Dear all,
I would like to ask how should I do if I have data (C/c0 v.s pore volume) Can I use inverse solution in CXTFIT to estimate D ?
If not, I use dimension approach I am wondering about "position of breaktrough curve" in "Inverse data structure" Is it mean length of column right ?
Another one question, I try to compare dispersivity from CXTFIT and HYDRUS 2D, I found that it 's too different values. I am not sure which one is better ?
Could anyone please suggest me more details in this ?
Best regards,
Srilert
normalized concentration and normalized time
Srilert
You need to look carefully at definitions of dispersion in both codes (they obviously were tested against each other). In CXTFIT you define directly the dispersion coefficient. In HYDRUS you define the components of the dispersion, i.e., dispersivity, which drives the hydrodynamic dispersion, and diffusion coefficient (molecular diffusion). The diffusion coefficient is by default (can be turned off) additionally multiplied in HYDRUS by the tortuosity function. You need to first read the manuals to see how dispersion is defined in each code before trying to compare them.
J.
You need to look carefully at definitions of dispersion in both codes (they obviously were tested against each other). In CXTFIT you define directly the dispersion coefficient. In HYDRUS you define the components of the dispersion, i.e., dispersivity, which drives the hydrodynamic dispersion, and diffusion coefficient (molecular diffusion). The diffusion coefficient is by default (can be turned off) additionally multiplied in HYDRUS by the tortuosity function. You need to first read the manuals to see how dispersion is defined in each code before trying to compare them.
J.
Srilert
In CXTFIT, you can use both dimensional and dimensionless time. The
program internally converts dimensional parameters to dimensionless
parameters, and returns to the original form in the output. If you
know the exact pore volume, you can use dimensionless time. However,
it is generally difficult to know the exact water content value even
for a saturated condition because of entrapped air. In this case, I
would recommend to estimate v using dimensional time. You can
estimate the volumetric water content, theta, from the water flux
value and the estimated v value.
>I am wondering about "position of breaktrough curve" in "Inverse
>data structure" Is it mean length of column right ?
CXTFIT can have multiple BTCs at different position as shown in
Fig.7.3 in the manual. If the BTC is for effluent, of course, the
position value is equal to the column length.
Srilert, please carefully read example problems in the manual. We
also provide all the projects for these examples. I think you can
solve most of questions from these examples,
Nobuo
In CXTFIT, you can use both dimensional and dimensionless time. The
program internally converts dimensional parameters to dimensionless
parameters, and returns to the original form in the output. If you
know the exact pore volume, you can use dimensionless time. However,
it is generally difficult to know the exact water content value even
for a saturated condition because of entrapped air. In this case, I
would recommend to estimate v using dimensional time. You can
estimate the volumetric water content, theta, from the water flux
value and the estimated v value.
>I am wondering about "position of breaktrough curve" in "Inverse
>data structure" Is it mean length of column right ?
CXTFIT can have multiple BTCs at different position as shown in
Fig.7.3 in the manual. If the BTC is for effluent, of course, the
position value is equal to the column length.
Srilert, please carefully read example problems in the manual. We
also provide all the projects for these examples. I think you can
solve most of questions from these examples,
Nobuo