Oscillations in water content output with time

A discussion forum for Hydrus-1D users.
Post Reply
Iozsef.Miskolczi
Posts: 2
Joined: Fri Oct 12, 2018 9:31 pm
Location: Canada

Oscillations in water content output with time

Post by Iozsef.Miskolczi » Mon Oct 15, 2018 5:32 pm

I completed over 30 of 1D runs of a 40 m tall column of fine soil (tailings) draindown with small incremental variations in material properties and computational constraints. I was expecting a monotonous and asymptotic decrease in water content to some lower value as computed by the final pressure head. The results are invariably showing oscillations in volumetric water content with time, in the sense that for a number of time steps the change is zero, then there is some decrease, then some zero values again. What would be the cause of these oscillations and how can it be fixed?
The top BC is constant flux=0, bottom BC is free drainage. IC is water table at the top, i.e. saturation water content. I tried changing time step to constant, increasing/decreasing grid density, adjusting soil properties parameters.
Attachments
TailingsDraindown_Coarse4.7z
(22.84 KiB) Downloaded 52 times

Jirka
Posts: 4799
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Oscillations in water content output with time

Post by Jirka » Mon Oct 15, 2018 7:47 pm

HYDRUS is a variably-saturated flow code. You cannot run the transient simulation when the entire profile is fully saturated. The initial conditions should be physically realistic. Also, you should use the iteration and discretization parameters as we recommend them. J.
Attachments
Bobadil_TailingsDraindown_Coarse4.zip
(356.8 KiB) Downloaded 54 times

Iozsef.Miskolczi
Posts: 2
Joined: Fri Oct 12, 2018 9:31 pm
Location: Canada

Re: Oscillations in water content output with time

Post by Iozsef.Miskolczi » Tue Oct 16, 2018 4:39 pm

Hello Jirka,
Thank you for the very quick response. I understand your point about the saturation profile. The reality is that tailings are fully saturated by virtue of their deposition method. I do however understand the limitations of modelling and agree that some simplifications are necessary and still lead to acceptable results. I also noted you removed the internal interpolation function. It all makes sense to me now.
Thanks again for you help.
Iozsef

zanacudi
Posts: 5
Joined: Sat Jan 19, 2019 12:15 am
Location: USA

Re: Oscillations in water content output with time

Post by zanacudi » Wed Feb 20, 2019 8:52 pm

My name is Zana Taher. I am a Phd student at Colorado State University. I guess Hydrus 1D uses ∂^2 u/∂y^2 =0 for saturated condition based on the assumption that there is no volume change at saturation. However, in geotechnical engineering (and mining geotechnics), this assumption will not work. I am currently working on creating a saturated and unsaturated modelling that can capture volume change at both saturated and unsaturated conditions like Hydrus 1D. I will post my results here when I got the modelling work.
Jirka, I appreciate all your works on Hydrus 1D, and am looking forward to your thoughts on this topic.

Jirka
Posts: 4799
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Oscillations in water content output with time

Post by Jirka » Wed Feb 20, 2019 9:36 pm

Indeed, HYDRUS assumes that specific storage is equal to zero and thus the hydraulic capacity for the saturated zone is equal to zero as well. I have my own version of HYDRUS-2D that can consider nonzero specific storage (and thus the hydraulic capacity for the saturated zone as well). However, I have never seen need to implement that also in HYDRUS-1D. J.

Post Reply