”I‘m mu; minim! M iii" llflllljlfl nu was. 4 ‘ . A . '— J LIBRARY L] Michir'an 5:371“: '_ “,5 Unix: i:t3' f .- oo v‘.“ H)- -.-."' "' ’ This is to certify that the thesis entitled APPLICATION OF MI CROPROCESSOR TECHNOLOGY FOR THE CONTROL OF CONTINUOUS FLOW COMMERCIAL GRAIN DRYERS presented by James Carl Borsum has been accepted towards fulfillment of the requirements for M.S. degree in A.E. Major professo/mr F“) 6' A ) Date 11/ fif 6’] 0-7 639 icrz‘nbus FINES: 25¢ per day per ite- Rnuaum LIBRARY mrsnms: Place in book return to renove charge from circulation accords APPLICATION OF MICROPROCESSOR TECHNOLOGY FOR THE CONTROL OF CONTINUOUS FLOW COMMERCIAL GRAIN DRYERS BY James Carl Borsum A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Engineering 1981 ABSTRACT APPLICATION OF MICROPROCESSOR TECHNOLOGY FOR THE CONTROL OF CONTINUOUS FLOW COMMERCIAL GRAIN DRYERS BY James Carl Borsum The feasibility of utilizing microprocessor—based technology for the automated control of continuous flow commercial grain dryers has been investigated. An indirect method of maintaining a desired outlet grain moisture content for a single concurrent flow drying stage, based on outlet air and grain kernel temperatures, is presented. The microcomputer control system was tested on a pilot-scale concurrent flow corn dryer. The microcomputer controller implements a modified Proportional—Integral (PI) control algorithm which adjusts the grainflow rate to maintain a pre-selected outlet grain temperature (moisture content). To compensate for a variable deadtime element, the microcomputer calculates a new set of PI controller constants whenever a correction in the grainflow rate is initiated. During limited concurrent flow corn drying tests the control system maintained the average outlet corn temperature within 0.5 C of the set-point value (corresponding to a simulated average outlet corn moisture content within 0.2 % w.b. of a set-point value). ACKNOWLEDGEMENTS The author wishes to express his sincere appreciation to Dr. Fred w. Bakker-Arkema for his guidance and inspiration. Thankful acknowledgement is extended to Blount, Inc. for making this work possible through their financial support. ii TABLE OF CONTENTS Page LIST OF TABLES 0.00.0000...0.0.0....OOOOOOOOOOOOOOOOOOOOv LIST OF FIGURES 0......OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO vii Chapter 1 INTRODUCTION 0 O O O O O O 0 O O 0 O O O O O O O O O O O O O O O O O O O 0 O O O O O O O l 2 OBJECTIVES O O O O O O O O I O O O O O O O O O O O O 0 O 0 O O O O 0 O O 0 O O O O O 0 O O O 6 3 BACKGROUND INFORMATION ............................. 8 3.1 Continuous Flow Grain Dryers ................... 8 3.2 Process Control Fundamentals .................. 12 3.3 Digital Control Systems ....................... 13 3.4 Microcomputers: A Brief Overview .............. 15 3.4.1 Information Transfer- .................... 15 Memory .................................. 16 Buses 0.00.00...OOOOOOCOOOOOOOOOOO0...... 17 Input/Output Circuitry .................. 18 Programming OOOOOOOOOOOOOOOOO0.0.0.000... 20 wwww bbhb o o o o mbww 4 LITERATURE REVIEW 00......OOOOOOOOOOOOOOOOOO0.00... 22 5 DRYER CONTROL PARAMETERS .......................... 31 . 5.1 Time .......................................... 31 5.2 Temperature ................................... 33 5.2.1 Air Temperature .......................... 34 5.2.2 Grain Temperature ........................ 35 Airflow Rate .................................. 37 Grain Moisture Content ........................ 38 5.4.1 On-Line Grain Moisture Measurement ...... 38 5.4.2 Indirect Moisture Monitoring ............ 42 5.5 Computer Simulation Analysis .................. 44 5.5.1 Grainflow ............................... 44 5.5.1.1 Concurrent Flow Dryers .......... 45 5.5.1.2 Crossflow Dryers ................ 52 5.5.2 Airflow Rate ............................ 54 5.6 Multi-stage Dryer Control ..................... 58 mm o 0 kW iii EXPERIMENTAL INVESTIGATION ......IOOOOOOOOOIOOOOOOO 61 6.1 The Contr01 system ......OOOOOOOOOOO0.0.000... 61 6.1.1 The Drying Plant ......................... 61 6.1.2 Microcomputer Controller ................ 64 6.1.3 Temperature Measurement ................. 66 6.1.4 Grainflow Rate .......................... 69 6.1.4.1 Motor Speed Control ............. 69 6.2 Dryer Control Algorithm ....................... 72 6.2.1 System Analysis ......................... 77 6.2.2 Controller Implementation and Testing ... 86 6.3 Airflow Rate ................................. 101 CONCLUSIONS ......OOCCOOOOOOOOO......OOOOOOOOOOOOO lgs SUGGESTIONS FOR FUTURE STUDY ......O.............. 107 APPENDIX A - Circuit Diagrams and Specifications for Equipment Used in the Experimental Investigation. ......... 109 APPENDIX B - ZIBL Code Implementing the Dryer Control Algorithm. .............. 113 LIST OF REFERENCES O0.0.0000.........OOCOOOOOOOOOOO 117 iv LIST OF TABLES Table Page 1.1 Data from a typical 12-hr run of a Ferrell-Ross 3-stage concurrent flow grain dryer at Saginaw, MI during November 1980. .............. 3 3.1 Microcomputer language levels. ................. 20 5.1 Concurrent flow dryer simulation, Input ConditionSO OO.......OOOOOOOOOOOOOOOOO.... 46 5.2 Outlet grain moisture content when maintaining three different outlet grain temperatures within a concurrent flow dryer. ................ 49 5.3 Outlet grain moisture content when maintaining a constant outlet grain temperature at three different inlet air temperatures within a concurrent flow dryer. ......................... 51 5.4 Crossflow dryer,simu1ation, Input conditions. .. 54 6.1 Input conditions and calculated controller settings for grainflow stepchange test from 4.0 to 2.4 (tonne/hr/Sq m) ......OOOOOOOOOOOOOOOO 82 6.2 Input conditions and calculated controller settings for grainflow stepchange test from 2.4 to 1.3 (tonne/hr/sq m). ......OOOOOOOOOOOOOO 84 6.3 Input conditions and calculated controller settings for grainflow stepchange test from 2.4 to 4.0 (tonne/hr/sq m). .................... 86 6.4 Input conditions and calculated controller settings for grainflow stepchange test from 1.4 to 2.4 (tonne/hr/Sq m). ......OOOOOOOOOOOOOO 88 6.5 Input conditions and controller constants for testing of the outlet grain temperature ContrOI SYStemo 0.000.00..........OOOOOOIOOOO... 93 Input conditions and controller constants for testing of the final outlet grain temperature contr01 SYStem. ....O.......OOOOOOOOOOO000...... Input conditions for airflow step change tests. vi 99 104 Figure 3.1 LIST OF FIGURES Page Air and grain temperature distribution in CtOSSflow dryers. 0............OOOOOOOOOOOOOOO 9 Air and product temperatures versus depth for a single-stage concurrent flow dryer (BrOOk' 1977). 00......OOOOOOOOOOOOOOOOOO0.0.00. 11 Control system block diagram (Bibbero, 1977). .. 13 Grainflow versus outlet grain temperature for a single-stage concurrent dryer (MSU computer SimUIation). 0.0.0.0.0.........OOOOOOOOOOOOO.... 47 Grain required to maintain a constant outlet grain temperature with a varying inlet moisture content in a single-stage concurrent flow dryer (MSU computer simulation). ..................... 48 Grainflow required to maintain a constant outlet grain temperature with a varying inlet grain moisture content at three different inlet air temperatures in a single-stage concurrent flow (dryer (MSU computer simulation). .............. S0 Grainflow required to maintain a constant outlet grain moisture content with a varying inlet grain ,cr moisture content at two inlet air temps in a single- stage crossflow dryer (MSU computer simulation). 53 Average product and exhaust air temperatures when maintaining a constant outlet grain moisture in a single-stage crossflow dryer (MSU) computer simulation). .................... 55 Effect of airflow rate on outlet grain moisture for a single-stage concurrent flow dryer (MSU computer SimUIation)o ....IOOOOOOOOOOOOOOO. 56 Effect ofairflow rate on outlet moisture for a crossflow dryer (MSU computer simulation). 57 vii Schematic of the MSU pilot-scale concurrent flow grain dryer (Dalpasquale, 1981). .......... Flowchart TEMPOUT - Program to read and convert thermocouple voltage signals. .................. Flowchart RPMSET — Program which controls the discharge auger motor speed. ................... Single-stage concurrent flow grain dryer control system - Block Diagram. ................ Determination of Ziegler-Nichols PI controller Constants (smith, 1979). ....................... Outlet grain in grainflow for a single Outlet grain in grainflow for a single Outlet grain in grainflow for a single Outlet grain in grainflow for a single temperature response to a step change rate from 4.0 to 2.4 (tonne/hr/sq m) stage concurrent flow corn dryer. . temperature response to a step change rate from 2.4 to 1.3 (tonne/hr/sq m) stage concurrent flow corn dryer. . temperature response to a step change rate from 2.4 to 4.0 (tonne/hr/sq m) stage concurrent flow corn dryer. . temperature response to a step change rate from 1.4 to 2.4 (tonne/hr/sq m) stage concurrent flow corn dryer. . Flowchart DRYERPI - Program which controls the outlet grain Experimental temperature. ...................... test results for the outlet grain temperature controller. ........................ Experimental test results for the final outlet grain temperature CODttOllero ......O........... Outlet grain temperature response to a positive step Change in airflow rate. .................. Outlet grain temperature response to a negative step change in airflow rate. .................. viii 62 68 71 81 83 85 93 96 98 102 103 CHAPTER 1 INTRODUCTION Recent advances in microelectronics have made it possible to put all of the elements of a conventional computer central processing unit into a single integrated circuit package known as a microprocessor. This study deals with the application of microprocessor technology to the control of continuous flow commercial grain dryers. Very little automatic control, of any type, is presently being utilized in the operation of such dryers. Continuous flow grain dryers provide an excellent opportunity for the application of microprocessor-based control. Much experience is necessary to correctly interpret and adjust the dryer parameters (such as grainflow rate, airflow rate and temperature) for optimum performance. An operator is required to determine the entering grainflow rate, the heating and cooling rates, etc. All this information needs to be blended together and used to modify the dryer parameters sometime in the future. Table 1.1 summarizes the operation of a continuous-flow three-stage concurrent flow dryer during a 12 hour period. An experienced operator was on duty at all times. The test was conducted during the 1980 fall drying season in Saginaw, MI [Bakker-Arkema et al., 1981]. Note that even an experienced operator has difficulty maintaining the desired outlet conditions. A microcomputer control system, while probably not completely replacing the operator, has the potential to maintain dryer parameters close to optimum values and to free the operator for other duties. Energy can be saved by operating dryers at higher inlet air temperatures or lower airflow rates. The source of the savings is the greatly increased ability of air to hold moisture at higher temperatures. When optimizing the design of corn dryers Brook and Bakker-Arkema [1980] placed primary importance on setting the air temperature in each stage of a concurrent flow dryer to the maximum value allowed. Minimal values of airflow and dryer length were then found which met moisture content and grain quality constraints. , Grain quality constraints include limits on the kernel temperature at the end of each stage [Westelaken, 1981]. The closer a dryer is operated to the product temperature limit, the more important is an accurate control of the drying conditions. Table 1.1: Data from a typical 12-hr run of a Ferrell- Ross CCF 3-stage concurrent flow grain dryer at Saginaw, MI during November 1980. AIRFLOW ** GRAINFLOW TIME M.C. in* AIR TEMPS (static pres.) (feedrbll) [% wb] [F] [in. H20] [RPM] top