Science: Energy: Solar: Collector
Light tube accumulator   (+1, -1)  [vote for, against]
Accumulate sunray energy in a line.

I wonder if this would work:

If I have an array of mirrors that bend the light to the side
so that light coming from above is redirected in 90 degrees (in an L shape)

1 2 3 4 5 6 7 8
L L L L L L L L


Would it be possible for the light from source 1 to pass through a one way mirror, serving on one side as a mirror to change the path of incoming "sunray" 2, but allowing light bent from source 1, to continue directly to the right in one line, through #2 and the rest...?

If this is physically feasable, then instead of a parabolic concentrator, a flat concentrator is possible. Always sending the light parallel to the ground. A circle around the peripheral would then send the light to the single collector location with no moving mirrors.

Also it could be used as a light "drainpipe" for gathering more light from a large area.

And by the way, is their a limit to the amount of light (and energy) that passes through fiber optic cables? Same question for "light tubes"...
-- pashute, Jan 30 2011

The problem with one-way mirrors... http://commons.bcit...oneway/onewmint.htm
...is that they let you build perpetual motion machines. [Wrongfellow, Jan 31 2011]

Infinite energy from vampires Infinite_20energy_20from_20Vampires
You could do incredibly cool things... [spidermother, Jan 31 2011]

Flat Panel Light Collector Flat_20Panel_20light_20collector
This is how I think this could work. [MisterQED, Jan 31 2011]

Optical isolator http://en.wikipedia...ki/Optical_isolator
An optical isolator, or optical diode, is an optical component which allows the transmission of light in only one direction [xaviergisz, Feb 01 2011]

This is how this works... http://www.physorg.com/news134917794.html
This stuff works better without mirrors (and smoke for that matter) [madness, Feb 02 2011]

Optical tile - flat light concentrator http://www.technolo...w.com/energy/22204/
So is this a hoax? Or is it just not as efficient as advertised? [pashute, Feb 14 2011]

better illustration of the LSO http://nextbigfutur...tratorlow-cost.html
[xaviergisz, Feb 14 2011]

US 2008/0271776 http://www.google.c...v=onepage&q&f=false
light-guide solar panel [xaviergisz, Feb 15 2011]

One way mirrors don't exist. A half silvered mirror under 2 would reflect half the light from source 1 and half from source 2, you would gain nothing over all.

I have thought about this sort of thing for a while - sadly, the reason for complicated sun-tracking systems is that there's no simple and elegant solution.

Fibre optic cables attenuate the light slightly, enough light heading down it would heat up and eventually melt the wire.

A mirrored tube is practically limitless, although the air, and the mirrors themselves, would absorb heat and eventually deflagrate/vaporise/explode if you sent enough down it.
-- mitxela, Jan 31 2011


//Would it be possible// No, for the reasons [mixtela] gives.

True one way mirrors can be proven not to exist precisely because if they did you could do incredibly cool things with them, which would break the laws of thermodynamics.

It is possible to create a highly concentrated beam of light using a combination of a large parabolic mirror and a smaller hyperbolic one, or various equivalent arrangements. This beam can then be directed down a light pipe. You still need to keep it pointed at the sun, though.

//gain nothing// Actually, with perfect, half-silvered mirrors, if the intensity of the incoming light is 1, the intensity reflected by successive mirrors is 1/2, 3/4, 7/8, ... , (2^(n-1)) / (2^n), i.e. approaching but always less than the incoming intensity.
-- spidermother, Jan 31 2011


//One way mirrors don't exist.//

What about mirrors made by total internal reflection? Admittedly this is only a one-way mirror at angles above a critical angle, but I'm interested as to what laws of thermodynamics are/would be broken by their existence.
-- xaviergisz, Jan 31 2011


See the link. Basically, if you separate a perfectly reflective cavity into two by putting a one-way mirror in the middle of it, all the light energy gets concentrated on one side, allowing you to build a perpetual motion machine.
-- Wrongfellow, Jan 31 2011


I think the link is more an argument against the possibility of "perfection" rather than the "one-way-ness" of a one-way mirror.
-- xaviergisz, Jan 31 2011


Even a perfect mirror (fully or partially silvered) does not break any laws of thermodynamics. Even an imperfect one-way mirror would, since it could change an equilibrium state into a non-equilibrium state without external input.
-- spidermother, Jan 31 2011


Surprised you missed this one pash. You're not "accumulating" any light along that horizontal path because to get it in there, you're bouncing it off a right angled mirror that must be partially mirrored to allow the light from the previous angled mirror to pass through. Partially mirrored also means partially attenuating so while you might bend 50% of the light into a right angle off the partially mirrored reflector, you lose 50% when it hits the next mirror along the "trough" then another 50% when it hits the next, and the next until you don't have any light coming through at all.

