Why Retaining Walls Collapse
practical.engineering
practical.engineering
A gabion is a galvanised steel wire cage say: 4' x 4' x 2' (HWD). They have a hinged lid and you fill them with stones and then wire the lid shut. Sounds stupidly simple, and it is but they have some rather useful properties. Each unit is an easy one man lift, place and fill. Once filled, each one nominally becomes a large single block with great drainage properties. You can wire these things together into long rows. They work very well with water courses because they are easy to fix in place and once filled, won't move. Pouring conc. into formwork is a right old pain in a fast flowing river and it is all too easy to lose the finer particulates before the stuff has gone off (set and cured).
You can finish the exposed surfaces in various ways. You can pour soil on top and grass them, pour a bit of low grade conc and gravel for a solid "path". The cages are not particularly pretty but neither are they particularly ugly.
If a single cell fails then it generally won't cause much surrounding failure and is easy to replace or repair. You can embed fancy anchors inside them if you have a lateral thrust to resist that can't be dealt with by sheer mass.
Sleepers and the like are quite convenient but you must consider drainage otherwise they will rot within a few decades. Block backed brickwork needs a decent brickie to lay them and if they fail it is usually rather bad. I'm no brickie but I've just repaired a broken 3' retaining wall at home and it looks a bit shit. I will be hiring a professional to sort it all out in spring. Here a gabion wall is overkill!
I did say home projects above but these things are used everywhere and that includes some pretty huge retaining structures. If you are not a Civil Engineer and need to build a decent sized retaining structure then I highly recommend that you consider gabions first because you are far more likely to get it right first time.
I have no idea what windbreaks and faux wotsits are let alone "decorative" means. I'm a bodger/Engineer.
If you need to stop a vast amount of stuff moving from A to B then gabions are a very decent solution with the added bonus that you don't need to be an Engineer. Civilians will generally get it right first time just by looking at the problem.
https://www.google.com/search?q=gabion+wall
What is going on here? This is a totally bizarre architectural element.
DURABILITY OF REINFORCED EARTH STRUCTURES: THE RESULTS OF A LONG-TERM STUDY CONDUCTED ON GALVANIZED STEEL. M DARBIN, JM JAILLOUX, J MONTUELLE, and ROMANOFF Proceedings of the Institution of Civil Engineers 1988 84:5, 1029-1057 https://www.icevirtuallibrary.com/doi/abs/10.1680/iicep.1988... (pay article)
Others here https://scholar.google.com/scholar?q=galvanized+soil
The following appear to be true: (1) HDG (or equivalent) has an order of magnitude greater resistance to corrosion in soil than uncoated steel, (2) corrosion rates are highly influenced by presence (or lack) of chlorides, sulfate, and citric acid in the soil, as well as temperature.
Given the above, the suitability of galvanized steel in soil, at standard steel diameters and coating thicknesses, is either suitable (for low aggressivity soils) or non-suitable (for high aggressivity soils) over an example span of 20 years. Tl;dr: test and characterize the soil it's going to be in.
Galvanized steel is surprisingly rust resistant and that's the key. These gabions do get quite a bashing when it rains hard. The stream in question is in Devon, Woodbury (Salterton), near Exeter (UK), which is popple land. Popples are rounded stones found in the streams thereabouts. They make ideal gabion fill. They also destroy things downstream if the flow is fast enough to rouse them!
That's one case study that I've chosen deliberately to show that they work for DIY.
Not sure these are permitted in LA Country. Having tangled with the permit office before, I know just how ugly it can get (brutally ugly!) to try to do something they don't have a checkbox for.
I can see them for decorative purposes (a firepit or bbq island). Even then, one of the problems we have out here is that you are guaranteed to have all kinds of insects make a home out of a pile of rocks. We have nice ones, like black widows, that you really don't want anywhere you are going to be walking around in flip-flops.
Interesting look though.
https://dpw.lacounty.gov/bsd/lib/fp/Building/Residential%20C...
