Rotated Segmental Retaining Wall 3

[PittEng88]

Hi All,

Our company recently received a submittal for a segmental retaining wall, that will be approximately 10-12 feet in height. Typically these submittals are pretty straight forward. However, the design engineer decided to rotate the entire wall, including the foundation base, 5 degrees back towards the retained soil on top of the 5 degree rotation he had from offsetting the blocks (see attached). Also, his calculations do not reflect this rotation properly, he is still using the equations as if the wall were vertical.

My main concern here is the global stability of the entire system. I feel that rotating the wall this way may cause the entire slope to give way. Is this thinking correct or am I way of base? Also, besides the fact that his calculations are incorrect can the wall even be constructed like what he is showing?

Also, there is no geogrid.

 

[hokie66]

PEinc is correct in all respects, IMHO.

 

[Doctormo]

Gentlemen,

I think I will be able to demonstrate and support my position next week when I return to the office. Was at a geotechnical conference today learning more about landslides and slope stability but I digress.

It is pretty amazing to me that that this can turn into a flame war like some sort of Facebook or Twitter feed. I would have not expected this response on this forum but live and learn.

PEinc - I trust you can wait until Monday for my information. It is just engineering, nothing personal. I think I can show where your analysis draws the wrong conclusion as it relates to the gravity wall described in the original post.

Oldestguy and RFreund - Thanks for being polite, I will post something next week that will support my position. It is not one of those things that is is obvious to the casual observer but that is why we ask questions and discuss.

For those who want to disagree, please hold your comments until next week or keep disagreeing about walls with no soil behind them, etc.

Doctormo

 

[Doctormo]

Sometimes a picture is worth a 1000 words. Happened to have access to some of these photos at home.

Note the pine trees acting as inclinometers due to the rotational failure (global stability). The wall in the photo will never tip over due to batter but can fail in other ways.

Doctormo

 

[RFreund]

See attached for different interpretations of going from a vertical wall to a battered wall. I don't think sections 1B and 1C are the same as 1A. I think we are all imagining a case for 1A, but I just wanted to post to make sure I was on the same page as others. I am curious to see Dr. Mo's response.

https://www.dropbox.com/s/oe9m6dd7f56wm7b/Slope Stability - Wall Batter.pdf?dl=0

EIT

http://www.HowToEngineer.com

 

[Doctormo]

RFreund - The differences you point out are worth noting when one gets into the specifics of the analysis and modeling. However, I think you are depicting a wall on slope where no wall is needed as an example. I would guess than 1C is the most unstable since you are adding soil to a slope and the others will not materially change the stability. One could certainly check all options but the stability of the original slope would be the most first analysis to see if the proposed construction is making it worse or not.

A wall is typically placed on a slope by a civil engineer because the grading does not work out and a change of grade is required thus the slope becomes more unstable due the change in grade constructed on it or into it. If you do not materially change the slope (other than 1C), the effects should be minimal.

Just to clarify something, slope stability is the analysis of soil slope stability whereas global stability is the analysis of the interaction of structures and slopes using slope stability methods. There are some differences in the approach and results when looking at structures vs. slopes. Checking structures for global stability deficiencies is a function of the type of structure and the soil/slope geometry above and below the structure.

I will go back to my previous example of tiered walls where the ground can be level in front of the tiers and level behind the top of the tiers but a steepened slope condition is created by the construction of the tiers that must be checked for the global stability condition even though there are no slopes involved.

Hopefully I can get something out early in the morning to discuss further before work takes over my day.

 

[SuperSandman]

Just off the main topic for a bit,

My understanding of this thread so far, is that there is a disagreement with the conclusions made by different engineers on the matter of global stability based on the tilting of a retaining wall.

My belief, as an engineer, we all have different approaches and conclusions to all engineering problems, the most important fact is that as engineers we look at all possible scenarios, think about them, calculate the outcome and proceed with the design accordingly. This maybe just a misunderstanding based on different experiences of different engineers. But from what I see, each one has a motivation for the reasoning and a possible solution to the problem (provided that nothing is neglected, that would be bad engineering).

So guys, I say this because recently I have been involved in arbitration, a large oil tank had deformed during a hydrotest, and during the arbitration process, I have read reports of numerous expert mechanical, structural and geotechnical engineers and its surprising how many different conclusions were made as to the reasons why the tank had deformed. Reports were reviewed and I have seen some nasty comments made between professionals, BUT, each professional had almost 40 years experience in their field and were part-time academics, and their evidence was certainly convincing, the poor arbitrator though!!!

Anyway that example is just taking things to the extreme...I am a junior in my field, and my experience is limited, but what I have learned from the aging generation of engineers is that our ideas, reasoning and theories may not be the same, but its the way we approach and solve the problem that makes us engineers.

Sorry for the long emotional post..

