Slope Stability of open channels

[geot88]

We are working on a project consisting of a trapezoidal open channel. Currently, the cross section is trapezoidal but it is not uniform along this channel. The height of the slope is between 2 - 3 meters (7 - 10 ft).
The project consists in excavating 1.5 to 2 meters (5 - 7 ft) deep, to increase the channel section. A slope stability analysis should be performed in static and pseudo-static conditions. The soil profile in summary shows a medium clay followed by a soft clay.

Here I attach an image of the slope stability analysis in static condition, where a FS of 1.00 is calculated. Stability analyses with similar soil profiles along this channel show the same critical failure surface (deep failure). This was expected since the foundation soil is a soft clay.

http://files.engineering.com/getfile.aspx?folder=fd3a453b-ba25-49c0-b7ff-91bbf5895026&file=slope.pdf

I have 2 questions:
1. The critical condition for the stability of this slope is in short or long term? Theoretically, when performing excavations, critical stability is presented in a long term. However, they are just trying to excavate 5 ft. deeper than the current section.
2. If the construction of this slope is considered UNSTABLE, what recommendations should I provide for this project since changing the slope of the channel is not a good way to increase FS when foundation soil is a saturated soft clay. They have recommended using geocells on the channel, do you think it will increase FS? I think it won't increase it if it's a deep failure.

Thank you!

 

[GeoPaveTraffic]

As for your questions:

1. Your analysis shows a FOS of 1 for long term conditions. You do not report the short term FOS. There is no one answer to whether short term or long term will be the critical condition, it depends on the soils involved.

2. As for how to increase the FOS, flatten the slope. Run different slope angles and you will see a change in the calculated FOS. Flatten the slope until an acceptable design FOS is calculated, then check the short term (undrained) FOS.

Mike Lambert

 

[geot88]

Thank you for your answer, GeoPaveTraffic.
Question 1 is clear now. However, for question 2, I have tried to flatten the slope to 2:1 or 2.5:1, but FoS have increased only up to 0.20, and the slope can't reach a FOS of 1.50. I've read in a book that flatten the slope works more for steep slopes, so FOS won't increase too much in this case, and that was proven on this one.
Do you think that a geocell on this channel (with tensors on each side) will increase stability? I attach a picture of the design given.

http://files.engineering.com/getfil...-0db3-4cb1-878b-9d60265f1d60&file=geocell.jpg

 

[cvg]

first of all, 1.25H:1V is generally too steep for any slope for any load condition and especially so for excavation below water table. second, not sure what a biaxial geogrid will do on the cut surface. it looks like the CELDA is some sort of block. Is that connected to the geogrid? you will need to incorporate that into your slope stability. it will provide erosion protection, but unlikely to improve your FS greatly. Think you need some sort of revetment that is a lot more substantial such as reinforced concrete, gabions or maybe a gravity wall. And flatten your slope, this is way too steep for a flood channel.

 

[geot88]

Thanks cvg. I agree with you when you say 1.25H:1V is too steep. We've worked with other flood channels and we have recommended at least 2H:1V and everything was fine. However, there is already a street on one side of the channel, so the slope can't be flattened, and if I flatten one side the failure will occur on the other side.
The street one one side is a 2-lane road (not big traffic) and I have adopted a distributed load of 20 kN/m2 for it, but I think I will decrease it to 10 kPa since the surcharge is really not substantial. Also, I will modify the water table to 3 ft above the bottom of the channel, and it will increase the FOS due to the weight of the water at the bottom of the slope. I was playing very conservative, but I will keep it more real now. I appreciate your comments.
The greogrid and geocell were recommended by some manufacturer to the director of the project (hydraulic engineer). I told him it will provide erosion protection but it won't help the stability because it will have a deep failure due to the soft clays and low shear strength. Now, do you think that a reinforced concrete revetment will also prevent a deep stability failure?

 

[GeoPaveTraffic]

The geogrid/geocell as shown in your attachment will have little if any affect.

If you can't flatten the slope further, which might even require a bench; then you will have to look at a more robust structural solution. The most likely option would be a retaining wall of some sort, maybe gabions or a a geogrid reinforced CMU wall.

