The ancient Roman secret to concrete resilience in seawater (2017) | Hacker News Reader

The ancient Roman secret to concrete resilience in seawater (2017)

als.lbl.gov

142 pointskvee76 comments

There actually IS a contemporaneous source on how to do it, without exact proportions, unfortunately. They didn't know why it worked, of course.

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

https://news.mit.edu/2023/roman-concrete-durability-lime-cas...

"pozzolanic material such as volcanic ash from the area of Pozzuoli, on the Bay of Naples." They shipped this all over the Empire.

Now we have a better idea why it got better in seawater, instead of deteriorating.

It woyld be funny to me if it turned out the secret was urine instead of seawater. One of the bits of received wisdom i got from my grandfather was peeing on your hands toughened them up. He claimed it as a bit of advice he was given on his first day chopping hardwood for a steam engine at a tin mine. I figured it as a first day prank like being sent to find headlight fluid in modern times but he said it was true.

Before major industrialisation, people used what was at hand. I think the peeing on your hands thing has been long debunked but in roman concrete it may have worked.

Urea is very effective at moisturizing and is an ingredient in many heavy-duty lotions, so peeing is probably overkill but not without benefit.

Practical questions abound. Do you wash your hands after peeing on them? Do you pat dry? Towel dry? How do you close up your pants without spreading a non-negligible amount of urine onto your pants? Would you want to pee on your hands every trip to the bathroom or more a once-per-day thing?

Water jackets for machine guns would often be filled with urine by soldiers if water wasn't available.

Thanks for making my day https://www.youtube.com/watch?v=s_0THARUP9A

Lots of baseball players used to do this.

You're in for a big surprise:

https://www.smithsonianmag.com/science-nature/from-gunpowder...

It's a bit off-topic, but the role of urine in teeth cleaning reminds me one of Catullus' poems, poking fun at someone who was a bit too eager to laugh/smile as trying to show off how much urine he drunk.

For context: Celtiberians used urine to brush their teeth. The Romans were aware of that, but they thought that it was disgusting.

(The poem is Catullus 39, if anyone is interested. Link: https://en.wikisource.org/wiki/Translation:Catullus_39 )

For most of human history pee urine was left to get stale, turning it into ammonia, and used for doing laundry.

Washing clothes in ammonia is really great at getting out oil and grease stains. Though now you just buy ammonia directly from a good hardware store.

Hi, nerdy hacker here, could use further advice:

Can I screw the ammonia up, like how bleach ruins clothes? Any recommendation on amount? Do I add it to a wash cycle?

I have a lot of grease-stained shirts that are some of my favorites.

Mixing ammonia and bleach gets you basically mustard gas, so be cautious about what is in any other detergent you use at the same time...

Specifically it makes chloramine gas, the same stuff that (in fairly low concentrations) makes swimming pools stink. It can be lethal so definitely take it seriously, although it's not nearly so gnarly as real mustard gas.

Here's an incident where a woman inadvertently made chloramine gas in an enclosed space while cleaning and was breathing it for 30 minutes. It nearly killed her and she was hospitalized for a week: https://www.nejm.org/doi/full/10.1056/NEJM199909093411115

Here's one where a restaurant manager died after someone mixed cleaners right on the floor.

https://www.cnn.com/2019/11/08/us/massachusetts-buffalo-wild...

It's not mustard gas, but it is really nasty. It produces a mixture of chloramines and hydrazine.

Thanks for the clarification.

Use in the laundry for regular stains: Soak stubborn stains on cotton, polyester, or nylon fabrics with a solution of 2/3 cup clear ammonia, 2/3 cup dish soap, 6 tablespoons of baking soda, and 2 cups warm water. Let it soak for about 30 minutes, then launder as usual. Never use ammonia on wool or silk.

Here's a step-by-step guide on how to wash clothes with ammonia:

1. Read the garment labels: Check the care instructions on your clothing items to ensure they can be safely washed with ammonia. Some delicate fabrics or certain colors may not be suitable for this method, so it's important to follow the manufacturer's recommendations.

2. Prepare the washing machine: Start by setting up your washing machine as you would for a regular load of laundry. Sort your clothes by color, fabric type, and level of dirtiness.

3. Measure the ammonia: For a typical load of laundry, use about 1/2 cup to 1 cup of ammonia. Adjust the amount based on the size of your load and the level of dirt or stains on the clothes.

4. Add the ammonia to the washing machine: Pour the measured amount of ammonia directly into the detergent dispenser or the main wash compartment of your washing machine. Be careful not to spill any ammonia on your skin or clothing.

5. Add detergent: Add your regular laundry detergent to the same compartment as the ammonia. The detergent will work in conjunction with the ammonia to clean your clothes effectively.

