CXTFIT Boundary Condition Question

dstan · Post by **dstan** » Tue Apr 20, 2010 2:43 pm

Hello,
I'm unsure of the boundary condition to select. I am using a constant injection NaCl concentration throughout my experiment (t=24). I've reviewed the posts on this forum and read the user's manual for CXTFIT. From this, I can see the dirac delta input is in mass while the step input is a concentration. I would like to use the dimensionless option for concentration units. This would lead me to think I need to use the step input, however, with my injection being constant, I'm unsure if this is a step input as the manual indicates that this implies a non-constant input concentration. Please advise me on this.

Thank you,
Dstan

Jirka · Post by **Jirka** » Tue Apr 20, 2010 3:32 pm

Dstan,

This should be the Pulse input (with constant Input concentration), with a duration (Application time) of 24 h. See Fig79a (or Fig79b) example in the Inverse Workspace, which is installed together with the STANMOD program.

Jirka

dstan · Post by **dstan** » Wed Apr 21, 2010 8:50 pm

Jirka,
Thank you for the reply to my first inquiry. Your advice was greatly helpful. I do have one other question regarding output structure. In the space discretization area, how would the spatial increment for output apply to a column experiment? I would think this would apply to experiments that have multiple sampling points in a soil column, but when sampling the effluent of column experiments, I'm not sure how this would apply. Currently, I have this set to the length of my column. Any suggestions on how to use this in my circumstance of column experiments?

I also have the same questions about the time increment for output. Would this be the time interval samples are collected? I've tried various values here when running my data and it appears to have quite the impact on the modeling. Any advice on how to use this for my column experiments is greatly appreciated.

Thanks,
Dstan

Jirka · Post by **Jirka** » Wed Apr 21, 2010 10:02 pm

Dstan,

This is too simple question. You need to figure this out yourself (with the help of the examples).

Time increment: You should choose something, that would give you a smooth analytical solution curve.

Jirka

dstan · Post by **dstan** » Wed Apr 21, 2010 11:39 pm

Jirka,
That was a very simple question, I figured it out.

Thank you,
dstan