mid bot top mid bot 8 00 pm 24.9 500 398 300 13.5 16.0 17.5 50 9 00 23.8 500 400 305 14.0 16.0 17.0 50 10 00 28.5 0 310 255 12.0 15.5 17.0 45 11 00 29.0 490 390 295 14.5 14.5 17.0 45 12 00 29.0 505 380 280 15.0 15.0 15.0 40 1 00 am 26.2 500 380 290 15.0 15.0 15.0 40 2 00 25.4 500 400 290 15.0 16.0 17.0 40 3 00 28.8 500 400 300 12.5 15.0 17.0 50 4 00 22.3 495 340 260 14.0 14.5 16.0 44 5 00 26.1 500 340 265 13.5 15.0 16.5 50 6.00 26.7 500 340 265 14.0 15.0 16.5 50 7-00 27.5 495 340 265 13.0 14.5 16.5 50 8:00 24.7 490 340 265 13.0 14.0 16.5 50 OUTLET GRAIN TEMP M.C. out [F] [% wb] top mid bot (SP=16.5%) 8:00 pm 138 129 102 15.6 9:00 139 134 87 19.7 10:00 138 139 87 17.6 11:00 129 130 93 17.4 12:00 134 100 79 19.2 1:00 am 112 136 71 22.0 2:00 133 123 76 20.6 3:00 147 128 77 20.7 4:00 119 151 70 19.0 5:00 141 138 102 11.9 6:00 130 132 86 15.5 7:00 133 124 85 16.3 8:00 114 124 80 19.8 AVG= 18.1 * Commercial corn, available at the terminal, was used in the test ** Static pressure is an indicator of the airflow rate As a result of inadequate control, elevator managers commonly underdry or overdry grain. Underdried grain can spoil and if maximum moisture content values are exceeded, a heavy penalty (dockage) results (e.g. for #2 yellow corn the shipping limit is 15.5% moisture or less; higher moisture contents place corn in the #3 or #4 catagory). Overdrying of grain is expensive. The costs involved include fuel costs, investment costs, maintenance costs, labor costs, ”handle" losses[1] and shrinkage losses. The cost of a microcomputer control system and its sensors can easily be recovered in less than one year by significantly reducing overdrying. Drying is an inherently stable process. If the feed to the dryer and its ambient surroundings do not change, there is no need for automatic control once the machine is properly adjusted. This is important to [1] The "handle” refers to the profit per bushel of grain handled. During the season elevators are restricted to the amount of grain taken in and by the limitations on drying, (wet grain can not be stored). The more grain an elevator handles in one season the more profit it can generate. consider for a responsible analysis of just how much control is actually justified. With microprocessor technology it is possible for a dryer to be overcontrolled [Zagorzycki, 1979]. CHAPTER 2 O B J E C T I V E S The objectives of this study include: to ascertain the parameters which can best be controlled in commercial grain dryers, in particular in concurrent flow dryers; to investigate the requirements for implementation of microprocessor based controllers for single and multi-stage continuous flow commercial grain dryers; to investigate the differences for controlling dryers of crossflow and concurrent flow design; and, to develop and test a microprocessor-based control system for a pilot—scale, single stage concurrent flow grain dryer. CHAPTER 3 B A C K G R O U N D I N F O R M A T I O N 3.1 Continuous Flow Grain Dryers Continuous flow grain dryers are generally of crossflow design. In a crossflow dryer, grain and airflow in perpendicular directions. The temperature of grain on the air inlet side of a crossflow dryer will approach or equal the maximum temperature of the heating air (see Figure 3.1) Overdrying and underdrying of the grain column occurs but mixing is used to eliminate wet grain pockets Continuous, concurrent flow grain dryers have recently become commercially available. In a concurrent flow dryer, grain and air both flow in the same direction. High rates of evaporation occur at the inlet of each drying stage where the hottest air encounters the wettest grain. The grain temperature remains considerably below the air Temp. °F Temp. °C _ 90 200 1 Heating Air Temperature 180 1 ... 80 160‘ - 70 Product Temperature 140 60 E 15 22.5 30 cm. I 1 1 I I I I I 3 6 9 12 in. Depth Figure 3.1 Air and grain temperature distribution in crossflow dryers. 10 temperature in this region and thus exceptionally high drying air temperatures can be used. As the air and grain move through the dryer their temperatures equilibrate. As shown in Figure 3.2 the air and grain temperatures normally become equal within the first few inches of a concurrent flow drying section [Brook, 1977]. The concurrent flow dryer has advantages over the crossflow dryer because of its favorable energy efficiency, grain quality characteristics, and pollution qualities. The high inlet air temperatures used in concurrent flow dryers result in high energy efficiencies. Energy efficiencies between 4185-5120 kJ/kg of water removed are common for concurrent flow dryers while crossflow dryers average about 7500 kJ/kg of water removed [Muhlbauer and Isaacs, 1975]. There is no moisture gradient among the grain kernels in a concurrent flow dryer and the continuous gradual decrease of the product temperature through the last portion of the drying section reduces drying stresses and helps lessen stress cracking and mechanical damage during subsequent handling. The amount of pollution given off by a grain dryer is a function of the quantity of air discharged to the atmosphere. The volume of air exhausted from a crossflow dryer is eight to ten times larger than from a comparable concurrent flow dryer [Bakker—Arkema et al., 1972]. 11 300 550 500 I 250 a 4504 200 1 400 . t) k. 350 - a. - E 150 —( E 300‘ AIR § § E11 [:3 a E4 250 . E21 E21 5* 100 _ E 200 . 150 q 50 - PRODUCT 100 -< T I I 1 2 3 DEPI‘H, ft. I 015 1.0 DEPTH, m. Figure 3.2 Air and product temperatures versus depth for a single- stage concurrent flow dryer (Brook, 1977). 12 3.2 Process Control Fundamentals In control theory, the physical hardware that is associated with the process being controlled is referred to as the “plant”. Elements that are added to effect control are designated as the “controller”. The task of a controller is to adjust the state of a process as measured by some variable - the process variable (PV) - to conform to a particular standard value termed the set-point (SP).[1] The difference between the SP and the PV is called the error(E). With knowledge of the error the controller acts on the process through a final control element to change the process variable in the desired direction. The way the controller responds to the error is called its ”control algorithm” [Bibbero, 1977]. The plant and the controller taken together constitute a control system (see Figure 3.3). The input-output relationship of a control system element can often be represented by a transfer function. The "transfer function” of an element is defined as the [1] The state of a process can usually be described by several process variables which may interact with each other. Although this study is only considering the case of independent process variable control, it should be pointed out that digital processors are particularly suited for interactive 'multivariable' control. l3 CONTROL umnruurso oumn VARIABLE F INAI. CONTROL POINT ELEMENT r-—----—-— _ ————_—fi I I rams rmmnv ' ------ <+ mmn amen? rv rv : Ins useoI Benson) : I M " I m L .. ._ -55.”. 26.1”331’1‘1'_ _. _. .I Figure 3.3 Control system block diagram (Bibbero, 1977). ratio of the transform of its output to the transform of its input. Laplace transforms are used for continuous systems analysis while the z-transform has the same role in discrete systems [Franklin and Powell, 1980]. 3.3 Digital Control Systems The control of physical systems with digital techniques is becoming more common [U.S. Dept. of Energy, 1979]. Traditional controllers have been of analog design 14 and variables such as temperature, pressure,and flow vary in a continuous analog fashion. Digital controllers (e.g. microcomputers) compute the correct output for a discrete instant in time. Therefore a digital control algorithm must be repeated periodically at a rate sufficient to make the control system behave as if it were under the influence of a continuous controller. Bibbero [1977] listed the following advantages of using digital control systems: 1. lower cost per function than analog control systems; 2. flexibility; 3. security (accuracy and stability); 4. human factors (digital control systems are easier to operate and maintain than analog control systems); and, 5. advanced control capability. Digital controllers allow a sampled data input. A sampled data input offers the greatest flexibility in compensating difficult plant processes, especially where the plant has considerable inherent deadtime.[l] [1] 'Deadtime', also known as reaction lag, is the time required for a change in controller output to be sensed as a change in the measured process variable. 15 3u4 Microcomputers: A Brief Overview The single most significant development in digital systems design in recent years has been the advent of the microprocessor [Short, 1981]. A “microprocessor” includes all the elements of a conventional computer central processing unit (specifically arithmetic, logic, and control) implemented on a single silicon chip. When coupled with the other major components of a digital computer (i.e., memory and input/output circuits), it forms a "microcomputer". A microcomputer is generally implemented as a collection of chips on a printed circuit board although single chip microcomputers have recently been introduced [Wise, K.D. et al., 1980]. 3.4.1 Information Transfer The basic memory element for digital systems is the ”bit". A bit can exist at one of two logic levels, represented by the binary digits '0" and "l”. A ”word" consists of N bits which can be stored, transferred, and manipulated by the microprocessor as a unit. If a word is divided into units of eight bits, these are usually termed "bytes”. The ”word length" of a microprocessor is the amount of information that can be manipulated at one time. 16 Word length determines the accuracy of data, the speed of calculation, and the cost of memory (or the trade-off between these factors) for a particular microprocessor. An ”instruction” is a unit of information that is used to indicate to the microprocessor which operation it is to perform. A set of instructions that are used together to accomplish a complete computational task is called a "program". A program or a collection of programs is, in general, called “software.” Strictly speaking, a microprocessor contains no software. All programs are located in memory and therefore enter the picture only in the discussion of microcomputers. 3.4.2 Memory Basically two types of memory are used in a microcomputer system, read-only memory (ROM) and random-access memory (RAM). ROM, as its name implies, can only be read. I Once its contents are set, they can be changed only by special equipment (e.g. ultraviolet erasers). ROMs are used to store sets of instructions and/or constants that need not be altered after the computer is installed. 17 RAM is the name commonly used for memory that can be both read from and written into (read/write). RAM is used for working storage, insertion of problem constants and program options, and as a data base. RAM is almost always volatile, i.e. with power cut off, the contents of the memory are lost. 3.4.3 Buses One or more electrical connections, known as "buses", are usually used to transfer information within a microprocessor system. The bus is designed to be N bits wide (N is the word length of the microprocessor) to facilitate high—speed data transfer, so that information can be transferred in parallel (e.g. one byte at at time). Typically, all data transferred within the chip and between the microprocessor and external input/output devices is passed on the "data bus”, and addresses of external—memory locations are passed on the ”address bus”. Since addresses are generally two words long, the address bus is usually two words wide. A third bus for control signals is sometimes employed, but more frequently these signals are sent serially over l-bit-wide paths [U.S. Dept. of Energy, 1979]. 