If there were such a thing as a mirror that reflected in only one direction this would work, but there's no such thing. The term "one way mirror" is used when you have a partially mirrored window between two rooms, one that's brightly lit and another, where the viewer stands, that's dark. This means light in the bright room bouncing off the glass is brighter than any light in the dark room passing through the glass, obscuring it so you simply see a reflection. The brighter light "drowns out" the lesser light source, the same way you don't see stars in the daytime because of the bright sun.

Reverse the lighting arrangement and it would reverse who can see who as well.

So if you see somebody wearing mirrored sunglasses at night, you'll know they're walking around staring at a reflection of their own eyeballs and therefore should probably be avoided.
-- doctorremulac3, Jan 31 2011


The idea as stated will not work for multiple reasons, many are listed above. I figured out how to make it work, and will submit it as a separate idea. (link)
-- MisterQED, Jan 31 2011


spidermother, I think we're having a "no true Scotsman" discussion; i.e. we're using different definitions of the term "one-way mirror".

If you define a one-way mirror as a mirror that reflects light from all angles, then I agree that it doesn't exist.

However, if you define a one-way mirror as a mirror that reflects light from shallow incident angles (i.e. angles greater than the critical angle), then one-way mirrors do exist (specifically, total internal reflection). If you want to play around with a one-way mirror, get a glass of water and look at the *underside* surface of the water from various angles.

I'm not convinced of the argument that *imperfect* one-way mirrors defy the laws of thermodynamics given in the first link. I can imagine a thought experiment similar to the one in the first link utilising total internal reflection; thus I think it is the perfection of a total internal reflector rather than the one-way-ness of the reflector that is impossible.
-- xaviergisz, Jan 31 2011


My link describes a mirror that reflects (perfectly) all photons that hit it from side A, while also transmitting (perfectly) all photons that hit it from side B, so that they appear on side A too.

That is, viewed from side A, it's a perfect mirror, while viewed from side B, it's perfectly transparent.

I claim that no mirror can do this, whether it uses total internal reflection, or any other technique.

As for critical angles, they have no relevance to my claim at all.

The original idea described it thus:

//Would it be possible for the light from source 1 to pass through a one way mirror, serving on one side as a mirror to change the path of incoming "sunray" 2, but allowing light bent from source 1, to continue directly to the right in one line, through #2 and the rest...?//
-- Wrongfellow, Jan 31 2011


I completely agree with you Wrongfellow - a one-way total internal reflection mirror is not a *perfect* reflector from one side and *perfectly* transparent from the other side.

But the real question is, do you consider a total internal reflection mirror a one-way mirror?
-- xaviergisz, Jan 31 2011


No, I don't. Total internal reflection reflects the photons from side A back to side A, and also reflects the photons from side B back to side B. Maybe it's not perfect, and some of them escape from one side to the other; but equal numbers of photons escape in each direction (assuming equal illumination on both sides) and thus the mirror is symmetrical, favouring neither one side nor the other, and hence I wouldn't describe it as "one-way".

A one-way mirror, per [pashute]'s description, allows more photons to pass from side B to side A, than it allows from side A to side B; this permits the construction of a perpetual motion machine, in violation of the laws of thermodynamics.

Perfection isn't the issue. Even an imperfect one-way mirror, that only allowed 1% more photons to pass it in one way than the other, would still break the rules.
-- Wrongfellow, Jan 31 2011


//Total internal reflection reflects the photons from side A back to side A, and also reflects the photons from side B back to side B. Maybe it's not perfect, and some of them escape from one side to the other; but equal numbers of photons escape in each direction (assuming equal illumination on both sides) and thus the mirror is symmetrical, favouring neither one side nor the other, and hence I wouldn't describe it as "one-way".//

I suppose we'll have to agree to disagree.

But I encourage you to have a look at a glass of water. View from the top at a particular angle, then view from the bottom (so you can see the underside of the surface of water) from the same angle. Does it look the same?

Alternatively, go to you local pool with a friend. While your friend stands at the side of the pool go under water. Your friend will be able to see you but you won't be able to see them.
-- xaviergisz, Feb 01 2011


I thought we were discussing this bit: //a one way mirror, serving on one side as a mirror to change the path of incoming "sunray" 2, but allowing light bent from source 1, to continue directly to the right//
-- Wrongfellow, Feb 01 2011


Ah, I thought we were talking about the concept of a one-way mirror more generally.

