LabVIEW Applications in Engineering Labs:

Controls, Chemical, Environmental

Jim Henry, Ph.D., P.E.

Professor
College of Engineering and Computer Science
University of Tennessee at Chattanooga
Chattanooga, TN 37403
Internet: Jim-Henry@utc.edu


Innovative Use of Computers in Laboratories

ASEE Conference
Anaheim, California
June 25 - 28, 1995


Abstract

This paper describes computer-based systems for data acquisition and control that have been installed in an engineering laboratories at the University of Tennessee at Chattanooga. The labs include automatic feedback controls systems, chemical engineering unit operations and environmental engineering experimental stations.

All systems have data acquisition and/or control programs using LabVIEW software. The software collects the data, massages it and presents it to the student users via a graphical window analogous to a hardware panel on traditional measuring equipment. The programs are used by students for data collections on the various systems. Steady-state, step response and frequency response experiments can be and are used for system identification and design.

The systems are implemented with LabVIEW on a variety of Macintosh products ranging from Mac SE's to various Quadras and on an IBM-clone '486-DX50.

INTRODUCTION

In an undergraduate teaching laboratories, we have installed a number of stations for data acquisition and control using computers. Each station is connected to a specific piece of engineering apparatus which contains sensors for inputs and in some cases, control elements for outputs.

AUTOMATIC FEEDBACK CONTROLS SYSTEMS

The objective for the stations in the laboratory is to give students a physical system which they could see, hear and touch as they carried out their experiments in system identification and controller design. Each system contains the apparatus called the "engineering system" and the computer hardware and software called the "controller." These parts fit into a control scheme that is depicted in the diagram below (Figure 1).




Figure 1. System diagram

All the engineering systems are inherently stable, single-input, single-output systems. They are non-linear and at least one is time-varying. All controllers are implemented on digital computers. The analysis and design in this laboratory is done with the assumption that the controllers are continuous signal controllers. The digital computer sampling time is fast enough (faster than the typical engineering system response) to make this assumption valid.

In the course of the semester, students (as operators of the equipment) are to conduct experiments to determine the steady-state and dynamic characteristics of their engineering system and then tune the parameters of the controller to achieve appropriately controlled behavior of the closed-loop, feedback-controlled system. These broadly constitute the actions of system identification and control system design, respectively.

The engineering systems include the following:
´ speed control on a motor-generator set with an electric light load

´ position control of a cart on a track

´ temperature control in a heat exchanger

´ air pressure control in a blower & duct system

´ water mass flow rate control in a pump & pipe system

´ water level control in a pair of interconnected tanks

The rest of this section describes how the systems are used.

CONTROL SYSTEMS PEDAGOGY

The laboratory experiences go in parallel with a course in which the principles of classical control systems are taught. Through the semester, the students spend 5 sessions running experiments on their systems intermixed with 5 sessions modelling their systems. (The modelling is done using STELLA II software; graphical presentation and analysis is done using Kaleidagraph software. These will not be discussed here.) Sessions of introductions, demonstrations and a plant tour complete the term. Reports and presentations by the students are also included.

A detailed laboratory manual has been written for the students to use. Copies are available on request. It is also available for down-loading on the Internet. See Appendix 1 for details.

In the five experiments, the students determine the following parameters for their systems:

With the controller in manual operation mode

Steady state operating curve

Region of normal operation and whether it is linear

With the controller programmed for step input

System gain

Dead time (transportation lag)

First order time constant

With the controller programmed for sine input

Lissajous plots

Amplitude ratios & phase shifts as functions of frequency

Bode plot

System order

Ultimate frequency

Gain margin

Phase margin

With the proportional feedback controller

Offset (steady-state error) as a function of Kc

Ultimate frequency

Tuning value of Kc for quarter decay response to step change in set point

With the proportional-integral feedback controller

Tuning values of Kc and tI for quarter decay response to step change in set point

INPUT-OUTPUT HARDWARE DESCRIPTIONS & CHARACTERISTICS

The procedure in implementing control systems such as these involves selecting (1) the sensor and transmitter for the system's signals (the "output variable"), (2) the device that is going to be the acting device (that receives the "manipulated variable"), (3) the computer platform, (4) the software and (5) the add-in board for the computer that handles the input and output signals.

