Lateral Pressure on Piers Installed in a Steep Slope 2

[moe333]

I am working on a project in Southern California where a road situated at the top of a narrow embankment (about 50 feet wide at crest) is being widened to construct a sidewalk. The embankment slopes are about 25 feet high, inclined at about 1.3:1 (h:v), and are comprised of medium dense silty sand fill with average N=15, constructed about 50 years ago, and no groundwater. Weak to moderately cemented silt/sandstome is present below the fill with a gentle slope (N>75, no groundwater). Preliminary stability analyses of the existing slope indicate a safety factor of about 1.3 but more analyses will be completed to confirm this.

Construction of a sidewalk will reuire widening the crest of the embankment by about 7 feet. The project boundaries are such that the slope geometry cannot be changed, and no walls can be constructed at the toe of the slope. Everyone concerned realizes that the safety factor of the existing slope is lower than what is normally considered acceptable, but provided no additional loading is added to the slope, the stability of the slope is not an issue.

One of the options in which to construct the sidewalk is to drill piers through the slope near the top of the embankment and support the slope on grade beams conctructed on the piers. Approximately 5-8 piers will be used at a spacing of about 15-20 feet. The piers will be extended at least 5 feet into the silt/sandstone.

I am completing an analysis using L-Pile Plus 4.0 to estimate the lateral capacity of the piers. This program does not accommodate sloping surfaces. I believe that there will be a decreased lateral capacity resulting from the sloped surface in front of the pier, and an increased lateral loading due to stresses resulting from a relatively low safety factor. I am aware that Module 4.0M in the L-Pile program has an option for soil movements, but I'm not sure this would be appropriate either. I am not counting on the piers to stabilize the slope, although there will likely be some stabilization effect.

Does anyone know of design methods I could use to assess these conditions using L-Pile Plus 4.0, or similar case histories. I am also trying to assess the depth of fixity for the structural engineer.

Any ideas would be greatly appreciated.

 

[GeoPaveTraffic]

I have done several analysis of piers to stabilize slope failures, however, I have not done an analysis quite like the one you are describing. First question, if you assume that the piers are not loaded by the slope, i.e. that the slope is currently stable and that the piers do not decrease that stability, what is the source of lateral load on the piers? If this load source is low, why not consider the piers flexible with respect to slope stability?

This should allow you to design the piers using just the structural loads and ignore the slope stability loads. Just let the piers move along with the slope, if it is moving.

With respect to handling the sloping ground surface within LPile. What I have done in the past is to draw up the geometry of the piers and ground surface to scale, and make an engineering judgment about where the slope no matter matters. Somewhere around where the amount of soil horizontally from the face of the pier to the slope is about 2 pier diameters. The soil above this level I give a very small resisting factor to. In other words, I say that effectively the pier is sticking up out of the soil. Conservative, yes, but then again how well do you really know the p-y properties of the soil?

Let me know what you think or if you need something cleared up.

 

[Focht3]

Yuck! You have a tough one, my friend. I'm licensed in California, practiced for a few years in the Inland Empire/Orange County/L.A. area - sounds like typical decomposed granite sands. Where is the site, roughly?

I did the foundation design analysis on the wind turbine relocation in San Gorgonio Pass in the early 1990's using COM624 - LPILE4's predecessor. Steady, high winds - steep slopes - similar to your problem. And my Master's study centered around the use of p-y curves for evaluating a pile group load test at Harvey, LA.

In reality, there is no really "good" answer. I'll offer up a few suggestions -[ol][li]Use "realistic" soil properties - don't factor them. Apply the factors to the loads.[li]Expect to embed the piers at least five(5) pier diameters below the fill. Not five feet.[li]Assume the slope will fail and add load to the pier; you will have to run slope stability analyses that include a force in the slope large enough to maintain a factor of safety slightly greater than 1; try 1.1 to start - 1.3 (or more) would be preferred. But the pier sizes may get out of hand -[li]Run your load case assuming 50% load, 100%, 150% and 200%. Plot the groundline and tip deflections vs the applied shear (or moment) and look for signs of instability.[li]Check with Ensoft to see if they have a p-y model for c-&[ignore]phi[/ignore]; soils. Use that instead of a c = 0 analysis for the native soils, but only include a small "cohesion" component of 50 to 100 psf. It's small, but significant.[/ol]Let us know how your analyses turn out -




Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.