The idea itself is quite impractical, but yes I think you could use total internal reflection to make it work (though very inefficiently).

If you had a glass rectangle with a sawtooth wave of grooves cut into the bottom you could re-direct light in a line. I'll try and draw it below:

light light
.|......|
.\/.....\/
_________
|..........|
|\|\|\|\|\| --> re-directed light
-- xaviergisz, Feb 01 2011


//fire hazard// Total infernal reflection?

[xaviergist] The problem here is //from the same angle//. Light is refracted (bent) at the water-air interface; a ray of light passing through the water's surface has two different angles. When this is taken into account, the mirror is symmetrical. As the angle inside the water approaches the total internal reflection angle, the angle in the air approaches 90 degrees (parallel to the water surface). Beyond that, in the region of total internal reflection, there is no path for a ray of light on the air side, so what you see from that side is of no consequence.

Here's an experiment you can do. Shine a laser pointer from beneath the surface of the water in a rectangular aquarium, and adjust the angle until you get total internal reflection (there is no need to get the pointer wet, you can shine it in through one of the sides). Take a second laser pointer and shine it at the same spot on the water surface, but from above. Try to make its beam in the water line up with the reflected beam from the first pointer.

*Spoiler alert* - you can't. That's why total internal reflection is not an exception, and why your glass sawtooth scheme won't work.
-- spidermother, Feb 01 2011


Ok spidermother, I think I understand where our difference of opinion lies.

I define a one-way mirror as: an object that reflects light from one side while transmitting light from the other side.

You define a one-way mirror as: an object that reflects light from one side while transmitting light from another side *and* it functions 'symmetrically'. Essentially your idea of a one-way mirror is an object that functions exactly like a window from one side and a mirror from the other side.

I am happy to concede that a one-way mirror according to your definition doesn't exist. But for the purposes of this idea, I think my definition is more suitable.

And I think my sawtooth groove idea would work. I'll try and explain.

Assume the glass has a critical angle of less than 45° and assume the sawtooth grooves are at 45°. A beam of light coming straight down will be totally internally reflected at the glass-air boundary and be reflected horizontally. The horizontal beam will exit the sawtooth perpendicular to a glass-air boundary. The beam will then meet the next sawtooth at a 45° air-glass boundary and will *not* be reflected (other than the usual 'glare' reflection which is reflected off any surface). At this point it will be deflected (refracted) then deflected (refracted) again on its way out of the sawtooth. And so on. The beam will not be horizontal after the deflections and this is why this is a very inefficient design.
-- xaviergisz, Feb 01 2011


[Ian Tindale] Your mirror is obviously broken or defective. Is it still on warrantee?
-- mouseposture, Feb 01 2011


Just to be clear, by 'symmetrically' I mean that for any given optical path, if you introduce a certain quality and quantity of light to either end of the path, the amount that makes it out the other end is the same regardless of which direction it takes along that path. It's a basic law of optics, and applies in all situations, not just with mirrors. An optical path need not be straight; it can be bent by reflection or refraction.

In the case of [Ian]'s mirror, if you consider an optical path that passes through the mirror, the proportion of light that makes it from front-to-back or from back-to-front is the same, close to zero in either case. Most of the light that strikes the front is blocked by reflection, while most of the light that strikes the back is blocked by absorption, but that makes no difference.

Even in the case of a mirror with a worn backing, if (say) 20% of the light that lands on the front makes it out the back, then 20% of the light that lands on the back will make it out the front (assuming equal wavelength distribution).
-- spidermother, Feb 01 2011


//the amount that makes it out the other end is the same regardless of which direction it takes along that path//

Exactly. All passive optics is symmetrical in this sense - reverse the direction of the beam (or, equivalently, reverse time) and the behaviour remains the same.

Any asymmetrical behaviour requires an external energy input, if we're to believe that perpetual motion is impossible.
-- Wrongfellow, Feb 01 2011


Now that that's sorted, it'd be really annoying if I mentioned a solar panel wired to light bulbs on its reverse side, so I'd better not.
-- spidermother, Feb 01 2011


//All passive optics is symmetrical in this sense - reverse the direction of the beam (or, equivalently, reverse time) and the behaviour remains the same.//

If I understand it correctly an optical isolator is an exception to this rule (link).
-- xaviergisz, Feb 01 2011


Is an optical isolator an example of //passive// optics, though? Anyway, it doesn't function as a one-way mirror, because the light that is blocked in the reverse direction is absorbed rather than reflected, so it can't be used in [pashute]'s idea. So I guess it's an exception to a simple statement of the rule, but not to the underlying principles.
-- spidermother, Feb 01 2011


