Theoretical way to levitate objects with light | Hacker News Reader

Theoretical way to levitate objects with light

caltech.edu

153 pointsrakkhi49 comments

As other commenters noted, the title is very misleading. The force exist but it is tiny, very very very tiny.

For radiation, force = power / speed of light

You can add x2 if the radiation bounce against the surface like in a mirror. For the calculations, let's pick a random value for power: 100W

* 100 W it's all the power that used an old incandescent lamp, most of it was wasted as heat instead of visible light, and in many directions.

* A modern LED lamp for a room has about 10W, so 100W are like 10 LED lamps.

* And 100W is HUGE for a laser. And 5W is huge for a laser. (Class 3R is up to .005W and Class 3B is up to .5W.)

So if we divide the force by the acceleration of the gravity https://www.wolframalpha.com/input/?i=2+*+100W+%2F+(speed+of... we get 7E-8Kg = 70ug that according to wolfram alpha is approximately 1/40 of a typical snowflake and 1/20 of a typical mosquito.

Note that if you suddenly put 100W into a snowflake or a mosquito, they will be destroyed instantly, and you need x20 or x40 more power to make them float, and they will absorb most of the energy so you need x2 the power.

(The article is about manipulating the light/material to stabilize the object, not to magically make the radiation force strong enough to levitate an elephant.)

Not to detract from your comment, but the most interesting thing about what you said (at least to me) is that a mosquito weighs half as much as a snowflake.

Snowflakes are non-dense more than they are light - they float down slowly due to air resistance rather than lightness per se.

Snowflakes have a pretty big variance in size, I'm not sure that a single weight could meaningfully be assigned to them.

I'm completely out of my element here. Hoping maybe for some clarification.

I've never heard of a laser, at any scale, having a force pushing backwards against the direction of light. It seems like one of two things is possible.

1) you could move a spaceship by shooting a laser. The force would be tiny, but it would keep accelerating as long as the laser is powered, with no reaction mass. i've never heard of such a thing being possible (but, again, way out of my element).

2) you could play the classic impossible machine game, with the fan on the back of a boat blowing into a sail. except with a laser and a sail. If 1) doesn't work, then 2) has to work, right?

1) is absolutely correct. This is also how solar sails work in proposed interplanetary spacecraft designs.

For more details, I can recommend this video from Because Science: https://m.youtube.com/watch?v=K6-q2edmiGk (hope I got the right one, didn't watch it because I'm on mobile)

That was the right one. I never got far along enough to learn about

    e^2=p^2c^2+m^2c^4

Thank you for the amusing explanation.

1) Yes, but it's easier to use the light from the Sun. More in https://en.wikipedia.org/wiki/Solar_sail

Also, it is possible to measure this in the lab. For example https://www.youtube.com/watch?v=Ng1X8mPJziA Beware that some popular experiments measure tiny difference of gas pressure instead of actual radiation presure https://www.youtube.com/watch?v=r7NEI_C9Yh0

2) If you put the laser an the sail in the same spaceship, and the sail absorbs 100% of the radiation, then the when the laser emit the photons it is pushed in one direction and when the sail absorbs the photons it is pushed in the other direction. So if the sail absorbs 100% of the radiation, the effect in null.

You can use a laser as a strange kind of thruster (if you remove the sail, and point the laser in the other direction). It's very inefficient, but it produces a force. It is not necessary to use a laser, you can use any kind of lamp (and it would be even more inefficient). The heat that escaper from the Pioneer is an involuntary example of this method https://en.wikipedia.org/wiki/Pioneer_anomaly

That made me think of a totally ridiculous idea for an interstellar ship: The Inchworm Drive.

Put a laser complex in space, which has a laser, a power source for the laser, and enough extra mass (if the power source isn't massive enough) to make the laser complex way more massive than your ship.

Use the laser to push our ship via a light sail.

Have our ship connected to the laser by cables, which we spool out from the ship as it travels, spooling them out fast enough that the cables are not exerting a retarding force on the ship, nor a pulling force on the laser complex (you can probably actually get some thrust from spooling out the cables). Yes, this will require several light-years worth of cables.

Go until the ship reaches the next solar system, and enters orbit around one of the system's outer planets. You need to pick a planet that has resources you can use to refuel the laser power source.

