Notes on obtaining empirical heat of mixing curve:

( For this part of the project, click here for a sample template.)

  1. Input the raw data.
    1. Exclude the missing data. (For example, in Appendix B-11, for r=20, there is no entry for , so there is no need to include it as part of the raw data.)
    2. Just use a large number for r = infinity. (For example, use r = 107 ).
    3. Include an extra data point for approximately pure H2SO4. (For example, for r = 10-3, enter = 0)
  1. Create a column for the estimates based on your model.
  2. Create a column for the square of the errors between data and estimates.
  3. Create a cell for the RMS (root mean of squared errors ).
  4. Minimize the RMS by using SOLVER to change the parameters.
  5. Generate another table to observe how close the model fits the data.
    1. Because of the particular behavior of heats of mixing around the dilution, it is advisable to plot  as a function of r with the r-axis in logarithm scale.
    2. To generate the model curve, create a column that varies linearly from say n=-2 to 7. Next, generate the r values via a formula such as r = 10^n. Using the model parameters obtained from step 5, you could now generate the corresponding specific heats of mixing.
    3. Now plot the model curve and the raw data in the same graph to observe how close the model fits the data.

      4.  
    ( To change the axis to logarithmic scale, first right-click on the axis and select FORMAT AXIS. A window should appear.
Under the SCALE tab, check the selection LOGARITHMIC SCALE.)

For this particular mixing process you should obtain a model that matches the data pretty closely as shown in the figure below.


 
This page is maintained by Tomas B. Co (tbco@mtu.edu). Last revised 4/20/2000.

          Tomas B. Co
          Associate Professor
          Department of Chemical Engineering
          Michigan Technological University
          1400 Townsend Avenue
          Houghton, MI 49931-1295

Back to Homepage