Why Bridges Don't Sink | Hacker News Reader

Why Bridges Don't Sink

practical.engineering

358 pointschmaynard114 comments

Favorite bit:

> The tagline of the Pile Driving Contractors Association is “A Driven Pile is a Tested Pile” because, just by installing them, you’ve verified that they can withstand a certain amount of force. After all, you had to overcome that force to get them in the ground. And if you’re not seeing enough resistance, in most cases, you can just keep driving downward until you do!

> in most cases, you can just keep driving downward until you do!

I feel that the cases in which that technique doesn't work are stories to be told. Do you just keep driving downward for a very long time? How long?

Satan, far below some very deep, soft, and slippery soil where an overpass needs to go:

> What the heck?

IANAE (No An Engineer), but I think he mentions both the issues of piles wandering off-course, and of unanticipated piling problems causing major budget & scheduling issues.

From a structural PoV, an extremely long piling in soft-ish soil will start having problems with lateral deflection - which it is too thin (relative to length) to resist. Then there's the case of "we think we finally hit bedrock...but what if it's just a big boulder?".

I can imagine cases of pilings running into large underground caverns, or penetrating strata containing water / gas / petroleum under pressure.

Edit: From a quick search...

In some locations, bedrock may not start until >1000' below the surface.

And here's a very quick & simple intro to the fact that "bedrock starts at depth D" is usually too simplistic: https://education.nationalgeographic.org/resource/bedrock/

> I can imagine cases of pilings running into large underground caverns

Example being the Lake Peigneur disaster.

https://64parishes.org/entry/lake-peigneur-drilling-accident

On the morning of November 20, 1980, the crew drilling near the salt mining operations reported that the tip of their drill shaft was stuck. After the crew removed the tip, they heard strange noises and abandoned the platform in the nick of time. A giant mud crater began sucking down barges, rigs, and almost some fishermen who escaped with moments to spare.

>Then there's the case of "we think we finally hit bedrock...but what if it's just a big boulder?"

Doesn't matter.

There's two types of pile support: noncohesive and cohesive. Which can be thought of as end (bearing) resistance and side (friction) resistance.

Most people only think of end resistance.

Most end resistance piles aren't driven to bedrock or even a boulder, but a strata of soil with sufficient strength. Usually a layer of sand under silt or clay, but a boulder could do it.

Here's the catch- if it's a one-off, then adjacent piles won't hit it, and you'll see the anomaly. Mitigation may or may not be required. If it's not, then you've hit a strong (noncohesive) layer of boulders.

Either way- it goes back to the point: each pile is resistance tested. And you know now not only the insitu soil strength, but also that of each layer to reach that depth.

Also side note- the act of driving and then post-drive settling both build addl strength. Eg, The force used to drive the pile, applied a few months later, usually won't be sufficient to drive it any further.

https://en.wikipedia.org/wiki/Kola_Superdeep_Borehole

Well, whatever you do, do not perforate an anhydrite layer!

https://en.wikipedia.org/wiki/Staufen_im_Breisgau#Geothermal...

This is fascinating, thankyou!

Another one to watch out for is mud diapirs. In coastal deltas where thousands of feet of infill has occurred over time the interaction between hydrocarbon formation and organic silts can create mud volcanoes.

https://www.researchgate.net/figure/Comparison-of-signatures...

These can be anywhere things that shoot liquid mud out of the ground to areas of very deep low seismic velocities where you could drive a pile thousands of feet to the bottom of hell and barely get any resistance.

Much the same, one should be careful when drilling into mud layers

https://en.wikipedia.org/wiki/Sidoarjo_mud_flow

Wow. I knew drilling could cause land to sink. I never imagined it could cause land to rise.

Fascinating, thanks!

Also don't accidentally let a lake--and the boats in it--start filling your tunnels through salt.

https://en.wikipedia.org/wiki/Lake_Peigneur

> Days after the disaster, once the water pressure equalized, nine of the eleven sunken barges popped out of the whirlpool and refloated on the lake's surface.

wow.

> there were no human deaths, three dogs were reported killed. All 55 employees in the mine at the time of the accident escaped

omF...g the mine was active!? and folks were just drilling on top of it!???

huh. good old 1980s.

Many have said similar about their code

What does this mean exactly?

Utilizing “Brute Force” as a testing and verification strategy.

