Hacker News new | ask | show | jobs
by class3shock 898 days ago
I like the ethos but to me it's pointed in entirely the wrong direction. Why not just make standard sections of 2x4 or 2x6 framing? Start standardizing window sealing details? Literally anything that will actually get some use? Instead we get opensource plywood boxes?

Using plywood for all of the structural components means if you get a leak, which all houses get, you are in danger of much more significant damage than lumber framing. And with this system you still need to do all the normal things to build a house (get a designer, route plumbing, electrical, get a civil engineer, install all that stuff, etc.) just instead of framing you have custom blocks you need made.

I just don't think these make sense.

10 comments

I think it's about this solution being easy to make with CNC routers. Traditional framing is not made for automation, so they went to precut sheets of plywood. It's similar to 3D printing: You could make a doorstop in 30 seconds from a block of wood and a saw, or you could print it in 1 hour from plastic. The second option is better if you have no saw and a printer.
It seems like it would be much, much slower to CNC all the plywood and insulation and assemble the boxes and then put the boxes together than it would be to just frame a wall, sheet it and put insulation batts in. Having access to and operating a CNC also seems harder than learning to use a skill saw and hammer a nail. It feels like they're trying to apply abstraction to construction where you just have one unit that can be adapted to work for everything but they don't realize that already exists and it's called dimensional lumber.
> It feels like they're trying to apply abstraction to construction [...]

As is tradition :)

We can fix anything with another layer of indirection!
It's funny how every country thinks their approach to building is good and everything else has huge problems. For example Germans tend to think wooden houses will not last.

The approach taken here is similar to what people are commonly doing in Scandinavia. Most generally almost all houses are pre fabricated in factories and assembled on site. The actual blocks vary quite a bit, mostly they are large wall sections, but there are approaches using blocks like this. There is also a manufacturer that makes all sections out of plywood. With all of the approaches I have never heard of issues with leakage. In fact they tend to be much more tight then houses build on site (often meeting passive house standards).

Structural Insulated Panels - SIPs (which I think is what you are talking about) have arguments in their favor as they actually benefit from many of the upsides of building things in factories vs on an individual/onsite basis. These do not seem like they'll realize any of those (since you are just prefabing a component comparable to a sheet or board as opposed to a whole wall).

SIPS absolutely do have issues with leakage, the same as any construction method. Joints, penetrations, windows, etc. can all be sources, same as in traditional framing, the difference is again that plywood / osb / glued together wood scraps are atrocious at handling moisture. I do think they are interesting though, more reading for the curious below.

https://buildingscience.com/documents/enclosures-that-work/h...

One of the benefits of SIPs for airtightness is that the layer that provides that is built in (or atleast can be) at the factory. This eliminates alot of the variables that hurt airtightness when you are building onsite (working in temps/humidity where seals don't get installed right, poor workmanship, etc.).

Marine plywood is a thing that obviates the problem somewhat. Admittedly though, a very very expensive thing.

When that's factored in, you may as well build with straw bales. (really, it is great, and much cheaper)

They are registered in England and Wales and they write about millimeters. I guess that they don't have 2x4 and 2x6 frames. Maybe houses there are built in a totally different way.
Houses in England are generally made from clay bricks, not wood
Yes, houses in England are not made from frames almost ever. They are made of solid brick. It was always so confusing to hear about Americans punching through walls. You can't do that here. There's no wood.
People talk about punching through /interior/ walls, which is typically drywall in the US.
Yeah even interior walls here are mainly brick. Dry walls only exist in conversions where walls have been added that did not exist in the original construction. Even then, making thin walls like that is highly frowned upon and seen as a bit of a bodge job because the standard for the sturdiness of walls is high here.
All good points, especially about the risk of leaks on the material. I've also heard anecdotally that framing is the cheapest part of the house.
It looks like framing usually takes two weeks to a month. Whole house takes 8 months. The lumber is a good chunk of money, but comparable to wallboard.
And what about recycled foamed PVC boards instead of plywood? They have decent mechanical properties (12-15MPa)
what about A frame housing? Is that easier? Don't know much about building house. Just heard it is an easy way to build house with friends and some help.
wait until you learn about 'engineered beams'

I shudder thinking about all the structures built with those things that're one undetected bathtub drain leak away from serious structural failure.