18 3.4.4 Input/Output Circuitry Input/output (I/O) circuitry allows a microcomputer to communicate directly with its surrounding environment. I/O devices can be connected to a microcomputer through either a serial or a parallel interface known as a port. A parallel interface moves information one word at a time, whereas a serial interface moves the information one bit at a time. A ”port" is nothing more than an integrated circuit that physically connects the device to each of the system's buses (data, address, and control). Each port is responsible for handling a number of complex tasks associated with the movement of data between the microcomputer and the I/O device. These tasks include error checking, parallel-to-serial conversion and formatting. There are three basic techniques used to accomplish the actual movement of information between the microcomputer system and an I/O device. The first technique, termed "programmed I/O", is normally used when a program in execution needs to read or write data. In programmed I/O the external logic responds to the microcomputer. If a data transfer is requested during execution of a program, the program code initiates the transfer by identifying the appropriate memory address or I/O port then through a series of microcoded instructions, 19 the microprocessor executes the transfer of data. The second method, "interrupt I/O', allows external logic to initiate a request for data transfer. Essentially, an interrupt is a subroutine call initialized by external hardware. Whenever a byte of data is ready for transfer, the I/O device sends an interrupt signal which causes the microprocessor to temporarily leave the program which is being executed in order to service the device. Once the microprocessor has completed the data transfer it resumes execution of the program where the interrupt occurred. The third method of executing an I/O transfer is called "direct—memory-access" (DMA). DMA differs from the other two methods because information can be transferred between memory and the external device without involving the microprocessor in the data-transfer logic. The current operation of the microprocessor is not interrupted. Of the three methods, DMA is the fastest, but it also is the most expensive to implement since the microcomputer logic requires a DMA controller (usually incorporating an additional microprocessor) for servicing the DMA requests. 20 3.4.5 Programming Programs can be written at any one of three language levels: machine language, assembly language, or high level language (see Table 3.1 for examples of each type). Only machine language is directly executable by a microprocessor. Programs written in assembly or high level languages must be translated to machine language for execution. Machine code is usually represented in Table 3.1: Microcomputer language levels. High Level Assembly Machine Language Language Language hexadecimal binary I = J+K LDA AUGND 3A 00111010 0C 00001100 00 00000000 MOV B, A 47 01000111 LDA ADDND 3A 00111010 0D 00001101 00 00000000 ADD B 80 10000000 STA SUM 32 00110010 0E 00001110 00 00000000 HLT 76 01110110 The above program adds two numbers obtained from memory, stores the results in memory, and then halts. The instruction set is for the Intel 8085A microprocessor [Short, 1981]. 21 hexadecimal notation for conciseness. However, it must be converted to binary to be loaded into memory. Programs of any appreciable length are not written directly in machine language because they are difficult to read and write. Assembly language uses mnemonic representations for operation codes, data, and addresses. These mnemonics are abbreviations of the names or descriptions of the instructions, addresses, or data and are used to aid the programmer's memory. Assembly language programs are more understandable than corresponding machine language programs, which makes writing and modification of the programs easier. Programming ease is further enhanced by a high level language, where a single instruction is equivalent to several machine language instructions. The high level languages now available for use with microprocessors include BASIC, FORTH, PASCAL, and FORTRAN [Short, 1981]. The question of whether a particular high level language is available for a specific microprocessor is actually a question of whether there is a compiler to translate the language to the microprocessor's machine code. CHAPTER 4 L I T E R A T U R E R E V I E W The following section summarizes the literature on grain dryer control pertinent to this study. Cloud [1957] listed three main objectives for automatic controls on crop drying equipment: 1. to provide or to utilize the best possible drying conditions consistent with the drying system; 2. to eliminate the human element and to reduce labor; and, 3. to provide safety in case of failure of the system. Matthews [1963a] developed a means for automatically controlling the moisture content of dried grain at the outlet of a continuous flow dryer. An 22 23 electronic controller unit adjusted the grain throughput rate in response to the signal from a capacitance-type moisture monitor. The monitoring system consisted of a permittivity sensing [1] electrode placed in the grain mass at the outlet of the dryer. A capacitive—sensitive bridge circuit was used to condition the signal from the sensor. An experimental low capacity cross flow dryer was used to determine the optimum dryer control action. The object in optimizing the action of the controller was to minimize the duration and extent of the outlet moisture content error when input moisture content variations were encountered. Proportional control action with automatic reset was found to give near optimum performance when drying both freshly harvested and artificially wetted wheat. The prototype control system was fitted to a tower—type farm grain dryer [Matthews, 1964]. Barley, wheat and oats were dried during the trial. The moisture content regulation over approximately 400 hours drying was shown to be generally satisfactory. The control system was not tested on a large scale commercial dryer. [1] Permittivity is also called the ”dielectric constant”. 24 Matthews noted that the sensing electrode introduced an obstruction to the moving grain, liable to collect straws, etc. with the subsequent possibility of partial blockage. The electrodes also caused considerable disturbance to the flow of drying air. Problems were encountered. with calibration of the equipment. Initial settings had to be obtained using a sample-type moisture meter for reference. High temperature rises in the commercial control unit caused the variable-speed drive, used to regulate the grain discharge rate, to malfunction. For these reasons the unit was not commercialized. Aguilar and Boyce [1966] stated that the control system developed by Matthews [1964] was also prohibitively expensive. They investigated various ratios, developed from the dry bulb and wet bulb temperatures of the drying and exhaust air, in order to determine whether such ratios could be used as a means of controlling grain drying processes. Specifically they investigated whether a drying process could be terminated when the average moisture content had reached some predetermined level. Aguilar and Boyce proposed the use of a ratio termed the Effective Heat Efficiency (EHE) defined as: EHE = (Ti-To)/(Ti—Tiw) (4.1) where: Ti = dry bulb temperature of the drying air 25 To = dry bulb temperature of the exit air Tiw = wet bulb temperature of the drying air The EHE ratio considers the sensible heat in the drying air as being the effective heat for drying. It was stated that EHE should always have the same value at a given average grain moisture content and grain depth, as long as the exit air is not saturated. Tests were conducted using a static bed dryer to dry barley. An EHE chart was developed for different airflows and temperatures using relatively simple instrumentation. It was concluded that development of an electronic device utilizing the BBB ratio was possible and that the device could be used to control the flow rate through a continuous flow dryer. As far as the author knows, the EHE has not been commercially adopted for control purposes. Zachariah and Isaacs [1966] described a simulation procedure used in the development of an automatic moisture control system for a continuous flow grain dryer. The control system regulated the flow rate of grain through the dryer. The principal objective of the control-system simulation was to select an optimum set of controller constants. It was established that a direct analytical approach to the synthesis of an optimal dryer-control system was not feasible. 26 Utilizing a mathematical model for the drying process based on the drying equation developed by Hukill [1954], Zachariah and Isaacs [1966] presented a means of simulating various control systems to select an optimum set of controller constants. Four different control systems were investigated including one based on the proportional-reset control algorithm which will be considered in detail in this thesis. A prototype proportional-reset control system was constructed and tested on a full scale dryer as a check of the simulation procedure. A modified capacitance grain moisture meter was used to monitor the moisture content of the grain leaving the dryer. The monitoring unit required periodic sampling resulting in a sampled-data system with zero-order hold. Thus, the electric signal representing the moisture content was established at the time of sampling and held constant until the subsequent sampling period. Several limitations of this study can be pointed out. The testing of the control system did not provide conclusive evidence of the accuracy of the simulation model. Difficulties were reported in achieving the required slow reset rate - in the order of one repeat per hour - using commercially available controllers (this problem can now be overcome by using a microprocessor-based controller). Perhaps the greatest limitation of 27 Zachariah-Isaacs' approach is that it requires a direct measurement of the grain moisture content. As will be discussed in a later section, on-line moisture measurement of grain, which is both relatively inexpensive and sufficiently accurate for control, is not commercially available. Holtman and Zachariah [1969a] used computer simulation to design and evaluate optimal controls for a continuous cross flow grain dryer. Utilizing a model which assumes moisture content is a linear function of time, computational methods for deriving the discharge pattern for any input-moisture pattern were developed. Performance indices for evaluation of the controls were based on the deviation from set-point (moisture content) of grain being discharged. A simulation of the process under control was used to obtain typical response data for the optimal controllers. The simulation requires as input the initial moisture of the grain flowing into the drier. Test results showed that the average error increased significantly with dryer size. A larger dryer size implies a higher discharge rate in bushels per hour. Thus when an error occurs in a larger dryer, more off-specification grain is discharged. Holtman and Zachariah did not test their controller on an actual dryer. They based their results on the assumption that a continuous grain moisture measurement 28 could be made. This assumption combined with the fact that a large time-shared computer is needed to implement the optimal control strategy casts doubt upon the usefulness of the approach for an on-line control situation. Harrell et a1. [1979] presented a design for a microprocessor-based control system for a solar assisted grain drying facility. The control system implements an optimization procedure utilizing a modified form of the critical path method of optimization [Colliver et al., 1979] and the logarithmic model of grain drying [Sabbah et al., 1977] to determine the optimum drying air temperature. The control network is configured around a Motorola M68MM01A single—board microcomputer which utilizes the M6800 microprocessor. The microcomputer is to monitor the grain moisture content and air temperatures (dry bulb and wet bulb) to determine whether the current moisture content is equal to or less than the desired final moisture content. If so, the drying is complete and the system is shut down. If not, the drying process is simulated for a one-hour interval with current temperatures and airflow rate. After the simulation is complete the moisture removal rate is computed and the microcomputer executes the optimization routine to determine the optimum drying air temperature. If more heat is needed a heater is turned on and the whole operation is repeated. 