- Email: [email protected]
Data acquisition and control system for experimental thin-layer drying study
Computers and Electronics in Agriculture, 3 (1989) 225-241 Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands
225
Data Acquisition and Control System for Experimental Thin-Layer Drying Study R.K. BYLER 1, J.B. GERRISH 2 and R.C. BROOK2 1Agricultural Engineering Department, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061 (U.S.A.) 2Agricultural Engineering Department, Michigan State University, East Lansing, MI (U.S.A.) (Accepted 11 November 1988)
ABSTRACT Byler, R.K., Gerrish, J.B. and Brook, R.C., 1989. Data acquisition and control system for experimental thin-layer drying study. Comput. Electron. Agric., 3: 225-241. A microprocessor-based system was constructed to control experimental conditions in the study of the moisture content of agricultural products. The equipment also recorded the condition of the air and the sample weights during the thin-layer drying tests for these products. This paper describes the hardware and software which were developed for use in the study of the properties of parboiled rice. Two samples were studied simultaneously at 13 combinations of relative humidity and temperature. The drying conditions were held constant enough that the variation in conditions in the drying chamber was shown to have no measurable effect on the thin-layer drying data, producing data repeatable to 0.001 moisture content dry basis.
INTRODUCTION E v a l u a t i o n of a l t e r n a t i v e d r y e r designs usir/g c o m p u t e r s i m u l a t i o n models requires far less t i m e a n d m o n e y t h a n building a n d t e s t i n g t h e actual equipm e n t . A model of t h e p r o d u c t a n d a c o m p a t i b l e model of t h e d r y i n g e q u i p m e n t being designed allow t h e e n g i n e e r to evaluate t h e design of d r y i n g e q u i p m e n t a n d the r e s p o n s e of t h e p r o d u c t - e q u i p m e n t c o m b i n a t i o n . F r o m the scientific s t a n d p o i n t , basic i n f o r m a t i o n can be o b t a i n e d a b o u t the d r y i n g process b y s t u d y i n g t h e c h a n g e in m o i s t u r e c o n t e n t of agricultural p r o d u c t s as t h e y dry. T h e p u r p o s e of this p a p e r is to describe t h e e q u i p m e n t w h i c h was used to collect t h i n - l a y e r d r y i n g d a t a for agricultural p r o d u c t s . T h e e q u i p m e n t in a d r y i n g s t u d y s h o u l d c o n t r o l t h e m a i n i n d e p e n d e n t variables, the d r y i n g air t e m p e r a t u r e a n d relative h u m i d i t y . It m u s t also m e a s u r e t h e m a i n d e p e n d e n t variable, m o i s t u r e c o n t e n t o f t h e p r o d u c t . All of t h e s e variables m u s t be measured at k n o w n t i m e s a n d recorded. T h e m o i s t u r e r e l a t i o n s h i p s of several p r o d u c t s have b e e n studied w i t h this p a r t i c u l a r set of e q u i p m e n t . T h i s p a p e r
Fig. 1. Block diagram of the information
I I
flow of the process controller
Temperature Sensor Interface
-7
DIgItal
Cassette Data storage
and data collection system.
Microcomputer
Output
Keyboard
Video
El&If--!Cl
Printer
r
Line
22"7
will discuss how the equipment was used to study the properties of parboiled rice. MICROCOMPUTER
Figure I shows the information flow in the data acquisition and process control equipment. There were two types of sensors in the system, weight sensors and temperature sensors. Temperature sensors were located in the water bath, near the electric air heater, in the air stream as it entered the study chamber, and near the samples in the chamber. The product was located in the chamber on trays which were attached to load cells so that the weight of the sample could be monitored. The microcomputer contained instructions to: use the information to control the air conditioning unit; display the data on paper through the line printer; and store the data on digital tape for subsequent analysis. The microcomputer consisted of four boards interconnected by an S-100, Institute of Electrical and Electronic Engineers (IEEE) standard 696, bus. The heart of the microcomputer was a Z-80 microprocessor on a Cromemco Single Card Computer, SCC (Cromemco, 1980) which also contained 8 Kb of read only memory (ROM), and 1 Kb of random access memory (RAM). Other functions contained on the Cromemco SCC were a serial port (for data transfer with the printer or video terminal), a parallel port (for control of the air conditioning unit), a second parallel port (for communication with the digital cassette deck), and one status port (for control of the communication with the digital tape deck). A California Computer Systems 16 Kb RAM board (California Computer Systems, undated) and a Vector Graphics 12 Kb ROM/RAM board which had 1 Kb of RAM and up to 12 Kb of ROM (Vector Graphics, undated) were used. The final part of the system ws the TecMar analog to digital (A/D) converter unit (TecMar, 1980). This two-board set contained a Data Translation 5712 module which provided software controlled gain, multiplexing of 16 input channels and 12-bit analog-to-digital conversion. Because the A/D converter was multiplexed, the eight transducer signals were connected directly to the module after signal conditioning. The 12-bit A/D converter provided a precision of 1 part in 4096, adequate for the range of values encountered in this study. This unit also contained an Advanced Micro Devices Am9513 timing controller chip, used as a 'real-time' clock which provided time in days, hours, minutes, and seconds as well as periodic interrupts to the microprocessor as required by the control and data acquisition software. TRANSDUCERS AND MICROCOMPUTER INTERFACE
Because only temperatures and weights were measured, there were only two types of sensors and two signal-conditioning circuits to consider. After the
228 signal-conditioning equipment was constructed and the transducers connected to the microcomputer system, the transducers were calibrated.
Temperature transducer The temperature sensor was the National Semiconductor precision temperature sensor LM335 (National Semiconductor, 1980). This sensor was a solidstate integrated circuit device in which the voltage output was linearly related with absolute temperature at about + 0.01 V / K . The rated operating temperature range was - 10 ° C to + 100 ° C and the corresponding typical nonlinearity over that range was 0.3 °C (National Semiconductor, 1980). Because the design temperature range for the laboratory equipment was at most 50 °C and the majority of the expected nonlinearity was at elevated temperatures (National Semiconductor, 1980) the expected nonlinearity in the range 5°C to 50°C was no more than 0.15°C. After the temperature sensor circuits were assembled they were tested against a laboratory mercury thermometer marked in 0.1 ° C. No nonlinearities were detected with any of the six sensors used. In all cases the correlation coefficient, r for a linear, two-parameter equation, was greater than 0.9999. Six temperature transducers were connected to allow for three to be used in the control algorithm leaving three to measure the air temperature around the samples. The use of three sensors allowed the temperature readings to be averaged and also provided backup in case of sensor failure.
Temperature transducer signal conditioning The circuit used with the temperature sensors is shown in Fig. 2. Because the sensors produce 0 V output at 0 K (nominal) and have a slope of 0.01 V / K (nominal) they produce at least 2.5 V at normal drying temperatures. Two and a half of volts were subtracted from the output of the sensors before the signal was converted to a digital signal. This circuitry allowed the gain of the A / D converter to be higher, producing a correspondingly higher precision. A National Semiconductor LM336 integrated circuit voltage reference (National Semiconductor, 1980) was used to produce approximately - 2 . 5 V to add to the voltage output of the temperature sensors.