 

[moe333]

Thanks for the advice. The site is just north of San Diego. My preliminary slope stability analysis is based on assummed conditions. We will be drilling the slope tomorrow in order to complete a specific analysis. Some additional information and comments:

Using very low strengths within the upper 2 pier diameters to model the sloping ground sounds like a good idea.

The slope has no signs of instability and has been in this condition for 50 or more years, but owing to it's steepness, and the fact that it is a fill, I have assummed it has a relatively low safety factor (1.1 to 1.3). The piers would probably increase the stability, but only to a relatively minor increase.

A grade beam will connect the tops of the columns which in turn will be connected to the piers. Since the head of the piers will not be restrained, allowing the piers to be flexible with the (potential) slope movement would result in lateral movement of the sidewalk, which is not desireable.

Any thoughts on what factor of safety a slope would need such that it would not induce any loading on piers? My estimate is greater than about 1.4 assumming granular soils.

Assumming that there is some loading on the pier due to the slope, I would estimate that the depth of fixity would be below the fill, 1 to 2 pier diameters into the sand/siltstone...sound reasonable?

Thanks again for your advice, I will keep you posted on the project.

 

[GeoPaveTraffic]

Piers spaced at 15 to 20 feet will not provide any significant improvement to the stability of the slope. I say treat the piers as flexible with respect to the load that may be induced to the piers from the slope. Not that the piers will be that flexible. In your followup post you did not discuss the source of the lateral load that needs to be resisted. Again, I would treat the slope stability problem seperately from the sidewalk support problem and design the sidewalk supports to just support the sidewalk.

Now if your detailed stability analysis indicates that the slope is actively moving, i.e. FS<=1, then you have other problems. However, based on your post this does not appear to be the case.

I have designed and used drilled piers to stabalize one slope in my carer. The slope was about 30 feet high with an average slope of 2.5 or 3(H) to 1(V). The final design consisted of 48-inch diameter piers installed at 8-feet on center. The piers were installed to depths of about 40 feet in a 24 hour a day operation for about 2 weeks. There was a parking structure at the crest of the slope supported on spread footings.

Point being you can stabalize a slope with piers, but it takes a lot of piers, large diameter and a whole lot of steel. If the goal is to simpliy add a sidewalk to the road with decreasing the stablility, then relatively short piers of small diameter spaced as you first indicated should easily support the sidewalk and not significantly decrease the stability of the slope. Assuming your initial estimate of stability is correct.

 

[Focht3]

I agree with [blue]GeoPaveTraffic[/blue] - well said.

It looks like you have two primary loading conditions: normal use with some &quot;high wind&quot; component, and earthquake loads. The risk for the normal condition is primarily one of the formation of a slope failure, which seems low given the materials and slope's history - particularly if the calculated FOS for this condition is above 2. Of course, a slope failure - triggered by an earthquake - is the very real concern for the earthquake condition, too.

I'm more concerned about the earthquake analysis for the structure; simply providing a single value of point of fixity to the structural engineer will result in a skewed analysis. Use of a single point of fixity assumes the pier's behavior is symmetric - which is a very bad assumption. Remember that the resistance to lateral load will be higher for top of pier movement into the slope, and weaker for movement away from the slope. Movement parallel to the slope will result in a third condition. If the SE is using point of fixity in a dynamic analysis of the system, use of a &quot;uniform&quot; point of fixity is fine for preliminary analyses. But s/he will need to modify the model to account for the changed resistance depending on the direction of movement for the final runs. This is a complicated structural analysis -

I'd also use 3 diameters rather than 2 when evaluating lateral behavior in the direction of the slope because of the steepness of the slope; or, put another way, I'd want the &quot;daylight&quot; intersection between the slope and assumed ground level to be at least 4 pier diameters away from the face of the pier. But that's a judgment call - 3 diameters of horizontal soil continuity is a bare minimum to me, which would be 3/1.3 = 2.3 pier diameters of depth.