6. Start the wash cycle: Close the washing machine lid or door and start the wash cycle using the appropriate settings for your clothing items. Follow the machine's instructions for temperature, water level, and cycle duration.

7. Complete the wash cycle: Allow the washing machine to complete the cycle as usual. Once finished, remove the clothes promptly to prevent wrinkling or odor development.

8. Rinse the clothes: If you prefer, you can run an additional rinse cycle to ensure all traces of ammonia and detergent are thoroughly rinsed out. This step is particularly important if you have sensitive skin or are concerned about potential residue.

9. Dry the clothes: Depending on the fabric type and garment care instructions, dry the clothes by air drying, using a clothesline, or by using a dryer.

10. Clean the washing machine: After completing the load with ammonia, consider running an empty cycle with hot water and a cup of white vinegar to help remove any residual ammonia smell or build-up in the washing machine.

Remember to handle ammonia with care, follow safety guidelines, and avoid mixing it with other cleaning agents, such as bleach, as it can produce hazardous fumes. Always keep ammonia out of reach of children and pets.

Additionally, if you have specific stains or heavily soiled items, it's advisable to spot treat them before washing with ammonia.

Don't mix it with bleach, otherwise you're o.k.

Can't really go wrong. After I do some bicycle repair, I toss my greasy clothes into the washer, and add a cup or two of ammonia during the wash cycle. I don't even add detergent, as the grease comes out without soap.

Found more professional instructions here, https://www.hunker.com/13422713/how-to-use-ammonia-in-the-la....

I only ever used it to clean floors, but word of warning: even a 3% solution is already noxious.

A whiff of Uranus so to speak.

Urine-us

Baseball player Moises Alou famously claimed he peed on his hands to ward off calluses. He's one of the few modern players not to wear batting gloves.

https://slate.com/news-and-politics/2004/05/can-peeing-on-yo...

Jorge Posada and Vlad Guerrero too. All from the Caribbean.

I would expect the opposite - urea is a keratolytic, so it should break down thickened skin, and make the hands less tough. That said, urine is 1-2% urea, and typical medical urea solutions are 5% or more, so i wouldn't expect much of an effect.

Lots of baseball players used to do this. Particularly Dominicans. It mostly stopped in favor of batting gloves but some players continued doing it such as Moises Alou.

We have a tendency to promote premature failure of seawater-exposed concrete structures by reinforcing them with rust-prone materials.

I don't think any amount of secret Roman concrete recipes is going to prevent crumbling apart due to the expanding forces of rotting rebar.

Old article from 2017. More advances were published earlier this year:

https://news.mit.edu/2023/roman-concrete-durability-lime-cas...

That's really cool research. Using quicklime heats up the concrete a bunch more than some other methods, and the result is concrete that fills its own cracks.

People who think A.I. will kill us should be worried that concrete will kill us.

There are many materials that could be superior in quality and environmentally benign than Portland cement-based concrete not least the Sulfur concrete which is commonly used in China today

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7660051/

and particularly Geopolymer concrete

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

but all of these face resistance from the Portland cement industry which has already standardized products for regulators.

Having seen the state of concrete buildings in China, I'd prefer the alternative.

It's not a conspiracy. Fly ash contains heavy metals that would leach out of the cement. There are more challenges remaining to be solved than the Wikipedia article suggests.

The problem with modern concrete falling apart prematurely is usually the use of rebar as a reinforcement.

And we use rebar because we like long, flat bridging spans that can hold hundreds of tons at a time — something Roman concrete has no hope of ever doing, unreinforced.

Not necessarily countering your point, but perhaps giving some perspective.

The City I grew in (Cordoba) has a roman bridge[1], originally built 1st century BC and still standing. It supported a 2-way street until 2004, when it was pedestrianised.

Caveats:

* It has gone over a great deal of renovations/partial rebuilds on its history, so its durability can not really be attributed to roman concrete alone. Only 2 of the original 16 arcades remain. One could argue that "it's not the same bridge" any more (Theseus' Bridge).

* I ignore wether they used concrete to build the original one. I'm not an architect. For all I know, the initial construction could have been mortarless stone, and cement (as well as reinforcement) could have been added in some of the later renovations.

That said, all bridges require periodical servicing/renovations anyway, in order to remain functional. And this one has stand for 2000 years.

[1] https://en.wikipedia.org/wiki/Roman_bridge_of_C%C3%B3rdoba

This doesn’t really give perspective, it’s basically what they’re saying. Roman bridges are gravity structures, the entire thing is in compression and heavy, the arches have really limited spans.

Compare to the Millau Viaduct for instance.

Well in order to compare things properly we would have to check if the Viaduct is still there 2000 years from now.

