture contents of grain and unload auger rpm are measured; 3. B1 is estimated by equation (4.18)and used to calculate Blc(equation 4.29); 4. the pseudo inlet MC is calculated using equation (4.28); 5. the required residence time is calculated using Mps and Blc; 6. the residence time is converted to its equivalent voltage and send to the SCR ; and 7. go back to step 2. 68 START l Set hutml Condltlons i N = N + 1 A i Cou For M and RPM V Prmt Measurements V Esflmote Parameters Bl 1 G) 69 3 ® 9 Blc(n)=(l-A)KBI +AHBlc(n-l) CoL Pseudo Inlet M.C. ‘7 Colc. Required Residence Time Dryer Control Panel - . Colc. RPM 1 Col. Voltage Voltage , N Y Time To Stop Figure 4.3 Control Algorithm for Crossflow Grain Dryers. 70 CHAPTER 5 5 EXPERIMENTAL INVESTIGATION 5,1 Eguipment i The crossflow dryer control system was implemented on two commercial crossflow dryers manufactured by Meyer-Morton, Inc.(P.O. Box 352 Morton, Illinois 61550) and Zimmerman, Inc.(P.0. Box 331, Litchfield, Illinois 62056). A schematic of the Meyer-Morton 850 dryer is shown in Figure 5.1. The dryer specifications are listed in Table 5.1 (Anderson, 1985). The heating section is 27.5 feet in length, the cooling section is 11.5 feet. The grain column thickness is 10 inches at the upper and 12 inches in the lower part of the dryer. The dryer is modified to incorporate a heat recovery enclosure for air recycling. Air to the heater is a combination of ambient air and recycled air. The recycled air is a mixture of air exhausted from the cooler and part of the air exhausted from the drying section. The rated capacity is 1400 bushels of wet corn per hour at 5 point moisture removal. The Zimmerman ATP 5000 dryer is shown schematically in Figure 5.2. The dryer specifications are listed in Table 5.2 (Anderson, 1985). The length of the of the heating section is 66.8 feet, of the cooling section 18.5 ft. The column thickness is 12 inches over the entire dryer length. A grain exchanger is located at the mid-point in the heating section; it splits the grain column to allow grain inside of the column to be moved to the outside of the column, and vice versa. The air flow in the cooling section is reversed compared to that in the heating section. Air from the cooling section is mixed with the ambient air before introduction to the burners. The rated capacity is 5000 bushels of wet corn per hour at 5 point moisture removal. 71 b . , . HEATING SECTION 4 _ , :////* -i *- /~l/j/’ 1 p .1 D .1 _._ —>* ~3— 10" GRAIN COLUMN -4 i- -l b -i H n cl r J i- , , -q 3-- 12" GRAIN COLUMN Ill I- 4 r. 27 5; a .-. BURNER ‘ : / -l i- / 4 >- 4 I- 4 i- .-l r 4 I- ‘ r « b TEMPERING 4 .7 ' 4 8 S 4 COOLING SECTION d in .1 ,~ '- ., DRYING FAN 4 L H I -i—« ~——-- « 2415—1 r M ’ 4 . . '5‘ <::::> COOLING FAN 11.2.54 _ 4 h // "L/// ‘ 7 ‘ ' 15' . T i- 4 i— + ~ < +- u H L====lg y__ : 18.5. J‘ Figure 5.1 Schematic Of The Meyer-Morton 850 Dryer. 72 Table 5.1 Dryer specifications for the Meyer-Morton 850 crossflow dryer. Airflow heat section, cfm/bu 122 Airflow cooling section, cfm/bu 142 Airflow heat section, cfm/ft2 102 Airflow cooling section, cfm/ft2 126 Static pressure heat section, in. of WC 3.0 Static pressure cooling section, in. of WC 3.0 Column cross sectional area, ft2 33 Column widths, in. 10 & 12 Grainflow, ft/hr at 5 point moisture removal 65.5 Recommended drying temperature, deg. F 230 Rated capacity at 20% - 15% MC, bu/hr 1400 Retention time at rated capacity, hr 0.63 Burner capacity, million of Btu/hr 8.7 Fuel type LP +————-23.25'—1 8.6' ' 2¥5' ///////}1\\\\\\\\4 1‘ Fl “ H -—aH+— 12" GRAIN COLUMN 1. 4 1 fl )-—- b i H w i‘ GRAIN , EXCHANGER ,r‘\/ 66.8' 103' A HEATING SECTION b- ).1 .3 ,, BURNER / >.. /Z /i—l )4 u BLOWERS (3) t. 44%»4 COOLING SECTION 18.5' i....=_—===.I _ l H L‘ 20 l [1 \ a 4.5' I t Figure 5.2 Schematic of the Zimmerman ATP 5000 Dryer. 74 Table 5.2 Dryer specification for the Zimmerman ATP 5000 crossflow dryer. Airflow heat section, cfm/bu 69 Airflow cooling section, cfm/bu 132 Airflow heat section, cfm/ft2 61 Airflow cooling section, cfm/ft2 111 Static pressure heat section, in. of WC 1.5 Static pressure cooling section, in. of WC 1.5 Column cross sectional area, ft2 70.1 Column width, in. 12 Grainflow, ft/hr at 5 point moisture removal 85.3 Recommended drying temperature, degree F 180 Rated capacity at 20%-15% MC, bu/hr ' 5000 Retention time at rated capacity, hr 1 Burner capacity, million of Btu/hr 54.2 Fuel type Natural Gas 75 'WW Figure 5.3 is a schematic of the control system for a continuous crossflow dryer. The system consists of: (1) a microcomputer, (2) a tachometer, (3) an automatic moisture meter, (4) A/D and D/A converters, (5) SCR (Silicon Controlled Rectifier) for. unload auger motor control, (6) system software (i.e. Basic), and (7) applications software (i.e. data collection and control algorithm). An Apple IIe microcomputer with 128 K RAM and floating point Basic language in read-only memory (ROM) constitutes the heart of the system. Operating data is displayed on a 12 inch screen for operator checking. A hard copy of the collected data is provided on an Epson dot-matrix printer. An AD/DA interface card is used in conjunction with the Apple. It enables the cOntrol and collecting of data from instruments that accept voltage as input or send voltage as output. The card contains a 12 bit analog to digital (A/D) converter and digital to analog (D/A) converter with an overall accuracy of 0.1%. The A/D and D/A converters can send or accept a voltage up to 4 volts. The specifications for the data acquisition system components are listed in Appendix C. An incremental optical encoder measures the unload auger rpm. The encoder outputs 500 cycles per revolution, and is powered by 5 volts supplied by the Apple microcomputer. The rpm is calculated by counting the number of cycles within a specified period of time, and then dividing the total number of cycles by 500 and by the specified period in minutes. A semi-continuous moisture meter, developed by Shivvers, Inc(P.0. Box 467, Corydon, Iowa 50060 ) automatically measures the inlet and Prmter I Monitor A 76 Apde He Micro- Computeq m ‘ Gram In +0 5 4. Y 5 Inlet Sample I U . m L 3 ”J. v g ,/// \\\‘\ Mmsture Meter <:>///~\\\\W 2 a E <3 1) RPM 9 ryer 33 +» *5 a, D Y Y E o L w -<-————— -<————————— SCR Auger Motor Voltage Figure 5.3 Schematic of the Control System for a Crossflow Dryer. 77 outlet grain moisture contents every 5-10 minutes. A microprocessor built into the moisture meter collects the moisture content data, and periodically transfers the information to the Apple. The moisture meter is calibrated with the use of a standard moisture meter. The calibration adjustment value is stored in the moisture meter microprocessor memory for adjustment of each moisture content,measurement. Figure 5.4 shows a comparison between MC values obtained with the Shivvers's moisture meter (COMP-U-DRY), a Motomco moisture meter, and an air-oven during a control test. Note that the Shivvers's moisture meter was calibrated using Motomco as the standard meter. The results show good agreement between the Motomco and Shivvers meters. The oven values slightly differ from the other two. The average outlet moisture content was 15.06% with a SD (standard deviation) of .67, 14.31 with a SD of .36, and 16.07 with a SD of .94, as measured by the Shivvers, Motomco, and oven methods, respectively. The variation in corn inlet moisture content as measured by the two moisture meters and air-oven is shown in Figure 5.5. Again the Motomco . and COMP-U-DRY meters show good agreement but are 2-3 points lower than the value obtained with the oven method. The average corn inlet moisture content as measured by the three meters, COMP-UrDRY, Motomco, and oven, are, 21.08% with a SD of .37, 21.53 with a SD of .36, and 23.88 with a SD -.2, respectively. The error in measuring the inlet moisture content is not as serious as the error in measuring the outlet moisture content because of the ability of the controller to account for the error through the dryer model parameter(s). The fact that the Shivvers moisture meter can be calibrated with an off-line moisture meter makes it attractive as an on-line moisture meter. Also, the sampling technique used in the moisture meter 78 251 244 13.1 21 J 204 194 18d 17-J 16-l 15-1 144 Ia-l 12-l o—c COMP—U-DRY 114 H OVEN H HOTOHCO INLET MOISTURE CONTENT (x we) O 1oTF§F§EFEFfiW£Efi£§6£EA SAMPLE NUMBER Figure 5.4 Corn Outlet Moisture Content Comparison (Oven, COMP-U-DRY, Motomco). 79 Ifl m 1H m 1% H HOTOMOO 11- H coup—u—om l-l WEN INLET MOISTURE CONTENT (% W8) .A ‘3 oriififiiififiafifiéfifififin SAMPLE NUMBER Figure 5.5 Corn Inlet Moisture Content Comparison (Oven, COMP-U-DRY, Motomco). 80 enhances the calibration process, since the same sample will be measured by both meters. The following parameters are measured and are used in the crossflow dryer control system: (1) the grain inlet moisture content; (2) the grain outlet moisture content; and (3) the unloading-auger rpm. The inlet and outlet moisture contents and the rpm are transferred through a shielded cable to the Apple microcomputer. The controller model determines the required residence time for the grain using one of the following two equations: M T = (In PS >/ file (5.1) W set or, Mps T-(W -32)/I33 (5.2) set Mp5 is given by equation (4.28), filo by equation (4.29), 62 by equation (4.26), and 63 by equation (4.27). The residence time of the grain in the dryer is achieved by sending .a'voltage, corresponding to the specific residence time, to the unload auger. The non-linear relationship between the auger rpm and the residence time varies with dryer design. The relationships between the residence time and the auger rpm, the auger rpm and SCR voltage for the Meyer-Morton 850 dryer were found experimentally and are given by equations 5.3-5.5, respectively: Residence time - 16.8 ~2.3 log(RPM) (5.3) RPM - 1524.5 exp(- .4343 Residence time) (5.4) Voltage - .565 + .002543 (RPM) (5.5) 81 The relationships between the residence time and the auger rpm, the auger rpm and SCR voltage for the Zimmerman ATP 5000 dryer are: Residence time - 1063*(RPM)*t(-.917) (5.6) RPM = 2085*(Residence time)**(-l.0954) (5.7) Voltage - .0058*(RPM)**(.8818) (5.8) The voltage to be send to the SCR is converted to its digital equivalent by equation (5.9), and then input to the D/A converter which sends it to the SCR in the dryer control panel. The SCR then adjusts the auger rpm accordingly: V - Volt * (2047) / 4 (5.9) where, Volt = analog voltage V - digital equivalent of Volt. 5.3 Procedure The following procedure was followed in conducting the controller tests performed on each of the two crossflow dryers: 1. the dryer is manually started with a constant rpm for a period of time equal to the residence time equivalent to the initial rpm. During this period moisture content and rpm data is continuously transferred to the Apple computer to be used by the control system during the subsequent automatic control; '2. after the start-up period has ended, the control system is switched to automatic; 3. at the end of each test the data is analyzed. 82 CHAPTER 6 6. RESULTS AND DISCUSSION The simulation and experimental results of this study on "Automatic Control of Crossflow Grain Dryers" are presented and analyzed in this chapter. First, the unsteady state differential equation model for crossflow grain drying is validated. This treatise is followed by the verification of the two empirical process models. Subsequently, the experimental data obtained from two commercial crossflow dryers, each equipped with the new automatic controller, are presented. Finally, several forms of a performance index for evaluating the dryer control system are analyzed. The chapter closes with a section highlighting the main results of this study. 6.1 Simulation The unsteady state differential-equation crossflow dryer model and the crossflow dryer control algorithm have been combined to. form the simulation model for the control system of a crossflow dryer. The grain outlet moisture content and the unload auger rpm as functions of time are the outputs of the automatic crossflow dryer simulation model. The computer program is implemented on a VAX/VMS minicOmputer system. The computer program uses excessive CPU time. Table 6.1 shows the CPU time used for different amounts of moisture removed and set points. The CPU time increases with an increase in the amount of moisture to be removed. The CPU time is longer than the drying time for all the inlet moisture content ranges used in the control system simulation model. 83 Table 6.1 CPU time for different amounts of moisture removed and set points. ’IAmount of Moisture Set Point CPU Time Simulated Drying Time Removed (%, w.b.) (%, w.b.) hrs hrs 2.54 14.5 11.67 8.00 4.70 14.5 17.40 8.00 7.20 14.0 21.40 7.73 8.00 15.5 23.13 11.18 Note: type of dryer is a Meyer-Morton (see Section 5.1); for Ax, Ay, At, etc. see Table 4.1. 6.1,1 Unsteady-State Model Verification To check the accuracy of the differential equation simulation model, two data sets obtained from tests #1 and #7 in the Meyer-Morton. 850 dryer were used as input into the simulation model. The simulation results of the two tests are shown in Tables B.1 and B.2 in Appendix B. 'Test #1 represents a moderate variation in corn inlet moisture content while test #7 represents a large variation. The comparisons between simulation and the experimental results are shown in Figures 6.1 through 6.4. (Figure 6.1 shows a plot of experimental versus simulated outlet moisture cOntents for test #1; the simulated results agree well with the experimental values. Theoretically, if the simulated values perfectly match the experimental values, a 450 angle is formed by the OUTLET M.C. (7.), SIMULATION 84 20.0 . 1905 18.0-j TEST #1 17.05 16.0—3 15.0% "‘ u 14.0-f "‘ ,, 13.0% 111 12.0-3 .1 110-] 10.0.IrEIVIIIIUIIU'IIIIU'IIIIUrI III III III 1C.O11.012.013.014.015.016.017.018.019.020.0 OUTLET M.C. (75), EXPERIMENTAL Figure 6.1 Simulation vs Experimental for Test #1 with the Meyer- Morton Dryer; differential-equation model used. 85 line connecting the data points and the x-axis. Furthermore, since the dryer is automatically controlled, the data points should converge to one point, the. set point. The simulated results of test #1 have the above characteristics which proves the ability of the differential- equation simulation model to simulate the dryer control system close to its actual performance. This is also shown in Table 6.2 in which the means of the simulated and experimental outlet moisture contents are tested statistically for equality. The test results show that the two means are equal and that the deviations are the result of random error. A plot of the differences between experimental and simulated outlet moisture content values versus time is shown in Figure 6.2. The differences are randomly scattered between i 2.5%; most of the data points show a difference of less than 1% point compared to the theoretical value. The simulation model predicts an average outlet moisture content of 14.49% with a standard deviation.of .38%; experimentally, values of 14.44% and of 0.63% were obtained. Figure 6.3 shows the simulated versus the experimental outlet moisture contents for experiment #7. The simulated points were calculated using the differential-equation crossflow drying model. lfiua data points1are not as close to the theoretical line as for experiment #1. One reason is the larger variation in the inlet moisture content in experiment #7 than in experiment #1. In an actual dryer, some mixing takes plaCe which reduces the variability between adjacent layers of corn at the dryer exit. In the case of the simulation model, the layers are accurately tracked until they exit from the dryer (iuea. no mixing is assumed to take place in the dryer control simulation). Figure 6.4 shows a plot of the differences between the experimental and simulated outlet moisture contents as a function of time for test 86 Table 6.2 Testing the hypothesis with the Student t-test that the mean of the experimental and simulated grain outlet moisture contents for test #1 are equal. 