Weight transducer The weight transducer was the 2-pound model 4850 built by GSE Inc (GSE, 1982). This device was a strain gauge device and its rated nonlinearity was 0.02% of full scale, or 0.2 g, with half of that being typical The maximum operating temperature range was - 18 ° C to 93 ° C (0-200 ° F); the temperature effect on rated output was 0.0014%/°C (0.0008%/°F) and the temperature
229
CHANNE !
LM 335
"
R5
~
"
CI
I
A/D CDNVERTER R5
I
'
LM 335
~
L
M
335
VPef
LM 336 '
~5V DC RB
5i !
~
i
/
i
i
R9
Fig. 2. Circuit diagram of the temperature transducer interface.
effect on zero was 0.0027%/°C (0.0015%/°C). Two of the devices were used to allow two samples to be studied simultaneously with the resulting data sets~ labeled A and B. The transducers were located outside the chamber to minimize temperature effects, and in the case of high-humidity conditions withi~ the chamber, to isolate the weight transducers from moisture. The transducers were designed to be insensitive to moments and forces other t h a n in the one direction of desired loading. The rated error per inch of off-. center loading at 1/2 capacity was 0.004% of full scale, or 0.04 g. The trans-. ducers were loaded with a 100 g mass alternately at each of the corners and at the center of the sample holder. No discernible correlation between the mas~,~ location and the number from the computer was noted. The weight transducers were loaded with known masses and the correspond.ing number in the microcomputer was noted. Since the relationship betweer, the weight and the number from the microcomputer should be linear (accord-ing to the manufacturer's literature) a linear lest squares curve fit was run to obtain the conversion factors from the number in the microcomputer to the actual weight in grams. In all cases the correlation coefficient, r, was greater t h a n 0.9999.
Weight transducer signal conditioning An Analog Devices (1982), model 2B31J, strain gauge signal conditioner was used as shown in Fig. 3. This device included an i n s t r u m e n t a t i o n amplifier with gain from 1 to 2000 and a low pass filter with a time constant of 0.5 s. The excitation of the strain gauge bridge was regulated at 6.9 V with a National
230 ADv~TIDHNAL ]
o -
[ADDITIONAL1
CHANNEL
] F~S~] '1--~..[~-~ ~ IGI~E~I !.RG~_~b~~, ~L~ /I/ 14-~~'.~EI 'l A/D CONVERTER L4B5o ~ INST~ATIONI BUFFER LOF~P L~RS WEIGHTTRANSDUCER
0 V e??
GDO
RI
Fig. 3. Circuit diagram of the weight transducer interface. Semiconductor (1980) LM399, a temperature-stabilized integrated circuit precision voltage reference. A second LM399 was used with a voltage divider to provide an offset voltage which produced nearly 0 V out of the strain gauge signal conditioner when there was no sample load on the weight transducer. A metal film resistor was used to set the gain on the strain gauge signal conditioner at approximately 1880. This gain was near the maximum allowable on the unit and produced a precision of about 0.016 g when used with the A / D converter; while the transducer itself had a worst case precision of 0.26 g. The total range with a gain of 1880 was about 65 g which was adequate for the testing of grain. AIR CONDITIONING UNIT An Aminco-Aire (Aminco-Air, 1967) model J4-5460, a commercial air conditioner unit, was used in the study. It can condition up to 8.5 m3/min (300 cu f t / m i n ) . This unit was designed to condition relatively small chambers of less than 1.1 m 3 (40 cu ft). The chamber used in this test had a volume of approximately 0.24 m 3 and about 2.5 m3/min of conditioned air were circulated through it. In the Aminco-Aire unit, the control of air moisture and temperature was achieved in to steps (see Fig. 4). First the air passed into the larger chamber where a mist of water droplets sprayed into the air. The droplets fell into a water bath the temperature of which was maintained by a refrigeration unit or an electric resistance heater. Each water droplet was surrounded by air and thus heat and moisture transfer occurred between the air and water droplet. The air then passed into a second chamber where it was heated by electric
231 AMINCO AIRE UNII ill
HI:ATE
-~=.--
REFRIGERATION UNIT
PRODUCT
I
::: . e l
I
SCRE~:IN
J[
i
4---
-
Fig. 4. S c h e m a t i c of t h e air flow in t h e s y s t e m .
heaters. The air passed through the test chamber and was recirculated into the first chamber. The microprocessor-based circuitry controlled the temperature of the water bath and the duty cycle of the electric heaters. Because the relative humidity was not measured directly, errors encountered in measuring humidity were avoided, A unique correlation between the Aminco water temperature, air dry bulb temperature and the relative humidity of the air was used to estimate the relative humidity. A chart of this relationship was provided by the manufacturer and the formula which was used was obtained by regression. This formula was: DB----P1 +P2
in (RH) +P3
WT in (RH)+P4/RH+P5
WT
(1)
where DB is the dry bulb temperature ( ° C ), RH the relative humidity ( % ), WT the temperature of the water bath ( ° C), P1 ----77.4, P2-- - - 16.98, P3 = - - 0.0821, P4 ----73.7, and P5 -- 1.377. The Aminco-Aire operated on 208 VAC three phase power, so voltage and power amplification were accomplished as shown in Fig. 5. The switches were operated from an 8-bt parallel output port available on the microprocessor board. An additional safety feature was added in case the microprocessor con-trol became inoperable for any reason. All power to the Aminco-Aire was re-moved if a signal was not received from the microprocessor at least every 10() ms. This control was implemented with a resettable, retriggerable, monostable multivibrator attached to a solid state relay.
232 24VDC ISOLATED FROM
180 VAC
OTHER DC POWER
?
?
REL 20
REL 4 ¢0 VATER HEATER DO
REL 21
REL 2 ¢o rATER REFRIGERATION B,L, RESET
AHINCO CIRCUITS
D•
SOLID STATE
G
D4~,~-['--'-D6 o NC
1
~
t
VALID
o
G
RELAY 2
G
SOLID STATE RELAY 3
A
AIR HEATER
i
°
-o
g
RETRIDDERABLE
MONOSTABLE
120 VAC AHINCO --oHAIN CDNTAOTDR
TC 0.I S{CI
i
)
+5V
+BY DCo
"
r
D7 o NO
GROUNDo
AIR HEATER i
SOLID STATE
95 o NC
DATA ~
RELAY L
C
TO ALL CHIPS
VOLTAGE REGULATOR
Fig. 5. Circuitdiagram of the air conditioner control interface.
STUDY CHAMBER A sketch of the study chamber is shown in Fig. 6. The chamber was 0.50 m by 0.30 m by 1.60 m and was constructed mostly of plywood, which was painted with several coats of shellac to reduce moisture absorption. It was lined in all but the viewing area with 0.025 m ( 1 in) of rigid foam insulation. A section of
Fig. 6. Schematic of the study chamber.
