That commenter is doing a lot of gymnastics to avoid talking about reinforced concrete's hidden failures. It's difficult to determine reinforced concrete failures before it gets bad, and repairing it is very expensive. Concrete can get really complex depending on how you make and apply it, and what it's exposed to.
I agree with the author that water is now a bigger deal than fire for wood construction, and we need way more testing for taller construction, perhaps even designs tailored to the material.
"The issue with CLT as a building system is exactly that -- its behavior as a whole when subjected to differing forces as statistically magnified by the great number of joints. There are a ton of connections in a CLT building. Thousands of them. One connection failure can create a cascade of failure that may destroy the building. Each connection has the fingerprints of a single worker. One bonehead can wreck a connection."
That's a non-negligible practical problem and a perfectly valid point.
And it's difficult to take the residential approach of "just over-spec" when you're building mid-rise+.
> That commenter is doing a lot of gymnastics to avoid talking about reinforced concrete's hidden failures. It's difficult to determine reinforced concrete failures before it gets bad, and repairing it is very expensive.
Care to elaborate? Concrete has already been extensively studied and has enjoyed widespread use, including under seismic loads. I don't understand how in this day and age where rc structures have been used for over a century there is any mention of "hidden failures".
Concrete in general can fail from simply being mixed poorly, which the slump and strength tests used on each batch are meant to address. But also additives can sometimes be added to concrete, which in some cases react with surfaces/chemicals over time and cause all sorts of problems.
For reinforced concrete failures, see here: https://en.wikipedia.org/wiki/Reinforced_concrete#Common_fai...https://www.usbr.gov/ssle/damsafety/risk/BestPractices/Chapt... By the time there's spalling the reinforcement is already rusted and breaking up the concrete. Of course you can say "Oh but we know all that now, so just design it right and you're good!" Sure. Unless you make a mistake, in which case you won't know it until the structure is crumbling. And remember, "it's a monolith", so good luck with those repairs.
I'm not saying CLT is better - but let's not kid ourselves, reinforced concrete isn't perfect. It's just cheap, strong, and easy to install.
I felt the same way. They said a lot of plausible things, but their motivation seemed more to be “don’t disrupt my life by pursuing alternatives to concrete” than “here is a reasoned and well-structured argument against CLT”. It read like a series of talking points. Glad it’s not just me.
Now I want to know how it will recover after fires. I read it will burn/collapse well after occupants have escaped. But will a building with some burnt CLT members be a total loss?
JLM's comments some days are better than Fred's. The guy has been a high ranking army officer, built a lot of the newer skyscrapers in Austin and now is retired from CEO of a Fortune 500 company.
Really? It seems like a lot of words to say "unfamiliar material bad." Surely the people advocating CLT have thought of some of these obvious downsides, and have ideas for how they're addressing them. If those are inadequate, then talk about that. Instead we get "In my professional judgment it is driven by a certain "gee whiz, isn't this cool" environmental state of mind." Excuse me if I want facts and not someone's un-sourced, un-backed-up "professional judgment".
The comment repeatedly states that CLT is neither new nor unfamiliar, one main issue is that CLT requires many joints with individual properties (many points of failure with manufacturing errors specific to the people that worked there) whilst concrete behaves / accepts load as a singular structure.
Admittedly, this isn't an argument against CLT but instead against anything that doesn't behave as a single structure under load - but it does point out that CLT isn't a suitable replacement for all building materials everywhere. High rise buildings need the safety of reenforced concrete, with it's load handling and simplified calculations (due to lack of joints), and until something can perform in a similar manner it's not likely to be replaced.
Here in San Francisco, CLT won't be legal for buildings over six floors until about 2023 because of dumb policies. California uses the International Building Code as its base building code and then modifies it based on particular laws we have here. This modification & ratification process takes about two years. Then San Francisco does the same thing, adding another two to three years (because surely _our_ engineers are better than a consortium of the world's best structural engineers...)
The 2018 International Building Code provides codes for tall CLT structures, which won't get adopted in the CA code until late 2020, which won't get adopted in the SF code until late 2022 or 2023. At that point, developers can start applying for permits for CLT buildings, which takes about four to five years. Then add two years of construction.
So SF won't see CLT structures over six floors until about 2030. Makes you realize how unseriously we're taking climate change when the only carbon-negative structural building material in existence won't exist here for another decade.
The process you describes is kinda a safety net. Buildings aren’t like software where you can continuously design and push updates. You get one shot and mistakes are very expensive.
If you want to change the environment right now, build buildings that let people live closer to work and commute less.
I mean, someone signed off on the footings of the millennium tower, too. SF is not taking a cautious, informed approach, they’re just burying their heads in the sand.
Friction piles are used around the world, including in far more seismically active areas like Japan. In fact, in many ways they can be advantageous. I don't think I've read a single article on the Millennium Tower where a professional claimed that the tower engineers did anything but apply proper, modern structural engineering. At best the few criticisms are more like, "but I wouldn't have done it."