1. Ho : p1 = p2 or pl-pz-O 2. H1 : pl # #2 or pl-pzflo 3. a - .l, .2, .4 4. Critical regions: a) T < -1.658 and T > 1.658 b) T < -1.289 and T > 1.289 c) T < -.845 and T > .845, where (x1 - x2) -do SF/(l/nl +1/n2) T- with u - n1 + n2 - 2 - 81 + 81 - 2 - 160 5. Computations ' . x1 - 14.44%, s 1 - .63%, n1 - 81 and - .38%, n2 = 81 x2 =- 14.49%, s 2 hence sp - J(.632(80)+.382(80))/(81+81-2) - .271 t - ((14.44—14.49)-0)/(.271/(1/81+l/81) = -.612 6. Conclusion : Accept Ho and conclude that the means of the two sets are equal. 87 K v ’7 m. g. N V 0: 2 O A _l D 2 In I. 0. E5. I l T l I l r” l T l I 01 2 3 4 5 6 7 8 9101112131415 TIME (HOURS) Figure 6.2 The Difference between Experimental and Simulated Grain Outlet Moisture Content vs Time for Test #1. OUTLET M.c.(%). SIMULATION 20.0 .. 88 19.0% 18.0-E 17.0% 16.0% 15.0% 14.0.3 13.0% 12.015 11.0-3- d 10.0 TESTin '3' I VTI'II I II? Fijrl’ rI’T TEE—1'77 10.0110120130140150160170180190200 OUTLET M.C.(z), EXPERIMENTAL Figure 6.3 Simulation vs Experimental for Test #7 Model was Used for the Simulation. ; the Differential— 89 .5 he JILLLl lflllllJlllluLlllL (d o-b (EXP.-SIMUL.) OUTLErM.c (%.w.3.) -5lUUUIIIUIFrrrIIIrIIr'IlrrrUrlllrllIIlIIU— 0 1 2 3 4 5 6 7 8 9 10 TIME (HOURS) Figure 6.4 The Difference between Experimental and Simulated Grain Outlet Moisture Content vs Time for Test #7. 90 #7. The differences are not randomly scattered but are negative for about two hours time, then become positive for four hours, and finally become negative again for three hours. The trend is similar to the inlet moisture content variation in test #7. This supports the argument that the inlet moisture content affects the accuracy of the simulation. Thus, a comparison of individual data points is not a good measure of how well the simulated and experimental results compare to each other when large variations in the inlet moisture content occur. A more objective measure is to compare the average outlet moisture content and the standard deviation. Assuming that the simulated and experimental outlet moisture contents are normally distributed with equal variances, the hypothesis to be tested is that the two sets of data have the same mean outlet moisture content. The Student's t test along with the average and SD of the outlet moisture content of the two data sets are used in testing the above hypothesis. Table 6.3 shows the calculations and the results of the test. There is no significant difference in the mean of the two tests at .l, .2, and .4 level of significance. Thus, the differential-equation simulation model accurately predicts the average outlet moisture content obtained experimentally, and any deviation within the data is the result of randomness. It can thus be concluded that the simulation model for crossflow dryers is acceptable for analyzing the effect of a load variable such as the inlet moisture content variation on the performance of the dryer control system. 6.1.2 Empirical Model Verifications Next,the unsteady state differential—equations model for crossflow drying developed in Section 4.1 was used to test the adequacy of the empirical models in describing the drying process of crossflow dryers. The outlet moisture content and residence time for a given inlet moisture content generated by the unsteady state model were used in 91 estimating the model parameters in the empirical equations discussed in Chapter 4 (eqns 4.16 and 4.17). Tables 6.4 and 6.5 show the values of the estimated parameters alongwith the estimates of the grain outlet moisture contents. 6,1.2,1 Exponential Figure 6.5 shows a plot of outlet moisture content predicted by the exponential mOdel and the unsteady-state differential-equation model. The values of the exponential model parameter vary with the inlet and outlet moisture contents, and residence time. The values of 61 (Table 6.4) decrease steadily with the increase in the inlet moisture content which suggest that 61 has some correlation with the inlet moisture content. An average Value - 0.42 is used for ,6]- when the exponential model is used to predict the outlet moisture content. The outlet moisture contents predicted by the exponential equation agree well with the unsteady-state model. A plot of the exponential model outlet moisture content vs the unsteady-state model outlet moisture content (Figure 6.6) proves the good agreement of the two models. This should be expected since drying can be considered as a chemical reaction process, and thus can be described by an exponential relationship (Berglund, 1987). 6.1,2,2 Lipear Figure 6.7 shows the outlet moisture contents predicted by the linear model and unsteady-state differential-equation. The parameters used in the, calculation are the average values for the parameters estimates excluding the first eight (values (see Table 6.5). The first eight values of the linear model parameters vary widely due to the nature of the sequential least square estimation method at the early stages of the estimation, and thus are excluded from the calculation. 92 Table 6.3 Testing the hypothesis with the Student t-test that the mean of the experimental and simulated grain outlet moisture contents for test #7 are equal. 1. Ho : ”1 - M2 or pl-pZ-O 2. H1 : p1 # #2 or pl-p2#0 3. a - .1, .2, .4 4. Critical regions: a) T < -l.658 and T > 1.658 b) T < -1.289 and T > 1.289 c) T < -.845 and T > .845, where (x1 - x2) -d T _ o sp/(i/ni +1/n2) with v - n1 + n2 - 2 - 55 + 55 - 2 - 110 5. Computations : xl — 14.03%, 51 - 1.2%, n1 - 56 and - 14.13%, s - .98%, n2 - 56 x2 2 hence sp = /(1.22(55)+.982(55))/(56+56-2) - 1.1 t = ((14.03-14.13)-0)/(l.1/(1/56+1/56) = -.48 6. Conclusion : Accept Ho and conclude that the means of the two sets are equal. 93 Table 6.4 Parameter estimates for the exponential model (eqn. 4.16 using data simulated by the unsteady state model (see Table 4.2) in drying shelled corn. Inlet Outlet Residence Bl Est. Outlet M.C. M.C. Time M.C. (%,w.b.). (%,w.b.) (hrs.) (1/hrs) (%,w.b.) 20.0 11.3 1.13 0.50 12.5 20.0 11.7 1.06 0.50 12.8 20.0 12.4 0.95 0.50 13.4 20.0 13.3 0.83 0.49 14.1 20.0 14.0 0.72 0.49 14.8 20.0 14.4 0.68 0.49 15.0 22.0 12.0 1.28 0.47 12.9 22.0 12.4 1.22 0.47 13.2 23.0 10.8 1.64 0.46 11.6 23.0 13.8 1.12 0.45 14.4 23.0 14.0 1.09 0.45 14.6 23.0 14.5 1.02 0.45 15.0 23.0 15.1 0.95 0.45 15.4 23.0 15.4 0.89 0.45 15.8 23.0 15.9 0.84 0.44 16.2 24.0 18.2 0.67 0.41 18.1 24.0 18.0 0.68 0.42 18.0 24.0 17.5 0.75 0.42 17.5 24.0 16.9 0.82 0.43 17.0 24.0 16.5 0.88 0.43 16.6 25.0 12.5 1.60 0.43 12.8 25.0 14.2 1.33 0.43 14.3 25.0 18.6 0.73 0.40 18.4 25.0 18.5 0.74 0.41 18.3 25.0 18.4 0.76 0.40 18.2 25.0 18.1 0.79 0.41 17.9 26.0 13.6 1.55 0.42 13.6 26.0 14.7 1.38 0.41 14.6 27.0 18.9 0.93 0.38 18.3 27.0 18.5 0.98 0.39 17.9 27.0 18.1 1.04 0.39 17.5 27.0 17.7 1.08 0.39 17.2 27.0 17.4 1.13 0.39 16.8 28.0 15.7 1.49 0.39 15.0 30.0 16.5 1.64 0.36 15.1 0 Average .42 94 Table 6.5 Parameter estimates for the linear model (eqn. 4.17) using data simulated by the unsteady state model (see Table 4.2). Inlet Outlet Residence 52 63 Est. Outlet M.C. M.C. Time M.C. (%,w.b.) (%,w.b.) (hrs) (l/hrs) (%,w.b.) 20.0 11 3 1.13 0.25 0.28 12.6 20.0 11 7 1.06 0.89 -0.29 13.0 20.0 12 4 0.95 0.92 -0.31 13.6 20.0 13.3 0 83 0.96 -0.35 14.2 20.0 14.0 0 72 0.96 -0.35 14.8 20.0 14.4 0 68 0.98 -0.39 15.1 22.0 12 0 l 28 0.91 -0.28 12.9 22.0 12.4 1 22 0.87 -0.25 13.3 23.0 10.8 1 64 0.83 -0.22 11.3 23.0 13.8 1.12 0.88 -0.25 14.5 23.0 14 0 1.09 0.89 -0.25 14.7 23.0 14 5 1.02 0.89 -0.26 15.2 23.0 15 1 0.95 0.91 -0.26 15.6 23.0 15 4 0 89 0.91 -0.27 16.0 23.0 15 9 0 84 0.92 -0.27 16.3 24.0 18 2 0.67 0.95 -0.29 18.1 24.0 18 0 0 68 0.95 -0.29 18.1 24.0 17 5 0 75 0.94 -0.29 17.6 24.0 16.9 0 82 0.94 -0.28 17.2 24.0 16.5 0 88 0.94 -0.28 16.8 25.0 12.5 1.60 0.94 -0.27 12.5 25.0 14.2 1 33 0.92 -0.26 14.4 25.0 18 6 O 73 0.95 -0.28 18.5 25.0 18 5 0 74 0.95 -0.28 18.4 25.0 18 4 0.76 0.95 -0.28 18.3 25.0 18 l 0 79 0.94 -0.28 18.1 26.0 13.6 1 55 0.96 -0.28 13.4 26.0 14.7 1 38 0.94 -0.27 14.6 27.0 18 9 0 93 0.97 —0.29 18.5 27.0 18.5 0 98 0.96 -0.29 18.1 27.0 18.1 1.04 0.97 -0.29 17.7 27.0 17 7 1.08 0.97 -0.29 17.4 27.0 17 4 1.13 0.97 -0.29 17.0 28.0 15 7 1.49 1.01 -0.31 14.9 30.0 16 5 1 64 1.00 -0.27 14.7 * Average 0.94 -0.28 * Note: average does not include the first 8 values. 95 ”T 30— . x U3. :. lNLET M.C. ymgggggxaé 33 25% ' xxxxxx 5: .. Mxxxxxxxmflem E '20-] xxxxxx QUE-fr “402° E ~ : . 3.5% M 8283 o g 15: 1336 _ 66666 Q ,8 21 .J A ¢ 0 1 80° 883 104 LL] . .1 G: i 1:3 51 U) i A EXPONENTIAL 5 1 w o UNSTEADY STATE 2 O FiitlrrrrlrrrrririirriTIITrrrrrrrrrrTrrl 0 5 10 15 20 25 30 35 40 SAMPLE NUMBER Figure 6.5 The Exponential Model vs the Unsteady-State Differential- Equation Model for the Drying of Corn in the 850 Meyer- Morton Dryer. 96 20.0 , 19.0-3 - 18.05 . ... 17.0-j . "‘ : ii 16.0-j as ' 15.0- 3111.: 311 as 14.04; 3" 13.0-j EXPONENTIAL MODEL it): 12.0.; 11.04: 1000d'rtliifirl 16" [WI III III III III—fr! 10.011 .012.013..014015.016017.018..019020.0 UNSTEADY STATE MODEL Figure 6. 6 Exponential MOdel Outlet MOisture Content vs Unsteady- state ' Model Outlet Moisture Content in Drying Shelled Corn in the Meyer- -Morton 850 Dryer. 97 23 . g: f iNLET M.C. ”mm?“ ~ 25: xxxxxx Fe : xxxxxxx***** 1‘2— 20— xxxxxx OUTLET M.C; LL] -: ‘denbd‘ lk‘nk‘k fiqkfifig ‘0 F- 15‘ 66° 3 E; -: Aaéee 8 836 a $8 A C) :30° 3 6 a 031 : i— 5—: 9 .. A LINEAR _ O 3 o UNSTEADY STATE 2 O rrrf[FIII[IirrrrrrrrTIirltrrrrrrrrfrrril 0 5 10 15 20 25 30 35 4O SAMPLE NUMBER - Figure 6.7 The Linear Model vs the Unsteady-State Differential-Equation Model for the Drying of Corn in the 850 Meyer-Morton Dryer. 98 20.0 1903 1 - e .. 18.01 1 17.0j at 16.0-E . at 15.0-l: 119‘ 14.0-f LINEAR MODEL 13.04: an: 12.0-f cl .1 11.0: .1 1000—III]fril'lr'IUFIYrrlUrr'[IUIrUrIrI—T'IUUT 10.0 11.0 12013.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0 UNSTEADY STATE MODEL Figure 6.8 The Linear Model Outlet Moisture Content vs the Unsteady- State Model Outlet Moisture Content in Drying Shelled Corn in the Meyer-Morton 850 Dryer. 99 The outlet, moisture contents predicted by the linear equation agree well with the values obtained by the unsteady state model. A plot of the linear model outlet moisture content versus the unsteady state model outlet moisture (Figure 6.8) proves the good agreement between the two models. In conclusion, the two empirical models evaluated in the tww (sections are simple in their formation, and thus efficient for-on-line calculations. They predict the grain outletImoisture content in crossflow dryers well. As process models for crossflow dryers control system they appear to have great promise. 6,1,3 Controller Stability Tests Table 6.6 shows eight inlet moisture content ranges used in.the theoretical analysis of the automatic control system of crossflow dryers. The exact nature of the inlet moisture variations along with the outlet MC and the rpm values are shown in Tables B.1-3.8. Sets #1 and #2 are actual inlet moisture contents encOuntered in the Meyer-Morton dryer in tests #7 and #1, respectively (see Section 6.2.1). The results are shown in Figures 6.9 and 6.10. The two sets are compared in Section 6.1.1 to their experimental counterparts. The controller predicts the experimental values well and was found to be stable. Innus, it can be used for the analysis of other inlet moisture content sets. Figure 6.11 shows the results obtained using inlet moisture content variation from set #3 (same inlet moisture content as thatin test #12, Section 6.1.2) . The inlet moisture content ranges between 19.9% and 23.5%.Tflmeaverage grain outlet moisture content after 10 hours of simulation.is 17.1% for a set point of 17.5%. Overdrying by .4% agrees with the .5% overdrying which occurred in test #12 (see section 6.2.2) . 100 Table 6.6 Inlet moisture content sets used as inputs in the simulation of crossflow grain dryers. Set Number Av. M.C. Min Max SD Period (%) (‘3) (’3) (’35) (hrs) 1 26 8 23.8 31 0 l 8 - 2 22 2 20.8 23 6 0 7 - 3 21 4 19.9 23 5 0 8 - 4 21 0 19.7 22 3 0 9 4 5 21 0 19.7 22 3 0 9 2 6 25 0 22.6 27 3 1 7 4 7 25 0 22.6 27 3 1 7 2 101 During the 10 hours of simulation, the dryer is operated at maximum auger speed 3/5 of the time, causing the grain outlet moisture content to remain below the set point. Figure 6.12 shows the results obtained from the simulation model rwith set #4 (Table B.4) as the inlet moisture content input. The inlet moisture content varies sinusoidally with an average of 21%, a maximum of 1.5% above the average, and a period of 4 hours. During the 8 hours of simulation, the average outlet moisture content is 14.4% (with a set point of 14.5%). The outlet moisture content is very close to the set point at all times. Figure 6.13 shows the simulation results obtained using a .sinusoidal variation in the inlet moisture content with anTaverage and maximum equal to that of set #4 and a period of 2 hours. The average outlet moisture content is 14.5% with a damped sinusoidal Shape with a period of 2 hours. Although the average outlet moisture1xnment is equal to the set point, the variation in the outlet moisture content of the individUal samples is larger than that of set #4. Thus, the period affects the damping in the outlet moisture content of a controller subjected to a variation in the inlet moisture content. The simulation results of a sinusoidal inlet moisture content variation with an average of 25%, a maximum of 3% above the average, and a period of 4 hours (set #6) are shown in Figure 6.14. The average outlet moisture content is 14.6% for a set point of 14.5%. The outlet moisture content has a sinusoidal shape with a period of 4 hours and damping ratio of .5. The rpm variation is also sinusoidal with a period of 4 hours. Figure 6.115 illustrates the simulation results obtained using the inlet moisture content variation given by set #7. Set #7 has a similar inlet moisture content variation as set #6 except for period which is 2 102 SET POINT = 14.0 7. AVERAGE OUTLET MC. = 14.1 % ,7 30.0 INLET M.C. 01! 3 ' .4 K: 25°C 3 '2 ~1ZOO . '- 20.o~‘ 5 "'1 ‘00 UZJ I OUTLET M.C. o. _1000 I— ‘ E z 15 o- m L900 :0 O ‘u o —800 1?. m 10.0 _ . DD: 700 {7) 5'0 R P M _600 _ . . o L— o 500 2 0.0 '1 I I I I rfl I I rrrt Frrrrrrrrr FrrrFrrrrr I rTrI O 1 2 3 4 5 6 7 8 9 10 TIME (HOURS) Figure 6.9 Simulation of the Automatic Control of the Meyer-Morton 850 Dryer(set #1). MOISTURE CONTENT %(W.B.) 103 SET POINT = 14.5 73 AVERAGE OUTLET M.C. =14.5% 25.0 INLET M.C. ‘- j W ‘ l ”T __ .. . . 5'5 --~-.. E'gtr-H'szf’ , . , <3}, ,5 E 20.0: 5 —-1200 J O. : OUTLET M.C. ta ”50 15.0- _ . , - .. . 9.9.. ,. Ur -1 100 #1050 go 10.0 91000 T0 a . —950 5.0-; . .. . 