233 the top and sides was removable for access to the chamber. The test chamber was connected to the air conditioning unit by insulated ducts. The center of the load cells was located 1.05 m from the front of the chamber and a honey-comb-shaped metal grid acting as a 'flow straightener' was located 0.29 m behind the load cells. The airflow was found to be parallel to the sides of the chamber in the vicinity of the grain sample and its velocity was measured with a hot-wire anemometer at 21 points in a plane normal to the side and b o t t o m of the chamber at the rear of the sample holders. The velocity was observed to vary from a minimum of 0.10 m / s to a maximum of 0.70 m / s with an average of 0.28 m/s. Henderson and Pabis (1962) found that, in this range, variations in airflow have an "insignificant' effect on the rate of drying. The weight transducers were located 0.25 m apart and were fastened to eL concrete block 0.18 m by 0.30 m by 0.35 m in order to reduce mechanical vibrations. The sample holders consisted of two trays 0.20 m by 0.30 m made o:t' aluminium screen with a lip of about 10 mm around the edge to hold the sample on the tray. There were two trays attached by an aluminum rod to each weight transducer. The b o t t o m tray was about 0.10 m from the b o t t o m of the chamber and the trays were about 0.10 m apart resulting in the top tray being located about 0.10 m from the top of the chamber. SOFTWARE The software for data acquisition and digital control was written in three languages: Z-80 assembly language, BASIC, and Pascal. A few primitive routines, which were used often and were critical to the timing were written in Z80 assembly language. The instructions for data display and storage, and the overall control of the equipment were written in integer BASIC. The data acquisition and air conditioning control software was written in Pascal and compiled. The code written in Pascal was called by an interrupt while programs written in BASIC were running. This software implementation made it appear that data storage and control were occurring simultaneously but actually the data acquisition and control functions had priority over data storage and display.
Data acquisition At the end of each second for the first 9 seconds, each of the eight analog input lines was sampled and converted to digital form. After the nine values for each of the eight input lines were obtained, they were averaged and the average stored in memory by the end of the 10th second. Also, before the end of the 10th second the average values representing the water temperature, heater temperature, and air temperature at the entrance to the test chamber were converted to engineering terms for use by the control algorithms.
234 The average of the digital representation of the eight transducer values was placed in memory at the end of the ten second period and a flag was set in memory. Software in BASIC responded to this flag and reset it before moving the averages to other memory and subsequently calculating averages over a 1min period. The 1-min averages were stored on cassette tape and printed at the terminal. The shortest period covered in the recorded data was one minute, with each data point representing an average of 54 observations. The averages of the weight on each transducer, the water temperature, and the five air temperatures were recorded with the reading of the clock in days, hours, minutes and seconds. The moisture content of the product on each sample holder was calculated from the sample weights and the sample dry matter content. The drying air relative humidity was calculated from the Aminco water temperatures and the air temperatures near the product.
Digital control The water temperature was controlled by an on/off algorithm. A dead band of 0.16°C was found to be the minimum which did not induce oscillations. Initially, the heating and cooling of the water was observed to overshoot the goal, especially when large changes in the setting were requested. Therefore, in addition to the simple on/of control algorithm, the set temperature was compared with the actual water temperature at the end of each second. If the two temperatures were within 0.05 ° C of each other, the heater or cooler was turned off for the remainder of the 10-s period. This adjustment allowed a narrower dead band without undesirable oscillations between the heater and refrigeration unit. The control algorithm for the air temperature was first designed as a simple velocity proportional-integral-differential (PID)controller (Bibbero, 1977). Initially temperature sensors near the product were used as the basis of control. It became immediately obvious that the heating elements, which had considerable thermal mass, became too warm before sufficient heat reached the temperature control sensors. The controller then reduced the duty cycle of the heater but because of the thermal mass of the heating element reduced it too much. This problem produced continual fluctuations of the chamber temperature. A simple solution to this problem was to base the control on a sensor located near the heater. It was assumed that under steady state conditions the heat losses would be relatively constant between the heater and the product so that the chamber conditions, while not directly controlled, would be constant. An algorithm in the form of equation (2) was adjusted using the method described by Smith (1979). Controlling the heater temperature did not allow the operator to choose the temperature in the chamber directly but a guess of
235
the temperature drop from the heater to the chamber would allow the chamber temperature to be set fairly closely. Locating the sensor near the source of the energy, the electric heater, made the control reasonably simple. The P I D velocity equation (Bibbero, 1977) was:
Ds= I (E o - E1) + JEo + H ( 3 E o - 4 E 1 - E2 + 2E3)
(2)
where DS is the change in setting of the control e l e m e n t , / , J, H are gains of the proportional, integral and differential parts of the algorithm, and E , the errors at time = t - n X At. The derivative action of the P I D controller was observed to be erratic, because of the noise in the temperature signal. Attempts were made to filter the signal both with analog filters of various orders and with digital filters. A simple RC analog filter with a time constant of 2.2 ms, was used in addition to the digital filter shown in equation (2). The derivative action was probably not: necessary in this application. At the end of each 10-s period the updated error in air temperature was computed and an adjustment to the duty cycle for the heater was calculated. During each interrupt, occurring every 40 ms, the current duty cycle was checked to see if the heater should be turned off or on. The duty cycle was based on 100 of the interrupts so that if the duty cycle were 46, for example, the heater would be on for 1.84 s and off for 2.16 s. This period of 4 s was short enough so thin the heater temperature did not change noticeably due to its thermal mass. This algorithm had two main weaknesses: it was difficult to repeat an experiment because it was difficult to obtain a given chamber temperature, and the heat load in the room was not constant so the chamber temperature tended to drift. Therefore, the control algorithm was modified between tests 9 and 10. Instead of the operator trying to guess what the heater temperature shouk! be to achieve a desired temperature in the chamber, the additional software adjusted the heater temperature setting. This two stage control process is sometimes referred to as cascade control. For this second level of control an integral controller was used. The goal of the original P I D controller was adjusted by 0.02 times the difference between the actual temperature at the entrance of the chamber and the desired temperature. This additional controller made the system much slower in responding to changes in the set point but the variation of temperature within the chamber was significantly reduced and the actual temperature within the chamber was usually within 0.5°C of the set temperature. DATA C O L L E C T I O N P R O C E D U R E
The parboiled rice which was tested was collected from a commercial parboiling plant and stored at 5 °C in sealed containers until needed for testing. T w e n t y one tests were conducted and were numbered consecutively. The first
236
two tests were considered to be 'warm-up' tests and were not analyzed completely. First the desired temperature and relative humidity of the test were determined and the equipment was given at least 1 h to stabilize at the new settings. The product which was to be tested was weighed on an analytical balance then loaded onto the trays and the data recording was begun. Separate samples were also placed into an oven for determination of initial dry matter content by the oven method. The tests were run for between 40 and 47 h at which time the dried product was again weighed on an analytical balance and samples taken to determine the final dry matter content. The temperature and weight integers were then converted to actual temperature, relative humidity and moisture content data and transferred to the mainframe computer for statistical analysis. The first step in the analysis of the moisture loss data was to plot the moisture content vs. time for each data set. Visual examination of these plots indicated poor data sets but no individual data points were removed from the data sets. Since all of the plots were at the same scale, the different data sets were also compared visually. Tests 3, 4 and 11 were eliminated because of mechanical problems with the equipment which were evident in the data plots. DATA COLLECTION AND SYSTEM VERIFICATION
The remaining sixteen data sets were kept for further analysis. These 16 tests on each of the two trays covered a total of 1308 h of data, with one data point every minute. This amounts to over 78 500 data points of moisture content vs. time. Because each data point represented an average of 54 observations the data sets resulted from over 4 million moisture content observations. Several of the data sets are represented by plots of the moisture content vs. time, Figs. 7, 8, 9 and 10. The data plots were drawn by positioning the pen at the first data point and then drawing a line to each following data point. The smoothness of the curves show the quality of the data in these data sets. The repeatability of the data can also be seen by comparing tests 20 and 21 since test 21 was a repeat of test 20 in all possible ways. The thin layer study equipment maintained constant conditions as shown in Table 1. The distribution of the temperatures and calculated relative humidities did not follow a Gaussian distribution because these two variables were controlled by the digital equipment. The control algorithm required measurable errors to occur. But after they occurred, substantially larger errors were unlikely to occur because of the feedback control. Therefore, the usual statistic of central tendency, standard deviation, was misleading. W h a t was calculated instead was the mean value and the range which includes at least 90% of the data points. The temperature and relative humidity each minute, after the first half hour of drying, were included in this analysis
23'7
Ii i i
I
\
\
i ! I
1
]
I j~
[ "4
•
i; ¸¸
,
Fig. 7. Moisture content data for test 18A.