Regards, SS

 

[Doctormo]

OK, I put something together quickly which is attached.

I created simple global stability comparisons between vertical and battered gravity wall for the various level, back slope, toe slope and combined conditions. The slopes are approximately 3h:1v but vary a little for the battered condition since I was not going to make the battered wall a little taller to accommodate the sloping ground conditions and the setback.

I think you will see that global stability factor of safety drops when a wall is battered for the level and back slope condition and does not increase as some may think. While the drop is not huge in this example, it may be greater for poorer foundation soil conditions where a true global stability problem may exist. The factor of safety stays about the same for the toe slope and "wall in the middle of the slope" condition which is to be expected given the minimal effect of the wall batter on the slope condition.

The most important thing to note is the progression of the factor of safety as the slopes are incorporated in the stability analysis. The examples start around 1.50 and end up around 1.0 in the worst case with both slopes (1.30 is usually considered the minimum). Keep in mind that a normal wall design (overturning, sliding, and bearing pressure) will never address the stability issues associated with slopes so these have to be addressed separately when required.

Global stability analysis is very sensitive to the long term effective soil strengths used in the analysis so just running the quick analysis like I did for these examples might be too simple for many conditions. For example, PA where the original post is apparently located has some really bad clay soils but also has a lot of rock to evaluate. A rock foundation will usually eliminate deep seated slide planes but a wall built on a man made fill slope may be highly unstable if it is a normally placed embankment clay. Geotechnical engineers are best suited to perform this analysis or at least provide the proper soil properties for the analysis by others.

SuperSandman - Yes, a little off topic but I understand your observation. Experts are hired and paid by lawyers who search out experts that support their client so that should not be surprising. Since most engineering problems are not black and white nor is the responsibility for said problem, the lawyers can spend lots of the client's time and money hiring experts to compete with the experts hired by the other parties. However, if someone misunderstands global stability and its application, the experts will be quick to point that out when investigating a failure that has global stability as part of the problem. Even if it was not your direct fault, you will spend a lot of time and money establishing that fact regardless of outcome. Legal discussions should be its own forum

 

[emmgjld]

Doc Mo;
Good job. Now I will define my terms a little better. In my 2nd posting, I referred to 3 conditions that I use in my analysis.
1. global, 2. for footing & 3. retained mass.
and the design for GeoGrid or Fabric inclusions.
and any required subgrade improvements or geometry.
Your conditions are what I refer to as a footing condition, as the failure is directly in the footing subgrade soils. I typically refer to global as incorporating all or most of the immediate retained mass or 'wedge'. It may not entirely fit a lot of the books but it sure has been important in my analysis and design, which leads to the next paragraph of my posting.

It should be obvious that increasing the face setback or even a face rotation should normally improve the global stability and the retained mass stability. The question often involves the footing stability. It is for the reason of footing (in)stability why I have had to place a mechanically stabilized fill matte beneath some walls. Now I commonly deal with collapsible soil horizons, which colors my perception.

 

[Doctormo]

emmgjld - I think you will just confuse everyone here as this is a global stability discussion vs. batter. I have described global stability analysis as it is defined in all publications so please start another thread if you want to compare bearing capacity to slope stability analysis (and yes, I understand the concept quite well). If you would like to remove the concrete block I modeled and replace it with blocks and geogrid, it is still called a global stability analysis just with more complexity and considerations since the failure planes can pass through the structure as well.

 

[Doctormo]

emmgjld - Sorry for being abrupt, this whole subject matter can be a long discussion that does not fit well in forum posts. All I really wanted to do is show that battering a wall does not reduce global stability concerns and in fact can make them worse. It is not obvious as leaning a wall back does make it more resistant to overturning and reduces applied bearing pressure but it does not help sliding resistance (less weight over the base) which shows up in the global stability analysis.

 

[RFreund]

Hmmm, very interesting. The problem I have is that at some point you batter the wall so far back that you get a very stable slope, say 4H to 1V or whatever it is. So it seems intuitive that you would go from a vertical wall being the worst case and as you batter the wall back you get more stable. However, the information that you have presented suggests that the wall gets less stable and then (I would think) at some point "reverses" and becomes more stable again. No?

Thanks for taking the time to present the analysis.

EIT

http://www.HowToEngineer.com

 

[Doctormo]

RFreund - If you keep laying a wall back, the wall has to get taller to catch the back or toe slope to make the geometry work out. At some point, the wall is gone and you end up with a much steeper slope which is why the wall was required in the first place. It would be nice if all slopes could be 4:1 but property lines seem to keep that from happening and this is why walls are required. Laying a wall back too much can have some unintended consequences as I have shown. I think the photos I attached show what can go wrong when counting on batter too much for stability.

 

[fattdad]

not being sure how to comment, I am confused by emmgjld's condition, "2. for footing." There's no footing on a segmental retaining wall. So, for me the stability conditions are just global, external and internal.

