Markets in everything the culture that is Japan

In theory the mechanism is really quite simple:

1. A sensor detects the rumblings of an earthquake.

2. Within .5 to 1 second an air tank pushes air in-between an artificial foundation and the actual structure of the home, lifting it as high as 3cm off the ground.

3. While the earth below violently shakes, the levitating home quietly and patiently waits, returning back to the ground once the tectonic plates have settled.

There are pictures and videos at the link, and the company claims it is implementing the idea at 88 sites.  The final video is especially fine.


Service Temporarily Unavailable!

Post-K some avant-garde architects tried to build houses in the Ninth Ward that had, essentially, floaties built into the foundation. The theory was when the floodwaters came, the house would float up (there was a tether so it didn't float away), and when the waters receded it would settle back down.

The mechanism, clever as it may be, betrays a startling lack of understanding of how most homes are actually damaged in real hurricanes.

If this damages the house less than an earthquake, why not keep compressed air around the house all the time? Google informs me that a typical house weighs less than 100 lbs per square foot; you could levitate a house with a pretty low air pressure or a higher pressure applied to a small fraction of the home's area (the fraction of the foundation you'd levitate would be lower since it would only have to serve the structural purpose and something else would deal with protecting from the environment). A passive system w/ separated air reservoirs at different pressures would be sufficient to level the house. Obviously you'd have to have a few architectural accommodations (space between the sides of the house and the hole it's in, entryways, garages, etc), but this seems superior in a couple ways to building the mechanism to levitate a house and only using it every once in a while: you wouldn't need such an awesome air compressor, no need to detect earthquakes and you'd know if there was a leak or a problem with the compressor if your house started to sink instead of finding out during an earthquake.

Note that this is a very poor idea in hurricane alley.

err Tornado Alley which you think I wouldn't screw up since I've lived there.

Maybe an oil-filled isolator would work. Or water filled? Might be cheaper and less leak problems. What are the necessary / optimum properties of an isolation fluid? i.e. why compressed air?

Because some leaks are inevitable and you'd probably want an open system to adjust overall pressure as the weight in the house changes, atmospheric pressure changes, gas temperature changes, etc. It seems like you could do this with a very weak pump. Using a viscous liquid as a buffer between the house and environment is a similar idea to lots of the quake proofing strategies, I think... buffer the shear force associated with a quake.

One thing is that my very limited understanding of quakes was that they produce different sorts of waves. Would this system work both for quakes that shift the ground laterally and vertically?

Vidoes/photos are unavailable so perhaps this is addressed:

What happens if the actual foundation is damaged by the quake? When the air's turned off and the house settles unto something uneven or broken, that could undermine the structure of the house, no?

It's really quite simple.

1. A sensor malfunctions, and suddenly your home is lifted 3cm off the ground, and then dropped. While you are showering.

2. You now need $50000 worth of repairs to the structural integrity of your home.

3. You are now paralyzed.

The sensor malfunctions but the mechanism that the sensor initiates functions smoothly as normal. You're safe.

a 3 cm fall is paralyzing you? You probably have bigger problems than whatever dropped you an inch.

....while he was shaving with a straight razor.

There is no TGS?

Yes! How was this not titled "There is no Great Stagnation, Japanese Earthquake Edition"?

I recall a mythbuster-esque competition show involving competitors building various rigs to accomplish some goal, and for earthquakes they built the foundation on a pair of curved rails which allowed the structure to rock and absorb the damage, since I believe similar technology is used in larger structures for wind it seemed like a pretty decent idea.

I would think that you'd have to know the exact direction of sway for that to work properly.

You would think they could have put the old man in the other chair that wasn't violently shaking and the young person in the shaking chair.

No, you would think that they would put the experience man in the shaking chair. Watch the video again. His torso was barely moving. When the shaking started, he simply lifted his body and used his arms to absorb the motion. I suspect martial arts or acrobatics training.

Base isolation - not involving in-event inflation/air cushioning, is a major mode of increasing earthquake resistance - and quite impressively has been installed in massive existing buildings in 'seismic retrofits' - e.g. the Salt Lake City and County Building and many major California city halls.