Reel the cables in, using the planet's gravity to keep the ship from getting pulled back to the laser. Doing this, the ship can now pull the laser after it, to the outskirts of the planet's solar system.

Then use the cables to pull the ship back to the newly positioned laser, and refuel it with resources you got from the planet.

Repeat for the next hope of the trip.

https://www.goodreads.com/book/show/34466963-why-we-sleep

The problem is that it's not easy to spool out the wires. Just disconnect the laser for a moment to simplify the discussion.

If the ship is traveling at speed V and the wires are at speed 0 in the usual reference frame [1]. If you imagine that you are over the ship, you will see that the wires are initially at rest and you have to accelerate them so they go away from the ship at speed V. So you will have to use a lot of energy to accelerate the wires [2]. And also the wires going away will act like the exhaust of a rocket. If you push the wires backwards, the wires will push you forwards. Now you have a very strange rocket that uses wires instead of some propellent.

[1] whatever it means, perhaps the initial reference frame of the space station that launched the ship, or the average of the local stars.

[2] Assuming that the wires are not so thin, because they have to pull a space station. Also, we are comparing them to a laser beam, so they are in comparison much heavier.

For (2) you can cut out the middle man (the mirror) and have the laser just point in the opposite direction of your desired acceleration. It amounts to the same thing, assuming the mirror reflects perfectly (by conservation of momentum - the momentum of the released photon is the same whether it bounces off a mirror first or just starts off in that direction).

Laser propulsion does use reaction mass. Mass-energy equivalence means that the emitted light has mass. Conservation means that the spaceship loses mass as it fires the laser. It’s basically a standard rocket with a ridiculously high specific impulse.

In an old comment throwaway_yy2Di made me notice that in spite it has a ridiculously high specific impulse, the momentum to power ratio is very small, it's only 1/c = 3.3E-6 N/kW.

I believe that’s an inherent tradeoff. The faster you make your reaction mass go, the more efficient your use of mass, but the less efficient your use of energy.

IIRC if you don't need to create the particles like in a ion thruster, it is better to accelerate them as much as possible. The problem with photons is that you need energy to create them.

If you double your exhaust velocity, you double your thrust but quadruple your power requirements. If you have a fixed amount of power, then you’ll have to cut thrust by one half.

Spacecraft are usually severely mass limited so less thrust and more energy consumption in exchange for better mass efficiency is usually the right tradeoff.

(2) isn't actually impossible with a fan and sail. The gas molecules recoil off the sail, imparting nearly 2x their original momentum (given by the fan) to the sail. Not as efficient as just using the fan as a propeller, but it does work. One of my college professors demonstrated it in class and published a paper about it in a physics education journal decades ago [0].

[0] https://aapt.scitation.org/doi/10.1119/1.2341931

5w isnt all that much for a laser these days. There are consumer laser "pointers" in the 5/10w range. 2w lasers are a norm for desktop engraving machines. So putting a 100/200w bank together wouldnt be a big deal.

(Ive personally verified some "legal" laser pointers in the 1.7w class. Weapons imho.)

> And 100W is HUGE for a laser. And 5W is huge for a laser. (Class 3R is up to .005W and Class 3B is up to .5W.)

It's kind of irrelevant, but I think it's worth pointing out that while those are legal/safety/sanity limits [0], there's a lot of companies that completely ignore them. Here's one that has quite a few 5W lasers [1], and some that are supposedly up to 50W which would be insane if true. As for 100W being huge... the US Navy is testing ones in the tens of kW [2]. Might eventually make for one hell of a point defence system.

That being said, photon drives still SUCK. Not having to carry the drive helps, but it's still in the Newton-per-100MW-power range.

0: http://www.funraniumlabs.com/2017/08/laser-products-hate/

1: https://www.laserpointerpro.com/attribute/power_5000mw-laser...

2: https://en.wikipedia.org/wiki/AN/SEQ-3_Laser_Weapon_System

A few years ago, I had for a year a green laser of 5W pointing in my direction over an optical table. It was redirected and used before it reach me, and there were a few security measures like a "wall" to stop the beam. Anyway, I was always slightly nervous ...