I . ] \\ %..
27. jll
-
i
J
.
]
Fig. 8. Moisture content data for test 18B. for the r e m a i n i n g period of the run. For m o s t of the tests there were at least 2400 data p o i n t s in this analysis. D u r i n g t h e first half hour the o p e n i n g of the chamber affected the drying c o n d i t i o n s and the results of an analysis including
238
{
I
5I
] {o
0.30 t 0,25
LI o.2o ; \ u
H
00.10
O.OS
0.00o
.,15
2Io DRYING
AS
30
•IME
,
35
L,O
]
I
7
U
3S
~3
~5
$9
H~I)i.{R5
Fig. 9. Moisture content data for test 20A.
3.30!2
I
l
l
I
I
213
T
i
o.2~';i
F b~ O0.]C•
I
0.05
O. '00 I 0
,
q
]1[,
!b
,
95 DRY I NG • 7 Vl[i
I
,
313 H,D,.!R!~
I
Fig. 10. Moisture content data for test 21A.
all data would not indicate the effectiveness of the control for the majority of the time. The improvement due to the cascade control introduced between tests 9 and
239
TABLE 1 Variation in chamber conditions Test
5 6 7 8 9 10 12 13 14 15 16 17 18 19 20 21
Relative humidity (%)
Average chamber temperature ( ° C )
Percentile
Percentile
5
50
95
5
50
95
23 24 39 50 46 27 49 50 40 40 51 27 26 31 46 47
24 25 41 51 47 27 50 51 40 40 53 27 27 31 47 47
24 26 42 52 49 28 50 52 41 41 54 28 27 31 47 48
34.4 30.9 32.0 32.9 39.5 40.3 24.9 17.1 24.8 32.3 39.5 40.2 39.5 37.8 26.1 25.8
34.8 31.5 32.5 33.4 40.2 40.6 25.2 17.3 25.0 32.6 39.8 40.6 39.8 38.0 26.3 26.1
35.3 32.2 33.0 33.6 40.7 40.8 25.4 17.6 25.3 32.9 40.4 41.0 40.2 38.2 26.5 26.4
10 can be seen. Before the implementation of the cascade control algorithm, the range of chamber temperatures was about 1.0 ° C; after the change the range was about 0.5 ° C. After the control algorithm was improved the temperaturefrequency distribution was much more square with a relatively uniform distribution over several tenths of a degree about the mean and a very rapid decrease in the number of observations at both higher and lower temperatures. The water temperature was less variable than the air temperature. The range which included over 90% of the water temperature observations was 0.1°C while the dry bulb temperature varied several tenths of a degree. Therefore the degree of variation in the chamber relative humidity depended on the degree of variation of the dry bulb temperature in the chamber. The average range of relative humidities before the algorithm change was 0.021 (2.1% RH ) and after the change was 0.010. Because the water temperature was virtually unchanged, the saturation vapor pressure did not change during the tests. The real question about the adequacy of control was whether the variations in the independent variables significantly affected the drying behavior of the product. The best measure of the effect of uncontrolled variation in the independent variables is to study the data from repeat tests. Tests 20A and 20B, were fit with one equation. Samples 20A and 20B had different initial moisture contents but since they were in the chamber at the same time there could be
240
no error due to differences in t e m p e r a t u r e or relative humidity during the test. An equation which was fit to the two actual data sets allowed for a different initial moisture content. T hi s equation fit the data with a residual mean square of 1.3E-6. Th is was the error a t t r i b u t e d to the difference in weight t ransducer plus the i n h e r e n t error. F r o m the calibration of the weight transducers, it was known t h a t th er e was no measurable difference between the transducers so this error was the i n h e r e n t error. Therefore, 1.3E-6 was used as the best estimate of the variance of these data. N e x t data from tests 20A and 21A were combined. T hese two tests were repeats as nearly as possible in all respects. T h e equation was fit to the actual data producing a residual m e a n square of 1.45E-6. Likewise the data from 20B was combined with the data from 21B resulting in a residual mean square of 1.19E-6. T h e source of the error from these combinations was due to uncontrolled variation in t e m p e r a t u r e and relative humidity between the two tests plus the i n h e r e n t error. Since t he error in combining tests 20 and 21 was virtually the same as the i n h e r e n t error there was no measurable effect on the drying data due to the variation in relative humidity and t e m p e r a t u r e during these tests. CONCLUSIONS
T h e microprocessor-based system was able to m ai nt ai n a relative humidity c o n s t a n t to + 1% RH and a dry bulb t e m p e r a t u r e c o n s t a n t to + 0 . 3 ° C for periods of over 40 h. T h e microprocessor-based system was able to collect and store a great deal of data, about 78 000 m o i s t u r e - c o n t e n t time data pairs in this experiment, with limited e x p e r i m e n t e r effort. T h e experimental conditions were well controlled producing moisture cont e n t data which was repeatable, with an estimated variance of 1.3E - 6.
REFERENCES Aminco-Aire, 1967. Aminco Climate Lab Assembly, Catalogues 4-5576 and 4-5577. American Instrument Co., Inc., Silver Spring, MD, 34 pp. Analog Devices, 1982. Data-acquisition databook 1982, Volume II: modules-subsystems. Analog Devices, Inc., Norwood, MA 02062, 723 pp. Bibbero, R.J., 1977. Microprocessors in Instruments and Control. Wiley, New York, 301 pp. California Computer Systems, undated. 16K static RAM module owner's manual, Model 2016B. California Computer Systems, Sunnyvale, CA, 25 pp. Cromemco, 1980. Single card computer instruction manual. Cromemco, Inc., Mountain View, CA, 23 pp. GSE, 1982. High performance single point load cell. Bull. 137, GSE Inc., Farmington Hills, MI, 28 pp.
241 Henderson, S.M. and Pabis, S., 1962. Grain drying theory, IV: the effect of airflow rate on the drying index. J. Agric. Eng. Res., 7: 85-89. National Semiconductor, 1980. Linear Databook. National Semiconductor Corporation, Santa Clara, CA, 1673 pp. Smith, C.L., 1979. Fundamentals of control theory. Chem. Eng., 86 (22): 11-39. TecMar, 1980. TecMar S-100 AD212. TecMar Inc., Cleveland, OH, 8 pp. Vector Graphics, undated. 12K PROM/RAM assembly instructions and users guide, 12 pp.