Global represents the rotational- or wedge-type failure and when you consider the loading of the reinforced soil mass is equivalent to foundation bearing failure. External is the stresses acting on the reinforced zone and internal is the stresses acting within the reinforced zone. NCMA (National Concrete Masonry Association) has a whole design book on this matter.

I just can't wrap my brain around the notion that if you take a vertical wall (i.e., 0:1) and redesign it for a batter (i.e., 1/12:1) the safety factor would decrease! This would only make sense if the landward side of the reinforced zone were vertical for both conditions.

I am currently involved in a slope failure along a remote secondary road. The road construction required a cut slope, which triggerred a failure. The scarp, however is over a hundred feet rearward from the top of the cut slope. On impluse I just thought, we'll just buy some right-of-way and lay the slope back. Modeling this, "Remediation" showed that the safety factor decreased. For this case, the removal of the cut slope reduced the normal load on the failure surface exactly where the greatest resistance was offerred.

I read this thread and thought of my ongoing project.

f-d



ípapß gordo ainÆt no madre flaca!

 

[GeoPaveTraffic]

I've been sitting on the sidelines and "listening" to this disucssion. It is interesting to hear all of the different opinions.

My response to the original post is that laying a wall or slope back can either increase or decrease the factor of safety FOS depending on the specifics of situation. In the end if the driving forces decrease more than the resisting forces decrease then laying the wall or slope back will increase the FOS. It depends on the specifics of the case being analized. One thing that has a significant effect on the calculated FOS is how thick the wall is modeled in the stability software.

Now a few observations on the examples posted by Doctormo.

Level Condition: Reported FOS are 1.528 for the vertical wall and 1.475 for the battered wall. The vertical FOS is likely lower, the software was attempting to move the center lower; but the analysis software prevented the calculation. The center should never be at the limits of your analysis. Additionally the analysis is being effected by the high unit weight being used for the wall material, 140 pcf may be on the high side. Additionally the wall is being modeled as about 5 feet wide, most blocks are between 12 and 24 inches wide.

Backslope Condition: Reported FOS are 1.426 and 0.887. Something doesn't look right with these numbers, the small change in geometery doesn't explain this big of a change.

Toe Slope Condition: Reported FOS are 1.109 and 1.091. These results are identical for all practical purposes.

Toe and Back Slope Condition: Reported FOS are 1.053 and 1.054. Again these are the same.

In summary, your analyses show that there is no significant difference in the calculated stability results for this wall. This will be particularly true if you narrow up the wall and make some adjustments to the unit weight. However, your conditions might justify this width/weight combination.

Remember - Slope stability is all a mater of resisting forces divided by driving forces. Geometery, soil properties, unit weights, and water levels all affect these forces.

Mike Lambert

 

[RFreund]

RFreund - If you keep laying a wall back, the wall has to get taller to catch the back or toe slope to make the geometry work out. At some point, the wall is gone and you end up with a much steeper slope which is why the wall was required in the first place.

I had the level top and bottom slope condition in mind. This is the hardest for me to accept. If you keep laying the wall back for the vertical case, the wall will not get taller and it will approach a more stable condition. So it is odd that you would approach a less stable condition before returning back to a stable condition.

I think the photos I attached show what can go wrong when counting on batter too much for stability

The photos are interesting and enlightening, thanks for sharing. This is actually a slope failure though? Not settlement related? Is the soil shown at the bottom of the wall, failed soil which has been pushed up due to the sliding/circular wedge failure or is that just from construction?

The scarp, however is over a hundred feet rearward from the top of the cut slope. On impulse I just thought, we'll just buy some right-of-way and lay the slope back.

This sounds kinda like a tiered condition? Like you have a large rear wall and try to batter the shorter wall in front.

In the end, this is enough for me to check it. The real 'problem' is, at least in segmental wall industry, global stability is not typically carried out (as mentioned by PEInc previously). I say 'problem' loosely because I'm not aware that it is a major issue and for most walls it is probably not needed. But the industry is getting better by requiring a internal compound stability analysis. I think it is hard because usually the owner doesn't want to pay for it, nor the borings required, the civil isn't typical aware it is required, sometimes there is not even a geotech involved in the project and the wall designer has such a small voice since they are not part of the design team.

EIT

http://www.HowToEngineer.com

 

[Doctormo]

You are wearing me out here.

fattdad - attached is a marked up global where I tried to show the slices and how the slices go from being a net driving force to a net resisting force in the slip circle analysis. I show the "missing batter triangle" whose effect is largely a resisting force as its weight is over base with little inclination one way or the other. When this triangle is removed, global stability is decreased as the driving forces remain essentially the same. This holds true for the level and back slope cases whereas the toe slope cases tend to have circles that move further away from the wall and the resultant is at an incline to cancel out the batter effect. Your on-going project sounds like an example of this. In my sketch, all you have to do is park your pickup truck in front of the wall and global stability is improved (toe buttress solution).