About duration, at least here (Italy) they are usually considered in the 100 years duration range, not unlike many other kinds of structures.
The inherent draining is the clearest advantage when compared to other solutions (make sure to use some geotextile behind them to "filter" small particles) , though - to be fair - they do need more space than other kind of walls and they rapidly become unusable when you go over 3-4-5 meters height.
I'm slightly puzzled it doesn't get mentioned here. Has this knowledge been lost? I've not been observing construction work first hand for a while but I don't recall the last time i saw a sheep's foot roller in use.
Either way, you can't really get to an angle that would be considered a "wall" without mechanical reinforcements, can you?
I think you may be mistaken - note that a "25 degree slope" is just a different way to say a 2 to 1 slope (2 feet across, 1 foot up). I've seen 2:1 slopes used for highway embankments, but earth fill usually specified as a 3:1 slope (18 degrees) - eg when I worked with an earth fill dam (holding back water, same as yours). I've only seen anything steeper (1:1, 45 deg) used a as a temporary (during construction) condition.
Sheepsfoot rollers are good for packing very fine material (silt, clay) but not very good for packing larger granular material (i.e. crushed stone/gravel). Silt and clay are very water sensitive materials: each has a very specific moisture content where it can be packed properly, if your material is outside of this narrow range it will not get to maximum compaction (and therefore it will eventually settle). Water moves very slowly through clay so if it is too wet or too dry it's very difficult to get it back into the proper moisture range, and if there's a bit too much sun or some rain between excavation and placement it will not get packed well. The only reason I've seen clay-ey earth intentionally used is for inhibiting water movement, IE an earth fill dam - and even there, it was a secondary barrier if anything happened to leak through the barrier membrane.
Crushed stone is (relatively) very easy to get to maximum compaction, and if it sits out for a while and gets too dry you can just hit it with a water truck before placing. It packs quickly and easily, and is stable at the ssame side slopes as earth fill
We had some crazy operators who would do things like chaining the dozer to a trackhoe at the top of the hill so it would not roll over, to do the final grade of the slope. There's more laws now, and/or fewer fools with earth moving equipment.
No worries, when I heard 25 degrees my gut reaction was "that can't be right". Then I did the calculation and saw that the common ratios used are a lot fewer degrees than I expected :) I was thinking of grades in percentage (rise divided by run expressed as a percentage - so 50% for a 2:1 slope, 33% for a 3:1, etc)
> There's more laws now, and/or fewer fools with earth moving equipment
I think stricter rules (or enforcement thereof) has led to the proliferation of long-reach excavators.
Reinforced earth is a different beast altogether for sure.
https://www.manualihoepli.it/media/doc/pr141.pdf
What I mean by 3:2 (base divided by height) is often referred to as 2/3 (height divided by base) and is roughly the "natural" angle of repose of - say - gravel.
By comparison to their source, our hills are pleasant soft sandcastles, all rubbery clays with veins of sand or gravel and never any excuse for explosives.
It seems several people are complaining that a transcript of a ~10 minute video isn't an entire engineering education in all possible details of how to create retaining walls. I think that's asking for an awful lot. But at the very least let's credit the things already discussed elsewhere.
I'm going to find "zero" very hard to believe, what with this being a 10 minute general audience video. I seriously doubt I'm just three additional minutes away from being an expert on the topic ready to take on any geoengineering task I could desire.
Edit: Let me put it a different way. While I watch and enjoy his video series, I've been showing them to my 10-year-old and 13-year-old kids, and I'm pretty sure they're not that far out of the target audience.
The top 2 videos from the following Youtube search results about "retaining walls" do not mention Proctor Tests.
https://www.youtube.com/results?search_query=retaining+wall+...
And halfway down those search results is a retaining wall video presented by geotechnical engineer Andrew Lees that's longer in duration than Grady's video and he doesn't mention Proctor Test either: https://www.youtube.com/watch?v=HGuX7rmzlzA
And after reading the wikipedia article about Proctor Tests, I think Grady made the right editorial judgement to omit that topic from a 10-minute video targeted at a general audience.
