Join 3,562 readers in helping fund MetaFilter (Hide)


Light powered by falling sand
August 19, 2014 5:50 AM   Subscribe

How does (or how might) Danielle Trofe's falling-sand-powered lights work?

I'm wondering if anyone can either find more details of the design, or offer a decent explanation of how it works. Is the sand driving a small generator inside, similar to how the GravityLight works? If so, is this a fairly novel mechanism (in that it's driven by sand), or does this kind of thing exist elsewhere?

The science and design press lets us down a bit since they seem to be satisfied with "kinetic energy" and ask no further questions. It is implied that there is a working prototype, but I couldn't find confirmation of this, or a video.

Another description from Danielle Trofe Design's website.

Thanks!
posted by theyexpectresults to Science & Nature (15 answers total) 4 users marked this as a favorite
 
I'm guessing a small turbine, like a little computer fan. Sand flows through it in the same way as a hydroelectric power station. It's going to be very low wattage.

So yes, it's going to be the same idea as the GravityWatt, which lifts the idea from a grandfather clock. Potential energy to kinetic energy ...
posted by scruss at 6:17 AM on August 19


The sand isn't falling very far, and there's not much of it. Not much potential energy there. My cynical guess is that what actually powers it is hidden batteries.
posted by Sophont at 6:27 AM on August 19 [2 favorites]


I did the math for this once, only for people on light-up floor tiles instead of sand. Reproduced here so you can find the errors if any:
A 150 pound person gets pulled by gravity with a force of about 650 Newtons.

650 Newtons moving one centimeter is 6.5 watt-seconds, according to google.

So if a 150 pound person stands on a tile and it sinks by 1 centimeter, the absolute most you can get out of that is enough to light up an LED light for a few seconds.
If I got that right for a 150 lb person moving 1 cm, then an ounce or two of sand moving 20 cm isn't gonna give you much light at all.
posted by moonmilk at 6:38 AM on August 19 [1 favorite]


You can light a single LED with extremely low voltage using a Joule Thief, but honestly I'd guess battery trickery as well.
posted by odinsdream at 6:50 AM on August 19


You might find some of the answers to this question illuminating (ahem).
posted by Chocolate Pickle at 6:53 AM on August 19 [1 favorite]


It looks like it's in the prototype stage, which means that it probably doesn't actually work.
posted by It's Never Lurgi at 9:10 AM on August 19 [2 favorites]


It looks like it's in the prototype stage, which means that it probably doesn't actually work.

The page on her website for the lamp says,

"When the hourglass is flipped over the falling sand passes through an internal mechanism that generates energy to power the light."

Which sounds suspiciously like the Underpants Gnomes: Phase 1-Collect Underpants, Phase 2-?, Phase 3-Profit.
posted by Beti at 9:25 AM on August 19


Note that they're only lit in the renders. In the photos of the prototypes, the lamps aren't on.

It doesn't work.
posted by mr_roboto at 10:19 AM on August 19 [1 favorite]


I'm going to come in on the side of "It doesn't work" and file under "design wank"

Firstly, as you point out, the mechanism is hand-waved.

Secondly, as moonmilk points out, the maths is borderline - novelty at best, not practical.

Thirdly, note in the lower pictures - they're so conceptual and focused on making slick renders instead of workable products that they streamlined the lower LEDs out of existence, so when you turn that thing upside down, there's no light any more, unless you want it to come out of the previous LED which is now upsidedown, pointing the wrong way. Useful.

Fourth, actually there are several more red flags, but at this point it doesn't seem worth entertaining further.
posted by anonymisc at 11:12 AM on August 19


Okay, let's look at the theoretical limits.

Assume the average sand height is 0.5 meters and the weight of the sand is 1.0 kilogram. The potential energy is mgh = 1.0 * 10 * 0.5 = 5 joules.

Assume the LED is 3.5 volts at 1 amp = 3.5 volt-amps = 3.5 joules per second. This is typical for a small LED flashlight.

So assuming 100% efficiency, this scheme could power the LED for about 1.5 seconds (5 joules divided by 3.5 joules per second). Given real world efficiencies I would expect less than 1 second , at best.

So you could say this whole thing is phoney.
posted by JackFlash at 11:19 AM on August 19


If the room is dim, you can get a reasonable amount of light (think: those uselessly dim solar garden lamps) at 0.03A (ie a single 5mm 0.1W (T1 3/4) LED at full brightness), so that 1 second of bright light could be 30 seconds of light that isn't very noticeable in a normally-lit room.
posted by anonymisc at 11:26 AM on August 19 [1 favorite]


You might get by with hand waving and a mention of triboelectric power generation, but I really doubt that is what is actually going on
posted by Dmenet at 12:00 PM on August 19


so that 1 second of bright light could be 30 seconds of light that isn't very noticeable in a normally-lit room.

So, in other words, the pictures are photo-shopped fakes.
posted by JackFlash at 12:24 PM on August 19 [1 favorite]


You might get by with hand waving and a mention of triboelectric power generation, but I really doubt that is what is actually going on

The precise mechanism - triboelectricity, dc generator, alien fireflies that use sand in their exotic metabolism, etc. doesn't matter - if there's a lump that masses m that can only fall a distance h, then you only have a total change in potential energy of mgh to play with.

As per JackFlash's explanation.
posted by sebastienbailard at 12:36 PM on August 19


Thanks all! Suspicions confirmed! Though I can't believe I didn't notice that I was looking at a rendering in the case of the big one, thanks for pointing that out. I had previously assumed that they were maybe using batteries or some other kind of shenanigans like you find in perpetual motion machines.
posted by theyexpectresults at 4:57 PM on August 19


« Older My friend is getting married. ...   |  YANMD. I have an appointment ... Newer »

You are not logged in, either login or create an account to post comments