9900 : R.P.M. L850 000 Trlttrl’rrrIII—Tl'trrrillirijliirrrrrrllirtrlrt 12 3 4 5 6 7 8 91011121314155 TIME (HOURS) Figure 6.10 Simulation of the Automatic Control of the Meyer-Morton 850 Dryer(set # 2). MOISTURE CONTENT %(W.B.) 104 SET POINT = 17.5 7:. AVERAGE OUTLET M.C. =17.1% 25.0.. INLET M.C. A -2000 1900 ~1800 L-17OO ~1600 P1500 ~14OO ~1300 SET POINT 'W'd'EI OOO I—rr‘I rT rrr r: If Ir II rI rt Writ rt— r1 IITTWTTTTITTTT 012.3 4 5 6 7 8 9101112131415 TIME (HOURS) Figure 6.11 Simulation of the Automatic Control of the Zimmerman ATP 5000 Dryer(set #3). MOISTURE CONTENT %(w.B.) 105 SET POINT = 14.5 % AVERAGE OUTLET MC. = 14.4% 250-, 9W E 200-; ~ Es —1250 :3 OUTLET M.C. I: —1200 15.0-3 . , m L4150 : ' 91100 1090-3 9-1050 3 RPM. “1000 5.0-j L950 3 L000 0.01...,...,--,,...,,,.,,..V-”merm O 1 2 3 4 5 6 7 8 9 10 TIME (HOURS) Figure 6.12 Simulation of the Automatic Control of the Meyer-Morton 850 Dryer(set #4). 'WTcTH 106 SET POINT = 14.5 73 AVERAGE OUTLET MC. = 14.5% TIME (HOURS) ax 25.0- m : INLET M.C. - . I— 3 - z :9 20.0-WW . .12... ‘ O. I-Z: ; OUTLET M.C. ta “1200- Lu 15.0- ...... ~ m 91150. I— ~ , z : -1100. o . o 10.0-_ 91050. LLI " I— m 1 1000. E 5041 R.P.M ~950. __ ‘ L o i 900. 2 000 IrIlI—I I I I I rrrrrlt II I I I I [TI I I FFTII I] 0 1 2 3 4 5 6 7 8 9 10 Figure 6.13 Simulation of the Automatic Control of the Meyer-Morton 850 Dryer(set #5). 'W'd'EI 107 SET POINT = 14.5 % AVERAGE OUTLET MC. = 14.6% ’“. 30.0 CD ; INLET M.C. V 25.0 N ‘ : ... -1400 . z E 20.0-3 E’. T1300 E 3 . OUTLET M.C. L—J _1200 .. . 00 L- o :W . w 91000 1%] 10.0-‘3 1.900 ,_ I 5.0.; R.P.M. L800 9 . 9700 O I 2 0.0 I—rl Fri I FrfrrrrrFFrIrt I rrfirrrrrft'rf[ I I I 0 1 2 3 4 5 6 7 8 9 10 TIME (HOURS) Figure 6.14 Simulation of the Automatic Control of the Meyer-Morton 850 Dryer(set #6). W'd'EI 108 SET POINT = 14.5 % AVERAGE OUTLET MC. = 14.6% A 30.0 INLET M.C. 25x) 2 201) 83 OUTLET M.C. t] 15.0 m -1400 L-1 300 P1200 -1 100 WJ WW .0 o 9" o HLLILI L1 LLLLI LLLLllLlll ' MOISTURE CONTENT %(W.B. .0 o LIOOO L900 L800 L700 ITTIITI'TrfrrrrirriilrrrT—TrrirrfrrrrrFF O 1 2 3 4 5 6 7 8 9 1 TIME (HOURS) 0 Figure 6.15 Simulation of the Automatic Control of the Meyer-Morton 850 Dryer(set #7). W'd'éd 109 hours. Although the average outlet moisture content is 14.6%, only .1% above the set point, the outlet moisture content is sinusoidal with a damping ratio of approximately ,one. A comparison of the simulation results of sets #7 and #6 shows that reducing the period of fine sinusoidal inlet moisture content by 50% can increase the damping ratio by 50%. Figure 6.16 shows simulation results for several step changes in the grain inlet moisture content. Step changes in the grain inlet moisture content are likely to occur in actual drying operations. The initial inlet moisture content is 28% for a three hours period. The average outlet moisture content during this period remains at the set point. After 3 hours of drying the inlet moisture content is suddenly decreased to 24%, and remains constant for 3 hours. The feedforward controllem'reacts to the change in the inlet moisture immediately when the.inlet moisture content change occurs. The reaction of the controller results in the outlet grain being partially underdried and partially overdried. After six hours of simulation, the grain inlet moisture content is increased by 8% and remains constant for approximately 3 hours. The controller reacts to the large change in the inlet moisture content by decreasing the auger speed, resulting in momentary overdrying and underdrying. Figure 6,16 proves the stability ofi the controller to control a crossflow grain dryer subjeCted to large variations in the inlet moisture content 6,2 Experimental Results The experimental results consist of results obtained by performing drying tests on two crossflow dryers fitted with the new control system. The two crOssflow dryers are, a Meyer-Morton model 850 and a Zimmerman model ATP 5000. The detailed descriptions of the two commercial crossflow dryers are given in Chapter 5. 110 SET POINT = 15.5 % AVERAGE OUTLET MC. = 15.6% ax 55.0 INLET M.C. Ill ,---J-w-- 5 50.0 1600 N ‘ C I— 1500 g 25'0 .. g L1400 E 20.0 OUTLET M.C. IL: L1500 Z 9. O 15 O - .- - ”A--.“ T o , . 33 1090 —800 D R.P.M. L700 E): 5.0 ‘ ‘ ‘ L L500 0 500 2 000 [Frrrl IIIIIIIIIIIIrI—TIII’FTI—rIr FI rrr rrr rI I 012 3 4 5 6 7 8 9101112131415 TIME(HOURS) Figure 6.16 Simulation of the Automatic Control of the Meyer-Morton 850 Dryer(set #8). 'W'd'érI 111 W1C . The tests were conducted with the Meyer-Morton dryer during the fall of 1985 and of 1986. Figures 6.17 through 6.28 and Tables A.l through A.10 show the experimental results obtained during the drying of corn. Figure 6.17 illustrates the controlability of manual control of the Meyer-Morton dryer. It shows the variation in the corn outlet moisture content, the corn inlet moisture content, and the unload auger rpm as a functions of time. The average outlet moisture content during the nine hours of drying was 13.6%(w.b)*at a set point of 14.5%. Overdrying by 10.9% point took place which is characteristic for manual dryer control. The variation in the corn inlet moisture content was typical for an on- farm dryer in Michigan in November. During the nine hours of drying, the rpm was changed three times which was insufficient to prevent the slight overdrying. A second example of manual control is shown in Figure 6.18. The outlet moisture content of the grain was .68% below the set point (14.5%). Only during the 1-3 hours drying period did a significant change occur in the inlet moisture content. Still the outlet moisture content varied considerably. Due to the limited information a dryer operator has during the drying process, it is difficult to control the dryer adequately even for the best operators. * all the experimentally determined moisture content values in this Section are expressed on a wet basis. MOISTURE CONTENT%(W.B.) 112 SET POINT =14.5 ’ AVERAGE OUTLET M.C. = 15.8 3090 INLET M.C. 25 0 1 W ' ‘1 ~1000 20.091 -900 3 O TLET M.C. L500 150— " , A .. A. :0 ~ WV x 9-700 1.0 10.0— 9500 Z: L500 5.0 9400 000lIrIrIII‘IIIIIIrIIIIIIIIIIIIIII'IIIFVIIII O 1 2 3 4 5 6 7 8 9 10 TIME (HOURS) Figure 6.17 Inlet and Outlet Moisture Contents vs Time During Manual Control of the Meyer-Morton 850 Dryer (1984). MOISTURE CONTENT%(W.B.) 113 SET POINT =14.5 AVERAGE OUTLET M.C.=13.6 30.0 25.0 INLET M.C. _1200 20.0 15.0 OUTLET M.C. 10.0 5.0 000'1'I[IIrrT—TIIIIT—[IrlrrrilrfTrIrrl'Il'rrIrI O 1 2 3 4 5 6 7 8 9 10 TIME (HOURS) Figure 6.18 Inlet and Outlet Moisture Contents vs Time During Manual Control of the Meyer-Morton 850 Dryer (1985). 114 Figure 6.19 shows test #1 with the automatic dryer control. The average inlet moisture content was 22.24% with an standard deviation of .67%, the average outlet moisture content 14.47%, and the desired outlet moisture content 14.5%. Although, the variation in the inlet moisture content was small, the controller auger rpm varied from 698 to 960 to keep the outlet moisture content as close as possible to the set point. Control of the average outlet moisture content to within .03% from the set point can be considered excellent. In Figure 6.20 test #2 is shown; the outlet moisture content at the beginning of the test was 2 1/2% above the set point. It slowly approached the set point as the test progressed. The high average out-let moisture content is due to the high values during dryer start up. This shows the importance of the start-up procedure. During the last six hours of the test the outlet (and inlet) moisture contents remained almost constant. However, over the total 7.5 hours of the test the average outlet moisture content was 14.94 , the set point 15.0% , and the average inlet moisture content 23.31%; the auger rpm was 717. The results of test #3 are shown in Figure 6.21. The average inlet moisture content during 6.5 hours of drying was 21.43. The inlet moisture content was almost constant during the test. The controller controlled the outlet moisture content very well to an average value of 14.53% and thereby deviated by only .03% point from the set point. The average auger rpm was 722 with an standard deviation of only 19, this is an example of the operation of the controller under conditions of only small variations in the inlet moisture content. In Figure 6.22 test #4 is shown; the inlet moisture content at the start of the test was 22.5%(w.b); it remained constant for two hours MOISTURE CONTENT %(W.B.) 25°03 - . INLET M.C. 115 SET POINT = 14.5 7. AVERAGE OUTLET M.C. =14.5% E 20.0~ 5 ~1200 3 O. 1 OUTLET M.C. [I] ~11OO 15.0-‘ All... -- - .. ..-- (I) ‘ ~1000 10.0 -900 J . —800 5.0: R.P.M. . ~7OO 000 ‘ rl I I i [ TV! I fr I I 3' r I rTI Fr I—FI rT‘Fr rFFI rrt r O 1 2 3 4 5 6 7 8 9 10 TIME (HOURS) Figure 6.19 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer-Morton 850 Dryer in Test #1. .70 TU 116 SET POINT = 14.5 75 AVERAGE OUTLET MC. = 15.0 % f) 25.0 a! INLET MC. 3 ‘2 iii 20.0 5 -900 D. I— . OUTLET M.C. Lin 15.04 ,___, A . _ . Fri L800 *- -I S i F0 0 10.0-E —7OO T0 m . a g I ,__ 5.0-j ~OOO 92 a o i 2 000 I III!rirrrTI’rrrrITITrrrl’TrIrIi{IIIIFT O I 2 :5 4 5 6 ' 7 8 9 IO TIME (HOURS) Figure 6.20 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer-Morton 850 Dryer in Test #2. MOISTURE CONTENT %(W.B.) 117 SET POINT = 14.5 7: AVERAGE OUTLET M.C. =14.5% 25.0 INLET M.C. E 20.0 '0‘ 900 O. OUTLET M.C. 15.0 -....- __ , [_f ...»- _ _ , OO .70 10.0 R.P.M. 700 2 5.0 600 000 FFI I ' I I I I l I I l I I l l 0 I 2 3 4 5 6 7 8 9 10 TIME (HOURS) and RPM vs Time During Figure 6.21 Inlet and Outlet Moisture Contents, on 850 Dryer in Test #3. Automatic Control of the Meyer-Mort 118 and then increased suddenly to 24%. It remained close to this value until the end of the test. Test #4 represents a step change in the inlet moisture content which is typical when a farmer moves from one parcel of land to another. The automatic controller reacted well to the sudden inlet moisture content change. The average outlet moisture content for 8 hours of dryer-operation was 14.94 at a set point of 15.0%. The auger rpm decreased steadily during the early drying time because the discharged grain was above the set point. The auger rpm decreased.fUIther as the inlet moisture content started.to increase. Towards the end of the test, the change in the auger rpm become small due to the relatively constant values of the inlet and outlet moistnnna contents. Figures 6.23 and 6.24 show the results of test #5 and test #6, respectively. In test #5 the auger rpm increased from 750 to 900 within three hours to reduce the overdrying at the earLy stages of the test. In contrastg in test #6 the auger rpm decreased from 850 to 700 within two hours due to underdrying during the early hours of the test. In both tests the grain inlet moisture content was fairly constant, and the average outlet moisture content was controlled to within .1% point from the set point. An accurate choice of the initial auger rpm would have resulted in even less underdrying or overdrying at the early stages of both tests. The controller controlled the drying process well in both tests and resulted in outlet moisture contents approximately equal to the set points. Figures 6.25 and 6.26 show results obtained with large variations in the corn inlet moisture content. The grain inlet moisture content varied in test #7 from 31% to 19.8% and in test #8 from 34.3% to 19.5%. In Figure 6.25, the variation in the inlet moisture content appeared to fluctuate randomly during the test. The fluctuation made it MOISTURE CONTENT %(W.B.) 119 SET POINT = 15.0 % AVERAGE OUTLET M.C. =14.9% 25.0-3 INLET M.C. F- 2: 290‘ 5 —900 I O. 2 OUTLET M.C. I33 ”850 15.0 “' ----.— “‘* -- _. "'7‘" _ . - m :00 I .750 ;O 10.0; -7OO .‘0 I z . RPM ~650 ° 5.0-j H600 3 ~550 4 0'0 FTTVIVFI rFI—I’rrrt‘trilrI—frlrtrrrt[Frlrtrt O I 2 3 4 5 6 7 8 9 TIME (HOURS) 10 Figure 6.22 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer-Morton 850 Dryer in Test #4. MOISTURE CONTENT %(W.B.) 120 SET POINT = 15.0 75 AVERAGE OUTLET M.C. =14.9% 25.0 . INLET M.C. *2 ‘ 20.0 6 ~1200 0. OUTLET M.C. t3 “00 15.0 ...-..V -... — . - . - m 1000 —900 10.0- —800 —7OO 5.0 —600 —500 000 rIIIIrrlfrrII—rrrrI—IlrrrIIIIlIrrrrrIlIrI O I 2 3 4 5 6 7 8 9 ‘IO TIME (HOURS) Figure 6.23 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer-Morton 850 Dryer in Test #5. 'W'cI'ézI ' MOISTURE CONTENT 121 SET POINT = 15.0 % AVERAGE OUTLET M.C. =15.1 % 'V‘I'cI'H 25.0 IDHJET AAJ3. 52 20.0 ' 5 ~1200 D. OULET M.C. Em: ““1100 15.0 ~- -..- _~ .--... _1000 —900 10.0 —800 5,0 ~7OO —600 0.0 riiirtIIIrII—rFFIrxxftlirffirixttlrrrrrrr O I 2 3 4 5. 6 7 8 9 10 TIME (HOURS) Figure 6.24 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer-Morton 850 Dryer in Test #6. 122 impossibleto control the instantaneous grain outlet moisture content to exactly the desired value. The auger rpm was increased momentarily ‘because of the overdrying and then decreased as the grain inlet umdsture content increased and the outlet moisture content drifted above time set point“ The average corn outlet moisture content (14.3%) was still acceptable to the set point. The controller performance can. be described as excellent based on the average outlet moisture achieved at the large variation in the inlet moisture Content encountered. The grain inlet moisture content for test #8 (see Figure 6.26) varied widely during the first two hours and remained fairly constant over the last 6 hours of the test. The auger rpm was changed frequently in.an.effbrt to control the grain outlet moisture content as close to the set point as possible. The resulting average grain outlet moisture contentwas only .5% above the set point(14.5%). The .5% underdrying was a direct result of the large and rapid change in the corn inlet moisture content. The level of the control obtained is excellent taking into account the large and sudden variation in corn inlet moisture content during the first two hours of drying. Tests #7 and 8 proved that if a controller encounters a large variation in the inlet moisture content, it is not be able to control the outlet moisture very close to the set point. This means that the grain inlet moisture. content change and the rate of change have to be considered in evaluating a grain dryer control system. The results of test #9 are shown in Figure 6.27. The inlet moisture content variation is small compared to the inlet moisture variation in tests #7 and 8. The average inlet moisture content was 19.2, the outlet moisture 14.6%, only .1% point above the set point. The variation in the auger rpm was small due to the limited variation in the grain inlet MOISTURE CONTENT %(w.a.) 123 SET POINT = 14.0 % AVERAGE OUTLET MC. = 14.3 75 'V‘I'd'éj 30.0 INLET M.C. 25.0 . 2 "Kit: 200-: no " I II P1000 15.0.1 ‘ .- OUTLET M.C. U) -900 ‘V _w-w-—.- -800 10.0 L700 50 ‘ . , I -600 \ ' R.P.M. ~500 000iTrrfrrrfrrrrrfrrrrrrrrrfrrrrrrrrrTr]rrr O ‘I 2 3 4 5 6 7 8 9 10 TIME (HOURS) Figure 6.25 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer-Morton 850 Dryer in Test #7. MOISTURE CONTENT %(W.B.) 124 SET POINT = 14.5 73 AVERAGE OUTLET MC. = 15.0 % 30.0 25.0 —1800 INLET M C E “'1 70° ' ° '5 -1600 20.0 = — 0- {1500 I: 1400 15,0 L A A. A..‘ _ . O.U-TETM‘. ‘0 “I300 " 1200 I—11OO 10.0 ~1000 ' HQOO ~ ~800 5'0 ~7OO ‘ —600 0.0 I I I I i I I I rI Frrrrr IrfI rr I I I I I I l I I FI I—I I I I I r 0 I 2 3 4 5 6 7 8 9 10 TIME (HOURS) Figure 6.26 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the MeyerHMorton 850 Dryer in Test #8. 