Sensors and transmitters

To begin selecting sensors and transmitters, it helps to know that many analog signal input converter boards are able to convert signal voltages that are DC voltages in the range of 0-10 volts. Thus, sensors and transmitters that have outputs in this range can be connected directly to the analog input board. If the sensor and transmitter have output different from this, a signal conditioner must be made or purchased that converts the signal to the range and type of signal that the analog input board can accommodate.

I have both situations in this lab. There are several piezo-resistive pressure transducers that have signals of 1-6 volts (when a proper power supply voltage is applied to the transducer) that are connected directly to the analog input board. For a motor-speed sensor, an interrupted photo cell signal is sent to a frequency-to-voltage converter before going to the analog input board. For a flow transmitter that puts out a 4-20 ma output signal, a 500Ż resistor is used to produce a 2-10 volt signal to be sent to the analog input board. For resistance temperature devices (RTDs), signal conditioners were acquired for that application that have 0-5 volt DC outputs.

Control action device

Analog output boards typically are capable of producing output voltages (DC) in the range 0-10 volts. So, the final action device needs to accommodate this type signal or, again, a signal conditioner can be used.

Again, both methods are used in this lab. Four of the stations have variable-voltage, variable-frequency speed controllers for three-phase AC motors. All of these accept the 0-10 volt DC analog output signal as the controlling voltage. On three of them, we have installed optical isolators between the analog output board and the speed controller because the control signal floats above ground potential by as much as 230 volts. Another station has an output signal conditioner to convert the analog output signal to 4-20 ma as required by a variable speed laboratory pump. The sixth station has the analog output signal go to a pulse-width modulator which modulates a 24 volt DC power supply for speed control on a DC motor.

On-off devices can be controlled with relays or solenoids that are operated by the digital output signals from the add-in board in the computer.

Computer platform

All stations described here use desktop computers, Macintoshes and a Windows 3.1 '486 machine. The stations have a variety of Macintoshes, including, Mac II, Mac IIci, Centris 650, Quadra 650 and a Quadra 950.

Software

All stations described here use LabVIEW software from National Instruments. This software takes advantage of the graphical user interface for developing and using the programs.

Input-output board

All stations described here use boards from National Instruments that work with the LabVIEW software. All boards have at least 8 channels available for analog input and 2 channels for analog output. They also have lines for digital input and output and for counters.

COMPUTER CONTROL DESCRIPTIONS & CHARACTERISTICS

The LabVIEW programs for these applications were built to simulate conventional analog controller hardware as is used in industrial settings. The features on the front panel of a controller include indicators or controls for the values of the output (process) variable, the desired set point and the manipulated (controller output) variable. Additionally, some industrial controllers have integral strip chart recorders. These features were included in the LabVIEW controller simulators.

The controller simulators can be operated in one of five modes: manual operation, programmed operation for step response testing, programmed operation for frequency response testing, feedback control with proportional control action and feedback control with proportional-integral control action. These five different modes were implemented by having five separate LabVIEW programs, one for each of the different modes. These will be described here.

Manual mode controller
The Front Panel

Figure 2 contains a portion of the front panel. The manual mode controller has a slide control knob for the "Motor input" signal (the manipulated variable). Adjacent to it is a pointer indicator that shows the "Speed Output" in this illustration (the output variable). Below these is a meter that indicates the "Controller output;" this is identical to the manipulated variable and directly tracks the manually input value of "motor input." At the top left is a "Stop" button to stop the program.

Shown in Figure 3 (next page) is a strip chart graph that provides a visual record of both the "Motor input" and the "Height." Above the strip chart graph is a window that shows how many data points have been taken and recorded.

Figure 2. Manual Control Panel

(Left side)

On the pressure and flow stations, there is one additional Boolean control on the front panel: a switch to open or close the valve on air duct or water line, respectively. This is true for all controller modes.