When unexpected things happen it doesn't always mean anything was done wrong, or even that anything should be done differently in the future.
Especially it seems SF is an area where it would seem wise to be a bit extra cautious...since you're both seismically active and in a marine environment.
You are more cautious than the SF city government - basically all of the recent (and future) construction is in the most seismically unsound parts of the city!
> At that point, developers can start applying for permits for CLT buildings, which takes about four to five years. Then add two years of construction.
I subscribe to HN for the insights into technical subjects I might not otherwise have exposure to. My primary takeaway has been "this is why we can't have nice things", whether it's safe air travel, affordable housing, or affordable healthcare.
Demolition on the Waldorf-Astoria began on October 1st 1929. Unforeseen financial difficulties slowed progress. Construction began January 22nd 1930. The steel structure of the Empire State Building was completed on September 19th 1930. Construction was completed April 11th 1931.
These days it's more difficult- impossible- to get permitting for low rise medium density housing buildings in a city with significantly more financial resources and significantly more demand for floorspace.
This is one of the reasons why housing prices in the Bay Area are always increasing. They literally can't build enough, and so what they do build is going to be expensive, because that's what's profitable.
Yeah every time there is an earthquake they learn how new modern materials and construction techniques fail and perform worse than the previous ones.
Kinda of like how joints in steel frame buildings held together with stupid janky rivets would shift and creek during earthquakes while strong modern welded ones wouldn't . Progress! No wait that last bit is bad. The steel at the welded joint buckles unless carefully designed. Which they weren't for the first 30-40 years.
My house has some LSL (Laminated Strand Lumber) beams which, like CLT, are engineered, but are built up from wood strands of and a lot of glue instead of from boards like CLT. As a result, they can be made from the remainder products of milling lumber.
They can even be made from rapidly harvestable plans like bamboo:
There are many forms of fungi (mold, various rot) that can feed on wet wood. There are not many forms of fungi that can feed on cured adhesive, which also functions to prevent the wood that is there from getting wet. Thus, I'd expect engineered/laminated timber to be better at repelling rot than standard untreated timber. Done well, it could last longer than steel.
It depends on how the material is designed. Some materials like commonly available particle board, can't get wet at all. They swell up and crumble. On the other hand there are grades of plywood that are rated for wet applications.
No, it's much closer to Glue Laminated Timber whereas plywood tends to be made of wide sheets of thin timber of varying quality (depending on the quality of the wood and number of sheets glued together). CLT and GLT/glulam aren't made in sheets like plywood, they're usually constructed for a specific purpose (beams, etc.)
I use this stuff when I design and build LEED-Gold certified hydroponic food production buildings. CLT is made out to panels and then cut to customer specifications. It has the same construction (including 90-degree off-orientation of each layer) and pretty much the exact same manufacturing process.
Glue Laminated Timber has the same orientation all the way through. It is meant specifically for one-direction loading (primarily structural support beam but can be done for joisting as well) whereas plywood and MDF and CLT will handle multi-directional loads (and is why it gets used for the outside skin of buildings framed with wood.)
But plywood is made out of veneer sheets that have been turned on a giant lathe. There is only a very thin layer of fibers running one way, you can't really see growth rings in a layer of ply. CLT layers can be at 2" thick or more, with a very definite grain structure and growth rings.
Which makes me wonder how CLT is supposed to behave with wood movement when atmospheric moisture changes. Is it made from a species of softwood with very little movement? Or is the adhesive supposed to be strong enough to withstand the stress under movement? Or has the wood been treated with some industrial process to reduce the movement? All of the above?
Cross laminated timber goes against everything I know about wood movement as a woodworker. But the only exposure I've had to this stuff was beating 6 inch thick (3 layers) CLT pieces with an axe for firewood. I'm really curious how this stuff can work.
"But plywood is made out of veneer sheets that have been turned on a giant lathe."
Ditto CLT. My class was making this almost 25 years ago back in high school wood shop, we just called it "Thick ply" and we primarily used it back then for constructing ultra-solid subwoofer boxes for vehicles. The lamellae were cut to 1/4" thick sheets on a lathe (they do make saw-based lathes, we had two in wood shop, one industrial one, one made out of a table saw with jigs for making dowels and such) planed, cut, oriented/glued/layered, veneer layer applied, sanded, then a final heat treating process, and done.
Your post shouldn't have been killed. Your comment is the same as the last sentence in the summary on wikipedia: "It is similar to plywood but with distinctively thicker laminations (or lamellae)."
Yes. Tricky thing is plywood uses veneers (very thin pieces) because wood expands at about a 10x different rate in width versus length, so cross-laminated joints can be subject to an incredible amount of stress. The thin veneers cannot develop enough stress to break the glue joint.
> so cross-laminated joints can be subject to an incredible amount of stress
OK, plywood overcomes the stress by having very thin laminates. So how does CLT overcome the stress due to different width vs length expansion rates? Shouldn't the problem be much worse with thick laminates?