'W'd'EI 125 moisture content. Test #9 demonstrates the ability of the control system to control a dryer encountering an inlet moisture pattern normally encountered in the Midwestern US (e.g. the States of Iowa, Illinois and Nebraska) to an average outlet moisture content close to the set point. Test #10 (Figure 6.28) is a run conducted using the linear model to describe the drying process in the control algorithm (see p 63, Chapter 4). The linear model is a two parameter model in which the two parameters are estimated by the sequential least square method. The set point in test #10 was 13.5%(w.b); the average outlet moisture content over 17 hours of dryer-operation was 13.7%. The .2% value above the set point is due to slight underdrying in the early stages of the test. The inlet moisture content varied slowly during the test. The linear model reacted slowly to changes in the inlet and the outlet grain moisture contents compared to the exponential model. Therefore, the exponential model controlled the drying process closer to set point than the linear model under similar inlet moisture content conditions. A further disadvantage of the linear model is that, the dryer has to run in manual for considerable time before it can be switched to the automatic mode because of the method used in estimating the parameters of the linear model. For the above reasons, the exponential model is preferred over the linear model and is used in the control algorithm of the second crossflow dryer tested in this study, the Zimmerman crossflow dryer. The summary of the results obtained from the tests conducted with the Meyer-Morton dryer is shown in Table 6.7. 126 SET POINT = 14.5 7.; AVERAGE OUTLET MC. = 14.6 % rs 25.0 91! 3 IN M.C. I— R7 20.0 g —1600 0' I-ISOO E OUTL M.C. I: LLI In 4400 g ~ISOO o ~1200 LLI 0:: I—IIOO ... HIOOO 9 L900 0 2 000 lIIIirrfiI I'I—IrrI—rrrrrrlIrIlIIIIrI—FFIIIIIIT OI 2 3 4 5 6 7 8 910_ TIME(HOURS) Figure 6.27 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Meyer-Morton 850 Dryer in Test #9. 'V‘I'd'EI MOISTURE CONTENT %(W.B.) 127 SET POINT = 13.5 73 AVERAGE OUTLET MC. = 13.7 7:. 25.0 . INLET M.C. 20.0 _. .. % E —1200 no. 41100 I5 0 .... .-.... ....-- Ali; HOOD “1"" .- ‘ —QOO 10.0 I H800 ‘ .H- -. . “700 5.0 R.P.M. _600 —500 0.0 0 2 4 6 8101214161820 TIME (HOURs) Figure 6.28 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control (Linear Model) of the Meyer Morton 850 Dryer in Test #10. 'W'd'H 128 Table 6.7 Summary of the results obtained from the different control tests (see Tables A.1-A.lO) with the Meyer- Morton 850 dryer. Test Number 1 2 3 4 5 6 Date :1985 12/9 12/13 12/14 12/15 12/19 12/20 Ave Inlet MC 22.2 23.3 21.5 23.2 20.6 21.0 Min Inlet MC 20.8 21.7 21.0 21.6 19.8 20.1 Max Inlet MC 23.6 25.2 22.0 24 6 21.3 21.9 Set Point 14.5 14.5 14.5 15.0 15.0 15.0 Ave Outlet MC 14.4 14.9 14.6 14.9 14.8 15.1 Min outlet Mc 13.0 12.5 13.4 13.9 13.1 13.8 Max Outlet Mc. 15.9 17.8 15.9 16.5 15.8 16.2 Ave RPM _ 821 717 721 695 848 775 Min RPM 698 629 673 609 752 711 Max RPM 960 893 752 817 960 914 Test Number 7 8 9 10 11 Date :1986 10/9 10/29 11/14 11/28 12/3 Ave Inlet MC 26.8 23.0 19.2 19.9 19.3 Min Inlet MC 23.8 19.5 .18.0 17.0 18.2 Max Inlet MC 31.0 34.3 21.0 20.9 20.9 Set Point. 14.0 14.5 14.5 13.5 14.5 Ave Outlet MC 14.0 15.1 14.6 13.7 14.5 Min Outlet MC 11.1 13.4 12.5 12.2 12.0 Max Outlet MC 16.3 17.1 17.5 15.1 16.5 Ave RPM 563 882 1125 765 1034 Min RPM 426 548 936 519 873 Max RPM 768 1224 1310 980 1213 129 6.2.2 Zimmerman Crossflow Dryer Figures 6.30 to 6.33 and Tables A.ll to A.14 show the experimental results of the dryer control system tests conducted on the Zimmerman dryer. Figure 6.29 illustrates the manual control of the Zimmerman dryer. The inlet grain moisture content varied between 18% and 25% during the 17 hour duration of the test. The set point was 15% (w.b.); the aVerage grain outlet moisture content obtained Was 15.5%. The .5% above the set point is due to insufficient corrective action carried out during the manual control. The operator reacted only to the outlet moisture content. During the 17 hours of operation the dryer operator changed the auger rpm only three times; each time the outlet moisture content was above the set point. Once, with the outlet moisture above the set point, the inlet moisture content dropped substantially, but still the operator reduced the auger rpm. This resulted in overdrying at the end of the test. In general, the operator controlled the drying process reasonably well considering the variation in the inlet moisture content during the test. It must be emphasized that during this test the dryer operator'luni additional information about the drying process supplied by thermfisture meter (inlet and outlet moisture content every 4-5 minutes), which helped him achieve the good result. Figure 6.30 and Table A.11 show the result of the automatic control of test #11 performed with the Zimmerman dryer. Theaaverage outlet moisture content during 14.9 hours of drying was 16.9%; tine set point was 17%. Based on‘the'average outlet moisture content, the controller was successful in controlling the drying process. The unload auger rpul Twas changed.from 481 to 1662 during the test; the large change was due to the large variation in the inlet moisture cOntent which varied between 19.1% and 31.5%. MOISTURE CONTENT %(W.B.) 30.0 25.0 20.0 15.0 10.0 5.0 0.0 INLET M.C. ' i I 'E' E OUTLET M.C. . ,, R1}IT_TZWL____‘ O 2 4 5 8 1O 12 14 15 18 20 130 SET POINT = 15.0 % AVERAGE OUTLET. MC.= 15.5 % TIME (HOURS) —2000 -1800 ~1600 ~1400 ~1200 61000 ~800 Figure 6.29 Inlet and Outlet Moisture Contents, and RPM vs Time During Manual Control of the Zimmerman ATP 5000 Dryer. ‘W'd'tI MOISTURE CONTENT %(W.B.) 131 SET POINT = 17.0 7.’ AVERAGE OUTLET MC.= 16.9 % I. , INLET M.C. 15.0' -—L 9 PI .0 CD C3 C3 1 LJ 1 ll 1 L111 UT M.C. .‘ ' , . . _ ~ I »‘ . l ‘ ‘Idl -.' ‘1 3‘s"? M“ “Affin‘vvn-oC—amx- 1" l'l .. SET POINT ~2400 ~2200 -2000 , R.P.M. 1800 -1500 P1400 -1200 L-1OOO L-800 L-600 0 2 4 6 8 YIIIIUIITIITTTT‘IIIIIUFIITII(TIIIIIIIUUITITITIIITTIIITI[IIITTIIITTTIIIUITIIIIIT 10 12 14 16 TIME (HOURS) 18 20 Figure 6.30 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Zimmerman ATP 5000 Dryer of Test #11. 'W'd'tI 132 Figure 6.31 shows the results of test #12. The inlet moisture content ranged between 19.9% and 23.5% which differs only by 2.4% and 6% from the set point, respectively. The average grain outlet moisture content after 13 hours of drying was 17% for a set point of 17.5%; the grain was overdried by .5%. The overdrying can be attributed partially to the maximum level reached by the controller (i.e. the controller had adjusted the auger rpm to its maximum value). The maximum rpm occurred because of the small amount of moisture removed. Manual control during start up contributed to the overdrying of the grain, the average grain outlet moisture content, excluding the first two hours of start-up, is 17.24%. Test #13 is shown in Figure 6.32. The inlet moisture content varied from 16.1% to 25.5% during the 12 hours of drying. The average grain outlet moisture content was .5% below the set point. The controller reached saturation four times for a minimum of. one hour duration. The 16.1% grain inlet moisture content was 1.4% below the set point, because the controller was not allowed to speed up the rpm of the unload auger to the desired value. Overdrying of part of the grain was thus unavoidable. The grain inlet moisture content sample in the Zimmerman dryer was taken at the wet leg conveyor at ground level and not at the dryer inlet. The location of the inlet sample port deleteriously affected the operation of the control system, because the controller reacted to an inprecise value of inlet moisture content. This had a significant effect on the controller performance when a large‘and sudden changes occurred in the inlet moisture content. To eliminate this time effect, a delay was introduced in the control algorithm. The main purpose of the delay is to delay the controller reaction to the measured inlet moisture content for a period equal to the time for the grain mass MOISTURE CONTENT %(W.B.) 133 SET POINT = 17.5 7» AVERAGE OUTLET MC.= 17.0 75 30.0 25 0 I *2- 5 NLET M.C. 5 .2600 - -4 .. - v - ”' 0- 92400 20'0 OUTLET M.C. t0 L-2200 Whit?" 91191352..“ ..-. xégynfiiflffi'u‘Yn-‘f("7“"- s“.- U) 2000 15.0 -1 800 I II. ‘ L—1600 . ~I4OO 0.0 Me..- 1 I-1200 ' R.P.M. L800 ~600 0.0 O 2 4 6 8 10 12 14 16 18 20 TIME (HOURS) Figure 6.31 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Zimmerman ATP 5000 Dryer Of Test #12. 'W'd'él 134 SET POINT = 17.5 75 AVERAGE OUTLET MC.= 17.0 % d 30.0 ‘I! 5 5 ° INLET M.C. 5 —2600 B\ T'— ’ 11,9215" .. D- ...-2400 Z " ‘ U ' ET M.C. t: ~2200 tfl jkfiflghflhfiwyhk%~‘e .E?fifién U) 2000 Z 15 0— *‘ ,q —1800 O 3 " 91600 O 10 0-1 “1400 E : v, —1200 D 1‘ P M t-I 000 I?) 5'01 ' ' ' ~800 5 '5 I—600 2 0.0 IITITITIIIIITIIITIITITIIIIIIIllIrIIIIWIIIITTITIITTUFITTTIIIITTIIWIIIIIITTIIII O 2 4 6 8 10 12 14 16 18 20 TIME (HOURS) Figure 6.32 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Zimmerman ATP 5000 Dryer of Test #13. 'W'd'tI 135 (represented by the present inlet moisture content sample) to reach the dryer entrance. The required time(delay) was calculated based on the current grain flow rate and the capacity of the wet leg. Test #14 with the Zimmerman dryer is shown in Figure 6.33. The test was conducted with the above modification incorporated in the control system algorithm. The grain inlet moisture content had a minimum of 17.9% and a maximum of 26.3%. At the start of the test the grain inlet and outlet moisture contents were almost the same. The average grain outlet moisture content during 12.5 hours of drying was 16.6%, only .1% above the set point. The auger reached the maximum rpm when very low moisture content grain entered the dryer. The limitation of the auger speed resulted in overdrying two hours later. The introduction of the delay in the control algorithm improved the control system performance of the Zimmerman dryer (compared test #13 and 14). The summary of the results obtained from the tests conducted with the Zimmerman dryer is shown in Table 6.8. 6.3 Performance Index The objective of an automatic control system is to control the process so that the output is close to the set point. In grain drying, the objective of the control system is to control the drying process of the grain entering a dryer to a set moisture content, regardless of the variation in the inlet moisture content or the drying conditions. To compare results obtained using different control system strategies or different tests for the same control strategy, a performance index needs to be established. The performance index should account for overdrying or underdrying, and for the rate of drying or the rate of inlet moisture content changes. Thus, it is reasonable to incorporate the deviation of the grain outlet moisture content from the set point,the change in the grain 136 SET POINT = 16.5 75 AVERAGE OUTLET MC.=—= 16.6 % r5 300-; 03 : 3 .1 INLET M.C. ._ R27 25'0 1 g ~2600 . ~ ,. ,. o .. E zoo . ' om M ,. g -3388 E B 111.34.. ‘ “" m L2000 z 15 0 "' ”... L-1800 8 - L1600 01 10.0 ...... . ~ " T1400 0: "“ ~— e1200 D i , R.P.M. L1000 O : 6 L600 2 000 YTIIIIY‘IIYTTTIIITTIIII[IFIIIYIIITIWIIIIIFIFIIIIIIIIWIIIFIIIIIIIIITIIIIIIIIII 0 2 4 6 8101214161820 TIME (HOURS) Figure 6.33 Inlet and Outlet Moisture Contents, and RPM vs Time During Automatic Control of the Zimmerman ATP 5000 Dryer Of Test #14. 'W'd'tl 137 Table 6-8 Summary of the results obtained from the different control tests (see Tables A.11-A.14) with the Zimmerman ATP 5000 dryer. Test Number 11 12 13 14 Date : 12/15/86 12/16/86 12/19/86 01/15/87 Ave Inlet MC 22.1 21.4 21.7 21.6 Min Inlet MC 19.1 19.9 16.1. 17.9 Max Inlet MC 31.5 23.5 25.5 26.3 Set Point 17.0 17.5 17.5 16.5 Ave Outlet MC 16.9 17.0 17.0 16.6 Min Outlet MC 14.3 14.9 14.6 14.4 Max Outlet MC 18.6 18.5 20.0 19.1 Ave RPM 1148 1390 1513 1294 Min RPM 481 811 848 554 Max RPM 1662 1749 1746 1765 138 inlet moisture content, and the grain flow rate into the performance index. Different performance indices are investigated in this section to evaluate the control strategies of certain automatic control systems. The performance indices are: T n P11 - f e2 (Gp/lOOO)dt - 2 ei2 (Gpi/1000) (ti- ti_l) (6.1) 0 1=2 T 2 -2 P12 - f e (Gp/lOOO)(Min-Mina) dt 0 n 2 -2 - iEZei (Gpi/1000)(Mlni-Mlna) (tiTti-l) (6.2) 1 2 . . -2 P13 - f e (Gp/1000)(Min+- Min) dt 0 n 2 -2 = ifzei (Gpi/1000)(Mlni- Mini_1) (ti-ti-l) (6.3) T 2 T 2 P14 = f e (Cp/lOO)dt/ f (Min-Mina) dt 0 0 n 2 n 2 - 15261 (Gpi/100)(ti-ti_1)/i§2(Mlni- Mina) (ti'ti-l) (6.4) 1 2 T 2 P15 - f e (Gp/100)dt/ f(Min+- Min) dt 0 0 139 .n n 2 . . 2 = '2 e1 (Cpl/100)(t1't1-1)/(.2 (Mini-Mlni-l) (ti-ti-l) (6 5) i=2 1-2 P16 = STD Of grain outlet m.c./STD of grain inlet m.c. (6.6) V 2 V 2 P17 - f e dv - f e Gp dt (6.7) 0 0 V2 -2 PI8 = f e (Min-Mina) dv (6.8) 0 where e = outlet m.c. - set point Gp - grain flow rate bushels/hr Min - inlet moisture content (%,W.B.) Mina average inlet moisture content (%,W.B.) Min - inlet moisture content at time t+dt P1 = performance index t - time, hours. STD = standard deviation Equations 6.1 through 6.8 have been evaluated for several tests conducted with the Meyer-Morton 850 dryer; the summary of the results is shown in Table 6.7. Table 6.9 gives the results of evaluating the performance indices of the tests in Table 6.7. A ranking of the different tests according to the different ‘performance:indices is shown in Table 6.10. P14 to P16 rank test #8 as the best, whereas PIl ranks test #8 as the worst. This is expected 140 since the inlet moisture content of the grain is not included in P11, and test #8 has the largest variation in the inlet moisture content of. the eleven experimentally conducted tests. The largecflmngezhlthe grain.inlet moisture content contributes to the overdrying and underdrying and results in the larger value of P11 for test #8. P13 ranks test #8 near the bottom due to the constant inlet moisture encountered after two hours of drying (see Section 6.2.1). The grain inlet moisture content is included in P12 and P13 by dividing the square of the error (outlet m.c. - set point) by the 'square of the difference between the present grain inlet moisture content and the average inlet moisture content or the previous inlet [moisture content. Dividing by the square of the inlet moisture content difference, creates a problem when the difference is zero or very close to zero. When division by zero takes place, the result is an indefinite value. Dividing by a small number results in over-penalizing the control system when there is no variation in the inlet moisture content or (P13 for test #8) the variation is very small. Therefore, P12 and P13 are inappropriate as performance indices. P14 and P15 are modifications of P12 and P13, respectively. In P14 and P15 the square of the present difference in the inlet moisture content is replaced by the overall sum of the square of the difference in the inlet moisture content either, the average inlet moisture content or the previous inlet moisture content. The ranking of the tests is similar. .P16 is simple and useful as a quick check of control system performance. The index indicates how the control system performs in reducing the variation in the grain inlet moisture. 