Or put another way, if cf100clunk made a video about retaining walls and was constrained to 10-minutes, you would also be forced to leave out some essential topic that other geotechnical engineers would criticize. You can't please everyone when you have time constraints.
EDIT add: thanks to other's links, I noticed that Grady already mentioned Proctor test in a previous June 2020 video "Why Does Road Construction Take So Long?"
Clearly not. He talked about soil compaction in earlier articles:
In "Why SpaceX Cares About Dirt"[1] he talks about soil compaction through surcharge loading.
In "What Really Happened At Edenville and Sanford Dams?"[2] he talks about how the lack of proper soil compaction was one reason behind the dam failures.
In "Why Does Road Construction Take So Long?" he identifies soil compaction as one of the most time consuming parts of road construction.
If anything he didn't talk about soil compaction in this article to avoid repeating himself. :)
1: https://practical.engineering/blog/2021/10/28/why-spacex-car...
2: https://practical.engineering/blog/2021/10/14/what-really-ha...
3: https://practical.engineering/blog/2020/6/1/why-does-road-co...
Also in use are rapid impact compactors, basically a crane pounding away. Vibro compaction, where 30m long vibrating needles are driven into to soil. And best of all, Dynamic impact compaction: lift a big chunk of concrete 50m in the air and let it free fall. Then do it again and again... See https://vimeo.com/415927984 @ 3:15
Nobody needs to understand soil compaction if they're willing to move and expend way more material than the bare minimum in order to solve the the problem. This is true in a lot of subject areas. You don't need to understand a lot of things if you're willing to copy what is tried and true and can tolerate some inefficiency.
For almost all personal and commercial projects the material is going to be cheaper than paying a real engineer to poke the soil with a calibrated poker and plugging the numbers into a spreadsheet that has some formulas.
Inefficiency doesnt work in packed urban environments that require to fit in a budget.
I salute you. My late father-in-law was a slide rule wrangler of a pipeline engineer who often did the same. By the time I got started in geotech computing support (Unix, Apollo Domain, VMS) we had the numbers stuff readily available on a CRT screen for guys like him.
That looks like geotechnics to me (UK). It's also not too useful here.
Compacting soil is for certain parts of paths and roads construction, not retaining walls.
And thats why running numbers has nothing to do with experience and judgement, something you don't seem understand. If you've never built a bridge you don't really understand anything about building a bridge, even if you've been to engineering school.
It's like putting out the firmware for someone's pacemaker because you can 'run the numbers'. You sure can, if you don't mind killing a few people and spending time in prison.
I think our world views diverge irreconcilably.
I recommend you take some time off the internet and do some mental self care. Whatever it is you think you're doing, it isn't good for you or anyone you interact with, throwaway or not.
These exact topics are covered in paragraphs 12, 13, and 14, respectively. 9:14 to 10:41 in the video
[0] https://www.cpr.org/2021/08/12/us-highway-36-collapse-poor-d...
As someone born in a completely flat city and now living in another completely flat city (600 m higher, but still flat), I always kinda liked slopes, especially sloped house plots - they force architects to come up with creative solutions instead of cookie-cutter boredom. But I didn't realise how far people's dislike for slopes can go until I saw this monstrosity near Nice (France) while on holiday there: https://goo.gl/maps/zf5H1jSA855bSa4XA (you can take a better look in the 3D view). That's right, they must have excavated a whole lot of rock there, and are putting up with a ~ 20 m sheer rock face right next to their houses, just so they can have nice flat plots of land! Ok, it's rock, so probably more stable than a retaining wall holding back dirt, but I would still be worried living next to that precipice (either above or below) - if not for my immediate safety, then for the long term value of my property...