225
Data Acquisition and Control System for Experimental Thin-Layer Drying Study R.K. BYLER 1, J.B. GERRISH 2 and R.C. BROOK2 1Agricultural Engineering Department, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061 (U.S.A.) 2Agricultural Engineering Department, Michigan State University, East Lansing, MI (U.S.A.) (Accepted 11 November 1988)
ABSTRACT Byler, R.K., Gerrish, J.B. and Brook, R.C., 1989. Data acquisition and control system for experimental thin-layer drying study. Comput. Electron. Agric., 3: 225-241. A microprocessor-based system was constructed to control experimental conditions in the study of the moisture content of agricultural products. The equipment also recorded the condition of the air and the sample weights during the thin-layer drying tests for these products. This paper describes the hardware and software which were developed for use in the study of the properties of parboiled rice. Two samples were studied simultaneously at 13 combinations of relative humidity and temperature. The drying conditions were held constant enough that the variation in conditions in the drying chamber was shown to have no measurable effect on the thin-layer drying data, producing data repeatable to 0.001 moisture content dry basis.
INTRODUCTION E v a l u a t i o n of a l t e r n a t i v e d r y e r designs usir/g c o m p u t e r s i m u l a t i o n models requires far less t i m e a n d m o n e y t h a n building a n d t e s t i n g t h e actual equipm e n t . A model of t h e p r o d u c t a n d a c o m p a t i b l e model of t h e d r y i n g e q u i p m e n t being designed allow t h e e n g i n e e r to evaluate t h e design of d r y i n g e q u i p m e n t a n d the r e s p o n s e of t h e p r o d u c t - e q u i p m e n t c o m b i n a t i o n . F r o m the scientific s t a n d p o i n t , basic i n f o r m a t i o n can be o b t a i n e d a b o u t the d r y i n g process b y s t u d y i n g t h e c h a n g e in m o i s t u r e c o n t e n t of agricultural p r o d u c t s as t h e y dry. T h e p u r p o s e of this p a p e r is to describe t h e e q u i p m e n t w h i c h was used to collect t h i n - l a y e r d r y i n g d a t a for agricultural p r o d u c t s . T h e e q u i p m e n t in a d r y i n g s t u d y s h o u l d c o n t r o l t h e m a i n i n d e p e n d e n t variables, the d r y i n g air t e m p e r a t u r e a n d relative h u m i d i t y . It m u s t also m e a s u r e t h e m a i n d e p e n d e n t variable, m o i s t u r e c o n t e n t o f t h e p r o d u c t . All of t h e s e variables m u s t be measured at k n o w n t i m e s a n d recorded. T h e m o i s t u r e r e l a t i o n s h i p s of several p r o d u c t s have b e e n studied w i t h this p a r t i c u l a r set of e q u i p m e n t . T h i s p a p e r
Fig. 1. Block diagram of the information
I I
flow of the process controller
Temperature Sensor Interface
-7
DIgItal
Cassette Data storage
and data collection system.
Microcomputer
Output
Keyboard
Video
El&If--!Cl
Printer
r
Line
22"7
will discuss how the equipment was used to study the properties of parboiled rice. MICROCOMPUTER
Figure I shows the information flow in the data acquisition and process control equipment. There were two types of sensors in the system, weight sensors and temperature sensors. Temperature sensors were located in the water bath, near the electric air heater, in the air stream as it entered the study chamber, and near the samples in the chamber. The product was located in the chamber on trays which were attached to load cells so that the weight of the sample could be monitored. The microcomputer contained instructions to: use the information to control the air conditioning unit; display the data on paper through the line printer; and store the data on digital tape for subsequent analysis. The microcomputer consisted of four boards interconnected by an S-100, Institute of Electrical and Electronic Engineers (IEEE) standard 696, bus. The heart of the microcomputer was a Z-80 microprocessor on a Cromemco Single Card Computer, SCC (Cromemco, 1980) which also contained 8 Kb of read only memory (ROM), and 1 Kb of random access memory (RAM). Other functions contained on the Cromemco SCC were a serial port (for data transfer with the printer or video terminal), a parallel port (for control of the air conditioning unit), a second parallel port (for communication with the digital cassette deck), and one status port (for control of the communication with the digital tape deck). A California Computer Systems 16 Kb RAM board (California Computer Systems, undated) and a Vector Graphics 12 Kb ROM/RAM board which had 1 Kb of RAM and up to 12 Kb of ROM (Vector Graphics, undated) were used. The final part of the system ws the TecMar analog to digital (A/D) converter unit (TecMar, 1980). This two-board set contained a Data Translation 5712 module which provided software controlled gain, multiplexing of 16 input channels and 12-bit analog-to-digital conversion. Because the A/D converter was multiplexed, the eight transducer signals were connected directly to the module after signal conditioning. The 12-bit A/D converter provided a precision of 1 part in 4096, adequate for the range of values encountered in this study. This unit also contained an Advanced Micro Devices Am9513 timing controller chip, used as a 'real-time' clock which provided time in days, hours, minutes, and seconds as well as periodic interrupts to the microprocessor as required by the control and data acquisition software. TRANSDUCERS AND MICROCOMPUTER INTERFACE
Because only temperatures and weights were measured, there were only two types of sensors and two signal-conditioning circuits to consider. After the
228 signal-conditioning equipment was constructed and the transducers connected to the microcomputer system, the transducers were calibrated.
Temperature transducer The temperature sensor was the National Semiconductor precision temperature sensor LM335 (National Semiconductor, 1980). This sensor was a solidstate integrated circuit device in which the voltage output was linearly related with absolute temperature at about + 0.01 V / K . The rated operating temperature range was - 10 ° C to + 100 ° C and the corresponding typical nonlinearity over that range was 0.3 °C (National Semiconductor, 1980). Because the design temperature range for the laboratory equipment was at most 50 °C and the majority of the expected nonlinearity was at elevated temperatures (National Semiconductor, 1980) the expected nonlinearity in the range 5°C to 50°C was no more than 0.15°C. After the temperature sensor circuits were assembled they were tested against a laboratory mercury thermometer marked in 0.1 ° C. No nonlinearities were detected with any of the six sensors used. In all cases the correlation coefficient, r for a linear, two-parameter equation, was greater than 0.9999. Six temperature transducers were connected to allow for three to be used in the control algorithm leaving three to measure the air temperature around the samples. The use of three sensors allowed the temperature readings to be averaged and also provided backup in case of sensor failure.
Temperature transducer signal conditioning The circuit used with the temperature sensors is shown in Fig. 2. Because the sensors produce 0 V output at 0 K (nominal) and have a slope of 0.01 V / K (nominal) they produce at least 2.5 V at normal drying temperatures. Two and a half of volts were subtracted from the output of the sensors before the signal was converted to a digital signal. This circuitry allowed the gain of the A / D converter to be higher, producing a correspondingly higher precision. A National Semiconductor LM336 integrated circuit voltage reference (National Semiconductor, 1980) was used to produce approximately - 2 . 5 V to add to the voltage output of the temperature sensors.