I'm confused. Do you have evidence that an engineered beam will fail faster due to water damage than a beam made from 2x4s?

Weatherproof LVLs are actually much more water resistant than a regular solid piece of pine. Even indoor engineered beams are quite water resistant.

The last time wikihouse showed up on HN (September '23) I did a deep dive on the process I still haven't come out of. 20 years ago as a grad student getting my masters in architecture, I definitely would have had similar comments to most of what's on this page. But I don't feel that way any more and I'm actually disappointed by the negativity toward this project – not because it's not warranted, but if this were a software project I feel like the criticism would be more constructive rather than so dismissive.

WH is not a perfect system, but the approach is commendable for its comprehensive take on building systems. And building systems are broken. For the past two years I've been trying to build my own house (using a contractor), and even as a trained architect the process is ridiculously opaque, costs are exorbitant, and quality control is difficult to manage.

I applaud the fact that WH has tried to tackle the entire structure (and for that matter the entire building process – their parent organization has additional projects [0] for design, local building codes, and innovative financing structures). Yes, in the US 2x framing is cheap and relatively inexpensive, but is still complex in its own ways. Floor systems are different from wall systems are different from roof systems. I like that WH has tried to make one unit type (plywood boxes) work for the different systems. Is it the most efficient way? Probably not, but it affords other opportunities.

From an ideological perspective I also like that they prioritize sustainability, low energy use, and accessibility of design information to laypeople. On top of that, they've organized their building information like a programmer – they use GitHub to track versions and even have the start of a components database. They are an API short of making this really something the HN community could easily play with.

That said they have a long way to go. Framing a house is the easy part. I think if they could standardize the interface for cladding and interior finishes they would be in a better position to disrupt the building industry, but for now contractors will still be a requirement – so there goes your budget and quality.

[0] https://www.opensystemslab.io/projects

I'm dismissive because I get tired of seeing shiny new projects that claim to solve issues when they don't seem to actually see what the problems are.

If the goal is to create opensource plans / design framework then doing that where the industry is (framing in the USA, maybe brick in the UK from other comments, etc.) is the way to benefit people, not trying to push them towards a unique "block" system that no builder has ever used.

If the goal is sustainability then a focus on locally sourced materials, energy efficiency, building longevity, or numerous other parameters would make sense. If I have a main criticism it's that their system seems to be based on the idea that using a block based design is a good starting point and not as a consequence of aiming to find a system that maximizes the aforementioned parameters.

Having worked in the industry I feel your pain and would agree that so much is broken. Here in the USA it is hard to find good contractors and harder to find ones familiar with building highly energy efficient buildings that normally require a very high level of attention to detail. Often the only projects I see that are highly energy efficient and/or green/sustainable are for very high end clients that are paying far more than a normal person could afford.

To offset that wall of negativity, if anyone wants to read about some (imo) good stuff check out the Pretty Good House (PGH) standard and Building Science Corporation (BSC).

That's really interesting about Github-based versioning. My company (bldrs.ai) makes a webapp that views IFC (soon STEP) and has github integration, so any *.ifc path on GH can be pasted into our search for viz.

I'm not seeing any IFC files on Wikihouse repos, but they look old (https://github.com/wikihouse). Do you know where they're hosting?

Oh wow, you just sent me down a really intriguing path. I was not familiar with IFC, but looking more into that now. No, it does not look like their repo [0] has IFC files, but I don't imagine it would be that difficult to export the additional needed file information.

[0] https://github.com/wikihouseproject/Skylark/tree/main/SKYLAR...

Their site does have IFC for that project tho and I've loaded it in Bldrs, so seems fine there.

Yeah, IFC is neat. It's actually a dialect of STEP, so able to carry the geometry and tons of BIM metadata. We've estimated the spec is ~7k pages printed. It's like an abstract toolkit for working with building codes

Happy to chat more too. We have our project Discord and DMs on Twitter

It's a flawed solution based on a misunderstanding of how houses are actually built/designed/engineered. Most of what you see being built today is dimensioned based on available timber. It's part of why all the houses look the same, because it's the most efficient way to build. Everything structural gets built with 2x4's or prefab trusses (which are also made of 2x4's), windows are all standard sizes, as are cabinets and flooring, and siding, and trim, and if you weren't terribly concerned with building homes that were interesting to look at, you could probably optimize all of this further on the design side of things and then prefab things like walls and door frames then lego them together and wind up with something much nicer and likely cheaper than what they're doing.