That said, the view is really incredible (we live in the woods) and we don't get water pooling in our place or flooding or anything like that thanks to some well-designed drainage, so, you know, pros and cons.
He clearly spends a lot of time building high quality small-scale versions of things to show how they work, but more often than not he just shows those things while the voiceover isn't actually talking about what's being shown. Or when he is talking about the thing being shown, it's a very brief comment and then he moves on.
I love his videos, but I very often finish them and think "I could have learned more if he spent some more time explaining in detail what's happening with the model, and replaying some component of it several times over as he explains it in more detail."
He needs a second youtube channel or something where he can show way more detail about these things.
Every time I watch his videos I'm always left feeling kind of empty...
Pretend that the retaining wall is a dam. Most of the time it has to only deal with soil/subsoil/rock pressure. The lateral force on a soil dam is not the same as water - most of the force is downwards (gravity) which does not affect a dam which worries about lateral forces. However there is a certain lateral pressure anyway. Think about a pile of sand being poured. The sand tries to move sideways as well as building upwards - the sideways bit is the lateral forces on our dam. Water always moves sideways and that's the main difference. However water is less dense than sand.
If we don't allow water to drain effectively then we have both a soil and a water dam.
Retaining walls that survive a day (ie can deal with initial loads) generally collapse later due to shit drainage or inappropriate "solutions". In my opinion - mostly drainage.
Maybe not the biggest deal, but I’m pretty sure the engineering algorithms assume a solid mass of concrete. Hopefully there’s enough safety margin that this doesn’t matter.
Organic material would be fine if it was underneath eg, a temporary pathway or something, but that's about it. Actually it might also be fine for secant pile cap (talked about in this video[1]): when you drill a secant pile, you generally pour a concrete cap (that then gets drilled away).
But my understanding is that steel always deteriorates over time, especially when in direct contact with moisture or cement.
So does that imply all such structures have a lifespan that's counted in decades, and that maintenance basically requires major surgery to replace the steel?
Using these as “permanent” support is usually not done for corrosion reasons as you point out. They can be made into long term support elements but it’s costly and only done when footprint is essentially like in the middle of a city in between buildings. The buildings basement walls are designed for the entire lateral load as if the support of excavation system is fully corroded and gone, so most basements are very over engineered against earth load type failures.
Long term is generally 30 to 100 years in civil engineering, depending on when the structure was built. So beyond that life yes it’s a tear down on paper to rehab below grade.
So what's the trick with something like the Hoover Dam, where (I'm assuming) major reworking every 100 years isn't acceptable?
Did they find some way to avoid steel, or maybe make it piece-wise replaceable?
There is some steel involved, in embedded cooling pipes that were used to speed the concrete curing process and later filled with grout.
The connections to the rock are usually reinforced - steel that's encased in concrete is generally thought to be relatively immune to most corrosion effects for the design life of any particular structure (plus a lot longer if it's maintained).
"Both mechanically stabilized earth and soil nails are commonly used on roadway projects, so it’s easy to spot them if you’re a regular driver."
That sentence would have made a lovely caption to the photo of the "easy-to-spot" soil nail that the article didn't include.
The image at the top of the article is a clickable url to the Youtube video which has the diagrams/illustrations/etc.
The page submitted to HN is really a pre-written script that Grady reads to narrate the video. The intended content for integrating visuals is really the video and not the script.
> Gravity walls and mechanically stabilized earth are effective retaining walls when you’re building up or out. In other words, they’re constructed from the ground up.
The soil behind this type of wall is supposed to have been stabilized by layering a tension material at intervals to keep it from moving. There's a demo of compacting sand with layers of cloth and then using the cube of 'sand' to support one wheel of a car. The material sags slightly and then holds.
My understanding is that those concrete puzzle pieces aren't predominantly load bearing. Mostly they are for erosion prevention and perhaps moisture control (not just keeping water out but keeping the hydration consistent over time and distance).
I am much more interested though to know why it was already 4 years behind schedule.