We looked into this with a completely open toolchain [0]. You could email either of the two first authors to get the code (not sure why it wasn't in the paper supporting information).
Building open carbon models isn't difficult, it's the input data that normally require licenses. I have built an open source life cycle assessment (LCA) software which has some traction [1], and there are alternatives for LCA [2] and integrated assessment models [3, 4]. However, data availability, especially on the level of completeness and detail you need to answer a specific question like carbon performance of a structure over a given period of time is a challenge. We are working on building a large open database to answer these kinds of questions[5], and Hacker News readers are welcome. Happy to chat via email if you want more info!
What I'm trying to do is, I believe, complementary to these LCA tools. I'm building a website https://futur.eco that bridges carbon models with our every day life as citizens.
Sorry, it's in french for now, but I have some hopes that you read french :-)
The process of making clinker uses heat to separate CaCO3 into CaO and CO2 [0]. This release of CO2 is normally about half the CO2 footprint of making concrete (the rest is mostly combustion of coal to generate heat).
CO2 can penetrate into the concrete and bond with CaO, over time reabsorbing all the CO2 released during the calcination process. In practice, CO2 doesn't penetrate deeply into concrete, so depending on the concrete type and environment something like 25% of the potential absorption of concrete is realized (this is based on a conversation I had with a colleague at work, the number is not exact).
Since this is basically plywood++, does it have similar failure modes? Like what happens when there's a leaky tub frequently wetting a structural piece of CLT?
That's just like plywood. I have to imagine other similar problems affect CLT. Your example of moisture damage essentially has to affect wood, though there are ways to engineer the CLT that will mitigate moisture damage.
Mild steel becomes ductile at relatively low temperatures, way before it actually melts, hence why you sometimes see steel beams deform and drape over timber beams in a fire. The ductile transformation happens relatively quickly once you get to the critical temperature on the iron carbon phase diagram for typical construction grade steels. Hence why Steel is usually protected with fire rated drywall like boards, intumescent paint or weird spray on insulation that looks like porridge.
Whereas with timber the rate at which it will burn is reliably predicted, so exposed timber structures are designed to be oversized so that the reduced beam sizes will still be effective. This is done using a charring calculation to give the required amount of structural integrity duration which varies depending on how tall the building is and what it is used for.
All types of structure will fail eventually in a fire, the assumption is that the fire brigade will tackle it before it goes too far.
In the UK, the main fire related problem with CLT is that the exposed surface doesn't meet the requirements for the speed of surface spread of flames so it has to be treated with a fire retardant or you need sprinklers if you want to expose the material. Both of which are expensive.
Today I learned why the bottoms of the staircases in the building where I work look like they've been sprayed with porridge. Thank you!!! (I had assumed it was insulation of some kind but could not figure out why they'd put it there)
I actually wonder if CLT/wood would perform better than steel during an all-out fire in terms of the structural integrity, at least for an initial period of time.
The steel structure would more likely survive in-tact, but the wood structure could guarantee a certain minimum bound in terms of evacuation time available to occupants (which might be superior to steel).
From my very novice perspective, I feel like wood does not attenuate in structural integrity as its temperature rises (at least not in the same way as steel), and it seems it would also absorb and conduct heat much more slowly than steel.
On the other hand, I feel that once a wood-based structure gets hit with a big enough fire, the entire building is guaranteed to be lost due to how hard it could be to fully extinguish the fire (e.g. if the steel in a skyscraper could burn openly, what would that be like?).
It burns slowly enough that it doesn't conflagrate like unprotected hardwood or plywood might, so there's much more time to evacuate. If the fire isn't put out quick enough, however, the building may lose its structural integrity
I would expect that the vast majority of concrete used in construction isn't poured into "arbitrary shapes," but flat walls. Even if only 50% of concrete is pored as flat walls, that would still be a huge market for this.
Re: acoustic isolation - having recently dealt w acoustics/concrete (laying a floor in a concrete apartment), I found out there’s not necessarily “concrete in general” so acoustic (and other) properties can vary from concrete to concrete.
The answer is basically that they don't know at this point. (Reference: I did extensive research on CLT as part of a project in LA a few years ago and talked to CLT experts from the Wood Products Council)
CLT is stiffer than plywood, and the construction techniques are closer to steel/concrete, so it'll likely need reinforcement or dampering. Traditional wood buildings these days are designed to flex in earthquakes.
But it hasn't been extensively tested for seismic properties yet
True, but worth noting plywood is spiral shaved veneers which are unrolled glued with the curves counteracting This is quite different from sawn dimensional lumber. Plywood is entirely cut tangent to the grain / tree rings.
Plywood is much weaker essentially because trees evolved to withstand wind and the veneers are cut perpendicular to wind loads.
As I understand it, the modern trend for buildings that are concrete on the first floor and wood frame for 2-5 floors above that started with earthquake-resistant designs. So they can probably play on that model.