141 P17 and P18 are not evaluated numerically because of their similarity to P11 and P12. The drawback of P11 and P12 is also applicable to P17 and P18. P14, P15, and P16 were used to compare tests #2 and #8 from the dryer control tests with the exponential drying model (see Table 6.11a and b); tests #10 and 11 were obtained using the control algorithm with the linear drying model. Using P14, P15, P16, the rankings of the above tests is shown in Table 6.11b. Test #8 is ranks at the top while test #11 is ranked at the bottom. Test #10 ranks second followed by test #2. The three perfOrmance indices are consistent in their ranking due to their-similarity. Test #11 is expected to be ranked low since the dryer was running for a shorter period of time compared to other tests. Thus, some consideration must also be given to the length of the test since tests with a longer running time have a better chance to be ranked high. It must be stressed that control systems of grain dryers should be grouped according to the similarity of conditions faced during the operations. I In conclusion, the idea of using a performance index to evaluate a control system has merit. However, it must be emphasized that a performance index must be closely examined so that a control system is not penalized when it encounters a difficult to control drying Operation. Finally, P14, P15, and P16 are recommended for measuring the performance of a dryer control system. 6.4 CONCLUSIONS In order to develop an automatic dryer controller, a dryer process— model is required. For this purpose, a basic heat and mass transfer differential equation crossflow-drying model has been developed, and validated with experimental data collected from a commercial croszlow 142 Table 6.9 Performance indices for the different control tests in Table 6.7. Test No. P11 P12 P13 P14 P15 P16 1 2.3 103.1 33.3 .6 .8 .9 2 6.9 205.9 150.6 1.5 .4 1.4 3 1.4 291.8 46.2 3.9 2.3 2.5 4 1.4 76.0 51.5 .4 .8 .8 5 1.7 175.9 60.0 1.4 2.4 1.4 6 4.1 77.0 122.3 4.8 3.7 1.7 7 4.3 21.6 31.4 .2 .3 .7 8 7.6 36.1 151.8 .1 .2 .2 9 6.9 314.9 147.6 3.3 2.7 1.5 10 3.1 117.4 82.2 .9 .6 1.0 11 6.2. 115.4 159.2 6.1 5.9 2.1 Table 6.10 Ranking of different control tests according to the different PIs. P11 P12 P13 PI4 "U H U1 P16 H H H (DONl-‘VO‘Of—‘U'l-F‘U) [—3 H H WHQQNLDOI—‘J-‘Nm P‘P‘ otehouwc>h¢haoxp~aww H tambouaoxc>unbcpra\q Idh>a\0tnu>bwac>\4a> Hrooxu>uwu>bwd<3~qcn H H H 143 Table 6.11a Performance indices for tests 2,8,10, and 11. Test No. PI4 " P15 P16 2 1.5 3.8 1.4 8 .1 .2 .2 10 .9 .6 1.0 1 5.9 2.1 11 6. Table 6.11b Ranking of tests 2,8,10, and 11. P14 P15 P16 8 8 8 10 10 10 2 2 2 ll 11 ll dryer. The new model requires excessive computer time but played a fundamental role in the development of two empirical process models. Both empirical models were employed, in conjunction with a dryer control algorithm, as essential components in the automatic control system. A series of successful tests were conducted with the feedforward moisture content-based controller on two commercial dryers over a ‘periodof two drying seasons. The controller controlled the outlet moisture content well even for large and rapid inlet grain moisture changes. The new automatic controller appears to be stable, durable and accurate. Commercial application of the controller by the grain: processing industrywill only be a matter of time. Adoption of the unit will result in improved grain quality, decreased energy consumption, and better record-keeping. 144 CHAPTER 7 7 SUMMARY 1. An unsteady state differential-equation simulation model for crossflow grain drying has been developed. 2. Two simplified empirical drying models for the automatic control of crossflow dryers have been developed. 3. A control algorithm has been developed for crossflow grain drying using a simplified empirical drying model and a feedforward with feedback trim control algorithm. 4. The control algorithm has been incorporated into a commercial on-line moisture-measuring system to form an automatic control system for crossflow grain dryers. 5. The control system.has been implemented and successfully tested on several commercial crossflow grain dryers; the control system consists of a microcomputer, a semi-continuous moisture meter, a ' tachometer, and the control/dryer-model software. 6. The average outlet moisture content in the commercial dryers was controlled to 10.6% of the set point during two drying seasons; the inlet grain moisture content variation in one hour was much as 9%. 7. The newly developed crossflow dryer automatic control system was found to be stable during all experimental and simulated tests. 145 CHAPTER 8 §DEQESIIQH§_EQB_EEIEEE_§IQDX 1. Test the control system on different dryer types (i.e. mixed-flow, concurrent-flow, fluidized bed dryers). 2. Test the control system on multi-stage grain dryers. 3. Test the contrOl system for different grain types (i.e. wheat, soybeans, rice, etc.). 4. Develop a control strategy which modulates the drying air temperature, and possibly the airflow rate, in addition to the grainflow rate. 5. Analyze the advantage of a control strategy based on the rate of change of the inlet grain moisture content. 6. Develop auxiliary software programs to complement the basic control software (i.e. plot the data, calculate and print summary of tests results, etc.). 7. Evaluate the effect of different pseudo inlet moisture contents on the performance of the control system. 8. Analyze the employment of different numerical techniques to reduce the CPU time of the control system simulation model. 9. Study the effect of variations in air and grain parameters on the simulation results of the unsteady state crossflow drying model. 10. Study the effect of MC valve location and the importance of the inlet MC measurement accuracy on the control system performance. 11. Evaluate the economical feasibility of use of the control system. 146 9. REFERENCES Anderson, J.C. 1985. Performance Evaluation of Commercial Crossflow and Concurrent-Flow Grain Drying. Unpublished M.S. Thesis. Department of Agricultural Engineering, Michigan State Universityy Eastlxuming, Michigan. Bakker-Arkema, F.W. Selected Aspects of Crop Processing and Storage : A Review. 1984. J. Agric. Engng Res., 30(1) 1-22. Bakker-Arkema, F.W.; Fontana, C.; Brook, R.C.; Westelaken,C.M. 1983 Concurrent-flow Rice Drying. Drying Technology, 1(2) 171-191. Bakker-Arkema, F.W.; Fosdick, S., Naylor, J. 1979. Testing of Commercial Crossflow Grain Dryers. 1979. Paperrhfl). 79-3521. Am. Soc. Ag. Eng., St. Joseph, MI. Bakker-Arkema, F.W.; Hall, C.W. 1965. Importance of Boundary Conditions in Solving the Diffusion equation for drying Wafer. 1965. Trans. ASAE. 8(3) 382-383. Bakker-Arkema, F.W.; Lerew, L.E.; DeBoer, S.F.; Roth, M.C. 1974. Res. report 224, Michigan State University, E. Lansing, MI, USA. Bakker- Arkema, F. W. Rodriquez, J. C. ; Schisler, I. P. Westelaken, C. M. 1982. A new Commercial Crossflow Dryer with Differential Grain Speed. Paper No. 82- 3007 Am. Soc. Ag. Eng. , St. Joseph, MI. Bakker-Arkema, F.W.; Rodriquez, J.C.; Brook, R.C.; Hall, G.E. 1981. Grain Quality and Energy Efficiency of Commercial Grain Dryers. Paper No. 81-3019 Am. Soc. Ag. Eng., St. Joseph, MI. Barre, H.J.; Baughman, G.R.; Hamdy, M.Y. 1971. Application Of the .lOgrithmic model to crossflow deep-bed grain drying. Trans. ASAE, 14(6) 1061-1064. Baughman, G.R.; Hamdy, M.Y.; Barre, H.J. 1971. Analog computer simulation of deep bed drying of grain. Trans. ASAE, 14(6) 1058-1060. Baumeister, T.; Avallone, E.A.; Baumeister 111, T. 1978. Markfs Standard Handbook for Mechanical Engineers. Eight Edition. McGraw-Hill Book Company, New York, N.Y. Beckg .JJV.; Arnold, K.J. 1977. Parameter Estimation in Engineering and Science. John Wiley & Sons. New York, N.Y. Becker, H.A.; Sallans, H.R. 1955. A study of internal moiSture movement in the drying of wheat kernel. Cereal Chem. 32(3) 212-216. Berglund, K.A. 1987. Personal Communication. Ag. Eng.lknr, MSU, E. Lansing, MI. Boyce, D.S. 1966. Heat and moisture transfer in ventilated grain. J. Agric. Engng Res., 11(4) 255-265. 147 Borsum, J.C.; Bakker-Arkema, F.W. 1982. Microprocessor control of drying processes. Paper NO. 82-6006. Am. Soc. Ag. Eng., St. Joseph, MI. Brooker, D.B.-; Bakker-Arkema', F.W.; Hall, C.W. 1974. Drying Cereal Grains. AVI Publishing Co., Westport, CT. Chittenden, D. H. Hustrulid, A. 1966. Determining drying constants for shelled corn. Trans. ASAE 9(1) 52-55. Chu, S.T.; Hustrulid, A. 1968 General characteristics of variable diffusivity process and the dynamic equilibrium moisture content. Trans. ASAE 11(5) 709-710. Ezeike, G.0.; Otten, L. 1981. Theoretical analysis of the tempering phase of a cyclic drying process. Trans. ASAE 24(6) 1590-1594. Flood, C.A. Jr.; Sabbah, M.A.; Meeker, D.; Peart, R.M. 1972. Simulation of a natural-air corn drying system. Trans. ASAE 15(1) 156-159, 162. Fontana, C.; Bakker-Arkema, F.W., Westelaken, C.M. 1982. Concurrent flow vs crossflow drying of long-grain rice. Paper No. 82-3569 Am. Soc. Ag. Eng., St. Joseph, MI. Forbes, J.F.; Jacobson, E.A.; Rhodes, E.; Sullivan, G.R. 1984. Model based control strategies for commercial grain drying systems. Can. J. Chemical Eng. 62(12) 773-779. Gustafson, R.J.; Morey, R.V. 1981. Moisture and quality variations across the column of a crossflow grain dryer. Trans. ASAE 24(6) 1621- 1625. Gygax, R.A.; Diaz, A.; Bakker-Arkema, F.W. 1974. Comparison of commercial crossflow and concurrent-flow dryers with respect to grain damage. Paper NO. 74-3021 Am. Soc. Ag. Eng., St. Joseph, MI. Hamdy, M.Y.; Barre, H.J. 1969. Evaluating film coefficient in single kernel. Trans. ASAE 12(2) 205-208. Henderson, S.M. 1974. Progress in developing the thin layer drying equation. Trans. ASAE 17(6) 1167-1168, 1172. Henderson, J.M.; Henderson, S.M. 1968. A computational procedure for deep bed drying analyses. J. Agric. Engng Res. 13(2) 87-95. Henderson, S.M.; Pabis, S. 1961. Grain drying theory: 1, Temperature effect on drying coefficient. J. Agric. Engng Res. 6(3) 169-174. Henderson, S.M.; Pabis, S. 1961. Grain drying theory: 2, A critical analysis of the drying curve for shelled maize. J. Agric. Engng Res. 6(4) 272-277. Holtman, J.B.; Zachariah, G.L. 1969a. Continuous crossflow modeling for Optimal control. Trans. ASAE 12(5) 430-432. Holtman, J.B.; Zachariah, G.L. 1969b. Computer control for grain driers. Trans. ASAE 12(5) 433-437. 148 Luikov, A.V. 1966. Heat and Mass Transfer in Capillary Porous Bodies. Pergamon, Inc., New York, N.Y. Manetsch, T.J.; Park, G.L. 1982. System analysis and simulation with applications to economic and social Systems. Part 1. Methodology, modeling and linear system fundamentals. Dept. of Elect. Eng. and Systems Science, Michigan State University, East Lansing, MI. Merchant, J. A. 1985. Control of high temperature continuous- flow grain driers. Agric. Engr. 40(4) 145- 149. Montgomery, D.C.; Johnson, L.A. 1976. Forecasting and Time series Analysis. McGraw-Hill Book Company, New York,_N.Y. - Nellist, M.E. 1982. Developments in Continuous flow grain Dryers. Agric. Engr. 37(3) 74-80. Nellist, M.E. 1976. Exposed layer drying of ryegrass seeds. J. Agric. Engng. Res. 21(1) 49-66. Nellist, M.E.; O'Callaghan. 1971. The measurement of drying rates in thin-layers of ryegrass seed. J. Agric. Engng. Res. 16(3) 192-212. Nybrant, T.G. 1986. Modeling and control of grain dryers. UPTEC 8625r. Upsala University, Upsala, Sweden. Nybrant, T.G.; Regner, P.J. 1985. Adaptive control of continuous grain dryers. Paper No. 85-3011. Am. Soc. Ag. Eng., St. Joseph, MI. Ogata, K. 1970. Modern Control Engineering. Prentice-Hall, Inc. Englewood Cliffs, N.J. Oleseul,IH.T. 1976. Automatic control arrangement for continuous drying plants. British Patent 1502710. British Patent Office, London, UK. Palmer, J. 1984. Selecting profitable automatic control systems. Paper No. 84-1635. Am. Soc. Ag. Eng., St. Joseph, MI. Parry, J.L. 1985. Mathematical modeling and computer_simulation of heat and mass transfer in agricultural grain drying: A review. J. Agric. Engng. Res. 32(1) 1-29. Pierce, R.C.; Thompson, T.L. 1981. Energy use and Performance related to crossflow dryer design. Trans. ASAE, 24(1),216-220. Rowe, R.J.; Gunkel, W.W. 1972. Simulation of temperature and moisture content of alfalfa during thin-layer drying. Trans. ASAE 15(5) 805-810. Routh, E.J- 1877. A Treatise on the stability of a given state of motion. MacMillan, Inc., London, Uk. Sabbah, M.A.; Keener, H.M.; Meyer, G.E. 1979. Simulation of solar «laying of shelled corn using the logarithimic model. Trans. ASAE 22(3) 637-643. Schisler, I.P.; Bakker-Arkema, F-W.; Brook, R.C. 1982. Control of concurrent-flow grain dryers. Paper No. 82-3513. Am. Soc. Ag. Eng., St. Joseph, MI. 149 Spencer,'H.B. 1972. A revised model of the wheat drying process. J. Agric. Engng. Res. 17(2) 189-194. Sharaf-Eldeen, Y.I.; Blaisdell, J.L.; Hamdy, M.Y. 1980. A model for ear corn drying. Trans. ASAE 23(5) 1261-1265, 1271. Steffe, J.F.; Singh, R.P. 1980. Liquid diffusivity of rough rice components. Trans. ASAE 23(3) 767-774, 782. Steffe, J.F.; Singh, R.P. 1980. Theoretical and practical aspects of rough rice tempering. Trans. ASAE 23(3) 775-781, 782. Thompson, T.L.; Peart, R.M.; Foster, G.R. 1968. Mathematical simulation of corn drying a new model. Trans. ASAE 11(4) 582-586. Watson, E.L.; Bhargava, V.K. 1974. Thin layer drying studies on wheat. Can. J. Agric. Engng 16(1) 18-22. Whitfield, R.D. 1986. An unsteady-state simulation to study the control of concurrent and counter-flow grain dryers. J. Agric. Engng. Res. 33(2) 171-178. Young, J.M.; Dickens, J.W. 1975. Evaluation of costs for drying grain in batch or crossflow systems. Trans. ASAE 18(4) 734-739. Young, J.H.; Whitaker, T.B. 1971. Numerical analysis of vapor diffusion in a porous composite sphere with concentric shells. Trans..ASAE 14(2) 1051-1057. ‘ Zachariah, G.L.; Isaacs, G.W. 1966. Simulating a moisture control system for a continuous-flow driers. Trans. ASAE 9(4) 297-302. 150 10. APPENDICES A. Experimental Results B. Simulation Results C. Specifications for the Data Acquisition Components 151 APPENDIX A ;E§perimental Results Table A.l Experimental results of an automatic control test on a crossflow grain dryer. Test Number : l Date : 12/9/1985 Dryer Type : Meyer-Morton Model 850 Set Point : l4.5%(w.b.) Sample # Time Inlet MC Outlet MC R.P.M. hr % w.b. % w.b. 1 0.00 21.7 13.9 817 2 0.14 21.1 14.3 834 3 0.27 21.2 14.3 834 4 0.39 21.2 14.3 817 5 0.52 21.5 13.7 817 6 0.65 ‘ 21.3 13.5 834 7 0.80 21.5 13.5 817 8 0.92 21.8 13.9 834 9 1.04 21.6 14.5 834 10 1.17 21.5 13.9 872 11 1.32 22.1 13.8 872 12 1.45 21.7 13.7 834 13 1.59 22.4 14.1 834 14 1.70 22.4 13.6 834 15 1.85 21.9 14.3 _872 16 1.99 22.1 13.0 914 17 2.12 22.1 13.8 914 18 2.24 21.7 14.6 893 19 2.37 21.5 14.6 914 20 2.50 22.0 15.0 872- 21 2.64 22.8 14.7 893 22 2.80 22.5 13.7 893' 23 2.94 22.1 15.2 834 24 3.07 22.4 15.3 834 25 3.20 23.0 15.4 853 26 3.34 22.0 15.7 872 27 3.47 22.1 14.9 800 28 3.59 21.8 15.5 783 29 3.72 22.6 15.5 817 30 3.84 23.5 13.9 783 31 3.97 22.3 15.6 738 32 4.10 23.5 14.7 738 33 4.22 22.5 15.9 724 34 4.35 22.4 14.7 724 35 4.49 22.8 14.6 738 36 4.62 23.1 14.4 752 37 4.74 23.4 14.7 724 38 4.87 23.2 15.3 724 39 5.04 23.1 14.0 711 40 5.17 23.5 14.2 738 41 5.29 22.9 15.0 711 42 5.40 22.3 14.2 698 43 5.55 21.9 15.2 724 44 5.69 20.8 14.1 711 152 45 5.82 22.9 13.6 738 46 5.94 23.6 14.6 738 47 6.09 23.5 14.8 724 48 6.20 22.1 13.7 738 49 6.34 23.0 14.7 724 50 6.47 22.5 13.8 711 51 6.60 23.0 13.5 768 52 6.72 22.6 13.7 752 53 6.85 22.5 13.6 800 54 6.97 22.3 13.3 800 55 7.10 23.0 14.3 834 56 7.24 22.8 14.1 872 57 7.37 22.4 15.6 872 58 7.50 22.6 14.7 834 59 7.64 22.3 14.5 834 60 7.75 23.3 14.6 834 61 7.90 22.4 14.1 817 62 8.04 22.3 14.3 800 63 8.17 22.6 15.0 834 64 8.30 23.1 15.2 853 65 8.44 22.3 15.3 834 66 8.57 21.7 14.7 834 67 8.70 22.2 14.5 834 68 8.82 21.9 14.7 834 69 8.97 21.6 14.0 834 70 9.10 21.0 13.8 834 71 9.25 21.1 14.5 893 72 9.39 21.2 14.7 893 73 9.52 21.6 14.1 914 74 9.65 22.1 15.2 914 75 9.79 21.2 14.4 893 76 9.92 21.6 14.3 914 77 10.05 22.3 14.4 914 78 10.19 21.9 13.7 914 79 10.32 22.1 15.1 960 80 10.45 21.8 14.8 936 81 10.60 22.2 14.7 914 Ave 22.2 14.4 821 Std 0.7 0.6 67 Min 20.8 13.0 698 Max 23.6 15.9 960 153 'Table A.2 Experimental results of an automatic control test on a crossflow grain dryer Test Number : 2 Date : 12/13/1985 Dryer Type : Meyer-Morton Model 850 Set Point : 14.5%(w.b.) Sample # Time Inlet MC Outlet MC R.P.M. hr % w.b. % w.b. 1 0.00 25.2 12.8 783 2 0.13 24.2 13.2 783 3 0.28 24.4 12.8 834 4 0.43 24.6 12.5 783 5 0.56 25.2 12.6 800 6 0.71 24.7 13.4 817 7 0.83 25.0 14.2 817 8 0.98 24.8 14.5 853 9 1.11 24.3 16.6 872 10 1.25 23.9 15.5 834 11 l 38 24.3 16.6 872 12 l 55 23.6 15.8 872 13 1.66 23.7 16.9 834 14 1.80 23.7 17.8 834 15 1.95 23.8 16.7 893 16 2 08 24.3 17.8 752 17 2 21 24.0 16.8 673 18 2 35 24.8 16.2 711 19 2.48 23.8 16.8 662 20 2 60 24.1 17.0 662 21 2 78 23.7 16.0 650 22 2 86 23.6 15.6 662 23 2 98 23.7 15.8 650 24 3 13 23.4 15 3 662 25 3 25 23.7 15.2 673 26 3 38 23.9 14.9 685 27 3.51 23.1 16.3 673 28 3.65 23.7 15.0 673 29 3.80 23.5 15.3 685 30 3.91 23.3 14.9 650 31 4.05 22.5 15.9 640 32 4.20 22.9 14.7 629 33— 4.31 23.4 14.3 650 34 4.46 23.0 14.9 650 35 4.58 23.4 15.3 673 36 4 73 22.9 13.9 662 37 4 85 23.2 15.8 673 38 5.00 _ 23.0 14.2 673 39 5.15 22.4 14.5 698 -40 5.28 23.2 14.6 662 41 5.41 23.1 14.9 685 42 5.56 23.2 14.2 685 43 5.71 23.2 14.6 711 44 5.85 23.1 13.3 673 45 6.00 23.1 14.8 698 46 6.13 23.3 13.8 698 47 6.26 23.0 15.3 711 154 48 6.40 22.9 13.5 698 49 6.55 22.5 13.9 698 50 6.70 22.1 14.9 724 51 6.83 23.6 13.9 724 52 6.96 22.5 14.9 724 53 “7.15 22.1 14.1 711 54 7.30 22.1 14.8 711 55 7.45 22.6 15.9 711 56 7.56 23.1 14.8 711 57 7.73 22.6 15.3 662 58 7.86 22.4 13.7 662 59 8.00 22.3 15.3 673 60 8.15 21.9 14.5 673 61 8.28 22.1 14.4 711 62 8.41 22.0 14.1 698 63 8.56 23.0 16.0 711 64 8.70 21.7 14.3 724 65 8.85 22.6 14.7 698 66 8.98 22.7 14.9 711 67 9.13 22.4 14.4 673 68 9.26 22.3 14.1 698 Ave . 