In operation, the operator can move the "motor input" control slide and observe the response on the output indicator, the dual curves on the strip chart and see, hear and/or feel the response on the physical equipment.

After running the controller program as long as desired, the operator clicks the "STOP" button on the panel. For most stations, this turns off the motor. For all stations, at operator option, it saves the data for the experiment to disk and/or plots graphs of the data.

Figure 3. Manual Control Panel (Right side)

Output graphs
Figure 4 shows the time response graph of an experiment. Time is on the horizontal axis; the values of the manipulated and output variables are plotted on the vertical axis. This format graph is available in all controller modes.

Figure 4. Time response graph of manual experiment
Figure 5 shows the input vs. output response graph of an experiment. This is the steady-state operating curve for the system. The manipulated variable is on the horizontal axis; the output variable is plotted on the vertical axis. This format graph is available in the manual and the frequency-response controller modes.

Figure 5. Input-output graph of manual experiment.

The program saves the data of run-time, manipulated variable and output variable(s) for each data point. Since this is saved in a array, the length of the experiment is limited by available memory in the computer. For the systems described here the limit is above 10,000 data points, more than anyone seems to care to analyze. If desired, this limitation can be overcome by reprogramming and writing the data values on a disk file at each point.

The graphs generated by LabVIEW can be printed for analysis later or the data saved on disk can be analyzed later with other software.

The LabVIEW diagram

The LabVIEW "diagram" is the program. It is a graphical programming language. It is very powerful and versatile. The initial learning curve can be steep. (It was for me--JH. I will be happy to share any of my programs with anyone who wants them. Contact me via e-mail and I can make the programs available on a Gopher server. See Appendix 1.)

Programmed step-response controller
The Front Panel

Three additional controls appear on this panel (Figure 7) that were not on the manual-mode controller. They are (1) the initial steady-state operating point (or "base line" or "bias point") of the "Motor input" variable, (2) the input step height and (3) a toggle switch to turn the step on or off. The "Motor input" signal (the manipulated variable) is changed into an indicator. All other front panel items are the same.

In operation, the operator sets the values of the "motor input" initial steady-state operating point and the input step height. During operation, the "Motor Input" indicator shows what the input value is. When the experimental test is run, the operator waits until the system reaches initial steady state operation, then clicks on the "Step" switch and then waits for the system to complete its response to the step.

Figure 7. Step response controller panel

Steady state operation can be observed on the output indicator, the dual curves on the strip chart and by seeing, hearing and/or feeling the response on the physical equipment.

After the step response has been observed as long as desired, the operator clicks the "STOP" button on the panel. Again, at operator option, the program saves the data for the experiment to disk and/or plots a graph of the data.

Output graph

Figure 8 (next page) shows the time response graph of a step response experiment. Time is on the horizontal axis; the values of the manipulated and output variables are plotted on the vertical axis. The students can determine the gain, first-order response time and system dead time (transportation lag) from this graph (Smith & Corripio, 1985, p 216). Further analysis could give second-order parameters if the system and the data warrant it (Coughanowr, 1991, p 296).



Figure 8. Time response for step-input

Programmed frequency-response controller
The Front Panel

Two different controls appear on this panel (Figure 9) from what were on the step-response controller. They are (1) the input sine wave amplitude and (2) frequency of the input sine wave. All other front panel items are the same, including the "base line" (or "bias point") of the "Motor input" variable.

In operation, the operator sets the values of the "motor input" base line, the input sine wave amplitude and the input frequency. During operation, the "Motor Input" indicator shows what the input value is. When the experimental test is run, the operator waits until the system reaches steady oscillation, he or she then runs the experimental test as long as desired to collect sufficient data and then clicks on the "STOP" button. Steady oscillation can be observed most effectively on the dual curves on the strip chart.

Figure 9. Frequency response controller panel (Left side)

After clicking the "STOP" button on the panel, at operator option, the program saves the data for the experiment to disk and/or plots graphs of the data.