Check out a paper titled &quot;Factor of Safety and Reliability in Geotechnical Engineering&quot;, by J. Michael Duncan, Journal of Geotechnical and Geoenvironmental Engineering, ASCE, Vol. 127 No 8, August 2001. This will give you a realistic - and usable - guide to selecting the appropriate factor of safety for your project. You are really dealing with a problem of reliability; this paper will give you a very practical method to evaluate reliability without taking a course in probability theory.

Let us know what you decide -



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.

 

[GeoPaveTraffic]

Very good points Focht3.

 

[BigH]

moe333 – this seems to be a nice interesting little project you have. I’d like to make a few comments and suggest what might be an alternative way at looking at your problem. I apologise to all if this seems long.

Major Concern #1: Slope Stability

You indicate that the slope consists of silty sand fill – material that has been there for 50 years with “no water”. I must say that a 1.3H:1V slope is a pretty steep slope, even for more robust fill materials. Most highway departments would likely not favour such a steep slope of fills much less constructed of a low compact (oops, medium dense to the Yanks) silty sand.

You have indicated a factor of safety of about 1.3 – then have later changed this to an assumed value of 1.1 to 1.3. You have estimated that the N value is 15. Based on a number of correlations – say by Terzaghi, Peck and Mesri - an N value of 15 corresponds to an estimated phi value of 32°. Your slope at 1.3:1 is at an incline of about 38° - the slope is steeper than the ability of the material to hold it up (B>phi). Now, roots and the like might have “strengthened” the slope – so if one were to assume a “required” breakout at about 10ft behind the crest (B’ of 32°) would give a computed factor of safety of 1. For you to have a FS in the slope of, say, 1.2 implies that the phi of the silty sand is about 37° which implies N of 28, say. Basically, I am putting forth that the silty sand likely has an N value in excess of your estimate of 15. A 1.3H:1V slope with N=15 is too steep. Finally, you are right, in my view, with pier spacing at about 15 to 20 ft, they will do diddly-squat for the factor of safety of the slope. One other aspect – for slope stability in your case, it is not the N values or phi values at depth that are important, but the phi values within the upper 5 to 8 ft (or so) of the slope face. (It may be that there is a clayey cover that would, then, increase the FS with a lower N value in the silty sand.).

As for the seismic events, etc. – you say the slope has been there for 50 years – it has already gone through a number of events, I would suppose and is still okay. To upgrade to an adequate slope FS might be more complex than you want to do.

Major Concern #2. Lateral Loads on the Piers

Of your current thought, I will pass to Focht3 and the others who are far more conversant in lateral load analyses as you attempt to gain support by drilled piers, grade beams and RCC wall; me, I would likely use a number of simplifications along with my Poulous and Davis.

Stepping back from the problem, though, I would suggest that there is other potential ways about this problem than just placing deep piers into the slope, put on a grade beam and wall, then fill for your sidewalk and require the pier to take all the lateral load of the sidewalk fill and, I see, your worry about the unbalanced slope fill on the pier. I give one and a half!

The sidewalk and fill is about 7ft wide by 4 ft high triangle. Why not consider the use of geofoam or Elastizel (though more expensive) for a lightweight fill? This would add insignificant weight to the slope at the top and hence would have little if any effect on its overall stability.

All you would need, then, is a facing wall with minimal lateral support requirements. In this, you might consider soldier pile and concrete lagging with one set of tie-backs. The piles could be taken to about 5 to 6 ft below the slope. You could dig out a couple feet of slope to get the lagging embedded. Use grouted rip rap for slope protection in this critical area of the slope. Geofoam could then be placed behind the wall. Place 1 ft of sand above the geofoam, a thin mat of crushed stone and then pour your sidewalk or use your paving stones.

Alternatively, you could use leap-frogging methods along your slope and dig locally 2 to 3 ft into the slope for 5 ft or so of width (use shotcrete on the back panel if necessary), place a layer of geofoam, then build the sidewalk embankment up using a facing panel with geogrids between the geofoam blocks. The toe of the geofoam would be protected by a grouted riprap blanket as before. In this way, you don’t even need the soldier pile and lagging. (I sort of like this one and would pursue it a bit.)