Mike Lambert

 

[EireChch]

A 2:1 slope is 26 degrees, that is too steep never mind a 1.25:1. Needs to at least 4:1 i would think. If you cant flatten it will need some sort of support retaining wall, soil nails etc. Soil nails extending beneath the road where services are, probably not a goer.

I was thinking that an in ground palisade wall could be an option. It could be designed for a 4m retained height a guess would be 350mm dia pile extending to 10m depth. This may not intercept your failure plane and the material in front of the wall would most likely fail over time.

That geocell sitting in the base of the channel wont do anything for stability as pointed out.

I would not rely on 3ft of water in the base of the channel for support. What if it runs dry for what ever reason and your slope fails.

The fill on your stability analysis, what material is it? It has quite a high friction angle of 30 degrees which would make me think its a sand/gravel but also 10kPa cohesion?

Also, OP/GeoPaveTraffic - is the analysis shown a short term (undrained) since there is zero phi angle for

 

[fattdad]

You can get a 1:1 slope to work just fine! Get a construction easement, overexcavate the slope and reconstruct the slope using a RSS (reinforced soil slope). Include some back-drainage behind the reinforcement zone. Uniaxial grid will work just fine.

I'd never consider this design without some laboratory strength data or in-situ testing.

You can make it work though!

f-d

ípapß gordo ainÆt no madre flaca!

 

[EireChch]

f-d, i thought about that but wasnt sure if over excavation was an option.

We use to specify a detail for small landscaping slopes at 1:1 <1.5m height. 300mm fill lifts with geogrid placed on the base then wrapped up the face and returned back in 3m.

If it was to be used then some support to the existing road would be required as it existing road is less than 5m from top of slope....sheet pilling? Taking that into consideration then a permanent retaining solution might be better?

 

[geot88]

Thanks for posting very good alternatives. I will make an analysis with them and check their feasibility.
Just making a summary of this thread: Now that flatten the slope is not an option, we have 2 alternatives:
1. Excavation and RSS (reinforced soil slope).
2. A substantial revetment or retention structure.

For option 2, could these 2 alternatives work?
- a concrete revetment (thickness: 10-15 cm)
- concrete duct (U-shaped) - this will change to a U-section

I will check bearing capacity and settlement of ducts if it is the case.

 

[EireChch]

the concrete revetment will provide erosion/armouring protection, for it to act as a retaining wall/structure it would need to be connected to a base slab of some sort - IMHO. Interested to hear others opinions

A u shaped channels will work as this will act as a retaining wall but it would be very ugly! Settlement and BC shouldn't be an issue if you are removing 1.5m of soil and which would be replaced (at times) with water. Not much increase in net loading.

Stone gabion baskets would be a nicer job I think. As with excavation and RSS it may require temporary support measures.

 

[geot88]

Thanks EireChch for your post. I believe we are thinking the same thing.
I was thinking about a concrete revetment on both slopes of the channels and at the bottom, so it can work as a unit. My common sense -I might be wrong- says that if the left slope tries to slide along a failure surface, it will be retained by the concrete on the bottom and the right slope because its rigidity... and vice versa.
I was trying to avoid excavations because the channel is around 3 Km long. The costs would be very high. We are in South America btw. I attach a picture of the current channel, and another one of a solution with a geosynthetic concrete mattress. I prefer the regular concrete revetment though.

http://files.engineering.com/getfil...3e1d-49e2-996c-f4f743c159cd&file=channel.jpeg

http://files.engineering.com/getfil...23afd&file=geosynthetic_concrete_mattress.jpg

I will appreciate any comments about it.

 

[Mad Mike]

The main observation here is the geometry of your failure circle- conventional retaining structures and even reinforced earth embankments will fail globally unless those failure planes are cut off completely and then some- this would require either excavating your road completely or else drilling anchors beneath it.

I suspect EireChch's concrete revetment- U-shaped channel with a connecting base slab is the correct line of thinking. Will probably require some serious construction dewatering...you will need to check against buoyancy.

All the best,
Mike