Maybe it's a parallel to "if it runs, it works" :)

works on my box

"if you’re not seeing enough resistance, in most cases, you can just keep driving downward until you do" -> "Many have said similar about their code" -> "works on my box"

How does this make sense or have any coherency?

User-side tests are the only tests that really matter.

Everything else you do is there just to reduce the odds of users tests catching anything. But you don't get any certainty before that step... that happens after your software is on production and people depend on it.

(Of course, that's a worldview that can be either very beneficial or incredibly harmful depending on what you are creating. It's not good to see it applied to bridges, but I believe the OP did it in jest.)

That's why we call it big test, not production.

Unless you are talking about fancy dependent typing you might misunderstand the quip. Writing code does not test it.

I can imagine that slow static loading could allow sinking whereas dynamic force would not. Soil liquification is a weird thing, analogous to silly putty where it can be soft when manipulated slowly but hard when impacted quickly.

Yeah, it also assume that the pile you're driving can be arbitrarily long and will last forever. They used to be made with trees, for which this is obviously false.

It depends. Fully soaked ground will actually preserve wooden piles (wood decay is aerobic and requires oxygen). This is why Venice and New Orleans are both built on them (sinking issues aside because they have other issues). The piles in both cases are quite stable because the ground is completely soaked. Where you run into issues is where water and air meet. I would imagine wooden piles in just water would have issues with shipworm (in appropriate venues). But the ones in fully soaked soil seem to last just fine.

I suspect, but don't have data on, that wooden piles may actually last longer in those exact circumstances due to galvanic issues with concrete and rebar or metal pilings.

> Venice and New Orleans are both built on them

Parts of Amsterdam as well.

But they don't last forever, so a lot of them are being replaced at the moment. Expensive operation as they have to be replaced in-place, but Amsterdam canal front houses are prime real estate.

> Fully soaked ground will actually preserve wooden piles (wood decay is aerobic and requires oxygen).

When building fences, the ground-air interface is often where rot occurs, and there are products to protect that area:

* https://www.postsaver.com/en-gb/products/pro-sleeve-fence-po...

Any idea how well those stand up to lawn trimmers?

It looks like a layer of plastics, so I would expect not well. Though, the part you'd hit would be above the ground, so it might still protect pretty decently for all the below ground stuff. You might consider putting a small set of stones around it so that the trimmer cord hits those instead of the wrapped wood.

> You might consider putting a small set of stones around it so that the trimmer cord hits those instead of the wrapped wood.

Stones could help with drainage and drying as well.

Should probably just do a layer of stones under the whole fence... then you'd get that benefit for all of the boards that are close to the ground and not just the posts.

Not to mention that wooden piles were often treated with an absolutely massive amount of creosote, to the point that a number of timber pile treatment yards are superfund sites.

Especially for trains. For one, they are much more susceptible to sinking, but also likely produce all kinds of resonance.

“the slow blade penetrates the shield“

how does that hold up to quake liquification?

Tested at the moment of installation != test valid 30 years later

>> Your guess is as good as mine why the same steel shape is an I-beam but an H-pile.

This is because the shapes are different. I beams are typically more slender through the web because the goal is to concentrate mass at the flange for moment capacity because they’re beams and geared towards bending. H piles are thicker in the web with the web thickness usually similar to the flange because the use case requires axial capacity and various constructability considerations. I beams turned into W (wide flange) and S sections in the standard shapes and H beams are called HP sections.

You’ll often see them cross-specified for foundation work but it’s rare that you’d choose an HP section over a more efficient section like a W or S for something “out of the ground.”

Thank you for this. In college, for some reason, i hung out with architecture majors instead of my fellow computer science people. They would talk about "w flanges" when, to me, they meant I-Beams. I never cared enough to ask but knew better than to try and correct them because that's pretty annoying heh.

It's good you didn't correct them because they aren't the same thing. Wide flange beams have wider flanges that don't taper or taper less than I beams. Your architecture friends were probably using the right lingo.

Important bit of background knowledge: "The horizontal elements of the I are called flanges, and the vertical element is known as the "web"."

Useful graphic: https://en.wikipedia.org/wiki/File:I-BeamCrossSection.svg

> But, what if you just keep loading it and causing it to sink deeper and deeper?

I believe this is the same fundamental engineering method used in a swamp by Herbert's father in Monty Python and the Holy Grail. [0]

[0] https://m.youtube.com/watch?v=w82CqjaDKmA

One day son, this will be all yours!