29 The control system has not been implemented. The major obstacle to be overcome is the measurement of the grain moisture content. An indirect method to determine the moisture content, based on dry and wet bulb temperatures, is being considered by Harrell et al. Another limitation of this study is that the optimization process must continually update itself, leaving very little computer time for heater control. Significantly more computer time must be available for monitoring and control of the variables in a multi-stage continuous flow grain dryer. Hinkle (1980) demonstrated the potential for microcomputer control of grain dryers. An Intel 8085-based microcomputer replaced the existing, electromechanical controls of an automatic batch grain dryer. Loading, purging, drying, cooling, and unloading functions were successfully scheduled by the microcomputer. An automatic batch dryer continues to cycle through the above five stages as long as wet grain is available. The drying cycle is terminated when the temperature of the grain reaches a preset level which is .indicative of the desired final moisture content. Hinkle noted that 98% of the computer operating time was spent in wait cycles. This point is particularly important when considering the implementation of a microcomputer to control the outlet moisture content of 30 continuous flow dryers because these control tasks are much more complex and may need considerably more computing time. Hinkle did not actually utilize the computational power of the microprocessor, but he did demonstrate that a microcomputer can replace existing dryer controls with the potential for implementing future complex control strategies. CHAPTER 5 D R Y E R C O N T R O L P A R A M E T E R S The primary objective of grain drying is to bring the moisture of the grain down to a particular level (set-point). Time, temperature, and airflow rate are the three primary variables that determine the moisture content of a particular grain discharged from each drying stage [Zachariah and Isaacs, 1966]. Measurement and control of each of these parameters is considered separately. 5.1 Time ‘ The drying time in continuous flow: dryers is determined by the rate at which grain is discharged from the dryer. Continuous flow dryers require a grain metering device so that the grainflow rate can be regulated and controlled. The metering device used on most continuous flow dryers is a feed-roll auger powered by an electric 31 32 motor. The grainflow rate can be considered constant at a given auger speed. In multi-stage dryers the grain metering device is located at the outlet of the last stage and thus the grainflow rate (in terms of dry bushels) is the same throughout the dryer. Auger speed can be regulated by controlling the velocity of the drive motor. A microcomputer can be programmed to put out a digital signal to a digital-to-analog (D/A) converter which in turn provides the input voltage to an electronic motor speed controller (usually replacing a manually operated potentiometer). This control strategy has been utilized successfully in the experimental investigation which is presented in Section 6.1.4 of this thesis. When interfacing a microcomputer to a motor speed controller, care must be taken to completely isolate the microcomputer from the current surges which occur in the controller during motor startup and when making abrupt changes in motor speed (the current surges are a result of back EMF generated by the motor at these times). This can be accomplished using optical or transformer isolation techniques [Burton and Dexter, 1977; Sheingold, 1981; and Morrison, 1977]. There are normally practical limits within which the grainflow rate can be controlled. Dryer capacity is decreased as the grainflow rate is lowered. When the grainflow rate is decreased to a certain point, rather than 33 further decrease the dryer capacity, an operator will adjust the inlet air temperatures and/or airflow rates to increase the drying rate. Also, when relatively dry grain is introduced into the dryer, an operator will increase the grainflow rate only to a certain point before adjusting the inlet air conditions. These practical limits should be incorporated into the dryer control algorithm. Controlling grain moisture by regulating the grainflow rate introduces a variable deadtime element into the control system. Also, because the grain velocity is the same throughout a continuous flow dryer, verying the grainflow rate to control the drying time for a changing inlet grain moisture content, also changes the drying time for grain which is already in the drying stage(s). These two consequences of utilizing grainflow rate as a controlled variable severely complicate the grain dryer control problem. 5.2 Temperature Drying efficiency is maximized by setting inlet air temperatures to the maximum value allowed [Brook and Bakker-Arkema, 1980]. Restrictions on grain temperature levels are imposed on the drying process in order to maintain the quality of the product [Westelaken, 1981]. 34 Both drying air and grain temperatures are important parameters when considering automatic control of commercial continuous flow grain dryers. 5.2.1 Air Temperature The heat input for the drying air of commercial grain dryers in the United States is usually supplied by direct-fired burners. Drying air temperature is controlled by regulating the fuel flow to the burners. Fuel flow is regulated using a throttling valve which is normally positioned by a modulating motor. Modulating motors are actuated via a potentiometer. A controller utilizing the signal from a temperature probe for a feedback signal may be used to regulate the fuel flow in order to maintain the drying air temperature at a particular set—point. A microcomputer based control loop can be used in place of the temperature controller on commercial dryers. The manual potentiometer used to control the modulating motor would be replaced with a digitally controlled potentiometer or a D/A converter operated by a microcomputer. The drying air temperature could thus be controlled in accordance with the overall grain dryer control software implemented in a microcomputer. 35 Accurate exhaust air temperature measurements are difficult to obtain in concurrent flow grain dryers [Bakker—Arkema, 1981]. Temperature measurements of exhaust air are difficult because of severe turbulence and mixing with ambient air in the proximity of the dryer exhaust vents. As shown in Figure 3.2 the air and grain temperatures at the exit of a concurrent flow drying stage are approximately equal. Thus, the grain kernel temperature at the exit of a concurrent flow drying stage may be monitored in lieu of the more elusive exhaust air temperature. Measurements of exhaust air temperatures from a crossflow dryer may entail calculating the average reading of a number of temperature probes (e.g. a thermopile) placed uniformly across the entire exhaust air vent. 5.2.2 Grain Temperature Grain kernel temperatures can be monitored using Type-T (copper-constantan) thermocouples. Type-T thermocouples are recommended for use in a moist, low temperature (less than 350.0 C) environment as found in a grain dryer. Accuracies of 10.5 C are possible with Type-T thermocouples which meet ANSI ”special limits of error" [Omega Engineering Inc. 1981]. The thermocouples should 36 be placed in sheaths directly in the grain mass. Aluminum sheaths are recommended for temperatures less than 350.0 C. Thermocouple signals must be amplified, compensated for ambient temperature, and converted to a digital value before being put into a microcomputer. The microcomputer converts the digital value to the corresponding temperature. There are several commercially available circuits which perform the required thermocouple signal conditioning. One such unit was used in the experimental investigation and is presented in Section 6.1.3 of this thesis. Copper-constantan extension wires should be used between the thermocouple leads and the signal conditioning circuitry to eliminate errors due to extraneous thermocouples along the signal transmission lines. Copper-constantan extension wire has a resistance of 2.46 ohms/m (0.75 ohms/ft), however voltage drops along the transmission lines are minimal because very little, if any, current is developed at the input of the signal conditioning circuitry [Sheingold, 1981]. Shielded thermocouple extension wire should be used to minimize the pickup of external noise along the transmission lines. 37 5.3 .Airflow Rate Ambient (and recycled) air is supplied to the burners in a commercial flow dryer by fans (usually of centrifugal design) powered by electric motors. The fans are normally operated at a constant speed. If a means for controlling the airflow rate is provided, it is usually a manually operated control damper located in the air duct between the fan inlet or outlet and the burner inlet. Airflow rate is normally determined indirectly by monitoring the static pressure in the air duct. Automatic control of the airflow could be accomplished by installing control dampers adjustable via a modulating motor. The modulating motor can be adjusted from the microcomputer as previously described in the discussion of fuel flow regulation. Using airflow rate as a controlled variable introduces a variable deadtime element into the control system, similar to that found with grainflow regulation [Shinskey, 1978]. Power requirements impose a practical limit on the maximum airflow rate obtainable and fans operate efficiently only over a small airflow range. 38 5.4 Grain Moisture Content Optimal control of a drying process requires knowledge of the product moisture content. The operator of a continuous flow grain dryer will typically conduct off-line moisture measurements of representative grain samples. These measurements are usually conducted with some type of electronic moisture analyzer, (the Motomco Moisture Meter has been accepted by the Standardization and Testing Branch 0f the Grain Division, U.S. Dept. of Agriculture [Henry, 1975]). 5.4.1 On-Line Grain Moisture Measurement An on—line method of monitoring grain moisture content would be useful for automatic control of the drying process. However, currently available continuous moisture-sensing instrument systems have proven to have serious limitations [Bakker-Arkema, 1981; Sen, 1981]. The following short-comings of currently available commercial moisture measurement systems are largely responsible for their limited use for control of drying processes [Yang, 1981]: 1. EXPENSE: High cost prevents wide use of moisture 39 sensor systems. Advanced systems cost from $5,000 to $20,000. LACK OF AUTOMATED READOUT: The popular dewpoint instruments commonly produce a temperature reading from which the relative humidity must be calculated by using a psychrometric table. This type of instrument does not generate a direct electrical command for automated process control. Other moisture sensors based on mechanical principals also lack an electrical signal output. NONLINEAR RESPONSE: A large class of instruments (e.g., hygroscopic chemical-electrical resistant-type sensors) produce a nonlinear response. Sophisticated electronics are required to compensate for the effects of nonlinearity in order to achieve a linear readout, (this is not a serious limitation when utilizing a microcomputer controller). SLOW RESPONSE: It is not uncommon for a moisture-sensing system to require a stabilization time in excess of 30 seconds (up to several minutes). 