See ~4:45 in the video on this page: https://www.thisoldhouse.com/landscaping/21018759/how-to-bui...
I'm confident it's his real name.
https://en.wikipedia.org/wiki/Aptronym
These things just happen.
This is like a geologist being named Rocky Fields.
And if Mr. Hillhouse has even the smallest sense of humor about himself I'm sure he enjoys the coincidence as well.
IMHO, Grady's YouTube channel is one of the best things on YouTube. He really is fantastic.
It really sucks to hear that, as well as every other oh-so-original joke about my name.
Usually retaining walls (higher than 2-3 meters) have an outer slope (something like 5%) otherwise they would seem (optical effect) to be falling in.
Which is exactly the point. Retaining walls are built for optical effects and looks rather than for being tough.
A tough retaining wall would be a slope. Ever seen an old church arch support structure? They are not standing vertically, they are sloped.
Sloped retaining walls would not be very nice to look at (they kinda look medieval) but would be much more resistant to lateral forces.
Retaining walls (generally speaking) are tough enough and do not fail.
Those that fail were either poorly engineered/calculated or badly built (or something else in the soils/terrain behind the was overlooked or external actions - like loads and/or vibrations from trafic were underestimated).
That is beyond and besides the actual shape and orientation of the wall.
Let's say Macbook keyboards fail much often than mechanical keyboards. Why is that?
One could say, the reason they are failing because "they were poorly engineered/calculated or badly built" and that might be true.
But another way to state this is the reason they are failing is because we like thin laptops. And making laptops thin pushes the engineering into territory where it is much more difficult to make the keyboard be designed and function reliably.
Just like our love of vertical walls makes it much more difficult to engineer retaining walls that do not topple.
You see, both of these do not necessarily conflict with each other. They are just different points of view on the same problem.
I think that blaming the problem only on engineering is shallow thinking. It is obviously right and at the same time it gives absolutely no understanding of what is happening.
It is the same kind of thinking that, for decades, blamed drivers for all accident. Obviously, it is the driver that caused the accident, but it gives no further useful understanding of the problem.
Once people understood there are other factors at play some countries started introducing solutions that make the traffic much safer without messing with the driver side of equation.
An owner/road department might decide to put some limits to the amount of land needed to build the wall and of course there may be particular cases of "narrow" areas, but we are talking of meters (and in my experience there is not that much difference in space needed for excavation between a "basic" and an "exceptionally heavy/thick" retaining wall, the difference may be at the most 1 or 1.5 meter in the base slab), and in these cases one could use gravity walls (that use much more concrete, though with less rebar and usually need less space "behind" for the excavation) or as the original article explaines anchored walls or walls on piles, or even temporary structures (micropiles and anchors usually) only to allow the building of the retaining wall.
The retaining walls (under road) in the first minute of the video are not conventional cantilever retaining walls, BTW, they are one of the common ways "reinforced earth" is made (the typical rebar concrete wall is sketched, as an example, at 10:26) compare these: https://en.wikipedia.org/wiki/Mechanically_stabilized_earth
https://reinforcedearth.com/products/retaining-walls/mechani...
with a "normal" retaining wall like the gravity and cantilever sketched here:
https://en.wikipedia.org/wiki/Retaining_wall https://en.wikipedia.org/wiki/File:Retaining_Wall_Type_Funct...
they work in a completely different manner, the MSE is more similar to an anchored wall, the "anchors" are in this case pieces of steel strips that resist the pull because of the weight and friction of the earth/soil, if the soil is not suitable or not properly compacted or you introduce a lubricant (water)there is the possibility of these strips to lose adherence, the thickness of the reinforced concrete outer slabs is irrelevant and usually fixed, independent from height of the wall or other local conditions, if I recall correctly some 14 cm.