The premise isn't terribly flawed though, prefab everything into 4x8ft panels, standardize your roof trussing, standardize your layout, and you can probably dig up a little more margin for builders to put directly in their pockets, and have perfectly cookie cutter houses with no regard to the environment they're placed in. Not really a neighborhood I'd like to live in, but, also not too dissimilar from what most of america lives in, so, I suppose it's tolerable.

That's not how we build in my country though
Yeah most countries in the world dont have 2x4s
We do in some sense but we call it something else like 45x120 or similar. That's not 2x4" but nor is a 2x4 actually 2 inches by 4 inches generally.

But what we definitely don't do is stick build homes using 2x4s apart from interior walls. An exterior wall where I live would probably use 45x220mm studs and probably more wood as well for a total depth of 350+ mm.

We don't use that wood to build houses. The US is special in that respect.
Realize we have a very, very large country, and it's chock full of forest, so plenty of lumber resources to do such things. Other places use different materials/techniques/etc which are mostly based on what resources they have easy (read: cheap) access to.

Also not all of the US uses timber framing, for instance, in costal florida, it's not uncommon to have concrete walls that are framed with sheerwall inside, so they can survive a hurricane.

2x6 is becoming more common for framing walls. The extra depth allows for more insulation.
To be fair, we don't technically have 2x4's either, they're a fair bit smaller in reality...
> Everything structural gets built with 2x4's or prefab trusses (which are also made of 2x4's)

Neither of these statements are true. 2x6's are commonly used for a variety of reasons in vertical placement. 2x8/2x10 and even 2x12 are still common for floor joists where manufactured beams (e.g. I-beams) are not used. Trusses are not always built with 2x4s either.

I'd also mention (as a recently certified firefighter) that the use of trusses in modern construction causes real headaches for the fire service: a minor failure in one part of one truss can lead the entire roof surface to become a collapse hazard. They are efficient uses of lumber but have few other merits as a construction technique.

If one wants to go the "lego things together" route, ICF and AAC are also options that exist already.
AAC is great, but

- I don't think it's super common in the US - it's a high-tech product that quickly loses its legolike value if quality controls get sloppy (you have to spend hours sanding down each row or switch to mortar) - it's just walls, you still need a system for building foundations, floors, ceilings and roofs.

I think AAC is pretty rare where I am (Oregon) but apparently there's a company in Bend that specializes in it. They're a Hebel distributor.

https://northwestaac.com/

AAC floor and roof panels do exist.

My impression is that ICF is a lot more common in my area than AAC, though it's still not super common compared to, say, stick frame construction. I'd guess that's probably because with AAC you'd need to add additional insulation, whereas with ICF you wouldn't -- but maybe there are other reasons.

AAC doesn't always require additional insulation. The rule of thumb is that a wall X mm thick made of AAC with density X kg/m3 has a metric R-value > 3, which is sufficient for Central Russian climate (which I am told is similar to Vermont, colder than the PNW). I have no idea what kind of blocks are sold in the US and how to convert between metric and imperial R-values, but online calculators suggest that a 300mm thick D300 block should have an imperial R-value of about 18.

Northwest AAC claims even higher values: "[f]or example, the R-value of a typical <...> 12-inch wall is R-29", so either I'm converting the wrong values or the measurement standards are very different across the countries. Xella (Hebel) claims their blocks are warmer than standard blocks, but 50% more sounds like too much.

> based on a misunderstanding of how houses are actually built/designed/engineered. Most of what you see being built today is dimensioned based on available timber. It's part of why all the houses look the same

It's worth pointing out that this is a UK-based project (where typical houses are concrete construction). A lot of the comments dismissing it are suggesting it should conform to American standards on the assumptions that those standards are universal.

As an open-source initiative it should certainly have a broader perspective than being purely UK-centric, but replacing that with a US-centric one isn't the solution.