This reminds me of a video where someone levitates diamond dust with lasers.

https://www.youtube.com/watch?v=Sq7GaO8iqu8

In newtonian physics you learn that momentum is calculated as mass x velocity. Then in modern physics you learn that light has zero mass so you naively assume light has no momentum. Then later you learn that the original equation you learned was missing a piece and light can actually transfer momentum. This is a cool application of that property!

Ionic Thruster + Light Levitation = Flying Drone with no moving parts. Wish i can see this in my lifetime.

Ionic thrusters for drones? Maybe. Using light? No. Orders of magnitude off even considering the absolute limits of physics. That's not what this article says.

I didn’t see anything in the article that changes the amount of momentum that would be transferred - as far as I could tell it’s all about the distribution of that force being modulated to make the object being accelerated tend to stay in the center of the beam

Editing to clarify my point: the article’s title seems wildly misleading by using the word levitation. The energy required to “levitate” a macroscopic object in earth gravity using light is unchanged, and while I don’t feel like doing the math right now i feel pretty confident stating that “impractical” is a monstrous understatement.

Ionic thrusters only work well at very low pressures. At atmospheric pressure they are no better than a cold jet, just like opening a bottle of compressed air.

Sorry, here comes probably the usual question asked whenever somebody talks about "light":

if I point a flashlight at an object, does the target object move (in theory, if no other external force exists) because of the light hitting it?

I could expand the question asking: when the light hits, does it lose energy (because some of its energy is transferred to the object that is hit? (e.g. target object heats up and/or moves and/or don't-know-what-else-could-happen) => if yes, then light would disappear if it would bounce around enough times (e.g. if I let light enter a box and then I close it and never open it again, is the light absorbed after a while by the box?)

Personally, I am totally confused. I think that the last time I did some investigations, the simplest statement I found said "light has momentum but no mass" (is this correct, in general?).

That "light has momentum but no mass"-statement confused me even more as I thought that gravity has influence only on things that have mass (what I keep thinking about is the example of the light being bent when it passes nearby a black hole).

Additionally, about "momentum", its definition needs mass, at least as per Wikipedia:

> momentum (pl. momenta) is the product of the mass and velocity of an object. ( https://en.wikipedia.org/wiki/Momentum )

Therefore if light does not have a mass then it shouldn't be bent by gravity nor have any momentum?

Thx :)

  if I point a flashlight at an object, does the target object move (in theory, if no other external force exists) because of the light hitting it?

Yes, the target object gains velocity due to conservation of momentum. The photon goes from momentum p to -p, so the target object needs to gain 2p of momentum to balance it out. For photons, this amount of momentum is very small.

  when the light hits, does it lose energy (because some of its energy is transferred to the object that is hit?

For perfect reflection, no. (What you're referring to is light absorption, which is a different effect that raises the temperature of the target object and red-shifts the incoming photon.)

  Personally, I am totally confused. I think that the last time I did some investigations, the simplest statement I found said "light has momentum but no mass" (is this correct, in general?).

Pretty much. Photons are usually described by their momentum and energy. For our peace of mind, we can derive an effective "mass" using p=mv, where p is the electron's momentum and v is its velocity; however, the notion is not very useful since it doesn't correspond nicely to our everyday experience of mass (whereas electron momentum does correspond nicely to our everyday experience of momentum).

> For perfect reflection, no.

That's not true, it wouldn't conserve energy. The kinetic energy of the object that's hit changes, so the light must lose or gain energy. I think for a reflection this must be because of a red- or blueshift.

> What you're referring to is light absorption, which is a different effect that raises the temperature of the target object and red-shifts the incoming photon.

Absorption usually means a whole photon is absorbed, not redshifted. A lower energy photon can sometimes be emitted after absorption (however the direction and phase relation with the previous photon are lost)

> That's not true, it wouldn't conserve energy. The kinetic energy of the object that's hit changes, so the light must lose or gain energy.

You mean that as the "path" of the photon is changed by hitting the target object, the vector of the target object must change accordingly? (if yes then how "can" the new vector of the target object computed, if the source object/photon has no mass (I would basically miss a variable for the equation)? IF it does have no mass...)

Yes, using its momentum and energy.

Here you can read the relations between a photon's momentum and things like wavelength. Without any mass :)

https://en.wikipedia.org/wiki/Photon_momentum#Physical_prope...

(You need to scroll down a bit for the formulas)

These results come from special relativity(and a bit of quantum mechanics), which you really need if things don't have mass.