Anecdotally, I have built both kinds of walls, and actually also once "saved" a piece (some 40 or 50 mt long) of MSE-like wall that was failing (though due to other reasons, the foundation was underdimensioned), the huge difference is that traditional retaining walls are "independent" from the soil you put behind it (as long as it is within a normal weight of 1,600-2,100 kg/m3) while the MSE stability depends greatly on the quality of the soil and the way it is compacted (besides a valid calculation of the length of the strips).
As a side note, MSE-like walls are usually perfectly vertical (and actually do seem like falling inward optically when they are high).
Old roads that formed on old paths are much more stable because people went where the area was dry, and then dirt roads were stable and didn't move down or cause rain erosion. The generalizing phrase, that terrain is just an obstacle that should be plowed through, is laughable. People lived very well without retaining walls and without such huge excavations, until transport engineers decided to please cars, and not slow them down, or not make them go too steeply.
Same for the sentence about buildings. There are plenty places where slopes are taken adavtage of. In Stockholm, there are some houses that have +2 storeys on one side, and entrances to 3 storeys from different sides. In my town, a mall built in 1960s has 2 storeys, the top one can be entered without steps at all. In Finland, they used sloped ground to build stadiums, and even to make basements with windows.
People like parks with slopes, children like to ride sleighs in winter on them. The most stupid thing you can do near housing is a flat surface and a concrete retaining wall. That generates more problems than benefits: the wall may float, it may create new concentraded areas of water running on the surface, and mini-waterfalls in rains. This never happens with natural slopes, because with grass they're very stable at 3-10 degrees, unless you concentrate water runoff on them.
Finally, the place becomes completely uncomfortable to stay at. It's pretty normal to lay down on a grass on a slope. But unthinkable if there's a retaining wall above you.
Unfortunately, nowadays I see the architects and clients prefer not to think and adapt to terrain, but just bulldoze the ground.
Otherwise, yes, there are engineering solutions to making retaining walls and making them stable.
Personally, I care a lot more about hills when I'm riding a bike or walking. When I'm driving a car, hills don't matter.
At the same time, building on slopes can be problematic. It depends on the local geology whether it is possible to do that or not.
FWIW, I'm a bit tired of "cranky old guy thinks everyone else is an idiot" posts on HN.
I believe that's called "survivorship bias"
As for buildings on slopes. Swedes and Finns enjoy rocky ground, to which bulidings stick perfectly, and they make all the nice things I mentioned. Here's an example: a mall & offices complex, the elevation difference is 5 m (15 ft): https://www.google.com/maps/@60.1687254,24.9389851,295m/data...
If there's rock under the ground, you bore and put the building on pillars. It won't matter if the surface around the buliding is flat or not.
If you have to build on a sedimentary (sandy) slope then a retaining wall won't fix this at all. You'll need many long piles to hold it to the ground by friction. Retaining wall -- no matter which end of the constuction site -- would have to be a huge and heavy structure to hold to the ground, otherwise it'll be slipping couple of cm/inches per year, which is enough to cause problems.
Even if you level the site on a slope and sedimentary ground, a building will still need long frictional piles to be stable.
But, first, many playgrounds are sloped up to 3%, people just don't notice that, and put tents on 3-5% slope without issues. (But if you mean tent for sleeping, I slept on 10% slope, without problems.)
Secondly, for playgrounds, you can put the pitch along level lines, level only the pitch itself and leave a steeper slope next to it to serve as a grandstand. Example from Finland: https://goo.gl/maps/CssidLvnTYCeZ6Qx8
Thirdly, a playground still does not require the entire lot to be leveled, which I see nowadays a lot. Earlier architects, if they wanted to build just a copy&paste standard project (rather than design something new with entries on different levels), they'd just put buildings and playgrounds along the level lines, and minimize levelling works. The lot would have 2-3 small slopes of 15-20% with grass on them between buildings or playgrounds. Nowadays, with digging machinery available, even if the same can be done, clients (mostly public agencies) order to level everything and build retaining walls at both ends.
My point is there's a plenty of ways to work around slopes in parks and land lots.