Output graphs

Figure 10 shows the time response graph of a frequency response experiment. The students can determine amplitude ratio (of the output variable to the manipulated variable) and phase shift (between the two signals) from this graph. The first few seconds on this graph demonstrate the transient start-up behavior for the system.




Figure 10. Time response for sine wave input function

Figure 11 shows the input vs. output response graph of a frequency response experiment. This is the Lissajous graph. From it, the students can also get amplitude ratio and phase shift. It's not as popular to use the Lissajous graph for these purposes, so mainly it's just for show.




Figure 11. Lissajous graph resulting from sine wave input signal.

Proportional feedback controller
The Front Panel

Two different controls appear on this panel (Figure 12) from what were on the frequency-response controller. They are (1) the "Desired set point" for the output variable and (2) controller proportionality constant, Kc. All other front panel items are the same, including the "base line" (or "bias point") of the "Motor input" variable.

In operation, the operator sets the values of the "motor input" base line, "Desired set point" and the value of Kc. When the experimental test is run, the operator runs the experimental test as long as desired to collect sufficient data and then clicks on the "STOP" button.

Figure 12. Panel for proportional feedback controller.

After clicking the "STOP" button on the panel, at operator option, the program saves the data for the experiment to disk and/or plot a graph of the data.

Proportional-integral feedback controller
The Front Panel

One additional control appears on this panel (Figure 13) from what were on the proportional feedback controller. It is the controller integration time constant, tI. All other front panel items are the same.

In operation, the operator sets the values of the "motor input" base line, "Desired set point" and the values of Kc and tI. When the experimental test is run, the operator runs the experimental test as long as desired to collect sufficient data and then clicks on the "STOP" button.

After clicking the "STOP" button on the panel, at operator option, the program saves the data for the experiment to disk and/or plot a graph of the data.

Figure 13. Panel for proportional-integral controller

CHEMICAL AND ENVIRONMENTAL ENGINEERING

The units with data acquisition and/or control include the following:
´ distillation column

´ water spray cooling tower

´ packed column absorption column

´ dehumidifier

´ flow through packed beds

´ plate-and-frame filter press (work in progress)

Distillation Column

The distillation column is a 12-tray, bubble cap column with a reboiler and condenser. The computer data-acquisition and control is instituted with LabVIEW and performs these functions:

Monitoring

tray temperatures with RTDs & signal conditioners

reboiler temperature with RTDs & signal conditioner

distillate temperature with RTDs & signal conditioner

feed temperature with RTDs & signal conditioner

electrical energy into the reboiler with a current transformer

liquid level in the reboiler

Controlling

feed flow rate

reflux ratio

electrical energy into the reboiler with a thyristor circuit

bottoms pump-out rate

In coursework, both undergraduate and graduate, the distillation column is controlled in a manual mode and the data is acquired via the computer. My research is extending a feedback control system to the distillation column. Cindy Wormsley completed an artificial neural network controller last year. Students now are working on that as well as a fuzzy logic controller.

Figure 14 shows the panel that the students interact with to operate the distillation column. On the left is the graphical representation of the equipment with some of the data readings shown. The temperatures are shown on a multiple-trace strip chart. On the right is the panel in which the operating parameters are entered for the operation of the column.


Figure 14. Panels for control of the distillation column

Water-Spray Cooling Tower

We use a residential-size (approximately 3-ton cooling capacity) cooling tower to demonstrate the principles and operation of water-spray cooling towers. This unit is connected to a computer with LabVIEW software to perform these functions:

Monitoring and data collection

temperature of entering and exiting water

temperature of entering and exiting air

humidity of entering and exiting air

flow rate of water

Controlling

the electrical energy into the heater with a relay circuit

the water flow rate (variable speed pump)

the air flow rate (variable speed fan)
Figure 15 shows the control panel for the cooling tower. There are two slide controls for the variable speed motors and digital indicators for the temperatures, humidities and water flow rate.