Thx a lot - next questions follow... :)

(I tried, but the wikipedia chapter is too complicated for me => trying to simplify here)

> ...it's momentum and energy.

Therefore, if the photon does not have a mass but is still able to have an effect against the target body it impacts with, then the photon must lose energy, right?

[puah - for some reason I cannot reply to the my post's - typothrowaway's - reply...]

> (I tried, but the wikipedia chapter is too complicated for me => trying to simplify here)

I can imagine. That stuff on the wiki is not even in every physics bachelor. It's definitely not for a general audience, but I thought it might help a little :)

> Therefore, if the photon does not have a mass but is still able to have an effect against the target body it impacts with, then the photon must lose energy, right?

Yes, or gain it if it slows down the object, if the object is traveling towards the light.

Or the internal energy of the object is changed. This happens when a photon is absorbed.

Thx!

So here are my next questions :) (I know nothing about physics etc... - I'm just a simple IT-employee)

> "Yes, the target object gains velocity due to conservation of momentum. The photon goes from momentum p to -p, so the target object needs to gain 2p of momentum to balance it out."

1) If the target object gains velocity, then the photon slows down? Meaning: a) are there then slow and fast photons and b) is the speed-of-light constant of ~300KM/s in vacuum valid only for not-sped-down-photons?

2) Then if I would put in orbit a bunch of mirrors and make them all reflect the sun's light to point to a fixed object, that fixed object would start moving because of the photons hitting it? (forgetting any other factor that does not have something to do with photons)

> For perfect reflection, no. What you're referring to is light absorption, which is a different effect that raises the temperature of the target object and red-shifts the incoming photon.

So, ignoring "perfect reflection", some photons will be reflected, some will be absorbed, which will raise the temp of the target object?

> Pretty much. Photons are usually described by their momentum and energy.

So, a remix of the original question, haha: if we say that a photon has no mass, but at the same time we say that momentum needs mass, then if we say that a photon has momentum we're screwed, no?

A photon's momentum is related to its wavelength, not it's speed. They all have the same speed in a vacuum.

Sure.

> So, a remix of the original question, haha: if we say that a photon has no mass, but at the same time we say that momentum needs mass, then if we say that a photon has momentum we're screwed, no?

We would be screwed, but momentum doesn't need mass.

1) So, the wavelength is basically its "energy"?

2) Therefore "momentum" needs only "energy" and velocity instead of "mass" and velocity?

Sort of, yes.

It's hard to put reasons and interpretations behind all this, because in the end we just use formulas and numbers and check them in experiments. It's usually simpler that way.

Be aware that a longer wavelength means a lower energy. Frequency is proportional to energy, which is inversely proportional to wavelength.

Can we kill photon by increasing its wavelength too much? Like put him between two mirrors so they would eat its energy multiple times (without absorbing)?

No, as far as I know, there is no limit to a photon's wavelength, so it will just keep on stretching.

From last year, a 3D volumetric display based on laser manipulation of particles in air: https://www.nature.com/articles/d41586-018-01125-y

Aren't those loud? I saw one that exploded air molecules to create the light.

Those are systems with pulsed lasers that ionize the air. This uses relatively low-power beams to trap and illuminate a small particle.

The new thing here is not pushing with photons. It's making the thrust self-stabilizing, so the target accelerates without spinning.

If they tried to do this with a spacecraft, wouldn't it need constant line of sight from earth with this technique? Seems interesting but pretty impractical for that application, unless I'm reading the text incorrectly.

You are right, can read more about one proposal here:

https://en.m.wikipedia.org/wiki/Breakthrough_Starshot

Avi Loeb described this on ep. 40 of After On [1], it didn't sound entirely unrealistic as he described it. It's around 19 minutes into the interview, I see a sibling links Breakthrough Starshot, in this podcast he describes it on a high level.

[1] https://after-on.com/episodes-31-60/040

Could we use this to make more haptic touch screens?

Surprised no-one's yet mentioned the XKCD blogpost on this exact topic: https://blog.xkcd.com/2008/02/15/the-laser-elevator/

Or the vaguely-related (because lasers) What-If scenario https://what-if.xkcd.com/13/

Holy shit, scaling this technology will change humanity. Amazing!

Reply on news.ycombinator.com