This method of lightweight fill such as geofoam means you have basically added no fill load to the slope, minimized lateral resistance requirements and have a functional sidewalk. I do not know about the costs of the above – but sketch it out and see – it might work for you – it may be just pie in the sky but I am hinting that there are more than one way to skin a cat.

 

[VAD]

Hello moe333

Interesting project. Many good comments have been presented by respondents. Would the sidewalk be a flexible pavement rather than a concrete one?.

If flexible pavement, There are ways to gain extra width. One is to cut down existing road about 5ft. This may not be economic in this situation unless the roadway requires grading as well.

If the sidewalk is to be of concrete then would it be possible to construct the piers on the inside of the slope (roadside) and cantilever the sidewalk slab.

I guess I am a bit unclear of the process you intend to use after reading your statements as follows.

1.&quot;One of the options in which to construct the sidewalk is to drill piers through the slope near the top of the embankment and support the slope on grade beams constructed on the piers. Approximately 5-8 piers will be used at a spacing of about 15-20 feet. The piers will be extended at least 5 feet into the silt/sandstone&quot;.

This is understandable and relates to comments made by Big H
regarding making up the void with lightweight fill etc

2.&quot;A grade beam will connect the tops of the columns which in turn will be connected to the piers&quot;.

How are theseb columns related to the piers? and where are they situated.

For 1 depending on the equipment you will be using you would have to notch into the slope to effect the construction of the piers.

Deadman anchors can be used to provide restraint to the top of the piles, if lightweight fill is not to be used.

If I have misread the concept of the proposed construction, please let me know.

Other insights/comments:

Can you notch out to a depth of 2 m from the crest on the slope side and rebuild with gabion wall tied back to the inside of the slope with Tensar geogrid.

Screw piles rather than drilled piers can be used in combination with the gabion baskets., with the gabion baskets placed through the screw piles. The piles will extend to below the top of the outer basket thereby hiding the piles.

Ther are many ways in which these piles can be used in this situation to effect the widening and are effective to install from the existing roadway without disrupting the slope.

 

[moe333]

Thanks for all the suggestions, I am in the process of analysing the slope stability, then will move on to construction options. The subsurface conditions were a little different than I thought. There is about 25-30 feet of dense to very dense fill (SM-SP, N=30-55), over 15 feet of medium dense alluvium with some perched groundwater(SM, N=18-25), over sandstone. Depending on the results of the slope stability (ie: location of slip surface and factor of safety), I may be able to set the piers into the dense fill. Although now there may be an issue of stability on a potentially liqufiable layer in the alluvium which I will have to look at.

Based on correlations with N1-60 (Terzaghi, & Peck, Hanson Thornburn), it appears I can use friction values from 38-43 for the very dense fine grained sand fill. I have never used friction values this high, and am not sure if I am comfortable with it, any thoughts?

I will keep you informed as the analysis progesses. Thanks again.

 

[Focht3]

Is the site closer to I-5 or I-15? Or close to Temecula?

I'd be reluctant to go that high for design, although those values could occur. If we use Fred Kulhawy's notation such that &[ignore]beta[/ignore]; = k * tan(&[ignore]phi[/ignore] I'd limit &[ignore]beta[/ignore]; to about 0.65. (&[ignore]beta[/ignore]; is nice in that you are selecting k and &[ignore]phi[/ignore]; simultaneously.) But you might want to look at some of Kulhawy's recent work in this area before you make any decisions.

The more I think about your problem, the more I like [blue]BigH[/blue]'s suggestion.



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.

 

[BigH]

moe333 - you've indicated the slope problem again. It is my experience in sands (c=0) that the infinite slope is the limiting FS. But, this is not actually reasonable so you need to address the &quot;depth&quot; of influence - or, as I pointed out in my earlier post, how much of the slope is judged significant is the critical point - I would suggest that a minimum of 3m behind the crest may be a starting point. It is only when you are resting on clayey foundations, does the above scenario change. Also - note how the geometry and the easy SF=tan(phi)/tan(beta) gives you a good point to see that the phi of the soil is much higher than the N=15 you had presented the first time.
Keep us informed!