Figure 15 -- Control Panel for Water Spray Cooling Tower Program

Packed Column Absorption Column

The packed column absorber is a 4-inch diameter by 5-foot tall packed column. Currently we only use it with air and water for fluid mechanics studies: pressure drop at various air and water flow rates and flooding conditions.

Interestingly, only after connecting this equipment to data acquisition could we understand some of its baffling behavior. The flooding behavior seemed to be random. After we had lots of data to graph, we realized the building air compressor (our source of process air) was cycling on and off, causing the air flow rate to vary above and below flooding conditions.

This unit is connected to a computer with LabVIEW software to perform these functions:

Monitoring and data collection

flow rate of water

flow rate of air

pressure differential between top & bottom of column

Controlling

the water flow rate (variable speed pump)

Dehumidifier

A standard domestic air dehumidifier is used to demonstrating psychrometric principles and to perform material and energy balances on a piece of equipment. We have placed temperature sensors on the inlet and outlet of the Freon condenser coil and the Freon evaporator coil.

At steady-state operation, this is a quite boring operation. It is start-up that provides interesting transient behavior. When starting up, the evaporator coils get colder than the freezing point of water and become frost covered. After a few minutes, they warm up and normal dehumidification proceeds.

This unit is connected to a computer with LabVIEW software to perform these functions:

Monitoring and data collection

temperature of compressor outlet

temperature of condenser coil outlet

temperature of expansion valve outlet

temperature of evaporator coil outlet

Flow through packed beds (work in progress)

We have two cylinders filled with sand through which water is passing by gravity flow. We can put a concentration of salt in the inlet as an "inlet signal" and measure the conductivity of the outlet stream as the "outlet" signal." We have the conductivity cells that provide a 0-10 volt DC outlet. We will connect those to a computer with a LabVIEW data acquisition program in the Fall.

Plate-and-Frame Filter Press (work in progress)

We have a laboratory scale plate-and frame filter press. We are going to put a flow meter and pressure transducer on the slurry inlet. We are also going to put a weigh scale transmitter on the collected filtrate.

We will connect those devices to a computer with a LabVIEW data acquisition program in the Fall.

EXTENSIONS & FUTURE WORK

In addition to the work described here for the teaching laboratory, work is on-going in further data acquisition and analysis in other laboratories and in controls research. Two other LabVIEW programs have been written to (1) automatically perform the analysis of first-order and dead time parameters from the various stations (┼strom, 1988) and (2) automatically run a series of sine wave inputs, find the amplitude ratio and phase shift with FFT utilities in LabVIEW and plot the Bode diagram.

Work in progress includes extracting the frequency response information from a finite pulse input and implementing a self-tuning controller (Shinsky, 1988).

Research extensions are in the area of application of artificial neural networks in model-based control of non-linear systems. The goal is to implement these also in the LabVIEW environment.

CONCLUSIONS

Several laboratory systems have been implemented with LabVIEW software for data acquisition and control. The systems have worked entirely satisfactorily for the purposes for which they were intended.

ACKNOWLEDGMENTS

Grateful acknowledgment is made for support of this lab by

Analog Devices, Inc.

Apple Computer

Chattanooga Armature Works

MicroMotion (Rosemount Instruments)

National Instruments

Plant Engineering Consultants

Taxpayers of the State of Tennessee

REFERENCES

┼strom, Karl J., and Tore Hagglund, Automatic Tuning of PID Controllers, ISA, 1988.

Coughanowr, Donald R., Process Systems Analysis and Control, 2 ed., McGraw-Hill, 1991.

Shinsky, F.G., Process Control Systems, McGraw-Hill, 1988.

Smith, Carlos A., and Armando B. Corripio, Practice of Automatic Process Control, Wiley, 1985.

Wormsley, Cindy, and Jim Henry, Neural Network Control of a Laboratory Distillation Column, presented at A.I.Ch.E. Annual Meeting, November, 1994.