That's a completely different axis is the point, you can not solve the issues the viaduc solved, let alone build anything even remotely similar, using roman materials (to say nothing of techniques).

does fiberglass rebar work? I remember reading that it works well for horizontal layouts but was not sure about use in pillars.

Yes, alternate forms of rebar work. You can also coat rebar with epoxy, or use stainless steel.

The two core issues are cost, and limited understanding of the long-term properties of the material (as well as experience e.g. epoxy-coated rebar needs to be handled carefully lest you scratch off the epoxy).

FWIW the Practical Engineering channel has an entire series on concrete which covers (the basics of) this and more: https://www.youtube.com/playlist?list=PLTZM4MrZKfW90PdaBFt70...

>... e.g. epoxy-coated rebar needs to be handled carefully lest you scratch off the epoxy).

That sounds impractical at scale. Whenever I pass by construction I see the materials haphazardly strewn about. How could you possibly ensure that all of the rebar used is never scratched when it has to pass through dozens/hundreds of hands before it is finally set?

I've always noticed it carefully packed and stacked on site. The epoxy coated rebar is usually formed and connected into structures that are placed/suspended in the forms or roadway. The epoxy is also very durable.

Isn't basalt the best non-steel option all around?

I don't think anyone's ever claimed that it was better for all modern applications.

Rebar reinforced concrete fails, I assume, because the rebar rusts? If so, isn’t the issue greatly reduced (potentially) if we were to use the quicklime process of the Romans? If the self-healing prevents water penetration, does that stop rebar from rusting?

Steel is really strong vs tension, not so much vs water / rust / chemical reactions.

It's also relatively cheap. Maybe carbon nanotube ropes could replace that after we have cheap mass production?

The point, is that the steel needs water to rust. The cracks allow the water, but supposedly the Roman concrete self heals and water does not continue to penetrate. One assumes, therefore, limited rust (perhaps). In practice this may not be the case, but surely someone has tested? Steel has the right properties + price; if you can keep the water out.

Concrete isn't waterproof, it might resist the flow of water, but as even most rocks are, there is at least a slow migration of water through the material.

There's already other ways to reinforce concrete, one I know of is embedding kind of fiber-glass mesh in it instead of rebar. I quick google suggests that this might be called Fiber Reinforced Polymer (FRP) reinforcing, or possibly that's a different thing I'm not really an expert.

I assume you could use a whole bunch of different materials really, including other metals that don't rust in the same way as steel does.

Remember that roman concrete often used iron for reinforcement. The Colosseum is a famous example of this. The only reason it's not still there today is because metal, particularly iron, was extremely valuable so after the fall of the Roman empire, people tore whatever iron they could out of the structure (akin to people today stealing wiring and copper piping from the walls of houses).

It is fascinating how some ancient techniques had remarkable properties even though they didn’t have the tools or science to explain it back then. I recently learnt that ancient concrete has self healing properties because of the reaction of air/water with lime and volcanic ash.

Makes you wonder how far we can get with trial and error.

> Makes you wonder how far we can get with trial and error.

And millennia long test periods.

Yes, the survivorship bias is strong in this case

Got us from cingle cell organisms. Just trial error and time.

It’s been amazing learning some knots these past few months, realizing what you can do with them once you have a good repertoire to draw from, and then realizing that many of them have been used since antiquity.

Stories about Roman concrete appear to flood the Internet about once per year. No idea why...

See also: seaweed reduces cow farts.

Maybe because 2000 years later, some of it is still standing?

And in 2,000 years there will still be some structures made of Portland cement still standing... although not any with rebar inside. The Hoover Dam could stand for 10,000 years (it's not even done curing yet, 100 years later). Don't overlook survivorship bias; the vast majority of structures made of Roman concrete are not still standing.

I think you are ignoring all the civil engineers who ARE impressed with it.

Getting better in salt water is a seriously impressive feat. Hoover Dam is fresh water.

See also "Why was Roman concrete so durable?" from January 2023:

* https://news.ycombinator.com/item?id=34280239

sadly no concrete evidence on the recipe - not surprised. Romans were excellent engineers but as so often, we're left with little documentation in this area (sciences). The humanities (law, politics, literature etc) has fared better, with way more surviving records.

I wonder how much documentation we have for the practicalities of modern construction methods, and how much of it is passed down orally by the construction workers who are actually building stuff, or locked within internal company manuals. I mean sure I can look up formulae for concrete, rebar etc. on wikipedia, but for actual step-by-step instructions with specific machines and materials? I bet that information is a bit more discombobulated.

Max dad worked for McDonald Douglas, later Boeing. Tons of little tricks to build things that weren’t documented. Made union strong. Replacement workers couldn’t reproduce parts to spec when following instructions. Things like adding strings to foam molds in just the way, etc.

