ME 368 Laboratory 13 Hot jet temperature control
Laboratory 13
Hot jet temperature control
Equipment needed
· Type T thermocouple with a bead of 1 mm diameter or smaller
· Convectronics 001-10007 resistive air heater in glass tube, 7.5 ohms (FRAGILE!)
· Ringstand for holding the two aforementioned devices
· Heat gun duct-taped to the ringstand base so that its nozzle tip is 15/16” from the side of the glass tube, for providing a heating disturbance to the glass tube
· Feno linear dimmer with 24 VDC power supply and control wire pairs attached, to power air heater
· 0.6 SCFM air supply (to each station) in ¼” o.d. polyethylene tubing
· LabVIEW / myDAQ for generating the dimmer command signal and logging data from the TC
Goals and Objectives
· Understand the distinction between open-loop control and closed-loop control
· Apply the control-related knowledge you have garnered from recent labs to a practical control problem
· Test controller response to a physical disturbance (auxiliary heating with a heat gun)
1. Fragile heaters
The glass-tube air heaters used in this lab are fragile and expensive ~ $200 each. Therefore:
i. Keep the heater near the back of your lab bench, where hands, cables, etc. are unlikely to hit it.
ii. Do not touch the glass. If you feel something must be repositioned, ask the lab instructor to do so.
iii. Avoid heating element burn-out. Air must always be running during this lab to keep the heaters from overheating. If you notice that the air flow to the heater is weak or off or that the heater is glowing, immediately unplug the power cord of the 24 volt power supply.
2. LabVIEW code
A controller is available as Lab_13.vi on the course website. This code is almost identical to the one you used last week.
3. Contest description
This lab will culminate in a “contest” in the week’s second meeting (Weds / Thurs) in which all groups will test their hot jet control experiments at the same time. The contest is broken into 3 tests:
a) Steady-state error: Each group will choose a set point in the 32 – 75 °C range. The set point will be set to a value 10 °C below the chosen value at t = -30 s. At t= 0, the set point will be increased to the chosen value. Data recorded by the group over the t = 20 to t = 50 s interval will be analyzed: the root mean square (RMS) temperature error [°C] during this period will be determined. The appropriate formula for RMS is given in the bottom-right entry of Table 9.1 of your equation sheet. The RMS can be found in LabVIEW using the RMS vi ( ). If you have temperature error data in column C in Excel, the RMS can be found using a formula like: SQRT(SUMSQ(C1:C600)/COUNT(C1:C600)). Your goal is to achieve the minimum steady state RMS temperature error [°C]. Note: the chart history length in Lab_13.vi has been set so that the most recent 60 s of data is displayed. Thus, for this test, it is convenient to hit the STOP button just before the step to the chosen set point reaches the left border of the chart.
b) Settling time associated with a set point change: For the exact same data as recorded in (a), the group will determine (and show the lab instructor) the time to settle to within 1 degree C of the new set temperature. Getting within 1 degree C temporarily and then over- or under-shooting does not count, the time will be measured to the point after which the results remain within 1 degree C of the new set temperature. Your goal will be to achieve the minimum settling time [s].
c) Immunity to a disturbance: In a separate test, each group will pick a set point in the 32 – 75 degrees C range, and set the set point at the chosen value at t= 0. During the t = 20 to t = 50 s interval, the heat gun must be ON, and during the t = 50 to t = 80 s interval, the heat gun must be OFF. The group will determine the maximum excursion [°C] from the set point over the t = 20 to t = 80 s. Your goal will be to achieve the minimum excursion. Stopwatches are available if you’d like to use one for timing. For this test, it is convenient to STOP the code at the t = 80 s mark; at this point, the 60-s history displayed represents the time interval of interest for evaluating the disturbance immunity.
Rules:
i. You must run the same code for all 3 tests.
ii. Your controller cannot use human intervention. For example, your control strategy cannot include a mouse or keyboard input during any of the tests.
iii. You must use the desktop computer for the contest. This rule is intended to eliminate an advantage a group might have because of owning a fast laptop.
The lab instructor will figure each student’s grades for this lab based on the following rubric:
group’s effort in this lab, including open-loop control trial….…………..20%
group’s demonstration of mastery of relevant lab skills……………….…20%
group’s performance in test (a)………………………………………….. 20%
group’s performance in test (b)…………………………………………...20%
group’s performance in test (c)…………………………………………...20%
There will be no mini report due but you will receive 2 (identical) lab completion grades for this lab (this is just to make it count double).
4. Open-loop control
Imagine that the desired gas jet temperature is simply “warm”. Not near the maximum temperature possible, and not so cold that it is difficult to distinguish from room temperature, but “warm”.
Starting from scratch, write a LabVIEW code that outputs a voltage of your choice from the Ao0 channel of the NI myDAQ.
Pick a value for the Ao0 channel output voltage. Run your code. Did you achieve a jet that feels appropriately warm on your first try? If you did, you made a successful “open-loop” controller. Let your instructor know whether you were successful on your first try (answering this figures into your grade, although it doesn’t matter whether you were actually successful on your first try). You may find you need to adjust the voltage to achieve the warm jet you had in mind. Once you feel the temperature and adjust the voltage based on what you felt, you are “closing the loop” to form what is called a closed-loop controller. This is true even though you aren’t assigning a number to what you felt. All of the control you have been working with so far in this course, up until this most recent test, is closed-loop control. Most controllers of interest are closed-loop controllers, for obvious reasons.
5. Preparing for the contest
Setup and experiment with a controller to use in the contest, and prepare one to the point that it will be ready for the contest in your second lab meeting this week. Your computer must also be in position to quantify the test results as needed. In building your controller, you may wish to try:
· Refining the setup, e.g,:
o Making sure the thermocouple is fully immersed in the hot jet, so that hot-jet-with-room-air mixing effects are negligible.
o Making sure the thermocouple reads near room temperature when the heater has been off for a while.
o Using the coolers to prevent issues associated with cold junction temperature drift
· Working with controllers / LabVIEW:
o Considering bang-bang control
o Considering P-only control in combination with a programmatic steady-state error correction
o Considering PID control tuned by the Ziegler-Nichols recipe, possibly followed by manual adjustment
o Developing a specialized control strategy, such as:
§ a controller that starts out in one mode and then transitions to another mode at some point
§ figuring out VSS, the Ao0 channel voltage that gives the desired set point temperature at steady state, then developing a controller that goes for max heat or max cool until it gets close to the set point, then switches to VSS.
§ a controller that includes a remedy for integral windup error. Integral windup error is described, for example, at http://en.wikipedia.org/wiki/Integral_windup
o Preparing different controllers for different tests
o Modifying the loop time in your LabVIEW code by tweaking the data acquisition settings
· Practicing:
o Running mock tests as in the contests, trying different set points, etc.
· Anything else you think might work (don’t be afraid to be a little creative)
7. The contest
Good luck and have fun! Contest winners: please save your winning results (screen capture or similar) to the memory stick you can get from the instructor.
1 of 4 4/27/20125:39:40 PM