23.3 14.9 717 Std 0.8 1.2 67 Var 0.7 1.4 4509 Min 21.7 12.5 629 Max 25.2 17.8 893 155 Table Au3 Experimental results of an automatic control test on a crossflow grain dryer Test Number : 3 Date : 12/14/1985 Dryer Type : Meyer-Morton Model 850 Set Point : 14.5%(w.b.) Sample # Time Inlet MC Outlet MC R.P.M. hr % w.b. % w.b. l 0.00 21.6 15.4 724 2 0.13 21.4 14.1 724 3 0.27 21.7 14.6 738 4 0.42 22.0 15.3 752 5 0.55 21.9 14.9 724 6 0.70 22.0 14.6 738 7 0.83 21.7 15.4 738 8 0.97 21.4 14.0 724 9 1.12 21.8 14.9 724 10 1.25 21.6 14.8 724 11 1.40 21.8 13.4 698 12 1.53 21.6 15.9 698 13 1.68 21.4 15.3 698 14 1.82 21.7 14.6 711 15 1.95 21.6 14.5 711 16 2.08 21.5 15.7 673 17 2.22 21.3 14.3 673 18 2.37 21.5 15.0 698 19 2.50 21.4 14.4 673 20 2.65 21.5 14.0 724 21 2.78 21.9 13.7 711 22 2.92 21.1 15.2 711 23 3.07 21.5 14.6 724 24 3.20 21.1 13.7 738 25 3.35 21.6 13.9 724 26 3.50 21.5 14.8 724 27 3.65 21.0 14.2 738 28 3.78 21.5 13.9 738 29 3.92 21.7 15.3 738 30 4.07 21.4 14.7 738 31 4.22 21.7 14.5 711 32 4.35 21.3 14.4 724 33 4.50 21.5 15.6 752 34 4.63 21.6 14.1 711 35 4.77 21.5 14.4 698 36 4.90 21.1 14.3 698 37 5.05 21.4 14.2 711 38 5.22 21.0 14.2 711 39 5.37 21.7 14.3 724 40 5.50 21.4 15.0 752 41 5.63 21.5 13.7 724 42 5.77 21.6 14.9 711 43 5.92 21.4 15.0 711 44 6.07 21.0 13.8 711 45 6.22 21.5 14.5 711 46 6.35 21.1 13.6 752 47 6.50 21.3 14.5 738 48 6.65 21.3 14.4 752 156 49 6.80 21.4 13.4 752 50 6.93 21.6 15.9 752 p 51 7.07 21.0 14.8 724 52 7.22 21.4 13.8 738 53 7.37 21.3 14.5 711 54 7.50 21.2 15.3 724 55 7.65 21.5 14.5 724 56 7.78 21.0 13.9 724 57 7.93 21.4 15.6 724 Ave 21.5 14.6 722 Std 0.2 0.6 19 Var 0.1 0.4 371 Min 21.0 13.4 673 Max 22.0 15.9 752 157 Table A.4 Experimental results of an automatic control test on a crossflow grain dryer Test Number : 4 Date : 12/15/1985 Dryer Type : Meyer-Morton Model 850 Set Point : 15.%(w.b.) Sample # Time Inlet MC Outlet MC R.P.M. hr % w.b. % w.b. 1 0.00 21.6 14.5 783 2 0.14 21.9 15.3 800 3 0.30 22.0 15.2 800 4 0.45 22.1 14.9 783 5 0.59 22.0 15.2 800 6 0.72 22.4 15.4 817 7 0.85 22.7 15.1 817 8 0.99 22.6 14.7 783 9 1.12 22.6 14.4 783 10 1.25 22.7 15.9 738 11 1.40 22.2 14.7 738 12 1.52 22.3 15.7 738 13 1.65 22.4 15.7 711 14 1.79 22.3 16.3 711 15 1.92 22.5 15.1 673 16 2.10 22.2 14.9 673 17 2.22 22.5 15.4 685 18 2.37 22.4 14.0 673 19 2.52 22.1 15.4 724 20 2.65 22.2 14.8 711 21 2.77 24.0 14.5 724 22 2.89 24.6 14.3 724 23 3.07 24.2 14.7 698 24 3.20 23.7 14.8 685 25 3.32 24.1 14.8 698 26 3.44 22.1 14.2 698 27 3.57 23.6 13.9 724 28 . 3.77 23.8 15.3 685 29 4.44 23.5 14.9 640 30 4.57 23.1 16.2 650 31 4.69 23.3 15.9 650 32 4.80 23.8 15.1 650 33 4.94 23.9 16.0 619 34 5.05 23.2 15.6 609 35 5.20 23.7 14.9 629 36 5.32 24.0 14.4 629 37 5.44 23.8 14.5 673 38 5.57 23.9 15.0 673 39 5.70 23.7 14.4 662 40 5.82 23.7 15.0 662 41 6.07 23.4 14.9 650 42 6.22 24.6 14.5 650 43 6.35 24.2 14.8 650 44 6.49 23.7 15.4 650 45 6.60 23.2 14.4 640 46 6.74 23.4 14.3 640 47 6.87 23.8 14.5 662 48 6.99 23.6 14.4 650 158 49 7.12 24.0 15.0 - 673 50 7.25 24.1 14.0 673 51 7.39 23.8 14.1 698 52 7.50 24.0 14.0 673 53 7.64 23.5 . 15.3 711 54 7.75 23.9 14.2 711 55 7.90 23.4 15.4 724 56 8.02 23.8 16.2 698 57 8.20 23.3 14.0 662 58 8.29 23.9 15.0 698 59 8.44 23.3 14.6 673 60 8.57 23.9 14.7 698 61 8.70 24.1 16.5 685 62 8.82 23.4 14.7 685 63 8.97 23.2 14.6 698 64 9.09 23.0 15.9 662 65 9.24 22.7 14.5 685 66 9.37 23.0 15.0 685 67 9.50 22.9 14.7 698 Ave 23.2 14.9 696 Std 0.7 0.6 49 Var 0.6 0.4 2426 Min 21.6 13.9 609 Max 24.6 16.5 817 159 Table ANS Experimental results of an automatic control test on a crossflow grain dryer Test Number : 5 Date : 12/19/1985 Dryer Type : Meyer-Morton Model 850 Set Point : 15%(w.b.) Sample # Time Inlet MC Outlet MC R.P.M. hr % w.b. % w.b. l 0.00 21.1 13.2 752 2 0.11 21.1 14.0 768 3 0.25 21.1 13.5 768 4 0.38 21.3 13.4 768 5 0.51 20.9 14.7 783 6 0.65 21.0 13.5 783 7 0.76 20.6 14.5 783 8 0.93 21.0 14.3 768 9 1.05 21.3 14.4 783 10 1.16 21.2 14.2 783 11 1.30 20.6 14.7 783 12 1.41 21.3 14.0 783 13 1.53 21.3 14.5 768 14 1.65 21.1 15.1 783 15 1.78 21.1 14.6 768 16 1.93 21.0 15.2 783 17 2.06 21.2 13.7 768 18 2.18 20.6 14.3 783 19 2.30 20.7 13.1 800 20 2.41 20.7 15.0 817 21 2.53 21.3 14.4 872 22 2.66 21.3 14.9 872 23 2.78 21.3 14.9 817 24 2.90 21.2 14.5 872 25 3.01 20.8 13.9 817 26 3.15 20.7 15.4 817 27 3.26 21.1 14.5 893 28 3.38 21.3 14.6 893 29 3.51 21.2 14.9 893 30 3.63 20.7 14.9 893 31 3.75 20.9 14.9 872 32 3.88 20.0 14.6 893 33 4.01 21.0 15.1 893 34 4.13 20.9 15.0 893 35 4.30 7 20.5 14.8 893 36 4.41 20.7 14.2 872 37 4.53 20.3 14.6 914 38 4.66 20.2 14.5 914 39 4.83 20.6 15.5 960 40 4.96 20.5 15.0 960 41 5.08 20.5 15.5 914 42 5.21 20.6 15.2 936 43 5.33 20.4 15.2 893 44 5.45 20.1 14.3 872 45 5.58 20.2 15.0 914 46 5.70 20.2 15.2 914 47 5.83 20.2 15.2 914 48 5 3 15.1 914 .95 - 20. 160 49 6.06 20.5 15.1 914 50 6.20 20.2 15.0 914 51 6.31 20.1 15.1 893 52 6.45 19.8 15.6 893 53 6.56 20.3 15.2 893 54 6.68 20.3 15.7 936 55 6.81 20.4 15.2 853 56 6.93 20.4 15.1 853 57 7.06 20.2 14.9 853 58 7.20 20.6 15.8 834 59 7.33 20.8 14.9 893 60 7.45 20.2 14.9 872 61 ‘7.58 20.5 15.0 834 62 7.71 20.5 14.9 817 63 7.90 20.4 14.8 834 64 8.01 20.2 15.1 834 65 8.15 20.2 14.9 872 66 8.26 20.1 14.8 817 67 8.40 20.4 15.0 834 _68 8.53 20.4 15.3 872 69 8.66 20.4 15.7 872 70 8.80 20.4 14.8 834 71 8.91 20.4 14.8 800 72 9.05, 20.4 15.4 800 73 9.18 20.3 15.0 834 74 9.35 20.0 15.0 834 75 9.55 20.2 14.7 800 Ave 20.6 14.8 849 Std 0.4 0.6 55 Var 0.2 0.3 2996 Min 19.8 13.1 752 Max 21.3 15.8 960 161 Table A16 Experimental results of an automatic control test on a crossflow grain dryer. Test Number : 6 Date : 12/20/1985 Dryer Type : Meyer-Morton Model 850 Set Point : 14.5%(w.b.) Sample # Time Inlet Mc Outlet MC R.P.M. hr % w.b. % w.b. 1 0.00 21.1 14.5 872 2 0.14 20.2 15.0 872 3 0.30 20.6 15.0 872 4 0.52 20.1 14.5 872 5 0.69 21.0 15.3 872 6 0.87 20.8 15.2 914 7 1.00 20.7 16.2 872 8 1.14 20.4 15.6 893 9 1.92 20.7 15.8 872 10 2.04 21.2 14.9 853 11 2.15 21.1 15.9 853 12 2.30 21.2 15.3 834 13 2.44 20.8 15.7 817 14 2.59 21.0 15.7 817 15 2.70 20.9 16.0 817 16 2.95 21.6 15.4 853 17 3.09 20.6 16.2 783 18 3.22 20.8 15.6 783 19 3.34 20.8 15.7 .768 20 3.49 20.6 16.2 800 21 3.60 20.7 15.0 738 22 3.74 20.5 16.0 738 23 3.87 20.8 14.4 724 24 4.00 21.0 15.9 738- 25 4.14 20.8 14.6 738 26 4.25 21.2 15.5 738’ 27 4.37 20.8 14.8 738 28 4.50 21.1 15.3 738 29 4.64 21.1 14.8 738 30 4.75 21.0 14.5 752 31 4.89 21.2 15.2 724 32 5.05 20.5 14.5 738 33 5.17 21.2 14.5 768 34 5.29 21.1 14.8 738 35 5.42 21.1 14.2 738 36 5.54 21.2 14.5 738 37 5.67 21.5 16.0 752 38 5.80 21.1 14.6 768 39 5.92 21.3 15.6 724 40 6.05 ‘ 21.2 13.8 724 41 6.17 20.8 14.9 711 42 6.32 20.9 15.3 724 43 6.47 20.8 16.0 752 44 6.59 21.0 14.6 752 45 6.74 21.0 14.7 724 46 6.87 21.0 15.3 724 47 7.00 21.1 14.5 768 48 7.17 20.7 15.0 724 162 49 7.30 20.8 15.0 738 50 7.44 21.4 14.6 783 51 7.57 21.9 14.1 752 52 7.70 21.2 15.3 752 53 7.84 21.1 14.2 800 54 7.95 21.4 15.2 752 55 8.07 21.6 14.4 752 56 8.20 21.3 15.1 752 57 8.32 21.3 14.6 738 58 8.45 21.2 14.2 768 59 8.57 21.7 15.5 738 60 8.69 21.6 14.2 752 61 8.82 21.0 15.2 738 62 8.97 20.7 14.9 738 63 9.09 20.7 15.6 752 64 9.24 20.7 15.3 783 65 9.37 21.3 14.9 752 66 9.50 20.5 15.6 752 67 9.62 20.8 14.4 768 68 9.75 21.4 15.5 752 69 9.89 21.1 14.5 783 70 10.00 21.2 15.3 752 71 10.14 20.9 14.3 800 72 10.25 20.7 15.4 768 73 10.40 21.3 15.5 817 74 10.59 21.2 14.5 783 75 10.72 21.2 14.8 752 76 10.84 21.0 14.5 752 77 10.97 20.7 14.9 768 78 11.12 21.3 15.0 768 Ave 21.0 15.1 775 Std 0.3 0.6 49 Var 0.1 0.3 2407 Min 20.1 13.8 711 Max 21.9 16.2 914 163 'Table A.7 Experimental results of an automatic control test on a crossflow grain dryer Test Number : 7 Date : 10/9/1986 Dryer Type : Meyer-Morton Model 850 Set Point : 14%(w.b.) Sample # Time Inlet MC Outlet MC R.P.M. . hr % w.b. % w.b. ‘ 1 0.00 25.9 12.4 518 2 0.15 25.2 13.1 533 3 0.28 25.1 13.0 640 4 0.41 24.9 11.1 650 5 0.55 24.5 11.6 711 6 0.70 25.0 12.3 752 7 1.23 24.4 12.9 619 8 1.38 24.8 14.6 619 9 1.51 24.4 13.1 724 10 1.65 25.0 13.7 685 11 1.80 23.8 13.2 711 12 1.93 24.2 12.8 698 13 2.08 27.0 14.0 768 14 2.21 26.2 13.9 768 15 2.35 27.3 14.8 698 16 2.50 28.2 14.4 711 17 2.65 27.9 14.1 619 18 2.80 28.1 13.5 600 19 2.95 26.3 15.0 619 20 3.10 28.7 13.5 650 21 3.23 27.0 15.0 581 22 3.36 27.1 14.9 590 23 3.51 26.7 15.8 540 24 3.65 26.0 14.9 533 25 3.80 27.7 15.8 511 26 3.93 27.5 15.6 511 27 4.08 31.0 16.1 441 28 4.21 28.4 16.3 426 29 4.63 29.4 15.7 498 30 4.78 27.3 16.0 451 31 4.91 27.7 15.6 474 32 5.20 26.3 16.3 462 33 5.35 31.0 14.8 462 34 5.48 27.0 15.8 457 35 5.61 27.1 13.8 436 36 5.76 28.9 14.9 441 37 5.91 28.7 14.5 457 38 6.05 29.9 13.9 468 39 6.20 28.9 13.7 446 40 6.33 26.9 13.6 441 41 6.48 726.1 13.2 457 42 6.63 - 30.2 13.5 462 43 6.75 27.1 13.5 468 44 6.90 27.6 13.0 486 45 7.03 25.9 13.2 511 46 7.20 29.2 12.5 548 47 7.38 29.6 13.6 548 48 7.51 27.1 13.7 548 49‘ 164 7.66 25.4 13.5 556 50 7.80 27.5 14.2 548 51 7.93 26.7 13.5 526 52 8.08 24.5 13.7 526 53 8.21 24.7 14.3 581 54 8.36' 24.3 13.3 581 55 8.50 24.9 12.1 650 56 8.66 24.2 14.6 619 Ave 26.8 14.0 563 Std 1.8 1.2 98 Var 3.4 1.4 9509 Min 23.8 11.1 426 Max 31.0 16.3 768 165 Table A.8 Experimental results of an automatic control test on a crossflow grain dryer. Test Number : 8 Date : 10/29/1986 Dryer Type : Meyer-Morton Model 850 Set Point : 14.5%(w.b.) ' Sample # Time Inlet MC Outlet MC R.P.M. hr % w.b. % w.b. 1 0.00 34.3 14.0 724 2 0.13 32.8 14.4 650 3 0.26 31.8 14.6 564 4 0.41 23.7 14.8 548 5 0.56 23.3 14.9 548 6 0.70 29.0 15.1 548 7 0.85 30.8 15.3 548 8 0.98 31.5 16.2 548 9 1.11 32.0 16.6 556 10 1.26 31.2 15.7 548 11 1.40 31.2 16.4 556 12 1.53 30.6 16.0 619 13 1.68 24.4 14.7 600 14 1.81 25.6 13.8 711 15 1.95 20.9 16.3 768 16 2.08 20.5 15.8 783 17 2.23 20.6 15.6 834 18 2.36 29.0 16.2 834 19 2.50 23.4 15.6 711 20 2.63 23.0 14.2 711 21 2.78 21.2 15.7 817 22 2.91 27.6 16.6 800 23 3.05 28.5 16.6 936 24 3.18 29.5 15.7 914 25 3.36 28.5 14.7 752 26 3.53 22.9 15.7 724 27 3.68 20.8 14.1 573 28 3.81 20.5 14.5 573 29 3.95 20.6 14.9 738 30 4.08 20.5 17.1 752 31 4.23 '20.7 14.6 783 32 4.36 22.6 15.4 783 33 4.50 21.3 16.3 724 34 4.65 20.6 16.7 724 35 4.78 20.2 16.8 893 36 4.91 20.3 16.1 893 37 5.05 20.6 13.8 1037 38 5.23 20.6 14.2 1010 39 5.38 20.5 13.9 834 40 5.55 21.4 13.6 800 41 5.70 20.7 14.3 936 42 5.83 20.3 13.9 936 43 5.96 20.8 15.3 984 44 6.11 20.6 13.4 984 45 6.25 21.1 14.1 1066 46 6.38 20.5 15.0 1037 47 6.53 20.5 15.4 1037 48 6.66 20.9 14.9 1066 166 49 6.81 20.8 14.3 984 50 6.98 20.4 14.4 984 51 7.15 20.5 15.7 1037 52 7.28 20.4 15.0 1037 53 7.43 20.5 14.4 1066 54 7.60 20.3 15.8 1066 55 7.76 20.1 14.1 1097 56 7.90 20.1 14.1 1097 57 8.03 19.9 14.1 1129 58 8.18 20.3 14.7 1129 59 8.35 20.5 15.1 1129 60 8.50 20.7 15.6 1129 61 8.63 19.8 15.7 1066 62 8.80 19.5 13.7 1097 63 8.96 20.1 15.6 1200 64 9.10 20.1 15.3 1129 65 9.23 20.0 14.8 1200 66 9.38 20.2 14.1 1129 67 9.55 20.6 14.9 1224 68 9.70 20.6 15.7 1163 69 9.86 20.6 16.3 1200 70 10.00 20.7 15.8 1200 71 10.16 20.7 16.3 1129 Ave 23.0 15.1 882 Std 4.1 0.9 213 Var 17.1 0.8 45200 Min 19.5 13.4 548 Max 34.3 17.1 1224 167 Table A.9 Experimental results of an automatic control test on a crossflow dryer. Test Number : 9 Date : 11/14/1986 Dryer Type : Meyer-Morton Model 850 Set Point : 14.5%(w.b.) Sample # Time Inlet MC Outlet Mc R.P.M. hr % w.b. % w.b. 1 0.00 19.6 12.5 936 2 0.16 19.3 12.8 960 3 0.36 19.3 13.4 936 4 0.53 19.4 14.0 952 5 0.73 19.4 14.6 960 6 0.90 19.1 14.4 1022 7 1.08 18.9 13.8 992 8 1.21 19.1 14.5 1010 9 1.38 19.2 13.5 1087 10 1.53 19.4 13.9 1028 11 1.70 19.0 14.5 1086 12 1.83 18.9 15.0 1066 13 1.98 19.5 14.4 1037 14 2.15 19.3 14.3 1107 15 2.28 19.5 15.2 1066 16 2.45 19.3 13.5 1066 17 2.58 19.1 13.7 1066 18 2.78 19.1 15.7 1066 19 2.95 19.4 13.7 1076 20 3.11 19.4 14.7 1056 21 3.25 19.4 14.9 1086 22 3.41 19.9 13.8 1119 23 3.58 19.5 14.8 1142 24 3.71 19.9 15.4 1086 25 3.85 18.9 15.4 1101 26 4.01 18.1 14.1 1077 27 4.18 18.2 15.1 1097 28 4.35 18.8 14.0 1086 29 4.48 19.7 13.9 1140 30 4.65 18.2 14.0 1151 31 4.81 18.7 14.4 1239 32 4.96 18.0 14.7 1163 33 5.13 18.6 15.3 1227 34 5.30 18.4 14.3 1151 35 5.45 18.7 14.1 1175 36 5.60 19.7 14.4 1163 37 5.76 19.4 14.8 1151 38 5.91 19.5 13.7 1163 39 6.06 18.9 14.7 1244 40 6.26 19.5 16.0 1164 41 6.41 18.6 13.8 1187 42 6.56 18.7 13.9 1175 43 6.73 19.5 13.7 1280 44 6.90 18.4 14.8 1310 45 7.05 18.6 16.0 1268 46 7.21 18.6 14.2 1280 47 7.41 19.0 14.3 1187 48 7.58 18.6 14.3 1200 168 49 7.73 20.5 14.5. 1253 50 7.88 21.0 15.3 1225 51 8.05 19.0 15.4 1200 52 8.20 19.8 14.9 1214 53 8.36 19.0 14.4 1238 54 8.50 18.7 16.7 1238 55 8.65 18.6 16.4 1200 56 8.81 20.4 17.5 1187 57 8.95 20.6 16.6 992 Ave 19.2 14.6 1125 Std 0.6 0.9 95 Var 0.4 0.9 8951 Min 18.0 12.5 936 Max 21.0 17.5 1310 169 Table A.10 Experimental results of an automatic control test on a crossflow grain dryer. Test Number : 10 Date : 11/28/1986 Dryer Type : Meyer-Morton Model 850 Set Point : 13.5%(w.b.) Sample # Time Inlet Mc Outlet MC R.P.M. hr % w.b. % w.b. 1 0.00 20.3 14.4 956 2 0.13 19.9 13.7 940 3 0.28 20.1 14.8 936 4 0.41 20.2 14.3 980 5 0.56 20.8 15.1 949 6 0.69 20.5 14.0 879 7 0.84 20.0 14.5 869 8 0.98 20.3 14.5 856 9 1.13 20.5 14.2 856 10 1.26 20.1 14.2 853 11 1.41 20.4 14.2 798 12 1.56 19.7 14.1 783 13 1.69 19.4 14.0 770 14 1.84 19.3 14.1 765 15 2.00 19.7 14.1 856 16 2.13 19.5 14.3 843 17 2.28 19.2 14.3 837 18 2.41 20.4 13.6 850 19 2.56 19.6 13.9 819 20 2.69 19.5 13.4 833 21 2.84 19.4 13.5 863 22 3.00 19.5 13.6 882 23 3.13 19.2 13.4 891 24 3.28 20.1 13.8 872 25 3.41 19.3 14.2 878 26 3.56 18.9 13.2 866 27 3.69 19.5 14.1 934 28 3.84 19.5 13.8 897 29 3.98 19.6 14.1 875 30 4.13 19.5 13.9 882 31 4.28 20.4 14.2 900 32 4.41 20.3 13.9 876 33 4.56 19.6 13.9 876 34 4.71 20.2 14.0 669 35 4.84 20.2 14.0 805 36 5.00 20.0 14.1 805 37 5.15 20.2 13.3 766 38 5.30 20.1 14.2 805 39 5.43 20.3 14.4 755 40 5.58 20.2 14.6 733 41 5.73 20.3 13.8 695 42 5.88 20.1 14.2 720 43 6.02 20.2 14.6 689 44 6.17 20.1 13.6 788 45 6.32 20.3 14.6 622 46 6.47 20.1 14.1 652 47 6.60 20.1 14.0 567 48 6.73 . 