EAS 199B Lecture 18:     6, 7 March 2013

Last updated at 02:23 PM on 07 Mar 2013

18. Model for thermal control of the fish tank

Class meeting on 6, 7 March 2013

Homework: Assignment 5 due 11, 12 March 2013

Learning Objectives

  1. Describe a procedure for estimating the smallest realistic value of the deadband for thermal control of the fish tank.
  2. Describe experiments for measuring the characteristic constant, K of the fish tank.
  3. Explain the physical significance of K.
  4. Describe a procedure for using K and the temperature error to determine the duration of heat input.
  5. Explain the role of salinity control deadtime in setting an upper bound on the duration of heat input.
  6. Explain the pseudocode version of a thermal control algorithm for the fish tank.

Notes and Supplemental Files

  1. Original notes on temperature control of the fish tank (Early AM, 7 March 2013)
  2. New! Improved! notes on temperature control of the fish tank (Early PM, 7 March 2013)
  3. Checklist 5 for calibration of the thermistor
  4. Arduino code snippet to measure and display time, heater status and fluid temperature.
  5. Arduino code snippet showing the structure of temperature and salinity control.
  6. MATLAB code to automate the measurement of K. Click on link below for more information

The plot_heater_test.m MATLAB program automates the measurement of the K, the slope of the temperature versus time curve. The code requires a plain-text, tab- or space-delimited text file with three columns: time, heater status, and temperature. The heater status is number, either 0 or 1. A value of 0 in the second column indicates that the heater was off. A value of 1 indicates that the heater was on.

The MATLAB code uses mouse input to identify two regions of the data that will be used to create least squares line fits. The code will wait until you click on the starting and stopping points for both segments.

Two data files heater_test1.txt and heater_test2.txt are provided as sample data for running the code. Do not use the results from the sample data to specify values for your control program*.

  1. Launch the code by typing plot_heater_test at the command prompt
  2. Click once at the start of the ramp in T versus t. This defines the start of the first line fit segment.
  3. Click again at the end of the ramp. This defines the end of the first line fit segment.
  4. Click once at a point where the steady state behavior appears to have asserted itself again after the heat input ramp has stopped.
  5. Click one more time a the end of the steady state behavior