Weight transducer The weight transducer was the 2-pound model 4850 built by GSE Inc (GSE, 1982). This device was a strain gauge device and its rated nonlinearity was 0.02% of full scale, or 0.2 g, with half of that being typical The maximum operating temperature range was - 18 ° C to 93 ° C (0-200 ° F); the temperature effect on rated output was 0.0014%/°C (0.0008%/°F) and the temperature
229
CHANNE !
LM 335
"
R5
~
"
CI
I
A/D CDNVERTER R5
I
'
LM 335
~
L
M
335
VPef
LM 336 '
~5V DC RB
5i !
~
i
/
i
i
R9
Fig. 2. Circuit diagram of the temperature transducer interface.
effect on zero was 0.0027%/°C (0.0015%/°C). Two of the devices were used to allow two samples to be studied simultaneously with the resulting data sets~ labeled A and B. The transducers were located outside the chamber to minimize temperature effects, and in the case of high-humidity conditions withi~ the chamber, to isolate the weight transducers from moisture. The transducers were designed to be insensitive to moments and forces other t h a n in the one direction of desired loading. The rated error per inch of off-. center loading at 1/2 capacity was 0.004% of full scale, or 0.04 g. The trans-. ducers were loaded with a 100 g mass alternately at each of the corners and at the center of the sample holder. No discernible correlation between the mas~,~ location and the number from the computer was noted. The weight transducers were loaded with known masses and the correspond.ing number in the microcomputer was noted. Since the relationship betweer, the weight and the number from the microcomputer should be linear (accord-ing to the manufacturer's literature) a linear lest squares curve fit was run to obtain the conversion factors from the number in the microcomputer to the actual weight in grams. In all cases the correlation coefficient, r, was greater t h a n 0.9999.
Weight transducer signal conditioning An Analog Devices (1982), model 2B31J, strain gauge signal conditioner was used as shown in Fig. 3. This device included an i n s t r u m e n t a t i o n amplifier with gain from 1 to 2000 and a low pass filter with a time constant of 0.5 s. The excitation of the strain gauge bridge was regulated at 6.9 V with a National
230 ADv~TIDHNAL ]
o -
[ADDITIONAL1
CHANNEL
] F~S~] '1--~..[~-~ ~ IGI~E~I !.RG~_~b~~, ~L~ /I/ 14-~~'.~EI 'l A/D CONVERTER L4B5o ~ INST~ATIONI BUFFER LOF~P L~RS WEIGHTTRANSDUCER
0 V e??
GDO
RI
Fig. 3. Circuit diagram of the weight transducer interface. Semiconductor (1980) LM399, a temperature-stabilized integrated circuit precision voltage reference. A second LM399 was used with a voltage divider to provide an offset voltage which produced nearly 0 V out of the strain gauge signal conditioner when there was no sample load on the weight transducer. A metal film resistor was used to set the gain on the strain gauge signal conditioner at approximately 1880. This gain was near the maximum allowable on the unit and produced a precision of about 0.016 g when used with the A / D converter; while the transducer itself had a worst case precision of 0.26 g. The total range with a gain of 1880 was about 65 g which was adequate for the testing of grain. AIR CONDITIONING UNIT An Aminco-Aire (Aminco-Air, 1967) model J4-5460, a commercial air conditioner unit, was used in the study. It can condition up to 8.5 m3/min (300 cu f t / m i n ) . This unit was designed to condition relatively small chambers of less than 1.1 m 3 (40 cu ft). The chamber used in this test had a volume of approximately 0.24 m 3 and about 2.5 m3/min of conditioned air were circulated through it. In the Aminco-Aire unit, the control of air moisture and temperature was achieved in to steps (see Fig. 4). First the air passed into the larger chamber where a mist of water droplets sprayed into the air. The droplets fell into a water bath the temperature of which was maintained by a refrigeration unit or an electric resistance heater. Each water droplet was surrounded by air and thus heat and moisture transfer occurred between the air and water droplet. The air then passed into a second chamber where it was heated by electric
231 AMINCO AIRE UNII ill
HI:ATE
-~=.--
REFRIGERATION UNIT
PRODUCT
I
::: . e l
I
SCRE~:IN
J[
i
4---
-
Fig. 4. S c h e m a t i c of t h e air flow in t h e s y s t e m .
heaters. The air passed through the test chamber and was recirculated into the first chamber. The microprocessor-based circuitry controlled the temperature of the water bath and the duty cycle of the electric heaters. Because the relative humidity was not measured directly, errors encountered in measuring humidity were avoided, A unique correlation between the Aminco water temperature, air dry bulb temperature and the relative humidity of the air was used to estimate the relative humidity. A chart of this relationship was provided by the manufacturer and the formula which was used was obtained by regression. This formula was: DB----P1 +P2
in (RH) +P3
WT in (RH)+P4/RH+P5
WT
(1)
where DB is the dry bulb temperature ( ° C ), RH the relative humidity ( % ), WT the temperature of the water bath ( ° C), P1 ----77.4, P2-- - - 16.98, P3 = - - 0.0821, P4 ----73.7, and P5 -- 1.377. The Aminco-Aire operated on 208 VAC three phase power, so voltage and power amplification were accomplished as shown in Fig. 5. The switches were operated from an 8-bt parallel output port available on the microprocessor board. An additional safety feature was added in case the microprocessor con-trol became inoperable for any reason. All power to the Aminco-Aire was re-moved if a signal was not received from the microprocessor at least every 10() ms. This control was implemented with a resettable, retriggerable, monostable multivibrator attached to a solid state relay.
232 24VDC ISOLATED FROM
180 VAC
OTHER DC POWER
?
?
REL 20
REL 4 ¢0 VATER HEATER DO
REL 21
REL 2 ¢o rATER REFRIGERATION B,L, RESET
AHINCO CIRCUITS
D•
SOLID STATE
G
D4~,~-['--'-D6 o NC
1
~
t
VALID
o
G
RELAY 2
G
SOLID STATE RELAY 3
A
AIR HEATER
i
°
-o
g
RETRIDDERABLE
MONOSTABLE
120 VAC AHINCO --oHAIN CDNTAOTDR
TC 0.I S{CI
i
)
+5V
+BY DCo
"
r
D7 o NO
GROUNDo
AIR HEATER i
SOLID STATE
95 o NC
DATA ~
RELAY L
C
TO ALL CHIPS
VOLTAGE REGULATOR
Fig. 5. Circuitdiagram of the air conditioner control interface.
STUDY CHAMBER A sketch of the study chamber is shown in Fig. 6. The chamber was 0.50 m by 0.30 m by 1.60 m and was constructed mostly of plywood, which was painted with several coats of shellac to reduce moisture absorption. It was lined in all but the viewing area with 0.025 m ( 1 in) of rigid foam insulation. A section of
Fig. 6. Schematic of the study chamber.