What, the curtains?

When I started here, all there was was swamp. All the kings said I was daft to build a castle on a swamp, but I built it all the same, just to show 'em. It sank into the swamp. So I built a second one. That sank into the swamp. So I built a third one. That burned down, fell over, then sank into the swamp. But the fourth one stayed up. And that's what you're going to get, Lad, the strongest castle in these islands.

The real world version of that: The causeway for the Lucin Cutoff across the Great Salt Lake.[1]

The Southern Pacific dumped in fill rock starting in 1902, and the rock sank into the sediment. But they didn't give up. They kept dumping in more rock. They still couldn't get above the water line. So they built wooden trestles on the foundation thus created. That worked, but the trestle was too weak and limited to slow trains. So eventually, the Union Pacific dumped in far more rock and built a solid rock causeway all the way across the lake. The causeway had to be raised in 1986 and strengthened.

Today, it carries long UP freight trains, part of the transcontinental main line.

[1] https://utahrails.net/pdf/UP_Great-Salt-Lake-Causeway_2007.p...

That's why there were five Babylon stations... JMS lifted the joke. Nice!

> Quote-unquote “bedrock” is a simple idea, but in practice, geology is more complicated than that.

There is always bedrock, but in some places your pile would have to be really long to reach it:

> The gravel deposits of 100 m (330 ft) are the deepest in the south of Munich and decrease towards the north.

(from https://en.wikipedia.org/wiki/Munich_gravel_plain - not saying this is anything really extraordinary, but it's the area I'm most familiar with)

The Otira Gorge Viaduct in New Zealand, that carries a highway that crosses the Southern Alps, has its foundations in a deep layer of talus that has fallen off Hills Peak over centuries - that movement of rock being why they built the viaduct to replace the road - as the slope eroded the road had to be moved higher up the slope, adding more switchbacks to the infamous Zig Zag [0]. Plus the falling rock that made the road dangerous.

They were determined to hit bedrock, but yeah, was buried too deep. [1]

[0]: https://teara.govt.nz/files/p-8788-gns.jpg

[1]: https://www.stuff.co.nz/national/117150792/awardwinning-otir...

TIL New Zealand have their own Alps!

Technically "Alps" is the plural of "alp", which means a very high mountain.

However "alp" comes from Latin "Alpes", which is the mountain range in Western Europe we now call the Alps.

The word has become genericized to a degree. One the other hand Alps used to be one very specific mountain range, and alp a mountain in that mountain range, so surprise at some other place calling their mountain range Alps is understandable.

>surprise at some other place calling their mountain range Alps is understandable

yes, if you come from Wellington in Suffolk and you fly to Wellington in NZ, and then encounter that the nearby mountains are called Alps, you would be shocked, shocked

The North Cascade Mountains in the Pacific Northwest are also sometimes called the "American Alps".

(Personally, I think it's a silly name. The Cascades are majestic enough in their own right and need no comparison to any other mountains.)

Australia, too: The Victorian Alps [0]

0: https://en.wikipedia.org/wiki/Victorian_Alps

The fascinating thing about the Munich gravel plain is that it's really just a (very slightly inclined) plain - it's almost 100% flat, and the gravel is covered by a (relatively thin) layer of soil, so you could easily mistake it for the typical "lowlands" alluvial plain, and you'll probably be surprised to learn that it's at a height of ~400-700 m. So you don't really have to worry about the gravel moving - but it's still there...

Some of them wont sink because they float...

https://en.wikipedia.org/wiki/Nordhordland_Bridge

There are three floating bridges on Lake Washington in the Seattle area as well. The Lacey V. Murrow Memorial Bridge[0], the Homer M. Hadley Memorial Bridge,[1] and the world's longest floating bridge the Evergreen Point Floating Bridge.[2]

[0] https://en.wikipedia.org/wiki/Lacey_V._Murrow_Memorial_Bridg...

[1] https://en.wikipedia.org/wiki/Homer_M._Hadley_Memorial_Bridg...

[2] https://en.wikipedia.org/wiki/Evergreen_Point_Floating_Bridg...

And one on the bottom, no longer floating.

(note -- was a bridge engineer in Seattle and did work on the old 520 bridge when we designed the retrofitted post-tensioning it in the late 90's. Among other tasks, I supervised a guy drilling holes in the bottom of the bridge with a concrete corer. )

I'm not an expert but I have seen Titanic (1997) and I would think a floating bridge is most vulnerable to sinking.