APPENDIX 1 -- FURTHER INFORMATION SOURCES

The Laboratory Manual for Control Systems is available as a Microsoft Word document on a Gopher server at UTC. Connect to the Gopher chem.engr.utc.edu/Controls Systems Laboratory; get (in binary) the file is

lab˝manual.word-5-mac (305kb). Once you get this file, it can be read with Microsoft Word 5.0 or later on Macintoshes or Microsoft Word 6.0 on Windows.

About a year ago, Gary Johnson published a good book about LabVIEW: LabVIEW Graphical Programming. Practical Applications in Instrumentation and Control, McGraw-Hill. ISBN 0-07-032692-4

There are two web sites that serve LabVIEW: http://k-whiner.pica.army.mil/info-labview/info-labview.html and http://www.natinst.com/labview/lvindex.htm

At the former you may subscribe to the Internet mailing list of LabVIEW users.

APPENDIX 2 -- CONTROL SYSTEMS' DESCRIPTIONS & CHARACTERISTICS

Speed

For the speed-control system, the "engineering system" consists of motor-generator set. The motor is a 5 hp, 480 volt, 3 phase, 1735 RPM electric motor. The motor is powered by a variable-voltage, variable-frequency power supply (Emerson Electric Accuspede 270). The motor is coupled to a DC generator that is connected to a bank of up to sixteen 300 watt, 120 volt light bulbs. The DC generator is a 3 kW at 2400 RPM generator. At full motor speed, the generator puts out about 80 volts. This does not fully energize the light bulbs and, more significantly, causes the light bulbs to have about a 3-second delay in coming on. This phenomena is observable in Figure 8 with the dip in the speed curve between 8 and 11 seconds.

The characteristic time for the motor-generator system is around 1 second, the steady-state operating curve (speed-vs-input voltage) is non-linear, the electric-light generator load is time-varying and the number of light bulbs connected to the load can be varied.




Figure A-1. Speed control system

The manipulated variable from the controller is a 0-10 volt DC signal that is supplied to the motor's power supply through an optical isolator. The output variable is a DC voltage in the range of 0-10 volts proportional to the speed of the motor. This signal is obtained by generating a voltage pulse train that has a frequency that varies with the speed by interrupting a light beam with a chopper wheel on the motor axle. Then this pulse train is converted to a DC voltage with a frequency-to-voltage converter.

The "controller" is implemented on a Macintosh Quadra 650 with LabVIEW software and a National Instruments' NB-MIO-16, multi-function input-output board. The controller operates at about 50 Hz sampling frequency.

Pressure

For the pressure-control system, the "engineering system" consists of motor-driven centrifugal blower. The motor is a 1 hp, 240 volt, 3 phase, 3450 RPM electric motor. The motor is powered by a variable-voltage, variable-frequency power supply (Accuspede 245). The motor is direct-coupled to the blower (Dayton 3N178G). The blower blows into a 4-inch diameter manifold that feeds three 3-inch ducts. The pressure at the manifold exit is sensed by a piezo-resistive silicon diaphragm transducer (Omega PX163-120 D5V).

The characteristic time for this system is also around 1 second, the steady-state operating curve (pressure-vs-input voltage) is non-linear, and one of the outlet ducts has a remote-controlled valve that can opened and closed at the discretion of the operator.




Figure A-2. Pressure control system

The manipulated variable from the controller is a 0-10 volt DC signal that is supplied to the motor's power supply through an optical isolator. The output variable is a DC voltage in the range of 1-6 volts proportional to the pressure at the manifold exit.

The "controller" is implemented on a Macintosh Quadra with LabVIEW software and a National Instruments' NB-MIO-16, multi-function input-output board. The controller operates at about 50 Hz sampling frequency.

Flow

For the flow-control system, the "engineering system" consists of motor-driven centrifugal pump. The motor is a 1.5 hp, 240 volt, 3 phase, 3450 RPM electric motor (Dayton 3N090J). The motor is powered by a variable-voltage, variable-frequency power supply (Accuspede 245). The motor is direct-coupled to the pump (Teel 1P793). The pump sends water from a 30-gallon reservoir into a 2-inch diameter PVC manifold that feeds three 2-inch lines. The mass flow rate in one of the lines is sensed by a Coriolis force mass-flow meter and transducer (MicroMotion D3). The three lines return water to the reservoir.