233 the top and sides was removable for access to the chamber. The test chamber was connected to the air conditioning unit by insulated ducts. The center of the load cells was located 1.05 m from the front of the chamber and a honey-comb-shaped metal grid acting as a 'flow straightener' was located 0.29 m behind the load cells. The airflow was found to be parallel to the sides of the chamber in the vicinity of the grain sample and its velocity was measured with a hot-wire anemometer at 21 points in a plane normal to the side and b o t t o m of the chamber at the rear of the sample holders. The velocity was observed to vary from a minimum of 0.10 m / s to a maximum of 0.70 m / s with an average of 0.28 m/s. Henderson and Pabis (1962) found that, in this range, variations in airflow have an "insignificant' effect on the rate of drying. The weight transducers were located 0.25 m apart and were fastened to eL concrete block 0.18 m by 0.30 m by 0.35 m in order to reduce mechanical vibrations. The sample holders consisted of two trays 0.20 m by 0.30 m made o:t' aluminium screen with a lip of about 10 mm around the edge to hold the sample on the tray. There were two trays attached by an aluminum rod to each weight transducer. The b o t t o m tray was about 0.10 m from the b o t t o m of the chamber and the trays were about 0.10 m apart resulting in the top tray being located about 0.10 m from the top of the chamber. SOFTWARE The software for data acquisition and digital control was written in three languages: Z-80 assembly language, BASIC, and Pascal. A few primitive routines, which were used often and were critical to the timing were written in Z80 assembly language. The instructions for data display and storage, and the overall control of the equipment were written in integer BASIC. The data acquisition and air conditioning control software was written in Pascal and compiled. The code written in Pascal was called by an interrupt while programs written in BASIC were running. This software implementation made it appear that data storage and control were occurring simultaneously but actually the data acquisition and control functions had priority over data storage and display.
Data acquisition At the end of each second for the first 9 seconds, each of the eight analog input lines was sampled and converted to digital form. After the nine values for each of the eight input lines were obtained, they were averaged and the average stored in memory by the end of the 10th second. Also, before the end of the 10th second the average values representing the water temperature, heater temperature, and air temperature at the entrance to the test chamber were converted to engineering terms for use by the control algorithms.
234 The average of the digital representation of the eight transducer values was placed in memory at the end of the ten second period and a flag was set in memory. Software in BASIC responded to this flag and reset it before moving the averages to other memory and subsequently calculating averages over a 1min period. The 1-min averages were stored on cassette tape and printed at the terminal. The shortest period covered in the recorded data was one minute, with each data point representing an average of 54 observations. The averages of the weight on each transducer, the water temperature, and the five air temperatures were recorded with the reading of the clock in days, hours, minutes and seconds. The moisture content of the product on each sample holder was calculated from the sample weights and the sample dry matter content. The drying air relative humidity was calculated from the Aminco water temperatures and the air temperatures near the product.
Digital control The water temperature was controlled by an on/off algorithm. A dead band of 0.16°C was found to be the minimum which did not induce oscillations. Initially, the heating and cooling of the water was observed to overshoot the goal, especially when large changes in the setting were requested. Therefore, in addition to the simple on/of control algorithm, the set temperature was compared with the actual water temperature at the end of each second. If the two temperatures were within 0.05 ° C of each other, the heater or cooler was turned off for the remainder of the 10-s period. This adjustment allowed a narrower dead band without undesirable oscillations between the heater and refrigeration unit. The control algorithm for the air temperature was first designed as a simple velocity proportional-integral-differential (PID)controller (Bibbero, 1977). Initially temperature sensors near the product were used as the basis of control. It became immediately obvious that the heating elements, which had considerable thermal mass, became too warm before sufficient heat reached the temperature control sensors. The controller then reduced the duty cycle of the heater but because of the thermal mass of the heating element reduced it too much. This problem produced continual fluctuations of the chamber temperature. A simple solution to this problem was to base the control on a sensor located near the heater. It was assumed that under steady state conditions the heat losses would be relatively constant between the heater and the product so that the chamber conditions, while not directly controlled, would be constant. An algorithm in the form of equation (2) was adjusted using the method described by Smith (1979). Controlling the heater temperature did not allow the operator to choose the temperature in the chamber directly but a guess of
235
the temperature drop from the heater to the chamber would allow the chamber temperature to be set fairly closely. Locating the sensor near the source of the energy, the electric heater, made the control reasonably simple. The P I D velocity equation (Bibbero, 1977) was:
Ds= I (E o - E1) + JEo + H ( 3 E o - 4 E 1 - E2 + 2E3)
(2)
where DS is the change in setting of the control e l e m e n t , / , J, H are gains of the proportional, integral and differential parts of the algorithm, and E , the errors at time = t - n X At. The derivative action of the P I D controller was observed to be erratic, because of the noise in the temperature signal. Attempts were made to filter the signal both with analog filters of various orders and with digital filters. A simple RC analog filter with a time constant of 2.2 ms, was used in addition to the digital filter shown in equation (2). The derivative action was probably not: necessary in this application. At the end of each 10-s period the updated error in air temperature was computed and an adjustment to the duty cycle for the heater was calculated. During each interrupt, occurring every 40 ms, the current duty cycle was checked to see if the heater should be turned off or on. The duty cycle was based on 100 of the interrupts so that if the duty cycle were 46, for example, the heater would be on for 1.84 s and off for 2.16 s. This period of 4 s was short enough so thin the heater temperature did not change noticeably due to its thermal mass. This algorithm had two main weaknesses: it was difficult to repeat an experiment because it was difficult to obtain a given chamber temperature, and the heat load in the room was not constant so the chamber temperature tended to drift. Therefore, the control algorithm was modified between tests 9 and 10. Instead of the operator trying to guess what the heater temperature shouk! be to achieve a desired temperature in the chamber, the additional software adjusted the heater temperature setting. This two stage control process is sometimes referred to as cascade control. For this second level of control an integral controller was used. The goal of the original P I D controller was adjusted by 0.02 times the difference between the actual temperature at the entrance of the chamber and the desired temperature. This additional controller made the system much slower in responding to changes in the set point but the variation of temperature within the chamber was significantly reduced and the actual temperature within the chamber was usually within 0.5°C of the set temperature. DATA C O L L E C T I O N P R O C E D U R E
The parboiled rice which was tested was collected from a commercial parboiling plant and stored at 5 °C in sealed containers until needed for testing. T w e n t y one tests were conducted and were numbered consecutively. The first
236
two tests were considered to be 'warm-up' tests and were not analyzed completely. First the desired temperature and relative humidity of the test were determined and the equipment was given at least 1 h to stabilize at the new settings. The product which was to be tested was weighed on an analytical balance then loaded onto the trays and the data recording was begun. Separate samples were also placed into an oven for determination of initial dry matter content by the oven method. The tests were run for between 40 and 47 h at which time the dried product was again weighed on an analytical balance and samples taken to determine the final dry matter content. The temperature and weight integers were then converted to actual temperature, relative humidity and moisture content data and transferred to the mainframe computer for statistical analysis. The first step in the analysis of the moisture loss data was to plot the moisture content vs. time for each data set. Visual examination of these plots indicated poor data sets but no individual data points were removed from the data sets. Since all of the plots were at the same scale, the different data sets were also compared visually. Tests 3, 4 and 11 were eliminated because of mechanical problems with the equipment which were evident in the data plots. DATA COLLECTION AND SYSTEM VERIFICATION
The remaining sixteen data sets were kept for further analysis. These 16 tests on each of the two trays covered a total of 1308 h of data, with one data point every minute. This amounts to over 78 500 data points of moisture content vs. time. Because each data point represented an average of 54 observations the data sets resulted from over 4 million moisture content observations. Several of the data sets are represented by plots of the moisture content vs. time, Figs. 7, 8, 9 and 10. The data plots were drawn by positioning the pen at the first data point and then drawing a line to each following data point. The smoothness of the curves show the quality of the data in these data sets. The repeatability of the data can also be seen by comparing tests 20 and 21 since test 21 was a repeat of test 20 in all possible ways. The thin layer study equipment maintained constant conditions as shown in Table 1. The distribution of the temperatures and calculated relative humidities did not follow a Gaussian distribution because these two variables were controlled by the digital equipment. The control algorithm required measurable errors to occur. But after they occurred, substantially larger errors were unlikely to occur because of the feedback control. Therefore, the usual statistic of central tendency, standard deviation, was misleading. W h a t was calculated instead was the mean value and the range which includes at least 90% of the data points. The temperature and relative humidity each minute, after the first half hour of drying, were included in this analysis
23'7
Ii i i
I
\
\
i ! I
1
]
I j~
[ "4
•
i; ¸¸
,
Fig. 7. Moisture content data for test 18A.