> McDonald Douglas, later Boeing.

Later Burger King :)

Today's tech content would fare worse than you might imagine. So much is tied into byzantine ERP/PDM systems . . or contains specialized processing instructions for a particular black box editor/processor . .OR juggles actual techdata with the entity peculiarities of a CCS (component content system) that itself can result from any number of nebulous engineering fads.

I guess, short version of what I'm trying to say, is that today's tech data doesn't have narrative. It's a bunch of little pieces floating around and needs a living org to put it together. If you froze it, Pompeii-style, future generations would find themselves facing dead ends all over the place.

with software, it's particularly problematic - the dependency hell: libs, api, versions, platforms, players, os. your software needs documenting, and so do the deps. and if you want to run it/replay it, good luck getting it operational 20years from now let alone in year 3000.

when a book, newspaper, magazine is published, a copy is archived with the National Library - narrative etc preserved.

Websites get some archiving, but pay-walling is threatening this. Other types of software is very tricky.

It is always interesting to me in science fiction how humans are able to repair/reverse-engineer ancient technology from aliens or pre-downfall humanity. If you stumbled across a non-encrypted WinXP hard drive, how would you possibly read the data? Figuring out the right power/voltage, creation of a SATA cable, constructing the SATA communication protocol, figuring out how to decode a NTFS file system, and then how to read bespoke binary format X.

Modern humans with a very good idea of how NTFS work struggled to reverse engineer it for years.

Our best hope would be exploit devs, cryptographers and infosec professionals with reverse engineering skills. It's incredible what some people can do with a black box system and no tools aside from a voltmeter and logic analyzer. First you probe the chip with the voltimeter to find where data is being interchanged, then you hit it with a logic analyzer to look for patterns, then you start breaking it down with cryptanalysis, which is really about finding patterns in noise. I think you'd be surprised what the best minds would be able to decipher, given the right hardware and unencrypted data. If the data's encrypted, it gets a lot harder, especially if the encryption is advanced enough that it basically looks like noise. But even then, there has to be some bootstrapping process, so we'd be able to reverse engineer that part of it, and then would need to hope to find an exploit in the encryption.

Regarding cables and such, you'd have to expect that you wouldn't just find a single isolated hard drive. Even if you didn't have a full system, surely there would be cables and auxiliary components strewn about. And if there weren't, their absence would probably be significant, since you'd know they used some material that didn't survive as long as the material used to make the hard drive, and could infer its properties from that. So you'd probably be able to piece together which cables plugged into which components and make educated guesses about voltages and the like.

I agree. just think home much effort went into braking the Enigma Code in WW2, a tech developed by a contemporary power, with similar scientific know-how etc. If you've got hints/docs from similar formats, AI could help develop a pattern.

Nominally patents are supposed to describe these technical details, but unfortunately the law ("enablement") has no teeth in the US. When I was a patent examiner, I wanted to do enablement rejections, but couldn't. The law is supposed to incentivize properly documenting the invention, but attorneys have gamed the system, disclosing little in exchange for a monopoly.

The story mentions volcanic ash. If the Roman recipe became popular again for marine use, I wonder how much ash is currently available without terrible mining operations.

Fly ash (i.e. ash from industrial chimneys) is being used for this same purpose, as pozzolanic (volcanic ash) deposits are not that big or common around the world. The key here is, by the opinion of a cement researcher close to me, the availability of non-cristalline (i.e. an amorphous or vitreous form) silicon dioxide.

I've been slightly following the papers on roman concrete for a decade or so, and as far as I remember, they've been discovering different cristalline aluminosilicates that would explain the resilience of these concretes, so maybe there still is some need for pozzolanic, or for adding aluminium to fly ash.

Relevant tidbit: hot mixing with quicklime was key for this self-healing property[1], as it allowed the creation of small lime clasts across the material.

[1] https://arstechnica.com/science/2023/01/ancient-roman-concre...

I'm completely new to ancient Roman civil engineering. So I ask completely innocently,could it be survivor bias?

That is a popular counter-argument against many things, but if we have visible (dare I say concrete) evidence that Roman concrete structures in seawater are still here, while now there's like a whole field of infrastructure/construction people who know that typical current concrete can't do it, then it has to be something there, right?

I hope they can perfect this and use it in places like coastal New Jersey.

We can build concrete that can withstand maritime conditions, we just choose not to due to cost. We also really don't know how much Roman concrete only lasted 100 years.

Obligatory mention of leading concrete-engineering YouTuber https://youtube.com/@TylerLey . He’s discussed Roman concrete in eg. https://youtu.be/U86tlUiFM1s .

Reply on news.ycombinator.com