40 5. LARGE SIZE: The main disadvantages of a large probe are inconvenience of installation and interference with the product flow. 6. ENVIRONMENTAL INSTABILITY: Certain types of probes deteriorate with time when exposed to contaminants such as smoke and chemicals or when subjected to an abrasive environment, such as rice. This can change the sensitivity factor, and requires frequent probe recalibration or replacement. 7. INCONVENIENT CALIBRATION PROCEDURES: The repeatability and the calibration factors of most sensors change with time, therefore periodic calibration is necessary. For the majority of exposed sensors, the calibration requires removing the probes and placing them in humidity-controlled chambers that can maintain a standard humidity. Several manufacturers have marketed systems fo continuous measurement of grain moisture content. The DICKEY-john Corporation (Auburn, IL) manufactures the CFMM-A Continuous Flow Moisture Monitor which is intended to provide continuous moisture monitoring of grainflow for grain handling operations. A capacitance-type detector senses dielectric properties of the flowing grain and 41 transmits an electrical signal back to a moisture display unit [DICKEY-john, 1980; Matthews, 1963b]. The system has been tested in the commercial drying of corn and soybeans. The performance of the system has been found to be unsatisfactory [Bakker-Arkema, 1981]. The limitations of the system which make it unsatisfactory for control purposes are: 1. detector plugging: although a scalper is used to deflect foreign debris from the detector, frequent plugging occurs due to the accumulation of wet grain, cobs, etc., this causes a wrong moisture reading and severely limits the system's use for control purposes; 2. initial cost: The initial cost of a CFMM system is in excess of $10,000. Diversified Engineering, Inc. (Richmond, VA.) manufactures the DM6 Digital Moisture Meter. The system consists of a probe head which utilizes a highly stabilized RF signal to detect product moisture. The signal from the probe is transmitted to a seperate unit where it is amplified and converted to a digital moisture display. The DM6 system is intended for sensing moisture in agricultural or cellulose products. The DM6 system has been tested in 42 grain drying processes [Sen, 1981]. The unit was found to be unsatisfactory for grain dryer control purposes because the probe's surface was worn down due to the abrasiveness of the grain. The worn probe heads had to be replaced at frequent intervals. The cost of the DM6 monitoring system is approximately $5,000. 5.4.2 Indirect Moisture Monitoring Indirect means of determining product moisture, based on ratios of inlet and exhaust air temperatures, have. been presented by Aguilar and Boyce [1966] (see Literature Review) and Fadum and Shinskey [1980]. Fadum and Shinskey presented a moisture control system embodying the technique of variable driving force. On-line measurement of product moisture or feed moisture is not required. The control strategy has not been implemented on a commercial grain dryer. To control moisture in a dryer, the driving force ratio at input and output must be kept constant by varying both input and output temperatures as the load varies, according to the equation: 43 Xp k * 1n(Ti-Tw/To-Tw) (5.1) where: Xp = product moisture Ti = inlet air temperature To = outlet air temperature Tw = wet bulb temperature k = a constant for a particular dryer and product. The equation assumes that the product dries solely in the falling drying rate region. Control algorithms based on the equation have been successfully implemented for steam-tube and direct fired dryers. As previously mentioned, the temperatures of exhaust air in commercial grain dryers are difficult to measure accurately. Also, wet bulb temperatures are difficult to obtain, especially at the high inlet air temperatures utilized in concurrent flow dryers [Shinskey, 1978]. The derivation of the equation assumes that product temperatures are limited to the wet-bulb temperature of the drying air. This assumption is not valid for continuous flow grain dryers [Bakker-Arkema, 1981]. For these reasons Equation (5.1) has limited use for control of continuous flow commercial grain dryers. 44 5.5 Computer Simulation Analysis Modified versions of the Michigan State University (MSU) crossflow grain dryer computer simulation and the MSU concurrent flow grain dryer computer simulation were used to assist in the analysis of the dryer control parameters. The accuracy of the simulation models has been verified [Dalpasqua1e, 1981; Brook, 1977]. The grain dryer simulations were treated as the objective function of a golden section search routine (modified Fibonacci search) [Beveridge and Schechter, 1970]. Given a set of input conditions, one parameter (e.g. grainflow, inlet temperature, or airflow rate) was searched until a prespecified outlet condition had been reached. The results of these simulations are presented in this section. 5.5.1 Grainflow Grainflow is the parameter most easily and often adjusted by commercial grain dryer operators attempting to maintain acceptable outlet grain moisture content-quality characteristics [Cloud, 1957; Zachariah and Isaacs, 1966; Hinkle, 1980]. In this study the MSU grain dryer computer simulations were utilized to evaluate the effect of variations in grainflow rate on the grain condition at the 45 outlet of a drying stage in a one-stage concurrent flow corn dryer. 5.5.1.1 Concurrent Flow Dryers Experience has shown that corn at the outlet of a concurrent flow drying stage which is maintained in the temperature range from 54.4 to 60.0 C (130.0 to 140.0 F) will have acceptable moisture content-quality characteristics [Westlaken, 1981; Bakker-Arkema, 1981]. Modified versions of the MSU concurrent flow computer simulation were used to obtain the results shown in Figures 5.1 — 5.4. These results are for steady-state . conditions (i.e. transient responses to changes in dryer parameters were not simulated). Corn was used for all simulations. Bakker-Arkema et a1. [1973] proposed the standard conditions listed in Table 5.1 for the drying of shelled corn. Table 5.1 also lists the input conditions used in the concurrent flow dryer analysis. These conditions (unless otherwise noted) are used in obtaining the results in the following sections. Figure 5.1 shows the steady-state grainflow vs outlet grain temperature relationship for a 3.0 ft (0.914 m) concurrent flow drying section for both 25% wb and 30% wb inlet grain moisture contents. Note that the 46 Table 5.1: Concurrent flow dryer simulation, Input conditions. Standard conditions for corn drying simulation Inlet Corn Moisture Content, % wb 25.0 15.0 Ambient Air Temperature, C Ambient Air Humidity Ratio, decimal db 0.006 Inlet Corn Temperature, C 15.0 Inlet Corn Foreign Material, % 3.0 Inlet Corn Breakage, % 10.0 Additional Dryer Input Conditions Inlet Air Temp., C 200 Airflow Rate, cu m/min-sq m @ Tamb 60 0.914 Dryer Length, m relationship is approximately linear in both cases over the range shown. Figure 5.2 shows the grainflow rates required to maintain grain outlet temperatures of 54.4, 57.2 and 60.0 C (130, 135, 140 F) over the inlet grain moisture content range of 20-35% wb with an inlet air temperature of 200 C (392 F). The outlet grain moisture content for each of these outlet temperatures was found to be essentialy constant at the values shown in Table 5.2. 47 .Acoavmgsam H3258 :mzv nomad 30H.“ pcmgocoo mmopmnmawcfim o no.“ onspmnomsop cagm pmapzo msmng :odfiamho H6. $de é um-fi\m§3 .zoézHéo 02m m .H 0.; n .o p — :m CH .9.2 RWN CH .n.3 Rom D 'EIHHLVHScINHl. NI‘V’HD .L'Ei'IlIlO 48 .Acoapmgfim Hopsmsoo szv Hog 30H.“ pzmgocoo omgmuoflwcam .m ca pcopcoo onspmaos gag pods." wcfinng .m 3.3: onsvmnomsop 5mg pogso #23300 d caflcams o... donguon :ofiwcawno N6 mag .s SLEEEB .zofizHéu m s m N H o 7 — p n - ON 0 .mmaémezms 2:5 page I mm won 1mm 'q‘M % ‘HHHLSION NIVHD .IE'INI 49 Table 5.2: Outlet grain moisture content when maintaining three different outlet grain temperatures within a concurrent flow dryer. Outlet Grain Temp. Outlet Grain Moisture [C] [% wb] 544138 """"""""" 57.2 18.0 60.0 17.2 Figure 5.3 shows the grainflow rates required to maintain an outlet temperature of 57.2 C (135 F) at inlet air temperatures of 250, 200 and 150 C (482, 392, .and 302 F) over the inlet grain moisture content range of 20-35% wb. The outlet grain moisture content for each of the inlet air temperatures was found to be essentially constant at the values shown in Table 5.3. Note that using a higher inlet air temperature results in a higher outlet grain moisture content for a given outlet grain temperature set-point. As the inlet air temperature is increased, the grain must be transported through the dryer at a faster rate in order to maintain the same outlet grain temperature. Because the drying time is reduced at the faster grainflow rate, the moisture content of the grain at 50 42033353 Hopsmeoo :93 Hohfio 30H.“ acogocoo owgmuoawfim .m 5.. mondpgomsov Hum pod: 8.30. pm vcopcoo ondpmdos cagm pods.“ gang m fin? onspmuomsop CHE poflbo 993.980 .m campcag ow cougwon :ofiflfiudno MK. 93mg .5 um.£\o§3 iofizHéo w m a m m H o P h P P F ON 0 .mmBémAEme 3.. EH oow 1 mm T on r mm 'Q'M % 'EHILLSIOW NIVHE) .LTINI 51 Table 5.3: Outlet grain moisture content when maintaining a constant outlet grain temperature at three different inlet air temperatures within a concurrent flow dryer. Inlet Air Temp. Outlet Grain Moisture [C] [% wb] 150162 """"""""" 200 18.0 250 19.4 the outlet of the drying stage increases. Changes in ambient conditions and inlet grain temperature do affect the grainflow rates required to maintain a constant outlet grain temperature. Also, the moisture content of grain at a given outlet temperature will change slightly with varyi g ambient and inlet air temperatures. However, the computer simulation did verify that for a given inlet drying air temperature, constant inlet grain temperature and constant ambient conditions, it is possible to maintain a fairly constant outlet grain moisture content by varying the grainflow rate to maintain a constant outlet grain temperature. An experimental, microcomputer based control system which automatically adjusts the grainflow rate to maintain a set-point outlet 52 grain temperature in a concurrent flow dryer is presented in Chapter 6. 5.5.1.2 Crossflow Dryers As shown in Figure 3.1, there is a temperature (and corresponding moisture content) gradient across the drying column of a crossflow dryer. Therefore average outlet moisture content and temperature values are used for this analysis. A modified version of the MSU crossflow grain dryer simulation was used to determine the relationship between inlet grain moisture content and the grainflow rate needed to maintain an average outlet grain moisture of 16.5% wb at inlet air temperatures of 93.3 and 104.4 C (200.0 and 220.0 F). Figure 5.4 shows the results of this simulation given the input conditions of Table 5.4. Note that the two curves are approximately parallel. Figure 5.5 shows the average outlet grain and exhaust air temperatures for the grainflow rates depicted in Figure 5.4. Note that the exhaust air temperatures are approximately equal for the two inlet air temperatures simulated. The product temperature curves are parallel. The correlation between outlet grain temperature and (moisture content found with the concurrent flow dryer simulation is not evident in the crossflow dryer simulation 53 mm _ .Acodpmdseam Hopsmeoo szv Hog zofimmmouo ommpmnmawcam m ca moflfifluomsop Ham pod: 23. pm pcopcoo ngmHos madam pods.“ musing .m 5?. 9:09:00 ofifimaos 5.3m 90.350 pcmpmcoo .w :aflhams op “commando.” zogagu :fi 93me 5.: R .9558: 22mm .533 on mm om _ L o .mmaémmzme mH< EH Afigx m.0—.I .0.5 93050 «0— #30 l \n I O H 1.3 'm bS-IQ/9UU01 'MOTJNIVHO 54 Table 5.4: Crossflow dryer simulation, Input Conditions. Inlet Grain Moisture, % wb 20-35 Inlet Grain Temp, C 14.9 Ambient Temp, C 15.6 Ambient RH, % 50 Inlet Air Temp, C 93.3 and 104.4 Airflow Rate, cu m/min/bu @ Tamb 2.32 Dryer Length, m (ft) 14.3 (47.0) Column Width, m (ft) 0.305 (1.0) Grainflow, tonne/hr/sq m variable results. Further research into the temperature-moisture content relationships for crossflow dryers should be conducted. 