Water Saturation on Retaining Wall 2

[Althalus]

I just got onto a project with a team with very different backgrounds than I'm used to. Different countries and different codes all seem to be clashing. So, I need some consensus on a particular issue.

We have a large, rectangular concrete tunnel. Don't ask why they didn't make it round.

The other engineers are postulating that when it rains the soil will become completely saturated to the point where the water will subject the soil particles to buoyancy. Thus, the EFP will change because of the fluid pressure. It will then be EFP + FP.from_water.

I could see that if it were pure sand that is completely pervious. But it is road base with about 10% to 15% fines. I just don't see that happening.

I tend to believe that the water will moisten the soil and it simply will not be "fluid". But it could add to the overall density of the soil. Thus, we'd still have the same (K.0)* new_gamma (with water).

Am I off on this? Could the water behave in a completely fluid manner through such soil?

 

[BridgeSmith]

We had to consider similar buoyancy on the ramp down to a pedestrian underpass, but that's the only time. Most of our box culverts daylight, which means the ground level comes down to the bottom of the box, and therefore it's assumed that water can drain away. If this is not the case for yours, and you don't have sufficient fill depth over the top of the box, you may have to consider the possibility of it floating out of the ground.

 

[EireChch]

BS I dont think the OP is talking about buoyancy of the tunnel itself. If I am right, he/she is talking about saturated and unsaturated unit weights.

I have never actually used EFP for design, I have alwasy used unit weights and water levels.

If measured groundwater level is not an issue then I would not consider that rain infiltration will be equivalent to a groundwater level at the surface of infiltration.

You may be using an unsaturated unit weight, so maybe just use a saturated unit weight. This is essentially increasing it by 1kN/m3

 

[BridgeSmith]

Rereading the post, I think you're right, EireChch.

For the pressure on the wall below the assumed water table, with any granular material, we would use effective soil pressure, which would be the effective unit weight of the soil (soil unit weight minus water unit weight) multiplied by Ka, and then add to that the water pressure.

 

[HTURKAK]

The active or at rest pressure of soil ( based on soil unit wt ) would be used and hydrostatic pressure would be added .
Can the use of filter material and drain pipe could be an option ? so that hydrostatic pressure will not build up.


...

He is like a man building a house, who dug deep and laid the foundation on the rock. And when the flood arose, the stream beat vehemently against that house, and could not shake it, for it was founded on the rock..

Luke 6:48

 

[EireChch]

Again, I think the OP is saying that there is no water table identified. The others in his design team are saying that they should consider full hydorstatic water pressure due to infiltration from rain above the tunnel. Which I think is overly conservative.

 

[SlideRuleEra]

I've though a lot about this question since seeing it yesterday. I've spent a career working with foundations and underground structures in a "flat" area with year-round high water table (South Carolina coastal plain). For example, at our electric generating stations, for design we assume water table is at ground level... which is literally true several times a year.

IMHO, ErieChch is right... saturated soil above the water table is unlikely to exert significant hydrostatic pressure on a structure. Newton's third law of motion; if hydrostatic pressure from elevated saturated soil were to try to "push" on the structure, the necessary opposing force would be "zero" because it is dissipated in surrounding elevated saturated soil.

 

[geotechguy1]

So SlideRuleEra, in your estimation if I somehow broke open Lake Mead at the north end of Overton arm and made it instantaneously so that the water would start flowing out that end, the force on the Hoover Dam would immediately drop to zero, before all the water had drained out (since at at that end there is now no 'opposing force'). Or less dramatically I build a barrier on 3 sides of a table but not the fourth, and start pouring water onto it with a hose, even if the height of water on the table reaches say a few inches, the force on the 3 sides with barriers would be zero because the water can flow off the 4th side?

Anyway the whole thing is actually basically a PHD-level partially saturated soil mechanics problem if you wanted to get a real answer. You'd need to be able to model the rate at which the water infiltrates and the wetting front moves down, how the suctions work, and how much water is running off the surface instead of infiltrating (assuming rainfall is the scenario you are dealing with).

As a sanity check / 1st pass, what you could do is a very conservative calculation using the void ratios of the soil around the tunnel, and credible rainfall (assuming of course that your tunnel doesn't have a big low point where surface water runoff is ponding at the top), and then assume that 100% of the water infiltrates, and see what thickness of soil could be saturated using design storms (eg. the 12h / 72h etc worse case rainfall). Eg, you might have a 500 year ARI 12 hour design storm of say 5 inches (or 125mm).

Using a very simple green-ampt infiltration model (flat ground, ignoring ponding and all the fancy stuff) you then have:

Z = I / (Omega1 - Omega0)

Where Z = depth of the wetting front (this is the depth where everything becomes saturated)
I = cumulative infiltration (we are going to assume that 100% of your design rainstorm is infiltration, here say 125mm)
Omega0 = initial volumetric water content before the rain storm (I'll asume it's 50% saturated)
Omega1 = final volumetric water content after the rain storm (I'll assume it's 100% saturated)

Lets say that Omega1 is about 0.3 (this is roughly volumetric water content for a saturated sandy soil) and that the initial condition is 0.10 (it's not totally dry, but it's not saturated either)

We then have

Z = (125mm) / (0.3 - 0.1) = 625mm.

Of course, the number varies greatly on many factors - If Omega0 is 0.25 you will get 2.5m, if 75% of your water runs off as surface runoff you will instead get 150mm depth of infiltration. Then you'd have to consider that in a 12 hour storm, the average rainfall rate is 11mm/hr so if the permeability of the gravel is greater than that it's going to drain out faster than it can accumulate (of course - then in the 'infiltrating fast' case, if the water doesn't have anywhere to flow to, it might start accumulating somewhere, say adjacent to the base of your tunnel wall).