Figure A-3. Flow control system

The characteristic time for this system is also around 1 second, the steady-state operating curve (flow rate-vs-input voltage) is non-linear, and one of the non-metered outlet lines has a remote-controlled solenoid valve that can opened and closed at the discretion of the operator.

The manipulated variable from the controller is a 0-10 volt DC signal that is supplied to the motor's power supply through an optical isolator. The output from the transducer is a 4-20 ma DC current that is converted (across a 500Ż resistor) into a DC voltage in the range of 2-10 volts proportional to the mass flow rate in the one line.

The "controller" is implemented on a Macintosh Centris 650 with LabVIEW software and a National Instruments' NB-MIO-16, multi-function input-output board. The controller operates at about 50 Hz sampling frequency.

Position

For the position-control system, the "engineering system" consists of motor-driven cart on rails with a swivel boom hanging down from the cart. This is to simulate, for example, an overhead crane. The motor is powered by a pulse-width modulated DC power supply (Copley 303P). The motor is fixed on the mounting base and drives the cart with a chain drive. The position of the cart and the angle of the boom are sensed with variable potentiometers.




Figure A-4. Position control system

The characteristic time for this system is less than 1 second, the pulse response operating curve (position-vs-pulse width voltage) is non-linear, and the load on the boom can be changed when desired.

The manipulated variable from the controller is a 0-10 volt DC signal that is supplied to the pulse-width modulator. The output from the potentiometers is a DC voltage in the range of 0-10 volts proportional to the position of the cart or angle deviation from vertical for the boom.

The "controller" is implemented on a KeyData 80486-DX-50 with LabVIEW software and a National Instruments' Lab-PC+, multi-function input-output board. The controller operates at about 30 Hz sampling frequency.

Liquid Level

For the level-control system, the "engineering system" consists of two tanks that have a water feed into the first tank which then drains into the second. Also, the first tank is connected to a variable speed pump (MasterFlex 7520-25) that can pump water out of the first tank. The pump speed is controlled with a 4-20 ma DC signal from a current output signal conditioner. The water level in each tank is monitored with air bubblers that measure the hydrostatic head above the bubbler port. The bubbler pressures are converted into voltage signals with piezo-resistive transducers as mentioned earlier.

The characteristic time for this system is several minutes, the steady-state response operating curve (level-vs-input voltage) is non-linear due to the square-root dependence of the effluent flow as a function of liquid depth, and the input water flow rate can be changed when desired.

The manipulated variable from the controller is a 0-5 volt DC signal that is supplied to the current output signal conditioner. The outputs from the transducers are DC voltages in the range of 1-6 volts proportional to the heights in the respective tanks.

The "controller" is implemented on a Macintosh II with LabVIEW software and a National Instruments' Lab-NB, multi-function input-output board. The controller operates at about 20 Hz sampling frequency, much faster than needed. The panel has provision for specifying a longer interval between data points.




Figure A-5. Level control system

Temperature

The temperature-control system consists of a water reservoir with two coils for heating and cooling fluids. The heating coil has hot water passing through it and the cooling coil has cooling water passing through it. The hot water comes from a closed-loop water heater circuit. The flow rate of the hot water is variable, controlled by a motor-driven centrifugal pump. The motor is a 1.5 hp, 240 volt, 3 phase, 3450 RPM electric motor (Dayton 3N090J). The motor is powered by a variable-voltage, variable-frequency power supply (Mitsubishi Freqrol Z024-0.75K). The motor is direct-coupled to the pump (Teel 1P793). The cooling water flow rate is variable, controlled by a manually operated valve connected to the city water line.




Figure A-6. Temperature control system

The temperatures of the reservoir and the inlet and outlets of each coil are monitored by resistance temperature devices (RTDs) connected to signal conditioners (Analog Devices 5B34).