Nah. You see the floating bridge is compartmentalized, so that it is impossible for an entire segment to flood at once and sink.

They are virtually unsinkable!

Seattle: hold my beer.

> Plans for a bridge had existed since the 1960s, and after the decision to construct the bridge was passed by the Parliament of Norway in 1989, construction started in 1991. The bridge opened on 22 September 1994

Pretty impressive timeline for an innovative idea.

And here in NY we've had 4 generations working on the 2nd ave subway line and only 3 of planned 15 stations have been opened so far.

"The last tolls were collected on 31 December 2005.”

Let that sink in. They paid for the project and then stopped taking everybody’s money.

That was the plan in Chicago, too...

> Tolls were reinstated on the bridge in 2019 to finance other road projects in the area

:-\

IMO it's probably a better idea to just keep on collecting them, and putting it away for the future. E.g. the I-5 bridge(s) across the Columbia River had tolls which stopped when Oregon & Washington bought the bridge, and now look where we are at. We have a 110 year old bridge needing replacement and no funds set aside for it. So what they will undoubtedly do is add tolls after spending a few billion to build a new bridge, and eventually it will get paid off. We could have been saving up for the cost and getting interest on it instead of the other way around. Even with a fairly modest toll, when you have a century to save.

This does require some legislative fortitude, however, to set aside the money for real and not just spend it on other things.

To me, the way they've done it seems correct. In your mind, where does the interest come from on the money saved for the bridge? The government has to be collecting interest from somebody no?

Local governments use banks. What are you getting at?

I assume the state can buy shares in index funds like the rest of us.

So, lets expand that idea a bit. Why doesn't the government just buy a lot of shares in index funds, kind of like an endowment, and then never collect any taxes? I can see a world where a large portion of private enterprise is held/managed by the government, and the proceeds of that is used to fund public works.

Well, the first reason is that most state governments[0] do not have anywhere near enough money in their accounts to get a return that will replace the revenue stream they get today through taxes. You might be able to build it over time by reinvesting budget surpluses.

The other problem I foresee is that the market is fickle. The S&P 500 reached a level in the second half of 2000 that it would not see again for over 14 years[1]. Any investment that needs to generate consistent revenue isn't going to have nearly the growth rate over the long haul that an index fund would provide. That makes the initial investment requirement significantly larger.

But otherwise, I am okay with the government owning a lot of private enterprise via index funds, so long as it has exactly the same voting power that I have. Which is to say, none.

[0] In case anyone needed the clarification, this whole discussion is about state governments; it does not really apply to the federal government for obvious reasons.

[1] Adjusted for inflation. The index did recover to the same number in 2007 just before dropping 50% in 2008.

Well, for the sake of argument, let’s say the federal government prints all of the money the state would need for an endowment and hands it to the state. There is inflation, but the state never has to collect taxes again.

I think there are some small optimizations we can make to this flow though, like what if the feds who printed the money held the endowment for the state and just distributed payments, instead of letting each state manage its own endowment.

Pretty much the same? Inflation, but no taxes.

There is a principle of governance that "you are taxing too much" is soo easy to build a platform off of that you will be DOA in elections if you manage any government that tries to invest like you say.

Investment restricted to "government responsibility foresight" infrastructure gets enough flac already. It only takes one down turn for the "golden goose" of the investment to be spent on the buddies of who just got into office

You can't trust politicians to just save money for the future. It will be abused by both sides. One side will gain votes by diverting it to something that has nothing to do with transportation, the other side will gain votes by repealing it and taking credit for lowering taxes.

I agree, it would be tough to implement, and I do not have an easy answer. Maybe something like a deposit-only account with a binding agreement on the earliest date the money can be withdrawn. No, I do not know how such a thing could be created in practice :). If it were possible, that would address the first issue, but not the second.

On the other hand it is more fair for the people using the new bridge to pay for it than for previous generations to pay for the old bridge and then keep paying for the new one which they will never see.

I consider infrastructure like bridges to be fundamental constructs of society, something we should build to promote the general welfare. I willingly pay for infrastructure today that will get more use by my children than I will ever see, and I think that is fair. My parents contributed to much of the infrastructure I am using.

> willingly pay for infrastructure today that will get more use by my children than I will ever see

Well, not many voters think that way...