5.5.2 Airflow Rate Figure 5.6 shows the simulated relationship between the airflow rate and the outlet grain moisture content for the second stage of a concurrent flow dryer. The grainflow rate and inlet grain moisture content are held constant. The relationship is linear over the range simulated. Figure 5.7 shows that the airflow versus outlet moisture content relationship is also linear for a crossflow dryer. Figures 5.6 and 5.7 show that outlet moisture content can be regulated by controlling the airflow rate. However, over' the normal ranges of control, variations in grainflow rate have more effect on the grain outlet conditions than .Acogflnsswm Hopsmsoo szv .3th 308393 omgmuofiwfim m ca pcopcoo mndpmaoe :Hmhm 96330 #2388 m madcampcams cos: wondpmnmmsmp Ham pmsmnxo and #26on mwgofiw m .m 933E 5.: R .mmEmHoz zHéu EH mm on mm ON 55 p P n.mm 3.33 'cIIAIEIJ. HIV .IEI’INI mam‘ 3.30.“ .11 .msz H: Emséxm .U>.<. 5.25m mdw .. 0.2 £90 “0330 on c: on cm on ow cm 0 ' HHHLVHMNIHJ. 56 .Acowpmassaw Hopsmsoo szv Hohho :on pcouudocoo ommpmuoamcflm m 90% pcopcoo ondpmfioe Camhw pofipso so damn :oamnam mo poommm w.m ondwfih .pm um\emo .3oqmmH< cod . com omfi owH 05H _ _ . u .pm am-nn\sp OH n zoqezH Grainflow 4Q; BUCKET ELEVATOR BURNER METERING AUGER ‘— DRIVE MOTOR DRYING SECTION (s EXHAUST AIR EXIT 0.61m COOLING <7 SECTION 0.61m Figure 6.1 Schematic of the MSU pilot—scale concurrent flow grain dryer (Dalpasquale, 1981). 63 used during this investigation. The cross-sectional area of the dryer is 1.0 sq ft (0.09 sq m); the drying section has a length of 3.0 ft (0.91 m). 5.9 bushels (0.21 cu m) of grain are required to fill the dryer [Kalchik, 1977]. A bucket elevator is used to transport grain to the top of the dryer and load the material into a wet grain holding hopper. The elevator is powered by a 1/3 horsepower capacitor start motor. Grain flows through the dryer by gravity. A 4 inch (10.16 cm) diameter metering auger unloads the grain at the bottom of the drying column. The auger is powered by a 1/2 horsepower shunt wound D.C. electric motor. A gear reduction shaft and interchangeable pulleys are used to arrive at a proper range of discharge rates for the type of grain being dried. Heating air is directed to the dryer via a tee section and elbow which connect to a propane burner. The burner temperature is adjusted manually by regulating the fuel flow using a series of throttling valves. The burner is equipped with an automatic shut-off valve to‘ guard against burner malfunction. Inlet air is supplied by a 22 inch (56 cm) diameter forward inclined fan powered by a 3/4 horsepower motor, via a 4 inch (10.16 cm) diameter wire-reinforced air hose. 64 6.1.2 Microcomputer Controller A Dynabyte, Inc. (Menlo Park, CA) model BC2, Basic Controller was used to implement the grain dryer control strategy. The Basic Controller is a self-contained, 280-based microcomputer. All system components except the power supply and analog-to-digital conversion circuitry reside on a single 12.4 x 14.8 in (315 x 376 mm) printed circuit board. For developmental purposes, a CRT display and alpha-numeric keyboard for programming were added. A casette tape deck was used to record process data and programs for future reference. A keyboard input port, a 16 line by 64 character composite video output port and a 300 baud cassette I/O port are provided to interface the peripherals. Specifications for the Basic Controller are presented in Appendix A. A high—level programming language called 280 Industrial Basic Language, or ZIBL, resides in ex of on-board ROM. Many traditional features of BASIC are incorporated in ZIBL. Looping operations can be programmed using both DO-UNTIL and FOR-NEXT formats. The IF-THEN conditional command can be used to set up conditional branching and looping operations. Subroutining is supported by GOSUB and RETURN commands. Standard arithmetic operations, using 24-bit triple-precision integers, as well as value comparison functions are 65 provided and 26 global and 26 local arithmetic variables can be defined. Constants can be programmed in decimal or hexadecimal form. The logical operators AND, OR, and XOR can also be employed. A summary chart of ZIBL is provided in Appendix B. Application programs may reside in the 16K of on-board RAM or in the 4K of user programmable EPROM (there is an on-board EPROM programmer) [Manufacturing Engineering, 1980]. During initial testing of the control system the Basic Controller was placed directly next to the grain dryer. When the grain dryer is operating, temperatures in the Processing Laboratory may exceed 38 C (100 F). These high temperatures combined with RF interference from the spark ignition system of the propane burner caused the Basic Controller to malfunction. To alleviate these problems the Basic Controller was moved ‘to an air conditioned room located 50 feet from the dryer. Shielded signal transmission lines were used to minimize problems due to electrical noise present in the Processing Laboratory. Power for the Basic Controller was taken from an isolated power line to protect it from noise present on the power lines supplying the grain dryer. Texts by Morrison [1977], Ott [1976], and Sheingold [1981] are excellent references dealing with noise reduction techniques. 66 6.1.3 Temperature Measurement Grain temperature at the outlet of the drying section was the process variable of interest for this investigation. A standard Type-T thermocouple with an aluminum sheath was mounted in the grain column at the outlet of the drying section (see section 5.2 for a discussion of grain temperature measurement). Copper-constantan extension wire was used to connect the thermocouple lead wires to the input terminal block of an Analog Devices (Norwood, MA.) AC-1215 mounting card for the Analog Devices 2854-8 four channel isolated thermocouple conditioner and 2856 high accuracy cold junction compensator (see Appendix A for specifications). The 2854 provides input protection, isolation and common mode rejection, multiplexing, filtering, and amplification of the thermocouple signals. The 2854 has a :5 v @ 5 ma output voltage swing which is transmitted to the input of the 2856. The 2856 operates with an external temperature sensor in thermal contact with the cold junction (a 2N2222 transistor is used as the cold junction sensor on the AC-1215). The 2856 is calibrated to compensate the cold junction to a reference temperature of O C. The 2856 may be digitally programmed to select compensation for seven types of thermocouples (including Type-T). 67 The AC-1215 uses a 74L8139 decoder for decoding digital signals from the computer. This allows the computer to read a particular channel of the 2854 and assign the appropriate cold junction compensation. Power for the AC1215 ( :15 v) is supplied by an external power supply. The inlet air temperature is not used in the control algorithm. However, it is monitored and displayed by the microcomputer to assist the operator in maintaining inlet air conditions. The inlet air temperature is measured with a Type-J (iron-constantan) thermocouple. All thermocouples were individually calibrated against a mercury bulb thermometer. Up to 4 thermocouples can be monitored using the 2854 multiplexing circuitry. Additional thermocouples can be used by installing a computer controlled multiplexing circuit ahead of the input to the AC-1215. Care is needed to insure that the multiplexing signal does not introduce any extraneous thermocouples into the measurement system. Figure 6.2 shows the flowchart for program TEMPOUT which implements the temperature monitoring routine for the Basic Controller. Output flags are used to set the control bits on the AC-1215 for the appropriate thermocouple. The amplified and compensated thermocouple voltage is converted to the corresponding temperature using a linear approximation. Note that the integer value of the outlet 68 CONVERT ET INT TEMP _ TO DECIMAL NOTATION _ I 1L [ SET OUTPUT FLAGS TO READ AND COMPENSATE OUTLET GRAIN THERMOCOUPLEr READ AND CONVERT AMPLIFIED THERMOCOUPLE VOLTAGE TO DIGITAL EQUIV. . III , CONVERT DIGITAL VALUE TO CORRESPONDING TEMP_ [PUT OUTLET GRAIN TEMP IN DECIMAL NOTATION ET LAG TO READ INLET AIR THERMOCOUPLE I READ AND CONVERT AMPLIFIED THERMOCOUPLE VOLTAGE TO DIGITAL EQUIV. * CONVERT DIGITAL VALUE TO C ORRESPQNDING TEMP r— RETURN TO _CALLING PROGRAM Figure 6.2 Flowchart TEMPOUT - Program to read and convert thermocouple voltage signals. 69 grain temperature is converted to decimal form for display and recording purposes. 6.1w4 Grainflow Rate The grainflow rate through the dryer was correlated with the auger drive motor RPM. Test data showed a linear relationship described by: GRAINFLOW (tonne/hr/sq m) = 0.004699*RPM + 0.032878 (6.1) L The 0.914 m by 0.093 sq m drying section has a volume 0.0849 cu m (3 cu ft) or 2.4 bu. The time for a kernel of grain to traverse the drying section (transportation time) can thus be related to the auger drive motor RPM by: t (min) = 2.4 * 60 * 0.2734/GRAINFLOW (6.2) 6.1.4.1 Motor Speed Control An SCR-DC motor speed controller was constructed to control the auger drive motor [Jakeway, 1981]. Because of instability in the motor control circuitry a software feedback loop is implemented using the Basic Controller. A 70 DC generator coupled to the motor shaft was used to measure the speed of the motor (1 volt output/100 RPM). The generator's voltage signal is scaled down using a resistive voltage divider and fed through a 2nd-order Butterworth low-pass filter. The conditioned signal is converted to a digital value in the RTI-1220 A/D converter. Figure 6.3 shows the flow chart for program RPMSET which implements a PI control algorithm (see Section 6.2 for a discussion of PI algorithms) to maintain a set-point motor speed. The set-point can be designated manually by the dryer operator or, when the controller is in automatic, by the dryer control software. The ZIBL code which implements the flowchart is shown in Appendix 8. There are several features of program RPMSET which show the versatility of a microprocessor-based controller. During tuning of the control algorithm it was found that even with a constant input voltage to the motor speed controller, the speed of the motor, as monitored by the DC generator, tends to fluctuate about a single RPM reading. The frequency and amplitude of the fluctuations vary with the speed of the motor. The low-pass filter (cutoff frequency of 1 Hz.) does not eliminate the wave pattern. It was found that by taking the average of ten successive RPM readings a fairly accurate speed measurement can be obtained. This type of signal correction is not possible with a continuous analog controller. It was also 71 (INITIALIZEl I = O SAMPLE PERIO US=~> MOTOR CONTROLLER (, CORRECTION (MIN) = KRPM x (ERPM + MIRPM) OR CONTR LLER INPUT (INPUT) = INPUT + MIN T5 = 0 IRUN TEMPOUTI OUTLET GRAI‘ TEMP ERROR UUTPUT CURRENT h \n ‘\ N CONDITIONS TO CRT DETERMINE TEMP AND TAPEPORT INTEGRAL CORRECTION (MI) FROM I A DESIRED SPEED (RPMNANT = RPMUANT + MI ‘ SAMPLE PERIOD T ) > o DETERMINE AVERAGE N MEO SE = M ' R r ED RP [RETURN TO DRYERPI] DETERMINE ERROR (ERPM) = RPMNANT - RPM , ,GAIN (KRPM) = HIGAIN @AIN (ARE-IT: LOGAINj J DETERMINE SPEED INTEGRAL CORRECTION __, (MIRPM) = ERPM x RPMSET Figure 6.3 Flowchart RPMSET - Program which controls the discharge auger motor speed. 72 determined that within a certain error limit, a higher gain enables the controller to better maintain the motor speed at the RPM set-point. When changes in the set-point or load occur (putting the error outside the limits) a lower gain (resulting in a 1/4 - decay ratio response [1]) is used in the control algorithm. This decision making capability is not possible with a conventional controller. The digital signal output from the Basic Controller was converted to an analog voltage using an optically isolated D/A converter. The optocal isolation was needed to protect the microcomputer from current surges which occur in the motor controller circuitry during startup and during abrupt changes in motor speed. Figure 6.4 gives the block diagram for the concurrent flow dryer outlet grain temperature control system. 