The characteristic time for this system is several minutes, the steady-state response operating curve (reservoir temperature-vs-input voltage) is nearly linear, and the input water flow rate can be changed when desired.

The manipulated variable from the controller is a 0-5 volt DC signal that is supplied to the hot water pump's power supply. The outputs from the transducers are DC voltages in the range of 0-5 volts proportional to the temperatures of the various RTDs.

The "controller" is implemented on a Macintosh IIci with LabVIEW software and a National Instruments' Lab-NB, multi-function input-output board. The controller operates at about 30 Hz sampling frequency, much faster than needed. The panel has provision for specifying a longer interval between data points.


APPENDIX 2 -- SUMMARY DESCRIPTIONS & CHARACTERISTICS

Table A-1. Equipment summary for the control lab stations
Engineering

System
Speed
Pressure
Flow
Position
Water level
Temperature
Significant

Hardware
Motor-

Generator
Motor with

blower
Motor with

pump
Cart & boom
Tanks, valves, pump
Heat exchanger, water heater, pumps
Motor speed

control
VVVF

Accuspede
VVVF

Accuspede
VVVF

Accuspede
Pulse-width modulated voltage supply
4-20 ma

control signal
VVVF

Mitsubishi
Manipulated

Variable
0-10 volts

into optical

isolator
0-10 volts

into optical

isolator
0-10 volts

into optical

isolator
0-10 volts

into modulator
0-5 volts

into current module
0-10 volts

into optical

isolator
Output

Sensor
Chopper wheel,

photo cell,

frequency-to-voltage converter
Piezo-resistive

pressure sensor
Coriolis effect mass flow meter with transmitter
Variable voltage across poten-tiometers
Static head by piezo-resistive

pressure sensor
RTDs

with signal conditioners
Output

Variable
0-10 volts

proportional

to speed
1-6 volts

proportional

to pressure
4-20 ma

proportional

to mass flow rate
0-10 volts

proportional

to positions
1-6 volts

proportional

to pressure
0-5 volts

proportional

to temperatures
Control

Computer
Mac

Quadra 650
Mac

Quadra 950
Mac

Centris 650
KeyData

'486 DX 50
Mac II
Mac IIci
Input-

Output

Board
NB-MIO-16
NB-MIO-16
NB-MIO-16
Lab-PC+
Lab-NB
Lab-NB

Table A-2. Equipment summary for the Chemical & Environmental lab stations
Engineering

System
Distillation
Cooling

Tower
Packed Column Absorber
Dehumidifier
Flow Through Packed Media
Filter Press
Significant

Hardware
12 tray distillation column
Heat exchanger, water heater, pumps
Packed column
Domestic dehumidifier
Packed columns
Lab-scale filter press
Motor speed

control
N/A
VVVF

Mitsubishi
VVVF

Accuspede
None
Lab Pump

4-20 ma
None
Manipulated

Variable
Reboiler heat

Reflux ratio

Feed rate

Feed composition
0-10 volts

into optical

isolator
0-10 volts

into optical

isolator
0-5 volts

into current module
Output

Sensor
RTDs

with signal conditioners

Current meter

Paddle wheel flow meter
RTDs

with signal conditioners

Paddle wheel flow meter

Humidity sensors
Paddle wheel flow meter with transmitter
Temperature sensors
Conductivity sensors
Pressure sensors

Weight cells

Paddle wheel flow meter
Output

Variable
0-5 volts

proportional

to temperatures

0-10 volts

proportional

to flow
0-5 volts

proportional

to temperatures and humidity

0-10 volts

proportional

to flow
0-10 volts

proportional

to pressure

Pulse train frequency proportional to flow
0-10 volts

proportional

to temperature
0-10 volts proportional to conductivity
Pulse train frequency proportional to flow

0-10 volts proportional to pressure & weight
Control

Computer
Mac

Centris 650
Mac IIci
Mac SE
Mac SE
Mac SE
Mac SE
Input-

Output

Board
NB-MIO-16
Lab-NB
Lab-SE
Lab-SE
Lab-SE
Lab-SE