There's a "Japanese-style" house near us that's said to be earthquake proof. In a quake it just falls in separate parts and you re-assemble it afterwards, The occupants will probably have been squashed, mind.

Is compressed air all that elastic?

Have you ever used a super-soaker?

What if the earthquake damages or destroys the foundation?

Then you are screwed anyway. At least with this system you will keep most of your valuables safe.

Somewhat related, people have designed high speed passenger trains based on a similar idea. Compressed air from the track (think air hockey) levitates, propels and stops the train.
There are evidently many advantages to this design. One is that the train can be light weight. With conventional designs it is hard to make a light fast train because you need a fair amount of friction between the wheels and the tracks to overcome the v^3 frictional force of air resistance at high speed.

I should have written above that drag force ~ velocity squared. It is the power ~ force * velocity that goes like velocity squared.
Another big problem with heavy trains is braking. One needs very expensive and elaborate brakes and even then it's hard to stop quickly because there is so much inertia. With the light train sitting on a compressed air tube idea, all the braking force is supplied by the air coming from the tube/track and one can stop fairly quickly.

That seems like a really inefficient way to move something... you'd propel the train via friction from directed, compressed air underneath? You'd need to seal the area below the train and it'd be quite inefficient, right? I guess the first problem can be overcome and the second can be canceled out by using a lighter train, though.

I think the idea is that you just need the compressed air to shoot straight up under the train when it passes. Basically when the train starts to pass, holes in the tube under the train open up and shoot air straight up. The train has the ability direct the air forward or back so it can use the air to accelerate in either direction. It's supposed to be quite efficient, but would obviously be expensive to build.

Earthquakes have different flavours (P, S and L waves) so not all of them are simply sideways motion.

Like California we get a lot of earthquakes here and as noted earlier, the lead/rubber base isolators are the shizzle.

Once I saw a tremor coming that was a simple "jump", just like shaking a kink out of an electrical extension cord. In cases like this that have a large vertical acceleration there may be some problems with this sort of solution but it's a typically great Japanese idea, though I note that it will entail building what amounts to two floors instead of one?

Do gas, water, and sewer connections typically have 3 cm of slack in them?

They are going to have to insert the whole system into the foundations anyways; can't be that hard to incorporate three flex bends into the piping.

The interesting part is that if you had to fill in the isolation "cushion" at 100 psi (say) and it was 3 cm high for a typical house area; you are looking at pretty large reservoir sizes or pressures.

Even if you had a 250 cubic feet air tank (that's pretty large) you'd have to fill it at 7000 psi. And this is ignoring how you'd get it all to fill within 1 sec with piping and valve resistances etc.

Is something wrong with my calculations? And I'm assuming there's no feasible compressor that can dynamically serve this sort of instantaneous cfm loads; so an air-tank must be what they use.

As stated above a typical house weighs 100lbs per square foot. At 144 square inches you would need less than 1lb sq/in of pressure to lift the house.

I wonder how many times your house can be lifted before maintenance is required. Many times, earthquakes have a number of shocks and it's going to be sad if the system protects you from the first shock but not the sometimes more powerful after-shocks.

lots of bitching, commenters have to read and think before writting, PLEASE!

engineers are NOT scientist looking for the ultimate truth. if the systems works, works & mission accomplished. you may bitch about "wrong design", "risk of sensor malfunction" or "best solution has been implemented in California". But this is real world engineering like how we ended up with 3-point seatbelts in cars. If 6-point seatbelts are better, why my car has only a 3-point seatbelt? And why plane seats only 2-point seatbelts? Cause it works =)

And if anybody have studied structural design (civil engineering stuff) may have not forgotten this facts:

a) resonance: sometimes it doesn't matter if the quake is trepidatory or oscillatory, the structure shakes and migh resonante along the seismic waves. an air cushion may help a lot.

b) in order to standarize earthquake simulations & building codes, trepidatory and oscillatory ground movements are simplified into an equivalent horizontal acceleration. that's why quake simulation is done in "shake tables". that "artificial" horizontal movement accounts for "real world" X, Y & Z displacements. remember engineering main goal is get the fucking job done, not finding the truth =)

Comments for this post are closed