20.0 13.8 571 ' 106 HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH U!91>bbl—‘L‘wwwwwwwNNNNNNNHHHHHHHOOOOOOO I-‘H mu: \OQOOOOmmmmmmmNVVNVVVO‘ 20. 20. 19. 19. 19. 20. 19. 19. 19. 20. 20. 20. 20. 19. 20. 20. 20. 20. 17. 20. 20. 20. 20. 19. 19. 20. 20. 20. 20. 20. 19. 19. 19. 20. 19. 19. '19. 19. 19. 20. 19. 20. 20. 20. 20. 20. 20. 20. 20. 20. 20. 19. 19. ‘OO‘QFU‘NPU‘NU‘IU‘OUImoNO‘OQO‘WOU’NO‘MNWHWJ-‘QVWWI-‘9OHN90‘NQONQNM\OO\NW\J\I\IHN 170 537 538 519 519 554 571 554 587 681 774 712 687 726 727 695 746 731 715 735 725 713 722 794 775 728 738 822 745 742 731 713 755 715 717 715 755 797 775 765 749 754 759 768 778 785 780 778 742 704 715 726 728 691 702 681 818 717 687 171 107 15.47 19.8 13.0 743 108 15.62 19.7 13.0 698 109 15.77 18.6 12.8 715 110 15.90 19.5 12.7 848 111 16.05 19.4 13.5 802 ' 112 16.20 20.3 13.3 802 113 16.35 20.5 13.7 800 114 16.50 20.0 13.7 786 115 16.65 19.9 13.7 745 116 16.80 20.1 13.7 747 117 16.95 19.9 13.7 752 118 17.08 19.8 13.7 749 119 17.23 20.0 13.4 762 120 17.38 20.2 13.8 852 121 17.53 20.2 14.3 817 122 17.68 20.6 14.1 735 123 17.83 20.0 13.9 741 124 17.98 18.8 13.9 721 125 18.13 19.7 13.2 754 126 18.41 19.2 13.6 757 127 18.57 20.0 14.1 863 128 18.72 19.5 14.1 870 129 18.87 19.5 13.7 894 130 19.02 19.5 13.7 819 131 19.17 19.6 14.3 843 Ave 19.9 13.7 766 Std 0.5 0.5 93 Var 0.3 0.3 8727 Min 17.0 12.2 519 Max 9 15.1 980 20. 172 Table A.11 Experimental results of an automatic control test on a crossflow grain dryer. Test Number : 11 Date : 12/15/1986 Dryer Type : Zimmerman Model ATP 5000 Set Point : 17%(w.b.) SAMPLE# Time Inlet MC Outlet MC R.P.M. ‘hr % w.b. % w.b. 1 0.00 20.6 16.7 1326 2 0.11 21.2 16.4 1326 3 0.19 20.2 17.9 1330 4 0.29 19.8 17.2 1330 5 0.37 20.5 17.6 1329 6 0.47 20.5 17.3 1329 7 0.55 21.0 17.3 1320 8 0.63 20.9 16.5 1320 9 0.73 20.9 16.6 1321 10 0.81 21.4 16.6 1321 11 0.92 20.9 16.8 1328 12 1.00 21.1 16.6 1328 13 1.11 21.2 17.0 1330 14 1.19 21.0 16.9 1330 15 1.27 20.6 16.3 1349 16 1.37 21.1 16.6 1349 17 1.69 20.9 16.2 1330 18 1.77 20.1 16.6 1330 19 1.87 20.1 16.4 1445 20 1.97 20.5 16.3 1445 21 2.07 20.5 16.6 1662 22 2.32 20.3 16.5 1662 23 2.42 20.7 17.2 1494 24 2.50 20.8 17.2 1494 25 2.58 20.5 17.1 1397 26 2.66 20.1 17.2 1397 27 2.79 20.1 17.6 1654 28 2.87 20.9 17.2 1654 29 2.95 20.7 17.5 1661 30 3.05 20.2 17.7 1661 31 3.13 22.4 17.2 1523 32 3.23 22.3 17.3 1523 33 3.32 22.5 17.5 1309 34 3.42 21.6 18.1 . 1309 35 3.50 22.1 17.6 1120 36 3.58 21.0 17.7 1120 37 3.68 21.8 17.5 973 38 3.76 21.7 17.5 973 39 3.87 21.9' 17.2 860 40 3.95 22.3 17.2 860 41 4.06 22.2 17.0 860 42 4.14 20.7 17.0 860 43 4.24 22.3 17.2 895 44 4.32 21.5 17.5 895 45 4.40 21.6 17.1 773 46 4.50 22.1 17.3 773 47 4.58 21.2 17.0 690 48 4.68 22.3 16.6 690 173 49 4.77 21.1 16.5 639 50 4.87 21.5 16.5 639 51 4.95 21.5 17.0 1331 52 5.04 22.6 16.8 1331 53 5.14 22.1 16.6 1325 54 5.22 22.3 16.6 1325 55 5.32 21.8 16.3 1325 56 5.40 22.8 16.0 1325 57 5.50 22.6 16.4 1325' 58 5.58 22.7 16.3 1325 59 5.68 21.9 16.6 1326 60 5.77 22.1 16.5 1326 61 5.87 21.2 16.5 1326 62 5.95 22.2 16.6 1326 63 - 6.04 26.1 16.5 1438 64 6.14 26.8 16.4 1438 65 6.22 25.1 16.5 1089 66 6.32 23.3 17.1 1089 67 6.42 21.6 17.1 1045 68 6.50 26.2 17.8 1045 69 6.60 25.7 17.5 891 70 6.68 27.7 17.7 891 71 6.77 24.3 17.4 699 72 6.87 23.7 17.9 699 73 6.95 26.4 17.2 662 74 7.04 24.5 17.2 662 75 7.12 19.1 17.3 615 76 7.22 21.3 17.5 615 77 7.30 21.3 17.1 726 78 7.40 23.3 17.5 726 79 7.48 23.1 17.7 650 80 7.58 23.5 17.9 650 81 7.66 22.6 17.1 495 82 7.77 22.9 16.9 495 83 7.87 24.1 17.4 569 84 7.95 21.6 16.9 569 85 8.04 21.0 16.3 569 86 8.14 27.2 16.8 569 87 8.22 24.5 16.8 563 88 8.30 26.5 16.2 563 89 8.40 24.6 16.6 501 90 8.48 24.1 16.7 501 91 8.58 22.6 16.4 481 92 8.66 23.8 16.0 481 93 8.77 24.7 16.7 501 94 8.87 21.9 15.5 501 95 9.19 31.5 15.6 1330 96 9.27 21.5 15.5 1330 97 9.35 21.7 15.2 1333 98 9.45 22.1 15.1 1333 99 9.53 22.6 14.4 1374 100 9.63 21.4 14.3 1374 101 9.71 21.8 14.5 1441 102 9.81 21.7 14.8 1441 103 A 9.90 21.4 14.6 1450 104 9.98 22.9 14.6 1450 105 10.08 22.5 14.7 1460 106 10.16 21.4 15.0 1460 174 107 10.26 22.7 15.2 1462 108 10.34 21.0 15.5 1462 109 10.44 22.1 17.0 1393 110 10.52 21.4 15.8 1393 111 10.62 21.6 16.1 1329 112 10.70 21.7 16.5 1329 113 10.78 22.4 16.3 1335 114 10.90 21.8 17.6 1335 115 10.98 21.5 17.9 1328 116 11.08 21.9 18.6 1328 117 11.16 21.9 17.8 1303 118 11.26 21.1 17.9 1303 119 11.34 22.0 17.9 1273 120 11.42 21.4 17.5 1273 121 11.52 21.9 18.1 1277 122 11.60 22.4 17.5 1277 123 11.77 21.3 17.3 1277 124 11.87 21.0 17.3 1277 125 11.95 21.5 17.3 1277 126 12.05 21.0 17.1 1277 127 12.13 21.9 17.3 1277 128 12.23 20.6 17.2 1277 129 12.31 21.4 17.5 1253 130 12.41 21.5 16.9 1253 131 12.49 21.3 16.6 1276 132 12.57 23.0 17.2 1276 133 12.67 21.5 17.3 1414 134 12.75 20.6 17.0 1414 135 12.87 20.7 17.3 1404 136 12.95 21.5 17.1 1404 137 13.05 22.0 17.3 1393 138 13.13 22.2 17.1 1393 139 13.21 22.3 16.9 1327 140 13.31 22.8 17.2 1327 141 13.39 23.3 17.4 1346 142 13.49 23.4 17.2 1346 143 13.57 24.1 17.5 1074 144 13.67 23.6 16.9 1074 145 13.75 23.2 18.0 951 146 13.87 20.7 17.3 951 147 13.95 20.8 17.8 868 148 14.05 20.8 17.7 868 149 14.13 20.9 17.9 878 150 14.23 20.7 17.4 878 151 14.31 20.4 17.4 927 152 14.39 20.7 17.8 927 153 14.61 21.5 16.7 984 154 14.69 21.3 16.6 984 155 14.90 22.0 17.1 984 Ave 22.1 16.9 1148 Std 1.7 0.8 321 Var 2.9 0.7 103054 Min 19.1 14.3 481 Max 31.5 18.6 1662 175 Table A.12 Experimental results of an automatic control test on a crossflow grain dryer. Test Number : 12 ‘Date : 12/16/1986 Dryer Type : Zimmerman ATP 5000 Set Point : 17.5%(w.b.) SAMPLE# Time Inlet MC Outlet MC R.P.M. hr % w.b. % w.b. 1 0.00 20.1 16.2 1265 2 0.10 20.5 16.2 1265 3 0.18 20.5 15.7 1320 4 0.28 20.9 15.7 1320 5 0.36 20.8 15.8 1323 6 0.44 20.9 15.5 1323 7 0.54 21.0 15.7 1328 8 0.62 21.4 15.6 1328 9 0.72 21.8 15.6 1331 10 0.80 21.4 15.5 1331 11 0.90 21.8 14.9 1334 12 0.98 20.2 15.0 1334 13 1.06 21.4 15.8 1319 14 1.18 21.3 16.1 1319 15 1.26 20.1 15.4 1326 16 1.36 21.1 15.2 1326 17 1.44 21.2 15.6 1318 18 1.54 20.8 16.1 1318 19 1.62 21.4 16.2 1323 20 1.72 20.2 15.9 1323 21 1.80 20.8 16.9 1318 22 1.88 21.5 16.7 1318 23 1.98 19.9 17.2 1326 24 2.06 20.9 16.9 1326 25 2.16 20.2 16.9 1329 26 2.24 21.0 16.8 1329 27 2.34 21.4 17.1 1729 28 . 2.42 21.1 16.6 1729 29 2.60 21.9 16.6 1736 30 2.70 21.8 16.9 1736 31 2.78 20.9 16.3 1686 32 2.86 21.4 16.3 1686 33 2.94 22.1 16.5 1738 34 3.04 21.1 16.7 1738 35 3.18 22.4 16.8 1740 36 3.28 22.0 16.9 1740 37 3.36 21.3 17.2 1712 38 3.46 21.0 17.2 1712 39 3.54 20.6 17.3‘ 1620 40 3.64 20.8 18.0 1620 41 3.72 21.2 18.2 1717 42 3.80 21.1 17.5 1717 43 3.87 21.0 17.4 1673 44 3.95 21.5 17.4 1673 45 4.05 21.3 18.2 1739 46 4.18 21.2 18.3 1739 47 4.28 21.6 18.4 1718 48 4.36 21.3 17.8 1718 49 50 51 52 53 54 55 56' 57 58 59 60 61 62 63 64 65 66 67 68 70 71 72 73 74 75 76 77 78 79 80 82 83 84 85 86 87 88 89 90 91 -92 93 94 95 96 97 98 99 100 101 - 102 103 104 105 106 \Owwxoosooxoxooooooooooooooooooooooooxlsl\txlxlxlxlxlwxlo‘oxoxoxo‘o‘oxmofloxmmwmmmmwmmbbbbbbb Ull-‘ifiC39waNWJ-\GDUN\D\OJ>NUI\10«P\OHCDwaNNOND-‘HwoopNU‘Ot-‘QHAO\OmNmNJ-‘NCDH‘PU'OC 176 NOWUNQL‘O‘FIOZPbwaQHmNHme\lpmomNbO‘mOO‘ONOl-‘OONUIQNmO‘NUJCDQO‘OOUINNNQ 177 107 9.95 22.0 17.5 1011 108 10.15 22.1 17.9 1011 109 10.23 23.0 17.4 965 110 10.33_ 22.4 17.7 965 111 10.41 22.9 17.3 829 112 10.51 22.7 17.8 829 113 10.59 22.6 16.6 912 114 10.69 22.3 17.0 912 115 10.77 22.7 16.9 1029 116 10.85 22.9 16.6 1029 117 10.95 23.2 16.6 838 118‘ 11.03 22.7 17.0 838 119 11.15 23.5 16.4 880 120 11.23 22.4 16.8 880 121 11.33 22.6 16.6 811 122 11.41 21.8 16.3 811 123 11.49 22.4 16.5 824 124 11.59 22.2 16.2 824 125 11.67 22.7 16.2 889 126 11.77 22.7 16.6 889 127 11.85 22.7 16.0 885 128 11.95 22.2 16.0 885 129 12.03 22.5 15.7 981 130 12.15 22.5 15.7 981 131 12.23 23.2 15.9 1157 132 12.31 22.7 16.7 1157 133 12.41 23.4 15.9 1200 134 12.49 22.3 15.9 ‘1200 135 12.59 21.9 16.4 1342 136 12.67 21.6 16.3 1342 137 12.77 21.9 16.3 1342 Ave 21.4 17.0 1390 Std 0.8 0.8 308 Var 0.6 0.7 94790 Min 19.9 14.9 811 Max 23.5 18.5 1749 178 Table A.13 Experimental result of an automatic control test: on a crossflow grain dryer. Test Number : l3 Date : 12/19/1986 Dryer Type : Zimmerman Model ATP 5000 Set Point : 17.5%(w.b.) SAMPLE# Time Inlet MC Outlet MC R.P.M. hr % w.b. % w.b. 2 0.00 23.2 18.1 1307 3 0.08 23.1 18.1 1307 4 0.18 20.4 17.9 1316 5 0.26 20.9 17.1 1316 6 0.36 21.6 17.3 1325 7 0.44 22.2 17.5 1325 8 0.52 22.2 17.1 1334 9 0.62 21.4 16.9 1334 10 0.71 20.6 16.6 1428 11 0.81 20.4 16.6 1428 12 0.89 22.9 16.9 1526 13 0.99 22.2 16.8 1526 14 1.07 21.8 17.0 1532 15 1.16 23.3 16.0 1532 16 1.26 21.9 16.9 1516 17 1.34 23.1 15.9 1516 18 1.44 24.2 15.8 1520 19 1.52 24.3 16.6 1520 20 1.62 25.1 16.1 1627 21 1.71 22.8 16.3 1627 22 1.79 23.2 16.1 1506 23 1.89 23.7 17.3 1506 24 1.97 23.3 16.4 1463 25 2.07 23.3 16.8 1463 26 2.16 23.5 17.5 1474 27 2.26 22.1 16.3 1474 28 2.34 23.0 17.3 1037 29 2.42 21.7 16.9 1037 30 2.52 21.2 17.6 1205 31 2.60 20.2 17.3 1205 32 2.71 21.1 18.2 1617 33 2.81 23.6 17.8 1617 34 2.89 24.9 17.7 1399 35 2.97 21.1 20.0 1399 36 3.07 21.1 18.0 1298 37 3.16 20.8 17.8 1298 38 3.29 20.8 17.2 1728 39 3.39 20.7 17.4 1728 40 3.47 22.2 17.5 1730 41 3.55 22.0 16.4 1730 42 3.66 23.7 16.7 1727 43 3.74 22.8 16.5 1727 44 3.84 22.9 16.8 1730 45 3.92 22.2 17.0 1730 46 4.02 23.2 17.0 1729 47 4.10 24.4 17.7 1729 48 4.20 22.9 17.7 1664 49 4.28 _ 25.5 16.9 1664 H \J \O N~4~JO\O\G\O\m¢m$‘¢‘9 U'J-‘kkDmNNNJ-‘O‘mwl-‘OOVU‘ObQWHNU’mO‘HO‘womP-‘w-PwWQNNle—‘bNHr-‘Nwflflwbwkimtflop WQOLDWUICXDUIWWU!\OKOO’\U\N\ImNONU‘O‘waw-L‘ChwamONOWOGDHNI—‘QNNNHQONQQNL‘VONHNW ooxoxoxoxoxoxooxoxooooooooooooooooooomxlxlwxlwxlx: H 1734 1734 1278 1278 993 993 982 970 970 988 988 848 848 1742 1742 1739 1739 1744 1744 1745 1745 1746 1746 1744 1744 1744 1744 1107 1107 1532 1532 1517 1517 1517 1517 1741 1741 1741 1741 1740 1740 1738 1738 1736 1736 1737 1737 1661 1661 1372 1372 1223 1223 1199 1199 1253 1253 H m 0 108 10.05 20.0 17.6 1036 109 10.10 20.7 17.8 1036 110 10.17 20.5 17.4 1144 111 10.26 21.6 17.8 1144 112 10.34 20.3 17.6 1638 113 10.41 21.1 18.4 1638 114 10.48 21.5 17.5 1741 115 10.53 22.4 17.3 1741 116 10.59 21.4 16.8 1739 117 10.66 21.2 17.3 1739 118 10.73 20.8 17.0 1652 119 10.79 21.1 16.7 1652 120 10.86 21.5 17.6 1727 121 10.93 21.3 17.0 1727 122 11.00 20.6 16.8 1736 123 11.05 23.6 16.3 1736 124 11.12 21.5 17.1 1739 125 11.19 20.9 16.2 1739 126 11 33 20.7 16.3 1743 127 11.38 21.8 16.2 1743 128 11.45 19.7 16.5 1741 129 11 52 20.5 16.8 1741 130 11 59 20.5 17.1 1742 131 11.64 21.5 16.7 1742 132 11.69 21.8 17.0 1743 133 11 76 21.1 17.1 1743 134 11.83 21.5 16.3 1733 135 11.88 21.0 16.9 1733 Ave 21.7 17.0 1513 Std 1.5 0.8 260 Var 2.2 0.6 67638 Min 16.1 14.6 848 Max 25.5 20.0 1746' 181 Table A.14 Experimental results of an automatic control test on a crossflow grain dryer. Test Number : 14 Date : 1/15/1987 Dryer Type : Zimmerman Model ATP 50000 Set POint : 16.5%(w.b.) SAMPLE # Time Inlet MC Outlet Mc R.P.M. hr % w.b. % w.b. 1 0.00 19.5 18.9 1322 2 0.10 21.4 19.1 1324 3 0.18 21.1 18.8 1317 4 0.28 20.1 18.3 1320 5 0.36 20.9 18.9 1308 6 0.44 21.4 17.7 1297 7 0.52 21.4 18.2 1299 8 0.62 21.6 18.3 1300 9 0.70 21.5 18.0 1306 10 0.80 21.7 18.3 1303 11 0.92 22.4 16.9 1298 12 1.00 21.9 17.2 1304 13 1.10 22.1 17.2 1303 14 1.18 23.2 16.6 1312 15 1.28 23.2 16.5 1299 16 1.36 22.1 16.4 1297 17 1.46 21.7 15.9 1293 18 1.54 22.2 15.1 1297 19 1.64 22.1 15.1 1303 20 1.72 21.7 15.3 1302 21 1.80 22.7 15.6 1280 22 1.92 23.0 15.7 1279 23 2.02 22.4 14.7 1295 24 2.10 22.2 15.0 1293 25 2.20 22.4 15.9 1365 26 2.28 21.4 15.4 1366 27 2.38 22.3 15.4 1363 28 . 2.46 21.3 16.1 1361 29 2.54 22.9 16.5 1469 30 2.64 22.4 15.7 1470 31 2.72 21.9 16.5 1478 32 2.82 22.4 16.0 1478 33 2.92 22.1 16.7 1520 34 3.02 22.5 16.3 1520 35 3.10 21.9 16.6 1566 36 3.20 22.2 16.7 1564 37 3.28 22.3 16.7 1587 38 3.38 21.9 16.6 1588 39 3.46 22.1 16.6‘ 1547 40 3.54 22.1 16.6 1546 41 3.64 21.9 16.8 1642 42 3.72 22.4 17.3 1643 43 3.82 21.7 17.4 1620 44 3.92 21.8 17.3 1620 45 4.02 22.1 16.9 1494 46 4.10 22.1 17.5 1493 47 4.20 21.5 17.4 1490 48 4.28 22.3 17.0 1491 \oxoxoxoxoxoxooooooooooooooooooooooo-xlwxlxlxl\quwuwmoxoxoxoxoxcxmoxoxoxmmmuwmmmmmmm4-‘4-‘1-‘4‘kph UlmwO‘mHmONWOVHHbHNt-‘HmflO‘PNNHWOWU‘I—‘NomwNOO‘NU‘IUQOI—‘OL‘WNOvaJ-‘O‘LDHWWO 182 waflNWNmNmN‘I-‘OWWVNHVHmbmeONmmHQ.W\Jw4I-‘OwwN®O\HmUIm-PH\IJ>O\NJ>QQOHHO\ 1421 1422 1427 1428 1440 1440 1519 1517 1589 1588 1656 1656 1758 1759 1760 1759 1760 1760 1760' 1761 1753 1762 1761 1760 1764 1763 1762 1762 1765 1329 554 1041 1132 1132' 914 914 981 979 930 930 961 958 909 907 604 601 651 657 702 699 724 935 935 728 728‘ 749 755 1150 183 107 10.02 21.8 15.5 1152 108 10.12 20.6 15.8 692 109 10.20 21.4 15.9 696 110 10.30 22.1 15.5 802 111 10.38 21.3 15.4 805 112 10.47 21.4 15.3 848 113 10.57 22.1 15.0 845 114 10.65 20.3 15.0 967 115 10.75 22.4 15.5 965 116 10.83 21.9 14.6 1044 117 10.93 21.9 15.1 1043 118 11.02 23.6 15.2 1205 _ 119 11.12 21.7 14.6 1205 120 11.20 22.1 14.4 1227 121 11.30 22.0 14.7 1227 122 11.38 22.7 15.6 1374 123 11.47 22.5 14.9 1374 124 11.57 22.1 14.9 1344 125 11.65 22.3 15.0 1344 126 11.75 21.4 15.4 1382 127 11.83 22.6 15.8 1382 128 11.93 25.8 15.8 1462 129 12.02 22.1 15.6 1462 6 130 12.12 22.1 16.5 1481 131 12.20 26.3 16.0 1481 132 12.28 23.7 16.3 1272 133 12.38 22.5 16.6 1273 Ave 21.6 16.6 1294 Std 1.1 1.0 325 Var 1.2 1.0 105707 Min 17.9 14.4 554 Max 26.3 19.1 1765 184 AEPENDIX B ; Simulation Results Table B.1 Simulation results of an automatic control test on a crossflow grain dryer. Set Number : 1 (Test #7) Dryer Type : Meyer-Morton Model 850 Set Point : 14.0%(w.b.) Sample # Time Inlet MC Outlet MC RPM (hrs) (%,w.b.) (%,w.b) 1 0.14 25.9 14.3 716 2 0.28 25.2 14.3 716 3 0.41 25.1 14.3 708 4 0.55 24.9 14.3 708 5 0.69 24.5 14.3 729 6 0.83 25.0 14.4 729 7 0.97 24.4 14.4 739 8 1.10 24.8 14.4 739 9 1.24 24.4 14.5 746 10 1.38 25.0 14.5 746 11 1.52 23.8 14.6 741 12 1.66 24.2 14.1 741 13 1.79 27.0 14.1 773 14 1.93 26.2 14.0 773 15 2.07 27.3 13.6 676 16 2.21 28.2 13.9 676 17 2.35 27.9 13.3 641 18 2.48 28.1 13.5 641 19 2.62 26.3 13.0 636 20 2.76 28.7 13.3 636 21 2.90 27.0 12.3 652 22 3.04 27.1 12.4 652 23 3.17 26.7 14.6 673 24 3.31 26.0 13.8 673 25 3.45 27.7 14.6 684 26 3.59 27.5 15.3 684 27 3.73 31.0 15.0 626 28 3.86 28.4 15.1 626 29 4.00 29.4 13.5 553 30 4.14 27.3 15.4 553 31 4.28 27.7 13.9 591 32 4.42 26.3 13.9 591 33 4.55 31.0 13.5 639 34 4.69 27.0 12.9 639 35 4.83 27.1 14.2 580 36 4.97 28.9 13.9 580 37 5.11 28.7 16.8 607 38 5.24 29.9 14.4 607 39 5.38 28.9 15.1 554 40 5.52 26.9 13.3 554 41 5.66 26.1 13.6 598 42 5.80 30.2 12.5 598 H 00 U1 43 5.93 27.1 16.6 597 44 6.07 27.6 13.1 597 45 6.21 25.9 13.3 618 46 6.35 29.2 14.7 618 47 6.49 29.6 14.5 610 48 6.62 27.1 15.6 610 49 6.76 25.4 14.7 577 50 6.90 27.5 13.0 577 51 7.04 26.7 12.5 647 52 7.18 24.5 16.0 647 53 7.31 24.7 13.6 677 54 7.45 24.3 14.1 677 55 7.59 24.9 12.9 728 56 7.73 24.2 15.9 728 Ave 26.8 14.1 654 Std 1.8 1.0 62 Min 23.8 12.3 553 Max 31.0 16.8 773 186 Table B.2 Simulation results of an automatic control test on a crossflow grain dryer. Set Number 2 2 (Test #1) Dryer Type : Meyer-Morton Model 850 Set Point : 14.5 Sample # Time Inlet MC Outlet MC RPM (hrs) (%,w.b.) (%,w.b.) 1 0 14 21.7 14.5 963 2 0 28 21.1 14.5 963 3 0 41 21.2 14.5 978 4 0 55 21.2 14.6 978 5 0 69 21.5 14.6 989 6 0 83 21.3 14.7 989 7 0 97 21.5 14.2 977 8 1 10 21.8 14.3 977 9 1 24 21.6 14.3 967 10 1 38 21.5 14.5 967 11 1 52 22.1 14.3 973 12 1 66 21.7 14.5 973 13 1 79 22.4 14.7 954 14 1 93 22.4 14.5 954 15 2 07 21.9 14.3 927 16 2 21 22.1 14.7 927 17 2 35 22.1 14.3 946 18 2 48 21.7 14.9 946 19 2 62 21.5 14.8 949 20 2 76 22.0 14.4 949 21 2 90 22.8 14.6 956 22 3 04 22.5 14.6 956 23 3 17 22.1 14.2 909 24 3 31 22.4 14.0 909 25 3 45 23.0 14 4 932 26 3 59 22.0 15.1 932 27 3 73 22.1 14.7 916 28 3 86 21.8 14.3 916 29 4 00 22.6 14.6 943 30 4 14 23.5 15.1 943 31 4 28 22.3 14.2 885 32 4 42 23.5 14.2 885 33 4 55 22.5 13.9 895 34 4 69 22.4 14.5 895 35 4 83 22.8 15.3 917 36 4 97 23.1 14.2 917 37 5 11 23.4 15.2 890 38 5 24 23.2 14.3 890 39 5 38 23.1 14.2 871 40 5 52 23.5 14.5 871 41 5 66 22.9 14.7 872 42 5 80 22.3 14.9 872 43 5 93 21.9 14.7 902 44 6 07 20.8 14.6 902 45 6 21 22.9 15.1 967 46 6 35 23.6 14.8 967 47 6 49 23.5 14.3 869 48 6 62 22.1 13 9 869 49 6 76 23.0 13 0 897 187 50 6.90 22.5 14.8 - 897 51 7.04 23.0 15.5 902 52 7.18 22.6 15.2 902 53 7.31 22.5 13.8 893 54 7.45 22.3 14.6 893 55 7.59 23.0 14.3 914 56 7.73 22.8 14.8 914 57 7.87 22.4 14.4 890 58 8.00 22.6 14.3 . 