Mike L - Sounds like you understand the essence of what I am trying to show. In response to your comments:

1. I am careful to not let the search go below the top of wall in this example as other problems creep into the analysis at that point and you can make the wall spin around a point if you go too low.

This is a simple gravity block example of a "big block" wall which is virtually all concrete and units to 5' deep are made. There was no point modeling a 12" deep unit at a height of 24" tall nor did I want to spend a lot of time with complexity as it was the principal that was being demonstrated.

2. The second FS is 1.367 but was obliterated by the contours in the PDF.

3. I agree that the differences are small but again the concept is what is important. The point is to understand the concept and to not think that a battered wall is more stable than a vertical wall when it comes to global stability issues. There has always been a tendency to increase batter as means of solving stability problems and there have been some spectacular failures when taken to far. My other point was to demonstrate how quickly global stability goes to hell with back and toe slope conditions as well and why the original poster was justified in asking about it.

RFreund - Hopefully my attachment helps a little. It is just the geometry and force resolution of the situation. A battered wall experiences less earth pressure than a vertical wall primarily due to the "missing triangle" in my sketch. That same triangle can be a stabilizing force in a global stability analysis where earth pressures do not matter, just soil weight and strength.

2. The soil is pushed out at the bottom as result of the wall movement and is quite classical in nature, a lot like my example shows only with more batter and a thinner wall section thus more unsafe.

3. Global stability failures are fairly rare so it does not make sense to check every wall or every situation. However, it is important to be able to identify where the risk situations are and get them addressed. Running global stability analysis with assumed soil parameters is kind of a waste of time other than for proportioning preliminary designs, etc. If there is a global stability situation, it should be addressed by the site geotechnical engineer as there is no one better qualified (usually) to assess the soils and do the analysis. In some parts of the US it is handled routinely. In other places, you would think that you invented a new technology that no one has ever heard of.

Thank you and good night.

 

[dcarr82775]

Doc,

I agree there are scenarios where the batter can lead to problems (bad soils, downward sloping toes, etc). For the example you showed the picture of, what would the FS of been if the wall was vertical v with the as-built batter? I just think you need a certain set of bad actors in the mix to run into the situation you showed (i.e. to make a big enough difference).

 

[Doctormo]

dcarr82775

I think in the photos I shared (I was not involved in this project), the wall is a fairly narrow section with a significant batter which can lead to this type of instability. I am sure the global FS would be higher for the vertical wall condition but then the wall would have just tipped over by calculation. It obviously takes certain soil conditions, moisture conditions, and possibly some poor construction to trigger failures but the key is that different failure modes can exist and that they be checked when applicable.

As a structural engineer, you know that most walls are just checked for sliding and overturning and to not exceed the allowable bearing pressures in the geotechnical report. This and the equivalent fluid pressure approach is fine for most concrete walls that structural engineers design although many are over-designed by a factor of 2X but nothing wrong with being too safe I suppose as long as someone pays for it.

Many "retaining wall systems" and "geo-structures" tend to take advantage of earth pressure theory and can manipulate the values to show adequate stability in sliding and overturning based on the same simple analysis done for a concrete wall. The common way to do this is to take narrow profile wall section (stacked blocks or short reinforcement) and batter it back a lot (>10 deg). The batter reduces the calculated earth pressure considerably and moves the center of gravity back from the toe to provide more overturning resistance, problem solved. It is also common to take the weight of the heavily battered wall structure and apply it over the base to calculate the sliding resistance even though only 50% of the weight is actually over the base. This is basically what my global stability example shows when the FS goes down due to batter. The example problem was a 11 ft wall with a 5 ft. base and a 10 deg batter as I recall. Run the same analysis with a thinner section and more batter and this effect is more noticeable.

Summary - A simple retaining wall analysis can not check global stability considerations. A global stability analysis can not check a wall for overturning. When the conditions are right, both methods of analysis should be employed to "check" a design. A heavily battered thin wall section can "fool" a simple analysis but will not "fool" a global stability analysis. Same holds true for tiered wall.

 

[dcarr82775]

I understand the differences in Global and local stability, thanks. Batter isn't bad as a general rule, taking things to the extreme is.. Cheers

 

[Doctormo]

dcarr82775 - Sorry, it was not meant to come across that way. I feel like I am talking to the group and it can read wrong to one person which I apologize for.

FYI - I have been involved in or observed many wall failures over the years (I am over 60) and a number of them were global stability related (did not tip over or fall apart). This causes me to be a little sensitive to those who dismiss, ignore, or misunderstand the concepts involved. I do not post often but this batter issue is more complicated than most people think based on this thread so I gave it my best shot.