6.2 Dryer Control Algorithm The computer simulation results presented previously gave no information concerning the transient response of the dryer to changes in input conditions. [1] A 1/4 - decay ratio response is when the ratio of the second overshoot to the first is 1/4. 73 .Emummwa xoon .HoLucoo Lo>uo cwmum soHu ucmuusocoo ommumtonCwm v.m ouswwm II- V it'lb'ibt .azme zHcmo ,_ oszoHeHozoo _mmemm>zoo aachm i y, ax< a an 9:0 V_maa:ooozmmme_ qr oszoHeHozoo aaonm .mozoo_+_mmaaazoo~. .azme zHamo 2Hpppmom m op uncommon onspmnomsop cpme poapzo. mfi.0 chumpm .GHE .mzHB 00. oo” 00“ p r _ upbp—bpb on" 0:” can _—P~F—~_-r——L—L op. L_Lp_p ”who 0m 0: N: D ‘SHHIVHEdNEL NIVHD IETIDO £65 :Hoo soap pcogocoo ommpw noamcpm o no.“ Aé wounds}: 33 an op 0: son.“ mpg 203.58 5.. 60:20 mopm 958.0me .m op oncommoh chapmnomsop spam poapso +210 930.00 .GHE .mzHe oma . n? 3. no” _ _ _ _ r p r p _ _ _ _ Lb _ p _ 103 ms .3 rem 1% IR :9 [an 06.6 IR 100 1R 100 1am too mequ H0 0 ‘IHHRIVHHdl/Jfll. NIVHD .IE'IIRO 104 temperatures were used during both tests. Table 6.7 summarizes the dryer input conditions for both airflow tests. As noted previously, using airflow rate as a controlled variable introduces a variable deadtime element into the control system [Shinskey, 1978]. However, note that the deadtime is approximately the same following the airflow rate step change in both tests. Also, noting that the grainflow rate is the same for both tests, one can assume that the deadtime of the outlet grain temperature response following a change in airflow rate, is more a function of the grainflow rate than the airflow rate for a concurrent flow dryer. It is important to consider this point when controlling both airflow rate and grainflow rate in a multi-stage dryer as proposed in Section 5.6. Table 6.7: Input conditions for airflow step change tests. Grainflow Rate, tonne/hr/sq m 2.4 Inlet Air Temp., C 230 Inlet Airflow Rate, cu m/min-sq m Test 1 34 to 45 Test 2 45 to 34 Ambient Temp., C 33.0 Inlet Grain Temp., C 6.0 Inlet Grain Moisture, % wb 20.5 Outlet Grain Moisture, % wb Test 1 17.0 Test 2 18.0 Est. Amb. RH, % 75 CHAPTER 7 CONCLUSIONS Automated control of continuous flow commercial grain dryers is needed. A microcomputer control system has the potential to maintain dryer parameters close to optimum values and to free the dryer operator for other duties. Time, temperature and airflow rate are the three primary variables that determine the moisture content of grain discharged from a continuous flow dryer or drying stage. Each of these primary variables can be controlled utilizing a microcomputer controller. Direct on-line measurement of grain moisture content which is both relatively inexpensive and 105 106 sufficiently accurate for control purposes, is not (yet) possible with commercially available instrumentation. The moisture content of grain discharged from a concurrent flow dryer can be satisfactorily regulated by controlling the outlet grain temperature. A microprocessor-based controller was used to maintain the average outlet grain temperature of a pilot—scale concurrent flow corn dryer within 0.5 C of a pre-selected set-point value with a standard deviation of 12.4 C (corresponding to a simulated average outlet grain moisture content within 0.2 % wb of a set point value with a standard deviation of i0.7 % wb). CHAPTER 8 S U G G E S T I O N S F O R F U T U R E S T U D Y The computer simulation used in the control parameter analysis of this study predicts steady-state responses to change in dryer input conditions. Development of computer simulation which can predict transient (real-time) responses to changes in dryer input conditions would greatly facilitate improved control algorithm development. Controller response can be greatly improved by shortening the process deadtime. Investigations should be conducted to determine if temperatures other than those at the dryer outlet (perhaps nearer the dryer inlet or the average of a series of sensors) can be used as indicators of outlet grain moisture content. The experimental investigation of this study was 107 108 conducted under laboratory conditions. A thorough investigation into the commercial grain dryer environment (temperature, electrical noise, etc.) and how it effects the microcomputer control electronics should be conducted. As the capacity of microcomputer systems increases it may become possible to utilize grain dryer simulation, similar to those utilized in this study, for on-line control purposes. Adapting a simulation model for a particular grain dryer may make this possible in the near future. APPENDICES APPENDIX A Circuit Diagrams and Specifications for Equipment Used in the Experimental Investigation 1(59 nunanm'la ~ MODEL BC2 BASIC CONTROLLER 0 SPECIFICATIONS 9 MICROPROCESSOR leog ZOO-CPU, operating at 2.58M: (crystal-controlled) MEIKDRY RAH: 16K bytes (dynamic) EPROH: Sockets for AK bytes (two THSZ716 type) with onoboard automatic EPROH programmer. RAM and EPROH are externally expandable in any combination up to “BK bytes total (exclusive of ZIBL in ROM). LANGUAGE 280 Industrial BASIC Language (ZIBL ) in 8K bytes of ROM. Fully memory-mapped variable and I/O structure with ZA-bit Integer arithmetic capability. 80 statements, commands, mathematical functions and operators. INTERRUPTS 6 user-available vectored hardware interrupts activated by negative-going LS-TTL level input signal. LS-TTL level interrupt acknowledge output signal. REAL‘TIME CLOCK 24 hour format with I second resolution (HHMHSS display), software accessible. Requires IPPS LS-TTL level interrupt input signal, 16 m5 nominal width (normally provided by BCX-Ol-OOI and BCX-Ol-ZOO Power Supplies). Timekeeping accuracy dependent on input signal accuracy and stability. KEYBOARD PORT 8-bit parallel LS-TTL level input port, accepts positive- logic ASCII code. Selectable positive or negative-going LS-TTL level Data Strobe input (strap option). hS bit-line Is used to select 8C2 "Auto-start" feature. VIDEO PORT RS-l70 composite video output signal. 75 Ohm Impedance. Memory-mapped video format of 16 lines by 68 characters per line. 6“ ASCII character display on a 7 x 9 dot-matrix. Underline cursor with X-Y cursor control. Additional "Supervisory" line continually displays current time and number of remaining free memory bytes. SERIAL PORTS 2 separate RS-232c serIal I/O ports with software- selectable baud rates from llO to 9600 (independently adjustable). CTS and RTS handshake signals; 5 to 8 bit word-length (software-selectable). Strap option on serial channel 2 for 20 mA TTY current-loop operation. PARALLEL PORT OUT: O-bit parallel output port, Schottky TTL levels. 500 nS negative-going LS-TTL level Output Strobe signal. IN: B-bit parallel input port, LS-TTL levels. LS-TTL level Input Strobe and Port Susy output signal. CASSETTE PORT 300 baud "BYTE” standard audio cassette interface for storage/retrieval of programs and/or data on cassette tape. On-board relay for automatic start/stop cassette recorder motor operation under software control. —-——a SENSE INPUTS 32 general-purpose LS-TTL level inputs with on-board 2K Chm pull-up resistors for contact-closure sensing. Externally expandable to 6b total SENSE inputs. FLAG OUTPUTS 32 general-purpose LS-TTL level outputs. dable to 6“ total FLAG outputs. Externally expan- RELJUY OLHWNITS 8 on-board electromechanical relays. to 64 total RELAY outputs. Externally expandable h SPOT power type; contacts rated 5 Amps Q 26 VOC, 3 Amps E 115 VAC, 2 Amps § 220 VAC (non-inductive loads). A SPST dry-reed type; contacts rated 10 VA (200 VOC maximum or 0.75 Amp maximum, non-inductive loads). LITE OUTPUTS 8 LS-TTL level outputs, each driving an on-board LED lamp (may be removed to access LS-TTL outputs for off-board use). Externally expandable to 6b total LITE outputs. LITEPORT 8-bit parallel LS-TTL level output port driving 'LEO lamps for display of a binary data byte (lamps may be removed to access LS-TTL outputs for off-board use). ANALOG INTERFACE Optional BCX-30-OOI Analog Interface Module (AIM I provides A/O and O/A conversion capability via interface with Analog Devices RTl-1220-12 Data Acquisition Board and RTI-lZZI-IO Analog Output Board. A/D conversion: 16 channels, lZ-bit resolution O/A conversion: 4 channels, lO-bit resolution One AIH supports mwo Analog Devices boards in any combi- nation. Up to 6 Modules may be paralleled to implement a simultaneous total of SA A/O and 32 O/A channels. POWER 5 SIGNAL- Maximum current-drain. Operating voltage tolerance: +/-5¥ +5 VOC 9 3.20 Amos +27 VOC ? 0.05 Amps Operating temperature range: 0 to +55 degrees Celsius Storage temperature range: -55 to +85 degrees Celsius REQUIREMENTS -5 voc a 0.10 n I +12 VOC 8 0.32 ” IPPS signal, LS-TTL I -12 voc a 0.10 " level (16 m5 width) I 1 Maximum dimensions and weight. Height dimensions increase ‘ by 0.13 inches when protective plexi-glass cover is added. I PC board only With Mounting Plate I PHYSICAL . Lenght: 15.75 inches 16.25 inches 1 CHARACTERISTICS Widfln 12.50 " 1b.25 ” I Height: 0.75 ” 1.50 " Weight: 2.00 pounds “.50 pounds I Mllfllfll'I'l-‘E 115 INDEPENDENCE DRIVE, MENLO PARK, CA. ghozs (his) 329-?921 111 SPECIFICATIONS (typical @ +25°C, Vs = :15V unless otherwise noted) MODEL 2856A COLD JUNCTION COMPENSATION Thermocouple Types' Internally Compensated J. K. T Externally Programmable B. E. R. 5. None Reference Temperature 0°C Compensation Accuracy Total Output Error 0 ozs°c' 202°C vs. Ambient Temperature (95°C to +4S°C)' t0.8°C max Compensation Error vs. Sensor Temperature (05°C to 445°C)3 vs Compensaror Module Temperature (o to o7o°c1’ Cold Junction Temperature Sensing Element INPUT SPECIFICATIONS Voltage Signal Range Input Impedance Signal Gain‘ vs. Temperature Input Offset Voltage vs. Temperature OUTPUT SPECIFICATIONS 5 t0.4°C max (:0.Is°c rypl :0 02"ch max (0.oi°c/°c typ) ADSOO or 2N2222 2101' 1001.9 eIV/V :iOppmfc zlmV max :1 sin/1°C max Output Voltage 210V @25mA Output Impedance 0.19 DYNAMIC RESPONSE Selection Settling Time 0 Sms Signal Settling Time, to 20.019. SOus DIGITAL INPUTS Select Inputs A 8: 8 POWER SL’PPLT Analog. Rated Performance Operating DIgIIII, VDD TEMPERATURE RANGE. Rated Performance TT L. CMOS Compatible :lSV dc “09600 ismA (212V to 118V dc) +5\' to OISV dc (Q ZmA max 010 o7o°c Operating -25°C to +85°C Storage -55°C to +125°C CASE 5m 1.5"xz"xo.4" PRICE. (1 - 9) $75 (la-24) $67 (100's) $49 'Total compensation error composed of errors of temperature sensor and module at - the same ambient temperature. ’Compensauon error contributed by ambient temperature changes at temperature sensor ’Compensauon error contributed by ambient temperature changes at the module. ‘ Signal gain of 2 is also available by )umper selection. 'Protected for shorts to ground or either supply voltage. Specifications nature! to change without notice. SPECIFICATIONS (up 112 ital @ +25°C, Vs= 115V and Vosc = +15V, unless otherwise noted) 2854A 2354B 2355A ANALOG INPUTS Number of Channels 4 ° ' Input Span Range 1$mV to ilOOmV ' :SOmV to 15V Gain Equation 6 - 1 + lOth/Rc ’ ' Gain Error 10.2% max (G = $0 to 300) ‘ 10.2% max (G .1 to 100) 11% max (G a 1000) ° NA Gem Temperature Coefficient 13 Sppme max easppm/‘c max flSppm/‘C max Gain Nonlineanty' 20.03% max (G - 50 to 300) 20.02% max(!0.012% typ) 10.02% max (G '1 10 100) 10.03% (C I 10%) ' NA Offset Voltage Input Offset. Initial (Ad). to Zero) 120uV max ' 1$0uV max vs. Temperature 12.5quC mu :lprC maxftOSpV/oC typ) tSquC max vs. Tune 11.5uV/month ‘ ' Output Offset (Adjustable to Zero) lemV max ° ° vs. Temperature zSOuVI°C max ° ' Total one: Drift (rm). max 2 (2.5. L8) ,1fo :(l .959»fo : (sum gym/c Input Noise Voltage 0.01M: -lOOHz. R5 - lkfl CMV, Channel-toChannel or OIannel-to-Ground Continuous. ac. 60H: Continuous. at or dc Common Mode Rejection luV P‘P 750V rms 11000V pk max R5< 1000. !