890 59 8.14 22.3 14.1 911 60 8.28 23.3 14.8 911 61 8.42 22.4 14.6 895 62 8.56 22.3 14.2 895 63 8.69 22.6 14.4 919 64 8.83 23.1 14.21 919 65 8.97 22.3 15.1 895 66 9.11 21.7 14.3 895 67 9.25 22.2 14.3 936 68 9.38 21.9 14.6 936 69 9.52 21.6 15.1 935 70 9.66 21.0 14.4 935 71 9.80 21.1 14.1 974 72 9.94 21.2 14.7 974 73 10.07 21.6 14.6 986 74 10.21 22.1 14.5 986 75 10.35 21.2 13.9 951 76 10.49 21.6 14.1 951 77 10.63 22.3 14.2 980 78 10.76 21.9 14.5 980 79 10.90 22.1 14.9 943 80 11.04 21.8 14.0 943 81 11.18 22.2 14.3 950 Ave 22.2 14.5 929 Std 0.7 0.4 35 Min 20.8 13.0 869 Max 23.6 15.5 989 188 Table 8.3 Simulation results of an automatic control test on a crossflow grain dryer. Set Number : 3 Dryer Type : Zimmerman ATP 5000 Set Point : 17.5%(w.b.) ‘ Sample # Time Inlet MC Outlet MC RPM (hrs) (%,w.b.) (%,w.b.) 1 0.07 20.1 16.8 1750 2 0.14 20.5 16.8 1750 3 0.22 20.5 16.8 1750 4 0.29 20.9 16.7 1750 5 0.36 20.8 16.6 1750 6 0.43 20.9 16.6 1750 7 0.50 21.0 16.5 1750 8 0.58 21.4 16.5 1750 9 0.65 21.8 16.4 1750 10 0.72 21.4 16.3 1750 11 0.79 21.8 16.7 1750 12 0.86 20.2 16.6 1750 13 0.94 21.4 16.9 1750 . 14 1.01 21.3 16.7 1750 15 1.08 20.1 16.8 1750 16 1.15 21.1 16.9 1750 17 1.22 21.2 17.3 1750 18 1.30 20.8 17.7 1750 19 1.37 21.4 17.3 1750 20 1.44 20.2 17.7 1750 21 1.51 20.8 16.2 1750 22 1.58 21.5 17.3 1750 23 1.66 19.9 17.2 1750 24 1.73 20.9 16.1 1750 25 1.80 20.2 17.0 1750 26 1.87 21.0 17.1 1750 27 1.94 21.4 16.7 1750 28 2.02 21.1 17.3 1750 29 2.09 21.9 16.2 1750 30 2.16 21.8 16.7 1750 31 2.23 20.9 17.4 1679 32 2.30 21.4 15.9 1679 33 2.38 22.1 16.8 1750 34 2.45 21.1 16.2 1750 35 2.52 22.4 16.9 1750 36 2.59 22.0 17.3 1750 37 2.66 21.3 17.0 1549 38 2.74 21.0 17.7 1549 39 2.81 20.6 17.6 1750 40 2.88 20.8 16.8 1750 41 2.95 21.2 17.2 1750 42 3.02 21.1 17.9 1750 43 3.10 21.0 16.9 1750 44 3.17 21.5 18.2 1750 45 3.24 21.3 17.8 1750 46 3.31 21.2 17.2 1750 47 3.38 21.6 16.9 1750 48 3.46 21.3 16.5 1750 49 3.53 21.0 16.7 1750 \1\|\I\A\t\1\l\lwwmoxoxoxoxmmma‘o‘mcao‘mmmmwmmmuunmmmmmbbbkbbkbpbbhbbwwwwww onn¢~m<38~wlpt>blnl~mxgnao~oh~bim\4<>b\ohamunc~wron>ororaraoaO\o¢~NLnc>oner~ooamnnn>broa>H4>uwo P“ 00 \O OH4—‘\1NHmmPNmNmL‘mONWDwD-‘HOHOONmmJ-‘NU'IQOCDOOmNHbHNoonwQQH‘F-‘HNDWCH 1750 1750 1750 1632 1632 1628 1628 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1750 1534 1534 1470 190 -108 7.78 22.1 16.8 1470 109 7.85 23.0 17.0 1594 .110 7.92 22.4 17.0 1594 111 7.99 22.9 17.1 1388 112 8.06 22.7 17.5 1388 113‘ 8.14 22.6‘ 18.0 1350 114 8.21 22.3 17.8 1350 115 8.28 22.7 17.8 1427 116 8.35 22.9 17.3 1427 117 8.42 23.2 17.4 1325 118 8.50 22.7 18.1 1325 119 8.57 23.5 17.5 1277 120 8-64 22.4 18.0 1277 121 8.71 22.6 17.7 1267 122 8.78 21.8 17.6 1267 123 8.86 22.4 17.3 1463 124 8.93 22.2 17.7 1463 125 9.00 22.7 17.8 1430 126 9.07 22.7 18.1 1430 127 9.14 22.7 17.6 1311 128 9.22 22.2 18.4 1311 129 9.29 22.5 17.4 1372 130 9.36 22.5 17.5 1372 131 9.43 23.2 16.8 1363 132 9.50 22.7 17.4 1363 133 9.58 23.4 17.2 1256 134 9.65 22.3 17.6 1256 135 9.72 21.9 17.6 1277 136 9.79 21.6 17.6 1277 137 9.86 21.9 17.2 1607 Ave 21.4 17.1 1655 Std 0.8 0.5 162 Min 19.9 15.9 1256' Max 23.5 18.4 1750 191 Table 3.4 Simulation results of an automatic control test on a crossflow grain dryer. Set Number : 4 Dryer Type : Meyer-Morton Model 850 Set Point : 14.5%(w.b.) Sample # Time Inlet MC Outlet MC RPM (hrs) (%,w.b.) (%,w.b.) 1 0.14 .21.0 13.2 911 2 0.28 21.3 13.2 911 3 0.41 21.5 13.4 1001 4 0.55 21.8 13.6 1001 5 0.69 22.0 13.7 973 6 0.83 22.1 13.9 973 7 0.97 22.2 14.0 952 8 1.10 22.3 14.3 952 9 1.24 22.2 14.4 941 10 1.38 22.2 14.5 941 11 1.52 22.0 14.6 939 12 1.66 21.9 14.7 939 13 1.79 21.7 14.7 948 14 1.93 21.4 14.7 948 15 2.07 21.1 14.8 968 16 2.21 20.9 14.7 968 17 2.35 20.6 14.8 996 18 2.48 20.3 14.7 996 19 2.62 20.1 14.8 1028 20 2.76 19.9 14.8 1028 21 2.90 19.8 14.8 1057 22 3.04 19.7 14.7 1057 23 3.17 19.7 14.7 1076 24 3.31 19.8 14.6 1076 25 3.45 19.9 14.6 1082 26 3.59 20.0 14.5 1082 27 3.73 20.2 14.4 1073 28 3.86 20.5 14.3 1073 29 . 4.00 20.7 14.2 1052 30 4.14 21.0 14.2 1052 31 4.28 21.3 14.1 1024 32 4.42 21.5 14.2 1024 33 4.55 21.8 14.2 994 34 4.69 22.0 14.2 994 35 4.83 22.1 14.3 968 36 4.97 22.2 14.3 968 37 5.11 22.3 14.4 950 38 5.24 22.2 14.5 950 39 5.38 22.2 14.6 941 40 5.52 22.0 14.7 941 41 5.66 21.9 14.7 943 42 5.80 21.7 14.7 943 43 5.93 21.4 14.7 957 44 6.07 21.1 14.8 957 45 6.21 20.9 14.8 982 46 6.35 20.6 14.8 982 47 6.49 20.3 14.8 1012 48 6.62 20.1 14.8 1012 49 6.76 19.9 14.8 1043 192 50 6.90 19.8 14.7 - 1043 51 7.04 19.7 14.7 1068 52 7.18 19.7 14.7 1068 53 7.31 19.8 14.6 1081 54 7.45 19.9 . 14.5 1081 55 7.59 20.0 14.5 1079 56 7.73 20.2 14.4 1079 57 7.87 20.5 14.3 1063 58 8.00 20.7 14.2 1063 Ave 21.0 14.4 1004 Std 0.9 0.4 53 Min 19.7 13.2 911 Max 22.3 14.8 1082 193 Table 8.5 Simulation results of an automatic control test on a crossflow grain dryer. Set Number : 5 Dryer Type : Meyer-Morton Model 850 Set Point : 14.5%(w.b.) Sample # Time Inlet MC Outlet MC RPM (hrs) (%,w.b.) (%,w.b.) 1 O 14 21.0 13 2 911 2 0 28 21.5 13 2 911 3 0 41 22.0 13 3 993 4 0 55 22.2 13 6 993 5 0 69 22.2 13.6 950 6 0 83 22.0 13.7 950 7 0 97 21.7 13.8 947 8 1 10 21.1 14.4 947 9 1 24 20.6 14.7 984 10 1 38 20.1 15.0 984 11 1 52 19.8 15 2 1038 12 1 66 19.7 15 3 1038 13 1 79 19.9 15 3 1072 14 1 93 20.2 15.1 1072 15 2 07 20.7 14.8 1056 16 2 21 21.3 14.4 1056 17 2 35 21.8 14.0 1006 18 2 48 22.1 13 8 1006 19 2 62 22.3 13 7 963 20 2 76 22.2 13 8 963 21 2 90 21.9 14.0 951 22 3 04 21.4 14.3 951 23 3 17 20.9 14.7 976 24 3 31 20.3 14.9 976 25 3 45 19.9 15 2 1027 26 3 59 19.7 15.3 1027 27 3 73 19.8 15.3 1069 28 3 86 20.0 15 2 1069 29 4 00 20.5 15.0 1065 30 4 14 21.0 14.6 1065 31 4 28 21-5. 14.2 1019 32 4 42 22.0 13 9 1019 33 4 55 22.2 13 7 971 34 4 69 22.2 13 8 971 35 4 83 22.0 13.9 950 36 4 97 21.6 14.2 950 37 5 11 21.1 14.5 967 38 5 24 20.6 14.8 967 39 5 38 20.1 15 0 1013 40 5 52 19.8 15 2 1013 41 5 66 19.7 15 3 1062 42 5 80 19.9 15 3 1062 43 5 93 20.2 15.1 1071 44 6 07 20.7 14.8 1071 45 6 21 21.3 14.4 1032 46 6 35 21.8 14.0 1032 47 6 49 22.1 13 8 981 48 6 62 22.3 13 7 981 49 6 76 22.2 13 8 952 I“ \O b 50 6.90 21.9 14.0 952 51 7.04 21.4 14.4 959 52 7.18 20.9 14.6 959 53 7.31 20.3 15.0 1000 54 7.45 19.9 15.1 1000 55 7.59 19.7 15.3 1052 56 7.73 19.8 15.3 1052 57 7.87 20.0 15.2 1074 58 8.00 20.5 14.9 1074 Ave 21.0 14.5 1004 Std 0.9 0.6 47 Min 19.7 13.2 911 Max 22.3 15.3 1074 195 Table 3.6 Simulation results of an automatic control test on a crossflow dryer. Set Number : 6 Dryer Type : Meyer-Morton Model 850 Set Point : 14.5%(w.b.) Sample # Time Inlet MC Outlet MC RPM (hrs) (%,w.b.) (%,w.b.) 1 0 14 25.0 13 6 716 2 0 28 25.5 13 6 716 3 0 41 26.0 13 7 761 4 0 55 26.4 13 7 761 5 0 69 26.8 13 8 725 6 0 83 27.0 13.8 725 7 0 97 27.2 13 7 699 8 1 10 27.3 13 7 699 9 1 24 27.2 13 7 687 10 1 38 27.1 13 6 687 11 1 52 26.9 13.6 689 12 1 66 26.6 14.0 689 13 l 79 26.2 14.3 703 14 1 93 25.7 14.6 703 15 2 07 25.3 14.8 725 16 2 21 24.7 15 1 725 17 2 35 24.2 15 3 756 18 2 48 23.8 15 5 756 19 2 62 23.4 15 6 794 20 2 76 23.0 15 7 794 21 2 90 22.8 15 8 830 22 3 04 22.6 15 8 830 23 3 17 22.6 15 8 856 24 3 31 22.7 15 7 856 25 3 45 22.9 15 5 864A 26 3 59 23.2 15 3 864 27 3 73 23.6 15.0 852 28 3 86 24.0 14.7 852 29 4 00 24.5 14.3 826 30 4 14 25.0 14.0 826 31 4 28 25.5 13 7 793 32 4 42 26.0 13 4 793 33 4 55 26.4 13 3 762 34 4 69 26.8 13 2 762 35 4 83 27.0 13 1 735 36 4 97 27.2 13 2 735 37 5 11 27.3 13 3 715 38 5 24 27.2 13 6 715 39 5 38 27.1 13.8 705 40 5 52 26.9 14.0 705 41 5 66 26.6 14.3 706 42 5.80 26.2 14.6 706 43 5 93 25.7 14.9 718 44 6 07 25.2 15.1 718 45 6 21 24.7 15 3 742 46 6 35 24.2 15.5 742 47 6 49 23.8 15 7 776 48 6 62 23.3 15.7 776 49 6 76 23.0 15 9 813 l—‘ \O 0“ 50 6.90 22.8 15.9 813 51 7.04 22.6 15.9 845 52 7.18 22.6 15.8 845 53 7.31 22.7 15.7 863 54 7.45 22.9 15.5 863 55 7.59 23.2 15.3 860 56 7.73 23.6 15.0 860 57 7.87 24.0 14.7 840 58 8.00 24.5 14.3 840 Ave 25.0 14.6 771 Std 1.7 0.9 61 Min 22.6 13.1 687 Max 27.3 15.9 864 197 Table 3.7 Simulation results of an automatic control test on a crossflow dryer. Set Number 2 7 Dryer Type : Meyer-Morton Model 850 Set Point : 14.5%(w.b) Sample # Time Inlet MC Outlet MC RPM (hrs) (%,w.b.) (%,w.b.) l 0 14 25.0 13.6 716 2 O 28 26.0 13.6 716 3 0 41 26.8 13.6 716 4 0 55 27.2 13.6 716 5 0 69 27.2 13.6 716 6 0 83 26.9 13.6 716 7 0 97 26.2 13.6 716 8 1 10 25.3 13 6 716 9 1 24 24.2 13 6 685 10 1 38 23.4 13 5 685 11 1 52 22.8 13 9 895 12 1 66 22.6 15 1 895 13 1 79 22.9 16 1 854 14 1 93 23.6 16 8 854 15 2 07 24.5 17 O 817 16 2 21 25.5 16 9 817 17 2 35 26.4 16 4 756 18 2 48 27.0 15.6 756 19 2 62 27.3 14.6 705 20 2 76 27.1 13 5 705 21 2 90 26.6 12 8 690 22 3 04 25.7 12 4 690 23 3 17 24.7 12 5 724 24 3 31 23.8 12 8 724 25 3 45 23.0 13.5 797 26 3 59 22.6 14.4 797 27 3 73 22.7 15 3 865 28 3 86 23.2 16 2 865 29 4 00 24.0 16 7 857 30 4 14 25.0 17.0 857 31 4 28 26.0 16.6 782 32 4 42 26.8 16.0 782 33 4 55 27.2 15.0 717 34 4 69 27.2 14.1 717 35 4 83 26.9 13 2 692 ‘36 4 97 26.2 12 7 692 37 5 11 25.2 12.5 713 38 5 24 24.2 12.7 713 39 5 38 23.3 13.3 778 40 5 52 22.8 14.0 778 41 5 66 22.6 14.9 854 42 5 80 22.9 15 8 854 43 5 93 23.6 16 5 870 44 6 07 24.5 17.0 870 45 6 21 25.5 16 9 805 46 6 35 26.4 16 4 805 47 6 49 27.0 15.5 732 48 6 62 27.3 14.6 732 49 6 76 27.1 13 6 695 H \O 00 50 6.90 26.6 12.9 695 51 7.04 25.7 12.6 704 52 7.18 24.7 12.6 704 53 7.31 23.8 13.0 760 54 7.45 23.0 13.6 760 55 7.59 22.6 14.5 838 56 7.73 22.7 15.4 838 57 7.87 23.2 16.2 876 58 8.00 24.0 16.8 876 Ave 25.0 14.6 770 Std 1.7 1.5 67 Min 22.6 12.4 685 Max 27.3 17.0 895 199 Table 8.8 Simulation results of an automatic control test on a crossflow grain dryer. Set number : 8 Dryer Type : Meyer—Morton Model 850 Set Point : 15.5%(w.b.) Sample # Time Inlet MC Outlet MC RPM (hrs) (%,w.b.) (%,w.b.) 1 0.14 28.0 15.0 647 2 0.28 28.0 15.0 647 3 0.41 28.0 15.0 647 4 0.55 28.0 15.0 647 5 0.69 28.0 15.0 647 6 0.83 28.0 15.0 647 7 0.97 28.0 15.0 647 8 1.10 28.0 15.0 647 9 1.24 28.0 14.9 614 10 1.38 28.0 14.8 614 11 1.52 28.0 14.9 684 12 1.66 28.0 15.0 684 13 1.79 28.0 15.0 684 14 1.93 28.0 15.1 684 15 2.07 28.01 15.1 680 16 2.21' 28.0 15.2 680 17 2.35 28.0 15.3 678 18 2.48 . 28.0 15.3 678 19 2.62 28.0 15.3 678 20 2.76 28.0 15.5 678 21 2.90 28.0 15.5 677 22 3.04 24.0 15.5 677 23 3.17 24.0 15.7 758 24 3.31 24.0 15.8 758 25 3.45 24.0 16.2 870 26 3.59 24.0 16.6 870 27 3.73 24.0 17.1 863 28 3.86 24.0 17.5 863 29 4.00 24.0 17.9 856 30 4.14 24.0 18.3 856 31 4.28 24.0 18.5 850 32 4.42 '24.0 18.7 850 33 4.55 24.0 15.0 846 34 4.69 24.0 14.9 846 35 4.83 24.0 14.9 861 36 4.97 24.0 14.9 861 37 5.11 24.0 15.0 871 38 5.24 24.0 15.0 871 39 5.38 24.0 15.1 877 40 5.52 24.0 15.1 877 41 5.66 24.0 15.2 882 42 5.80 32.0 15.3 882 43 5.93 32.0 14.9 669 44 6.07 32.0 14.6 669 45 6.21 32.0 14.2 553 46 6.35 32.0 13.8 553 47 6.49 32.0 13.5 553 48 6.62 32.0 13.1 553 49 6.76 32.0 12.7 554 200 50 6.90 32.0 12.3 554 51 7.04 32.0 12.0 557 52 7.18 32.0 11.6 557 53 7.31 32.0 18.0 561 54 7.45 32.0 17.6 561 55 7.59 32.0 17.1 554 56 7.73 32.0 16.9 554 57 7.87 32.0 16.7 546 58 8.00 32.0 16.6 546 59 8.14 32.0 16.6 540 60 8.28 32.0 16.5 540 61 8.42 28.0 16.5 535 62 8.56 28.0 16.5 535 63 8.69 28.0 16.7 671 64 8.83 28.0 16.9 671 65 8.97 28.0 17.1 669 66 9.11 28.0 17.3 669 67 9.25 28.0 17.5 668 68 9.38 28.0 17.7 668 69 9.52 28.0 17.9 667 70 9.66 28.0 18.2 667 71 9.80 28.0 18.4 665 72 9.94 28.0 15.0 665 73 10.07 28.0 15.3 665 74 10.21 28.0 15.3 665 75 10.35 28.0 15.3 665 76 10.49 28.0 15.3 665 77 10.63 28.0 15.3 666 78 10 76 28.0 15.3 666 79 10.90 28.0 15.3 666 80 11.04 28.0 15.3 666 81 11.18 28.0 15.3 667 Ave 28.0 15.6 686 Std 2.8 1.4 108 Min 24.0 11.6 535 Max 32.0 18.7 882 APPENDIX C : 201 Sneoifioat‘ions for Data Acauisition (" ts C.1 AZD Converter Specifications Integrated Circuit: Resolution: Full Scale Voltage Maximum Conversion Time: Minimum Conversion Rate: Maximum Input Voltage: Input Impedance: Input Current: Temperature Coefficient: Overall Accuracy: Intersil 7109 dual-slope A/D converter 12 bits plus sign bit and over-range bit i0.5V, il.0V, i2.0V, or i4,0V, jumper selectable 50 milliseconds 20 samples per second i12V without damage minimum 8 megohms maximum 0.5 microamperes 100 ppm/degree C adjustable to better than 0.1% of full scale range Differential Nonlinearity (maximum deviation from ideal step size): i2 counts (0.5%) Integral Nonlinearity (maximum deviation from ideal straight line): i4 counts (0.1%) C.2 DZA Converter Specifications Integrated Circuit: Resolution: Full Scale Voltage: Maximum Conversion Time: Minimum Conversion Rate: Output Current: Nonlinearity: Accuracy: Monotonic: Temperature Coefficient: Software Interface: Analog Devices DACSO 12 bits i0.5V, i1.0V, 12.0V, or i4.0V, jumper selectable 20 microseconds up to 50,000 conversion per second, limited only by software speed. sources or sink 10ma t1 least significant bit Adjustable to better than 0.2% of full scale range over entire 0 to 70 degree C range 100 ppm/degree C via output of two data bytes; the most significant 4 bits are stored until the least significant 8 bits are output and then the 12 bits of data are presented simultaneously to the D/A converter 202 C.; Digital 110 SpeCifications Integrated Circuit: MOS Technology 6522 Versatile Interface Adapter 16 bidirectional lines (usually used as 8 bits in and 8 bits out) Latching capability on input or output Four handshaking signals accommodate positive or negative logic Interrupt register and interrupt enable registers are available for each handshake signal Input Characteristics: High Voltage: 2.4 to 5.0V Current: -100 t0 -250 microamperes .Low Voltage: -0.3V to +0.4V Current: -1-0 to -1.6 milliamperes Leakage Current: $1.0 to $2.5 microamperes Off-State Current: $2.0 to $10 microamperes Capacitance: 10 pF Output Characteristics: High Voltage: 2.4V minimum Current: -0.1 to -1.0 milliamperes (PAO-PA7, CA2 -3.0 to -5.0 milliamperes (PBO-PB7, 081, C82) Low Voltage: 0.4V maximum Current: 1.6 milliamperes Leakage Current: 1.0-10 microamperes Capacitance : 10 pF 203 C,4 Real Time Clock and CounterZTimer Specifications Integrated Circuit: Timers 0 and 2 Timers 1 and 3 Shift Register Interrupt Control: Two MOS Technology 6522 Versatile Interface Adapters 16 bit countdown timers can be used as: * one-shot interval timers with optional pulse output on PB7 * continuous frequency generator with optional square wave output on P87 16 bit countdown timers can be used as: * one-shot interval timers * frequency counter that counts a predetermined number of pulses on PB6 * shift register rate generator Inputs or outputs 8-bit serial data with timing pulses supplied by Timers 1 or 3, the 1.023MHZ processor clock or external clock. Interrupt flag and interrupt enable on all functions. Signal Characteristics: TTL compatible signals (one TTL load or service) "I1111111111111111111113