Over the depth of the wetting front you lose the beneficial effect of suction on strength (no issue - we all already ignore this), and I'd argue over the 600mm you do have hydrostatic pressure acting at the sides against a wall.

 

[SlideRuleEra]

geotechguy1 - I'll accept your explanation since my depth of geotechnical knowledge is no where near PHD level. Actually, it is much lower than that since my formal engineering education is in mechanical / electrical (which I've mentioned many times is various forums, but maybe not here.) What structural / civil / geotechnical insight I have I chose to teach myself over the past 49 years... because I like it, and my employer gave me full opportunity to put it to use... even though the company was well aware of my ME / EE background.

The (mechanical) analogy that I believed applied was hydrostatic testing of a tank or pipeline. Tank / pipe walls contain the water. As long as the volume of water is constant, pressure is maintained. If a leak occurs, pressure drops rapidly when the water volume decreases just a small amount. I assumed that saturated soil, above the water table would lack containment, resulting in no hydrostatic pressure on the structure.

I stand corrected.

 

[geotechguy1]

Maybe this is a case of different disciplines using the same word to mean something different? Geotech's use 'Hydrostatic' to refer to a condition where the pressure from the water is simply related to the height of water above you. Eg. If you have 5m of water, you have 5m * ~10 kPa/m = 50 kPa of pressure in all directions (at 5m depth) and 20 kPa at 2m depth and so on and so forth. Whereas in your explanation, 'hydrostatic' testing sounds like something different - i.e., isolating part of a pipeline and pressurizing the fluid in it to test for leaks.

 

[geotechguy1]

Although I'm not going to pretend to be an expert in fluid behaviour...as geotechs our understanding is quite limited on this topic

 

[SlideRuleEra]

geotechguy1 - You may be right, never occurred to me that the definition of hydrostatic could vary. Many of my projects were a blend of several disciplines (that's why I got the assignments), so I never had a fixed set of terms.

 

[geotechguy1]

Do you have any good references that explain how hydrostatic testing works and the theory behind the pressure drop. Curious to learn about it now

 

[SlideRuleEra]

The principle is basic, water is almost incompressible. A small change in the volume of pressurized water, for even a very large volume of water, results in a rapid drop in pressure.

Pumps or just the gravity head of water is used to provide the pressure. Then the tank, pipeline, etc. is closed. If there is no or minimal drop in pressure, there are no leaks. Sometimes dye in the water is used to help find a leak. The hardest part is getting all the air out of the system. With air being compressible, it takes a much larger leak to cause significant pressure drop.

Safety is another reason to use hydro testing. If there is a catastrophic failure of the vessel while pressurized with water, there is just sort of a "poof" as hydraulic pressure drops suddenly without a big change in volume. If the vessel had be pressurized with air, the sudden expansion would be similar to an explosion.

Hydro testing works well with steel and ductile iron vessels and pipelines which are fairly rigid. Not as well with PVC and especially HDPE, which expand as they are pressurized.

There are all sorts of systems, large and small, to be hydro tested at a new generating station. The most interesting one I did was a project to design, construction manage and test a 1.5 mile long underground pipeline. I settled on 20 inch diameter HDPE pipe. It took two days to hydro test since the pipeline expanded slowly over time, even though operating pressure was very low (about 15 or 20 psi).

I don't know of any references, but there is plenty of info on the internet.

 

[PEinc]

Althalus, a sketch would be helpful. I do not know how big or deep your concrete tunnel is; but I assume the tunnel bottom would be at least 3 to 5m deep. If there is no normal ground water level at or above the bottom of the tunnel and unless it is monsoon season with very permeable soil, I can't imaging there being enough rain and enough time to fully penetrate and saturate the soil from the ground surface to the bottom of the tunnel.

http://www.PeirceEngineering.com

 

[EireChch]

I think its far from a phd problem

We dont design retaining walls for hydrostatic pressure created by retaining walls getting soaked through days of rain! well, at least I have never done so or seen it recommended in a standard!

C760, which is the retaining wall gospel in the UK, mentions 'rain' once in its 470pages and that is in relation to leaching of chlorides and sulphides and concrete protection. Nothing on rainwater soaking the ground and inducing a lateral earth pressure.

 

[geotechguy1]

We don't design walls for hydrostatic pressure from rainfall because...

1. We design walls with drains so that the water doesn't build up
2. Frequently we seal the area around the top of walls to prevent / minimise infiltration
3. We ignore the benefical effects of suction which means we don't have to account for losing those effects progressively during rainfall
4. We often consider implicit effects like increasing the static GWL 1m or 2m which is in essence implicitly considering a system-level groundwater table rise cause by seasonal weather effects (rainfall)
5. It's really complicated, and we like to pretend all of our partially saturated soils are dry or saturated when they are in fact not

 

[EireChch]

GG1

1 - Not always
2 - not always
3 - agree
4 - agree, but in the OPs case, there is no groundwater within the tunnel depth. So no need to consider any potential effects (in terms of lateral pressure) of rainwater, IMO. What would you do in the OPs position?
5 - agree.

 

[geotechguy1]

To make one more point before I give up on overcomplicating it:

We're prone to thinking of these problems as 2d but really it's 3d - and a big concrete box is a pretty nice barrier to groundwater flow that might be prone to causing groundwater mounding arounding it depending on what the surrounding terrain and geology is like.

 

[Smoulder]

https://www.google.com/search?sca_e...HdbKAJYQ0pQJegQIBhAB&biw=412&bih=702&dpr=2.63

Navfac has a method for extra force due to water flowing from ground to drain through the active zone.