And that’s how we ended up with the selfish Boomer meme. Previous generations set them up for success. Rather than paying it forward, they kept those funds along with their own economic growth and left the rest of us in the lurch.

There is a huge difference in being taxed to create and maintain infrastructure that you and your children will both use, and being taxed to fund future hypothetical infrastructure that you will never use.

I don't like that we as a country put off the costs of maintenance as long as possible thus creating debt for future generations, but asking the current generation to create a savings fund to pay the future generation's (non emergency) expenses is a bridge too far.

"A society grows great when old men plant trees whose shade they know they shall never sit in"

Still is the plan, they just keep building and rebuilding the roads.

Man in Year 10: See, this is how you do it. No tolls.

Man in Year 50: We need funding for much needed maintenance that has been neglected through sheer incompetence

We did that here in Seattle, where we have the longest floating bridge in the world, SR 520 across Lake Washington: tolls stopped in 1979 after construction was paid off.

Alas, tolling resumed in 2011, to pay for the complete reconstruction of the bridge. This time we are probably stuck with it, since WSDOT has grown inordinately fond of tolling as a traffic-management tool.

How do they pay for maintenance?

I live in Seattle. _Our_ bridges sink!

See: https://www.youtube.com/watch?v=gm0YQ3vuyyY

Ken Burns' documentary [0] about the construction of the Brooklyn Bridge was really fascinating discussing the innovative (at the time, late 19th century) engineering methods and challenges. It's pretty short, only 1 hour. Highly recommended.

]0] https://kenburns.com/films/brooklyn-bridge/

Read this a few years back.. highly recommended

The Great Bridge: The Epic Story of the Building of the Brooklyn Bridge by David McCullough

https://www.amazon.com/Great-Bridge-Story-Building-Brooklyn/...

Oh yes McCullough narrated the Burns doc. Brilliant historian

The Japanese Pile driving company, Giken, has some great engineering videos of their silent pile driving technology and some novel use cases, such as stopping a lava flow, or making underground bike storage in cities: https://www.youtube.com/@GikenGroup/videos

TIL that "Piledriver" wasn't invented by WWE.

The WWE didn't invent it, they perfected it.

Some do sink, on purpose https://en.wikipedia.org/wiki/Submersible_bridge, and some are mis-designed https://en.wikipedia.org/wiki/Lacey_V._Murrow_Memorial_Bridg...

Author: Grady Hillhouse

Videos: https://www.youtube.com/channel/UCMOqf8ab-42UUQIdVoKwjlQ

Blog: https://practical.engineering/blog/

Feed: https://practical.engineering/blog?format=rss

I have a guess to why H Pile i stead of I Pile: pronunciation. The initial I seems almost silent when saying I Pile. Whereas with H there’s a more distinct sound.

Isn't that the same with I-Beam as well?

I-Beam rolls of the tongue H-Beam doesnt, I guess thats the reason for adopting I term over H term

The context was the silent-ness of I vs H, not what rolls off the tongue. I agree that I Pyle rolls better, though.

Too close to Apple's trademarked iPile

My pet peeve along this line is O-ring. Is there any other conceivable shape for a ring?

Doesn't that refer to the cross-section? https://www.allorings.com/x-ring-seals

good call, thank you

They're just different profiles. I beam flanges taper more and are less wide than h piles or wide flange beams.

The bridge in the thumbnail is the Coronado bridge, rumored to have a floating hollow-box mid-section so that in the event of a collapse Navy ships can easily push debris and clear a channel. I remember hearing this "fact" on a San Diego harbor cruise long ago. Alas Wikipedia says it's a myth...

Tangential but does someone know the animation software used to display extract of the FHWA report, starting from 1:45? It seems to be used a lot by journalists and looks fantastic.

this was such a great way to spend 17 minutes thank you for posting! I feel like I learned so much about foundations that I never would have otherwise on my own lol

This guy’s videos are consistently great, they get a lot more technical than most other edutainment without getting bogged down.

The design, materials, construction techniques of bridges are so fascinating for me

Shame that website does not have RSS

I think all of the blogposts are just transcripts of videos from his youtube channel. You could use the feed from youtube to tell when there's a new release.

Wait, I just checked the page source after writing the above to confirm and it looks like it does have one: https://practical.engineering/blog?format=rss

Thanks. Miniflux could not find it. I did not know that this can happen.

Even authors are probably not aware of this.

Reply on news.ycombinator.com