I . ] \\ %..
27. jll
-
i
J
.
]
Fig. 8. Moisture content data for test 18B. for the r e m a i n i n g period of the run. For m o s t of the tests there were at least 2400 data p o i n t s in this analysis. D u r i n g t h e first half hour the o p e n i n g of the chamber affected the drying c o n d i t i o n s and the results of an analysis including
238
{
I
5I
] {o
0.30 t 0,25
LI o.2o ; \ u
H
00.10
O.OS
0.00o
.,15
2Io DRYING
AS
30
•IME
,
35
L,O
]
I
7
U
3S
~3
~5
$9
H~I)i.{R5
Fig. 9. Moisture content data for test 20A.
3.30!2
I
l
l
I
I
213
T
i
o.2~';i
F b~ O0.]C•
I
0.05
O. '00 I 0
,
q
]1[,
!b
,
95 DRY I NG • 7 Vl[i
I
,
313 H,D,.!R!~
I
Fig. 10. Moisture content data for test 21A.
all data would not indicate the effectiveness of the control for the majority of the time. The improvement due to the cascade control introduced between tests 9 and
239
TABLE 1 Variation in chamber conditions Test
5 6 7 8 9 10 12 13 14 15 16 17 18 19 20 21
Relative humidity (%)
Average chamber temperature ( ° C )
Percentile
Percentile
5
50
95
5
50
95
23 24 39 50 46 27 49 50 40 40 51 27 26 31 46 47
24 25 41 51 47 27 50 51 40 40 53 27 27 31 47 47
24 26 42 52 49 28 50 52 41 41 54 28 27 31 47 48
34.4 30.9 32.0 32.9 39.5 40.3 24.9 17.1 24.8 32.3 39.5 40.2 39.5 37.8 26.1 25.8
34.8 31.5 32.5 33.4 40.2 40.6 25.2 17.3 25.0 32.6 39.8 40.6 39.8 38.0 26.3 26.1
35.3 32.2 33.0 33.6 40.7 40.8 25.4 17.6 25.3 32.9 40.4 41.0 40.2 38.2 26.5 26.4
10 can be seen. Before the implementation of the cascade control algorithm, the range of chamber temperatures was about 1.0 ° C; after the change the range was about 0.5 ° C. After the control algorithm was improved the temperaturefrequency distribution was much more square with a relatively uniform distribution over several tenths of a degree about the mean and a very rapid decrease in the number of observations at both higher and lower temperatures. The water temperature was less variable than the air temperature. The range which included over 90% of the water temperature observations was 0.1°C while the dry bulb temperature varied several tenths of a degree. Therefore the degree of variation in the chamber relative humidity depended on the degree of variation of the dry bulb temperature in the chamber. The average range of relative humidities before the algorithm change was 0.021 (2.1% RH ) and after the change was 0.010. Because the water temperature was virtually unchanged, the saturation vapor pressure did not change during the tests. The real question about the adequacy of control was whether the variations in the independent variables significantly affected the drying behavior of the product. The best measure of the effect of uncontrolled variation in the independent variables is to study the data from repeat tests. Tests 20A and 20B, were fit with one equation. Samples 20A and 20B had different initial moisture contents but since they were in the chamber at the same time there could be
240
no error due to differences in t e m p e r a t u r e or relative humidity during the test. An equation which was fit to the two actual data sets allowed for a different initial moisture content. T hi s equation fit the data with a residual mean square of 1.3E-6. Th is was the error a t t r i b u t e d to the difference in weight t ransducer plus the i n h e r e n t error. F r o m the calibration of the weight transducers, it was known t h a t th er e was no measurable difference between the transducers so this error was the i n h e r e n t error. Therefore, 1.3E-6 was used as the best estimate of the variance of these data. N e x t data from tests 20A and 21A were combined. T hese two tests were repeats as nearly as possible in all respects. T h e equation was fit to the actual data producing a residual m e a n square of 1.45E-6. Likewise the data from 20B was combined with the data from 21B resulting in a residual mean square of 1.19E-6. T h e source of the error from these combinations was due to uncontrolled variation in t e m p e r a t u r e and relative humidity between the two tests plus the i n h e r e n t error. Since t he error in combining tests 20 and 21 was virtually the same as the i n h e r e n t error there was no measurable effect on the drying data due to the variation in relative humidity and t e m p e r a t u r e during these tests. CONCLUSIONS
T h e microprocessor-based system was able to m ai nt ai n a relative humidity c o n s t a n t to + 1% RH and a dry bulb t e m p e r a t u r e c o n s t a n t to + 0 . 3 ° C for periods of over 40 h. T h e microprocessor-based system was able to collect and store a great deal of data, about 78 000 m o i s t u r e - c o n t e n t time data pairs in this experiment, with limited e x p e r i m e n t e r effort. T h e experimental conditions were well controlled producing moisture cont e n t data which was repeatable, with an estimated variance of 1.3E - 6.
REFERENCES Aminco-Aire, 1967. Aminco Climate Lab Assembly, Catalogues 4-5576 and 4-5577. American Instrument Co., Inc., Silver Spring, MD, 34 pp. Analog Devices, 1982. Data-acquisition databook 1982, Volume II: modules-subsystems. Analog Devices, Inc., Norwood, MA 02062, 723 pp. Bibbero, R.J., 1977. Microprocessors in Instruments and Control. Wiley, New York, 301 pp. California Computer Systems, undated. 16K static RAM module owner's manual, Model 2016B. California Computer Systems, Sunnyvale, CA, 25 pp. Cromemco, 1980. Single card computer instruction manual. Cromemco, Inc., Mountain View, CA, 23 pp. GSE, 1982. High performance single point load cell. Bull. 137, GSE Inc., Farmington Hills, MI, 28 pp.
241 Henderson, S.M. and Pabis, S., 1962. Grain drying theory, IV: the effect of airflow rate on the drying index. J. Agric. Eng. Res., 7: 85-89. National Semiconductor, 1980. Linear Databook. National Semiconductor Corporation, Santa Clara, CA, 1673 pp. Smith, C.L., 1979. Fundamentals of control theory. Chem. Eng., 86 (22): 11-39. TecMar, 1980. TecMar S-100 AD212. TecMar Inc., Cleveland, OH, 8 pp. Vector Graphics, undated. 12K PROM/RAM assembly instructions and users guide, 12 pp.