> 50H: ISOdB min (G - 1000) ° 14508 min (5 u 100) R5<100I2J>50Ha 128d! mlniGsSO) ° ”0.13 min (5.)) Normal Mode Input. Without Damage 130V rms. 60R: ° 0 Normal Mode Rejection. O 60H: SSdB min (6 - 1000) ’ 55:10 mm (C a 100) Input Resistance. Power On 100M!) ‘ ° Power Off ”Itfl min ° 74kg min Input Bias Current +0nA max ' 0 Open Input Detection Tune3 6 set (C I 1000) ' NA 120 sec (G I 50) ' NA Open Input Response Negative Overseale ° NA ANALOG OUTPUT Output Voltage Swing, 25v 0 ”ma - . Output Noise. dc - lOOltHz 0.8mV p-p 0 e Output Resistance Direct Output 0.19 ' ' Switched Output 359 ‘ ° CHANNEL SELECTION Channel Selection Time to 10.01% FS 2.5ms max ' ° Channel Scanning Speed 400 than/sec mln ' ° Olannel Select Input Reverse Voltage Rating JV max ‘ ° POWER SUPPLY Voltage Output 1V5 (Rated Performance) :ISV dc 110% ° ‘ (Operatirg) 112V to 118V dc max ° ' Oscillator evosc (Rated Performance) +13.$V to +24V ° ‘ Absolute max evosc +26V ' ' Current Output 1V5 - 115V 14mA max ° ° Oscillator +Vosc - +15V 40mA max ° ° Supply Effect on Offset Output :vs lOOuV/V RTO ' ' Oscillator 9V0; luV/V RTI ' ° ENVIRONMENTAL Temperature Rated Performance 0 to 00°C ° ' Operating -zs°c to «35°C ° ° Storage 45°C to 45°C ° ' Relatlve Humidity NON-CODdChSIm to «0°C 0 to 85% ‘ ° CASE SIZE 2" x 4' x 04" ° ' PRICES 1-9 3225 3275 $220 10-14 8205 $2 $6 8200 ‘00" 3144 $180 8140 'Speeafoeauona same as 1054A. ' Gain nonlinearity is petrified as a percentage of output signal wan representing peak deviation from the best Ira'ght line. eg nonlinearity at an output man of 10V plupli (25V) 'u 30.02% or :2mV. ' New time can be reduced by addition of external resistors. More than one open input may cause output to saturate on all channels. To prevent this. use external echoes for a positive overseale "pot-ac (F'ure ll. 'Proteeted for irons to ground and/or either napply voltage. Spec-meado- object so change without notice. APPENDIX B ZIBL Code Implementing The Dryer Control Algorithm ZIBL ' STATEMENTS FOR NEXT DO UNTIL IF (THEN) GOTO GOSUB RETURN REM (LET)VAR=exp LINK exp DTOA exp,exp TURNON LITE, RELAY or FLAG exp TURNOFF LITE, RELAY or FLAG exp READ DATA RESTORE DELAY exp PRINT PRINT°/oexp PR PR°/oexp LIST LIST°/oexp NEW CLEAR CLEAR GLOBALS RUN RUN NAME NAME NAME RENAME NAME INPUT lOAD END ' END NAME DIR STAT=exp TRACE ON TRACE OFF TIME=exp IN°/oexp OUT°/oexp ZIBL»! OPERATORS +, -, ', /, =, >, <, NOT >,=, < =, <> , AND, OR, XOR ZIBLS FUNCTIONS MOD(exp,exp) RND(exp,exp) ATODIex p) SENSE(exp) TOP STAT FREE(o) 113 ZIBL SUMMARY CHART ZIBL ' CONSTANTS l3 HEX DECIMAL “. . . STRINGS ZIBL“ VARIABLES A°/o TO 2% GLOBAL ARITH AS TO ZS GLOBAL 32 CHARACTER STRINGS A TO 2 LOCAL ARITH ZIBL " STRINGS ll = CHARACTER MASK IN STRING COMPARISONS 8- = REMAINDER OF STRING MASK IN STRING COMPARISONS ZIBL“ SPECIAL ADDRESSES VIDEO SCREEN FAOO TOTO" SENSE 0 TO 63 FEOO TO FE3F ATOD 0 TO 63 FE40 TO FE7F FLAG 0 TO 63 FEBO TO FEBF LITE 0 TO 63 FECO TO FEFF RELAY 0 TO 63 FFOO TO FF3F DTOA 0 TO 63 FF40 TO FF7F PORTS FFBO TO FFBF USER DEFINED FFCO TO FFFF PARALLEL PORT FFBI LITE PORT FFB2 TAPE PORT FF83 KEYBOARD PORT FFB4 SYSTEM INPUT PORT AT FFBOH BITS: BITO RXDO SERIAL RECEIVE CHANNEL 0 BITI CTSO CLEAR TO SEND, CHANNEL 0 BIT2 RXDI SERIAL RECEIVE CHANNEL I BIT3 CTSl CLEAR TO SEND, CHANNEL I BIT4 UNUSED BITS PIPST PARALLEL INPUT PORT STATUS BIT6 RTC REAL TIME CLOCK AS INPUT BIT7 ORXD CASSETTE RECEIVE CHANNEL SYSTEM OUTPUT PORT AT FFBO BITS: BITO TXDO SERIAL SEND CHANNEL 0 Bl" RTSO READY TO SEND CHANNEL 0 BIT2 TXDI SERIAL SEND CHANNEL l BIT3 RTSl READY TO SEND CHANNEL I BIT4 FLAGO INTERNAL CONTROL BITS FLAGI INTERNAL CONTROL BIT6 FLAG? INTERNAL CONTROL BIT7 RTCLR REALTIME CLK CLEAR DRYERPI 50 GOSUB 3000 70 Q%=0 : I% = 0: 0:0 100 RUN TEMPOUT 300 U5=l66 500 QS=QS+1 600 E%=P%-T% 620 E=E%/10 625 PR:PR 'ERROR= ”,E,“ [DEG. C*l0]" 650 IF E%>=150 GOTO 750 660 IF E%>-150 GOTO 2400 670 F=29 : GOTO 800 750 F=l00 800 GOSUB 4000 900 K%=-lS0000/(L%*F) 950 I%=I%+(E%*l00/U% 1100 M=K5*E%/1000 1300 R5=R5+M 1350 IP R5<=100 OR >=1200 GOTO 2000 1400 PR:PR ”NEW SPEED= ”,R% 1500 GOSUB 4000 1600 B%=((4*L%)+5/10 1700 GOTO 2500 2000 B%=8 2005 IF R%<=1000 R5=l00 :IF R%>=1200 R%=1200 2010 PR:PR "GRAIN FLOW LIMIT REACHED' 2020 PR:PR "CHECK INLET AIR TEMP." 2030 GOTO 2500 2400 B%=4 2410 PRzPR "NO SPEED CHANGE" 2500 RUN RPMSET 2600 GOTO 500 3000 PUT 12:PR 3010 PR “ENTER DESIRED OUTLET GRAIN TEMP., C 3030 PR : INPUT P%:P%=P%*10:RETURN 4000 Y=(1718*R%+12024)/100 4100 y=144000/y 4200 IF P%>T% GOTO 4500 4400 L%=(60*Y)/l00 4500 RETURN 5000 END 114 115 RPMSET 10 20 21 20 25 26 27 30 31 35 37 39 40 45 50 100 110 120 122 124 126 128 132 135 140 150 160 170 180 190 200 210 230 500 510 520 530 540 550 560 575 600 610 620 630 640 650 660 700 800 OUT%127:LINK #F003:OUT%0 B=0 : T=0 REM -- R=RESET RATE x 100000 REM —- R=RESET RATE x 100000 R=50 M%=I%*K%/l000 H-R%*100 IF Q%<=1 N%=0 DO : A%=0 : X=0 IF B%<>4 GOSUB 2000 W=W+M% R%=(H+W)/100 DO A%=0:X%=0 REM -- PI CONTROL LOOP FOR MOTOR SPEED DO GOSUB 500 E-R%-A IF E) 50 GOTO 132 IF E<-50 GOTO 132 K=48 GOTO 135 K=20 A%=A%+A E=E*10 N%=N%+(E*R/100000) M=K*(E+N%)/l0000 S%=S%-M IF S%<=35 S%=35:IF S%>=240 S%=240 @iFF81=S%:@#FF82=S% X=X+l REM -- C% IS USED FOR TIMING OF THE LOOP GOTO 575 C=0:S=0 DO:A=ATOD(3) A=(((A*1000)/4095)*25)/l0 S-S+A:C=C+1 UNTIL C=l0 A=S/C RETURN A%=A%/C% PUT 12 RUN TEMPOUT PR:PR "COMPUTED SPEED= ',R%,"[RPM]" PR ”SPEED= ”,A%,'[RPM]" PR:PR PR "SETPOINT TEMP= ',S$,'[DEG. C1" PR:PR “OUTLET GRAIN TEMP= ',O$,'[DEG. C]" PR ”INLET AIR TEMP= ",X%,'[DEG. C1" RUN TAPEOUT GOTO 45 1000 2000 2010 2020 2030 2040 2050 116 END Y=P%-T% IF Y)=50 GOTO 2050 IF Y>-50 GOTO 2040 GOTO 2050 B=B%-1 RETURN LI ST OF REFERENCES LIST OF REFERENCES Aguilar, C. S. and Boyce, D. S. 1966. Temperature ratios for measuring efficiency and for the control of driers. Jour. of Agr. Eng. Res. 11(1):19-23. Bailey, S. J. 1980. Moisture sensors 1980: On-line. roles increase. Control Engineering 27(9):112-117. Bakker-Arkema, F. w. 1981. Personal communication. Professor, Dept. of Agr. Eng., Michigan State University: East Lansing, MI. Bakker-Arkema, F. w., Brooker, D. B. and Hall, C. W. 1972. Comparative evaluation of crossflow and concurrent flow grain dryers. ASAE Paper No. 72-849. Am. Soc. Agr. Eng.: St. Joseph, MI. Bakker-Arkema, F. w., DeBoer, S. F., Lerew, L. E. and Roth, M. G. 1973. Energy conservation in grain dryers: I. Performance evaluation. ASAE Paper No. 73-327. Am. Soc. Agr. Eng.: St. Joseph, MI. Bakker-Arkema, F. w., Rodriguez, J. C., Brook, R. C. and Hall, G. E. 1981. Grain quality and energy efficiency of commercial grain dryers. ASAE Paper No. 81-3019. Am. Soc. Agr. Eng.: St. Joseph, MI. Beveridge, G. S. G. and Schecter, R. S. 1970. Optimization: Theory and Practice. McGraw-Hill: New York, NY. Bibbero, R. J. 1977. Microprocessors in Instruments and Control. John Wiley and Sons, Inc.: New York, NY. Blount Agribusiness. 1981. Commercial brochure. Montgomery, AL. Brook, R. C. 1977. Design of multistage grain dryers. Unpublished Ph.D. dissertation, Michigan State University: East Lansing, MI. 117 118 Brook, R. C. and Bakker—Arkema, F. W. 1980. Design of multistage corn dryers using computer optimization. Transactions of the ASAE 23(1):200-203. Burton, D. P. and Dexter, A. L. 1977. Microprocessor Systems Handbook. Analog Devices, Inc.: Norwood, MA. Cloud, H. A. 1957. Accuracy and limitations of automatic controls for crop drying. Presented at the 1957 winter meeting of ASAE. Colliver, D. 6., Peart, R. M., Brook, R. C. and Barret, J. R., Jr. 1979. Comparison of minimal energy usage management procedures for low temperature grain drying. ASAE Paper No. 79-3027. Am. Soc. Agr. Eng.: St. Joseph, MI. Coon, G. A. 1956. How to find controller settings from process characteristics. Control Engineering 3(5):66-76. Dalpasquale, V. A. 1981. Drying of soybeans in continuous-flow dryers and fixed-bed systems. Unpublished Ph.D. dissertation, Michigan State University: East Lansing, MI. DICKEY-john Corporation. 1980. CFMM-A Continuous flow moisture monitor, installation and operating manual. DICKEY-john Corporation: Auburn, IL. Fadum, O. and Shinskey, F. G. 1980. Saving energy through better control of continuous and batch dryers. Control Engineering 27(3):69-72. Franklin, G. F. and Powell, D. J. 1980. Digital Control of Dynamic Systems. Addison-Wesley Publ. Co. Inc.: Reading, MA. Harrell, R. C., Allison, J. M. and McLendon, B. D. 1979. Microprocessor-based control system for solar assisted grain drying. ASAE Paper No. 79-5524. Am. Soc. Agr. Eng.: St. Joseph, MI. Henry, 2. A., ed. 1975. Instrumentation and measurement for environmental sciences. ASAE Special Publication SP-0375. Am. Soc. Agr. Eng. St. Joseph, MI. 119 Hinkle, C. N. 1980. 8-Bits for a grain dryer. ASAE Paper No. 80-5021. Am. Soc. Agr. Eng.: St. Joseph, MI. Holtman, J. B. and Zachariah, G. L. 1969a. Computer controls for grain driers. Transactions of the ASAE 12(4):433-437. Holtman, J. B. and Zachariah, G. L. 1969b. Continuous crossflow modeling for optimal control. Transactions of the ASAE 12(4)430-432. Hukill, W. V. 1954. Storage of Cereal Grain and their Products. Am. Assoc. of Cereal Chemists: St. Paul, MN. Kalchik, S. J. 1977. Drying of soybeans in a pilot scale concurrent flow dryer. Unpublished M.S. thesis Michigan State University: East Lansing, MI. Jakeway, J. J. 1981. Personal communication. M.S. candidate, Michigan State University: East Lansing, MI. Manufacturing Engineering. 1980. A different type of control — industrial control microcomputers. Manufacturing Engineering 84(5):54-55. Matthews, J. 1963a. Automatic moisture content control for grain dryers. Jour. of Agr. Eng. Res. 8(3):207-220. Matthews, J. 1963b. The design of an electrical capacitance-type moisture meter for agricultural use. Jour. of Agr. Eng. Res. 8(1):17-30. Matthews, J. 1964. Performance of an automatic moisture control unit. Jour. of Agr. Eng. Res. 9:180-187. Melsa, J. L. and Schultz, D. G. 1969. Linear Control Systems. McGraw Hill: New York, NY. Morrison, R. 1977. Grounding and Shielding Techniques in Instrumentation. Second edition. John Wiley and Sons, Inc.: New York, NY. Muhlbauer, W. and Isaacs, G. W. 1975. Warmeruckgewinnung bei Kornerfruchttrocknungsanlagen. Landtechnik 24:161-164. 120 Omega Engineering, Inc. 1981. Temperature Measurement Handbook. Omega Engineering, Inc.: Stamford, Conn. Ott, H. W. 1976. Noise Reduction Techniques in Electronic Systems. John Wiley and Sons, Inc.: New York, NY. Sabbah, M. A., Keener, H. M. and Meyer, G. C. 1977. Simulation of solar grain drying using the logarithmic model. ASAE Paper No. 77—3012. Am. Soc. Agr. Eng.: St. Joseph, MI. Sen, R. K. 1981. Personal communication. President, Diversified Engineering, Inc.: Richmond, VA. Sen, R. K. 1980. Personal communication. President, Diversified Engineering, Inc.: Richmond, VA. Sheingold, D. H., ed. 1981. Transducer Interfacing Handbook. Analog Devices, Inc.: Norwood, MA. Shinskey, F. G. 1978. Energy Conservation through Control. Academic Press, New York, NY. Short, K. L. 1981. Microprocessors and Programmed Logic. Prentice-Hall, Inc.: Englewood Cliffs, NJ. Smith, C. L. 1979. Fundamentals of control theory. Chemical Engineering 86(22):11-39. U. 8. Dept. of Energy. 1979. Computer technology: Its potential for industrial energy conservation. Technology Applications Manual DOE/CS/2123-T2: Technical Information Center, U. S. Dept. of Energy: Washington, D. C. Westelaken, C. M. 1981. Personal communication. Blount Inc.: Montgomery, AL. Wise, K. D., Chen, K. and Yokely, R. E. 1980. Microcomputers: A Technology Forecast and Assessment to the Year 2000. John Wiley and Sons: New York, NY. Yang, L. C. 1981. Developments in moisture sensors. M & C Measurements and Control 15(1):98-108. Zachariah, G. L. and Isaacs, G. W. 1966. Simulating a moisture-control system for a continuous-flow drier. Transactions of the ASAE 9:297-302. 121 Zagorzycki, P. E. 1979. Automatic control of conveyor dryers. Chemical Engineering Progress, 19(82):50-56.