Lab 3 - FEM II. - parameter extraction
Lab 3 - FEM II. - parameter extraction
LAB: FEM II. - parameter extraction
Goal: learning how to estimate flow parameters of a simulated microfluidic chip
Tools needed: COMSOL Multiphysics -> see FAQ
Video tutorial:
- Not possible in MOOC version
If you would like to register and get graded for the course, please see FAQ-1.
OVERVIEW
This
lab is based on top of Lab2. In order to follow the instructions, the completed
Lab2 COMSOL file is needed. All tasks are explained as edits to that file. Both
chip geometries (with and without micropillars) will need to be used for calculations.
Both CAD drawing files (.dxf) can be found on Moodle under Lab2.
REPORTING GUIDELINES
Not relevant in MOOC
REPORTING GUIDELINES
Not relevant in MOOC
H-MICROREACTOR WITHOUT MICROPILLARS
First, we will go through the necessary additions to your COMSOL model to extract parameters such as ionic concentrations at the outlets and electric conductivity (EC). We will also learn how to perform parametric sweeps for your model to get outputs for different input values of defined parameters. Based on your experience, you can do this alone or together with the practice leader.
2.2.1 Input vs output concentration
In previous lab we went through the use of Boundary Probes and Globally Defined Parameters. We will now introduce parametric sweep (Study), which is a convenient tool to run various input values (concentration in this case) while executing the simulation once.
a) Open your COMSOL file from Lab3
b) In the Model Builder under Global Definitions > Parameters, set a fixed value for the input velocity: 3 [ml/min] (to enable us to see the effects of only the input concentration)
c) Right Click on Study 1 in the Model Builder and select Parametric Sweep.
d)
In the Parametric Sweep
settings window, add our first parameter with the 
[Add] button under Study
Settings table which initially is empty.
e) In case such as ours, we have multiple globally defined parameters, change the “Parameter name” to c_in from the dropdown menu if not already so.
f) Continuing in the Parametric Sweep Study Settings, specify the “Parameter unit” as mol/L
g) Using the values given in the table below, add each c_in value to the “Parameter value list” (example: 0.01, 0.1, 1.0 …)
h) Run the simulation
Under the Graphics/Plot window a Probe Table will start to accumulate output concentration values while the simulation moves through each defined input.
Table 1:
|
Nr. |
c_in (mol/L) |
c_out1 (mol/L) |
c_out2 (mol/L) |
|
1 |
0.382 |
|
|
|
2 |
0.764 |
|
|
|
3 |
1.528 |
|
|
|
4 |
3.057 |
|
|
|
5 |
6.113 |
|
|
Effect of velocity
Since we are trying to sweep multiple parameters it is more efficient to add a separate Study for each of our parameters, a new Probe Table and extra probes in order to not lose time to change our simulation settings for another parameter every time and to prevent overwriting our simulation data in one COMSOL file.
a) In the Model Builder under Global Definitions > Parameters, set a fixed value for the input concentration: 3.057 [mol/L]
b) In the top of the COMSOL layout, on the ribbon bar next to File select Study, then Add Study
c) In the opened Add Study window normally next to the Graphics/Plot window, choose Stationary as with our already existing Study and add it to our Model.
d) In the Model Builder add Parametric Sweep to the newly added Study.
e) Edit the Parametric Sweep Study Settings to have v_in as our changing parameter with the values and units given in the table below.
f) Moving to the Results in the Model Builder, expand “Tables”. Here you’ll see your dataset Tables which can be specified for individual studies in order to separate data and not overwrite them. Right click on Tables and select Table to add a new one. Rename it to “Probe Table 2” for example.
g) In the Model Builder under Component > Definitions duplicate your two outlet Boundary Probes and rename them to be tied to the effects of velocity.
h) Under the settings of the duplicated Boundary Probes, in the Table and Window Settings section, change the Output table to be “Probe Table 2”. Do this for both duplicated Boundary Probes.
i) Under our first Study (Study 1) Parametric Sweep, in settings at the Output While Solving section where you have the dropdown menu for “Probes:” change that to Manual and delete the duplicated Boundary Probes from the list.
j) Under our second Study (Study 2) Parametric Sweep, in settings at the Output While Solving section where you have the dropdown menu for “Probes:” change that to Manual and keep the duplicated Boundary Probes and delete the rest from the list.
k) For this example, compute both Studies one after another.
l) Once the simulation has ended, for exporting the dataset look to the Model Builder, under Results Right Click on Export and select Table, in the settings you can change which Table data you want to export as also the name and location for the output file. Do so for both Probe Tables.
Table 2:
|
Nr. |
v_in (ml/min) |
c_out1 (mol/L) |
c_out2 (mol/L) |
|
1 |
1 |
|
|
|
2 |
2 |
|
|
|
3 |
3 |
|
|
|
4 |
6 |
|
|
Estimated EC from model
In the first experimental lab we measured salinity of sample saltwater mixtures. The goal there was to obtain a set of electrical conductivity values which when plotted and compared to molarity a linking relationship formula would be possible to be obtained from the plotting tool i.e. Excel using logarithmic regression function. For the purposes of the first example, we can use the pregiven formula of y = 0.4497*ln(x) + 7.2734, where y = EC and x = Boundary Probe in COMSOL
In this simulation we will be sweeping only the input concentrations as given in the table below.
Table 3:
|
Nr. |
c_in (mol/L) |
EC1 (μS/cm) |
EC2 (μS/cm) |
|
1 |
0.382 |
|
|
|
2 |
0.764 |
|
|
|
3 |
1.528 |
|
|
|
4 |
3.057 |
|
|
|
5 |
6.113 |
|
|
H-MICROREACTOR WITH MICROPILLARS
Change your geometry for the one including micropillars and conduct the simulations for the tables below as was done for the example without the micropillars.
Input vs output concentration (fix input velocity)
Table 4:
|
Nr. |
c_in (mol/L) |
c_out1 (mol/L) |
c_out2 (mol/L) |
|
1 |
0.382 |
|
|
|
2 |
0.764 |
|
|
|
3 |
1.528 |
|
|
|
4 |
3.057 |
|
|
|
5 |
6.113 |
|
|
Effect of velocity (fix input concentration)
Table 5:
|
Nr. |
v_in (ml/min) |
c_out1 (mol/L) |
c_out2 (mol/L) |
|
1 |
1 |
|
|
|
2 |
2 |
|
|
|
3 |
3 |
|
|
|
4 |
6 |
|
|
Estimated EC from model
Table 6:
|
Nr. |
c_in (mol/L) |
EC1 (μS/cm) |
EC2 (μS/cm) |
|
1 |
0.382 |
|
|
|
2 |
0.764 |
|
|
|
3 |
1.528 |
|
|
|
4 |
3.057 |
|
|
|
5 |
6.113 |
|
|