soil-water characteristic curve for transient stability analyses

by Nate Hekman on July 28, 2008

In an earlier comment on transient stability analyses, Ong said:

We (engineer) face a problem when we want to define the soil-water characteristic curve.  We have no idea what is the maximum suction can develop during the time of analysis, so we normally give up.

I asked Lori, one of our support engineers, if she could address this problem.  Lori says:

It’s important to understand that you only need to define the soil-water characteristic curve (also known as the volumetric water content or VWC function) if you are doing a transient analysis.   The VWC  function describes how water is stored or released due to changes in matric suction or negative pore-water pressure during the transient process.    

 

However, once you have a VWC  function defined, you can also use it to estimate the hydraulic conductivity function using established estimation techniques that are built into the GeoStudio software.  The hydraulic conductivity function is required for both a steady-state and transient analysis, so often a VWC function is defined regardless of the type of analysis that is being solved.

 

It’s true that if you expect the negative pore-water pressures to reach a particular value, then the function needs to at least be defined up to this value.  However there is a rule of thumb that you can use to determine what this value might be.   Unless you are considering evaporation and using a negative flux boundary condition at the ground surface, the maximum negative pore-water pressure that will develop above the phreatic surface will be hydrostatically negative.   

 

To estimate hydrostatic pressures above the phreatic surface, take the maximum distance between the ground surface and the anticipated phreatic surface or P = 0 contour, and multiply it by the unit weight of water (9.807 kN/m^3 or 62.4 pcf). 

 

So if the maximum distance between the phreatic surface and the ground is 10 m, then the maximum negative pressure that would develop  would be approximately 100 kPa. 

 

Please note that if you are going to use the volumetric water content function to estimate the hydraulic conductivity function, you will need to ensure that it has been fully defined, from saturation through to residual water content. 

 

There is no problem with defining the water content function over a range that is larger than the anticipated computed negative pore-water pressures.  Just make sure you have defined a smooth function that has a nicely defined air-entry value and a smooth transition to a residual water content if necessary.  

 

Within the GeoStudio software, you can estimate the volumetric water content  function using grain-size data, so you may be interested in watching the movie on function estimation that is available on our website at http://www.geo-slope.com/support/geostudio2007/tutorials.aspx?id=SepFnEstMov.

Does this help?  Please respond with your own comments or questions!

{ 11 comments… read them below or add one }

1

Bo 09.15.08 at 12:03 am

I really hate this option when I was doing excavation analysis – coupled and asked to input the soil-water charactistic curve, which I consider very unnessary. So I have to switch to Plaxis to carry on the analysis. This end up with the result that we are not interested in Geostudio product other than SLOPE/W.

2

Nate 09.16.08 at 10:25 am

I presume that by “very unnecessary” you mean that the SWCC does not affect the results. If changing the SWCC dramatically changes your results, then it seems very necessary to me. And if it does not much affect the results, then it doesn’t much matter how you define the function, right?

Not knowing the engineering aspect I can’t really comment on why we feel it’s important to have a SWCC. Does any other reader have a thought?

3

Alice 11.25.08 at 1:35 pm

Hello, I am doing an analysis and would ideally like to be able to switch between the wetting and the drying path of the soil water retention curve within the analysis. Is there a way of doing this? Currently I am stopping the solver when time period reaches a day of no rainfall and then changing all the parameters to the drying path of the curve then resetting the parameters to the wetting curve when rainfall begins again. This is quite long winded when I am looking at years within my analysis. I have thought about trying to find the best approximation for the drying and wetting curve but this would produce less accurate results.
Thank you for your help
Alice

4

Alice 11.25.08 at 6:02 pm

I think I have just worked out a way of doing this by using parent and child analyses.
Thank you
Alice

5

Nate 11.25.08 at 6:13 pm

Ahh, you’re too quick for me. 🙂 I was in the middle of writing a more detailed answer. Going with a hierarchy of analyses will definitely be better than what you were trying earlier. I have a couple of other suggestions too, which I’ll post tomorrow. I’m just double-checking them with others in the office before I say something wrong.

Glad you were able to figure it out in the meantime.

6

Nate 11.26.08 at 10:16 am

GeoStudio does not have a built-in way to do hysteresis, but I may be able to suggest a few things to make it easier for you.

1. As you discovered, rather than continually stopping the solver, changing the parameters, and restarting it, you can save some effort by creating a hierarchy of analyses. Give each analysis the same q-time boundary condition; make each analysis get its initial pwp conditions from its parent; and assign different materials in each analysis with the appropriate properties.

That approach will save you from having to “babysit” the solve, and will allow you to more freely make changes and still be able to re-run the entire series of analyses.

However, I discussed this with Curtis, one of our engineers, and he cautions:
“Keep in mind that some parts of the domain may still be ‘wetting’ long after the boundary condition switches to no rainfall (e.g. prorogation of the wetting front towards the water table). Conversely, while the near surface is ‘wetting’ when subjected to rainfall other parts may still be drying.”

2. It should be possible to implement hysteresis by writing your own Add-In, but naturally that’s a fairly advanced topic. You can find more information about writing Add-Ins on our web site.

If programming Add-Ins is beyond your comfort zone, we occasionally offer our services to implement Add-Ins or we may be able to recommend a third party that we know is experienced in the area you need. You will have to bring that up with the technical support team at support [at] geo-slope [dot] com.

7

N B N 11.13.10 at 7:36 pm

I think I have just worked out a way of doing this by using parent and child analyses.
Thank you

8

M. Ahmadi 05.29.11 at 2:10 pm

Dear Sir,
Hi,
First of all let me thanks because of your kind assistance. I have a question, how do I cal calculate the draw down rate of my reservoir in order to the us-stability of up-stream right abutment of my dam reservoir? The mentioned position has the landslide due to first impoundment (300 meters up stream of my dam body on right abutment).
All the Best
M. Ahmadi
Dam Department of BANDAB Consulting Engineers, Tehran, Iran.

9

Nate 05.30.11 at 8:54 am

Hello M. Ahmadi. Your question is outside my area of expertise. I suggest you email support [at] geo-slope [dot] com, and one of our engineers will be able to help you.

Nate

10

Chet Raj Joshi 07.12.11 at 12:40 am

Dear sir,
I have a question, how do I define the hydraulic conductivity function for fractured schist (more confusion about the relation between pressure and hydraulic conductivity)

Chet Raj
Nepal

11

Nate 07.12.11 at 7:39 am

@Chet: Your question is also beyond my ken! Please email support [at] geo-slope [dot] com to have an engineer assist you.

Nate

Leave a Comment

You can use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>