CONTROL EXPERIMENT
"Control of a Laboratory Heat Exchanger"
Chemical Engineering and Environmental Engineering
Laboratories I, II
CONTROL EXPERIMENT
Objectives
It is the purpose of this experiment to illustrate the theory and application of digital process control. The process comprises a concentric-pipe heat exchanger in which water is heated by condensing steam. The outlet water temperature serves as the measured variable while the steam-valve position acts as the manipulated variable. Step tests on the manipulated variable will be performed to determine the process gain, time constant, and deadtime as a function of the operating region. Cohen-Coon and time-based PI and PID tuning rules will be used to tune the velocity-form, digital PI(D) controller as well as the analog controller. Qualitatively, the effect of model uncertainty due to process nonlinearity will be determined.
Procedure
Start Up: Trace all piping lines in the heat exchanger and determine the function of each component (valves, thermocouples, etc.). Water flows through the inner tube and steam flows through the outer tube of the concentric pipe heat exchanger. Remember that turning the valve clockwise closes it and turning the valve counterclockwise opens it (The opposite is true for the air pressure regulator.).
The Heat Exchanger
1. Open the main water flow valve and adjust the valve such that the orifice plate pressure drop reads six in-Hg on the manometer.
2. With the main stream valve closed, open the bypass valves to the steam trap and to the control valve. Open all of the valves on the steam line so as to give clear passage of the steam through the exchanger.
3. Be certain that the analog/digital switch next to the manometer is turned to analog. There is also an analog/digital switch in the back of the exchanger near the control valve that should be turned to the analog position.
4. Crack the main steam valve and allow the system to warm up slowly. Gradually open the main steam valve further until steam blows through the exchanger, flushing out the air. Open the vent valves of the top of the exchanger to vent the air. Steam should blow out of these vents.
5. Open Air Valve 1 to allow air flow into the instrument at tubing.
6. Open Air Valve 2 to allow air flow to the control valve. NOTE: This air valve is opposite to most valves in that turning it clockwise opens it. Open the valve so that the meter reads 20 psig.
7. Shut the bypass valve on the trap and open the main steam valve all the way. Adjust the steam pressure to 60 psig with Steam Valve 1.
8. Switch on the analog controller, set it for an outlet water temperature of 135°F, and close the bypass valve around the control valve. (Note: For analog control, adjust the controller with the knob in the upper right hand corner so that the air pressure to the controller is between 3-15 psig.)
9. Open the appropriate valves to make the system countercurrent.
Electrical and Computer Preparation
1. Plug in the two electrical cords in the back of the exchanger. These turn on the thermocouples, temperature readout box, and power strip behind the computer.
2. Make sure the line from the white box on the control board is plugged into the power strip.
3. Turn on the power strip and switch on the back of the computer cabinet.
4. Turn on the hard drive. The prompt should read C on the computer. Type run.
Preparation for Digital Control
1. If 15 minutes has not passed since the steam was turned on wait until this time has passed.
2. Turn both of the analog/digital switches to digital.
3. Open the bypass to the controller only slightly until a few seconds before data are taken, to make sure that steam is maintained. DON'T FORGET to close the bypass all the way before taking data.
Rules of Thumb
1. The first prompt on the computer is regarding resolution of the monitor. High resolution is the best. It allows you to monitor temperatures from 0-50°C. If later you want to look at a narrower temperature range (for better accuracy) extra high resolution will monitor 20-45°C. Often this is too narrow and the data will go off the screen.
2. The next prompt is for the sampling rate. This is the number of data points read in a certain amount of time. Adequate control is maintained with a sampling rate of one Hz.
3. The next prompt is for voltage and the duration of this voltage. The control valve is opened proportional to a transmitted voltage. Zero volts means the valve is entirely shut (hence the need to open the bypass a bit until the experiment begins) and five volts means the valve is entirely open.
4. Step changes are best made at about 80 seconds. This will be obvious after making a few runs and seeing how the screen works. Eighty seconds is enough time to make sure steady state has been reached before the step change and allows the maximum amount of the screen to be filled with data for printing.
5. To print a page with data on it, hit the print screen button on the computer. When printing the screen will stop showing new data. The system will catch up with itself when the printing is finished.
Experiment
1. Start the experiment by performing open-loop response tests, using the computer.
a. Voltage step changes of 0.5 volts should be used.
b. Keep the set point temperature constant (30°C works well) and observe the temperature response to the step change, performing several trials at the specified step change in the 2 and 4 volts range.
c. Determine the deadtime, process gain and time constant for each trial and discuss the linearity of the process response. Use the values of the deadtime, process gain, and time constant from the 3-4 volt range for the remainder of the experiment.
2. Determine the controller gain and integral time using IMC-based and Cohen-Coon tuning parameters for both PI and PID control. The IMC-based parameters are a function of the desired closed-loop response time.
3. Run the closed-loop PI control for a temperature set point change of between 3 and 5°C. (25 to 30°C range works well). Do several runs and adjust the parameters calculated above. (Tuning K
4. Run one or two trials using PID control and discuss the response.
5. Make a load change by changing water flow rate (a quarter turn of the valve is sufficient; make sure the mercury does not flow out of the manometer) while the system is running and notice the response. Make certain that all other operating variables are constant for this section.
6. Use the analog controller for feedback control and the computer for data acquisition. Realize that the analog controller uses proportional band rather than proportional gain, and that these units are different than for the digital control case.
Shutdown
1. After all the necessary runs have been made, the bypass valves should be opened and the steam valve should be closed.
2. While the system is cooling off, shut down the computer by exiting the program and typing "system" to return to the root directory. Shut off the hard drive and the power strip.
3. After 15 minutes the system will be cool enough the shut off the water. Make sure the cold water is shut off at the manometer as well as the main valve.
4. Set the pressure regulator for the air to the control valve to zero psig.