Not so! An optical isolator reflects just as much light back on itself as it allows to pass. The first reference link from Wikipedia is interesting reading, for example this:

//The second half, which after passing through Nicol 2 is totally reflected at Nicol 1 and then returned upon itself, on its arrival at Nicol 2 is not transmitted (as Wien seems to suppose) but is totally reflected. When again returned upon itself by a perpendicular reflector and again rotated through 45 ̊ by the magnetized medium, it is in a condition to be completely transmitted by Nicol 1, and thus find its way to body 1 and not to body 2 as the argument requires. The two bodies receive altogther the same aount of radiation, and there is therefore no tendency to a change of temperature.//

This bit kind of parallels something I mentioned, too:

//I suggest correcting HRK by striking out “its direction” and replacing those words with “time” since reversing time reverses both the magnetic field and the direction of light propagation//

Well, yes, quite. ;)
-- Wrongfellow, Feb 02 2011


Thanks, [Wrongfellow]. Nonetheless, the last sentence of your first quoted paragraph clearly states that an optical isolator does not work as a one-way mirror in the thermodynamics-busting sense (which we knew already, of course.)
-- spidermother, Feb 02 2011


My "Not so!" was a response to [xaviergisz].
-- Wrongfellow, Feb 02 2011


After all this discussion I think I'll summise what we all agree upon:

1. thermodynamic-defying one-way mirrors do not exist

2. total internal reflection non-thermodynamic-defying one-way mirror do exist

And here's the things we still need to reach agreement upon:

3. This idea is completely impractical, but not impossible

I think I can convince spidermother of how my sawtooth design would work. First, consider the simplest form of this idea: only two sawteeth with a vertical laser hitting each sawtooth. (note that it is easier to think of the sawteeth as separate right-angle prisms).

The two beams will travel as follows:

laser 1 reflects off sawtooth 1 to make a horizontal beam 1 that get deflects off sawtooth 2 and ends up at photovoltaic cell at an angle.

laser 2 reflects off sawtooth 2 to make a horizontal beam that end up at a photovoltaic cell.

Basically each sawtooth will create produce its own beam. The two beams *will not be parallel*: when the beams come out the side of the light tube accumulator they'll be travelling at different angles. Thus if the beams were reversed, each beam would travel back to its source. Thus it complies with the 'reversible optics' rule. Convinced?
-- xaviergisz, Feb 02 2011


Yes, I'm convinced. But the beams will be wildly non-parallel, not just slightly. Impractical, yes, but it might only be a moderate amount worse than just putting the solar panel directly in the path of the lasers/sunlight.

If the angles of the prisms varied appropriately, you could make a total internal reflection fresnel concentrator; possibly even one that produces an intensely bright spectrum. But that's a horse of a different colour.
-- spidermother, Feb 02 2011


This idea reminded me of something that actually works (apparently).

The idea is to coat the surface of a piece of glass with fluorescent dye so that light is captured within it. I guess you could say reflected internally....

As far as I can tell etching the surface of a piece glass might have the same effect -- but I think that two pieces of glass are required one on top to deflect and one below it to transmit.

... check out the link there is a diagram ...
-- madness, Feb 02 2011


According to all you guy's summarized calculations this cannot be done.

So how (see link) has it been (bummer for me) baked?
-- pashute, Feb 14 2011


It's a fresnel lense. The first one was used in 1823 so if this was you're idea, yea. It's baked.

My understanding was that you were talking about re-directing the light at right angles and in series so the light accumulated along a linear path. As stated, this wouldn't work because the reflecting elements would be located within that light path thus attenuating it.

So shot or hanged, take your pick.
-- doctorremulac3, Feb 14 2011


The 'optical tile' is not quite the same as your idea, nor is it a Fresnel lens.

I have linked to US patent for the 'optical tile' application number 2008/0271776. As you can see, it is not really based on one-way mirrors. Rather, it is about mirrors that redirect light into a light guide.

It is a nice design, but it requires light that is exactly perpendicular to the surface. This requires accurate tracking of the sun which can be difficult.

The 'optical tile' could be used in reverse thus spreading a beam. This could be used as a low-profile torch or car headlight.
-- xaviergisz, Feb 15 2011


I stand corrected.

It's a bunch of mirrors that looks like a fresnel lens, but it's not directing the light along a straight line which I gather is the core of the described idea.
-- doctorremulac3, Feb 15 2011



random, halfbakery