1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

B Methods of Trapping Laser Light

  1. Nov 7, 2017 #1
    A) Can the light of a laser be trapped by a spherical mirror. B) would it help to have a liquid in the sphere so as to distribute this light equally throughout the volume of the sphere. Assume laser is shown in the internally spherical constantly. The key is that we want this light to remain in the sphere as long as possible

    Thank you
     
    Last edited: Nov 7, 2017
  2. jcsd
  3. Nov 7, 2017 #2

    davenn

    User Avatar
    Science Advisor
    Gold Member

    if you mean that you give a pulse of laser, or any other light for that matter, then the simple answer is no
    the reason, no mirror is perfect and light will escape

    what is your definition of as long as possible ?
    milliseconds ?
    micro seconds ?

    Others better than me may be able to give some ideas of duration


    Dave
     
  4. Nov 8, 2017 #3

    tech99

    User Avatar
    Gold Member

    I suppose that this is a spherical cavity resonator. The dimensions would need to be exact for resonance. The light will decrease exponentially, persisting for a time dictated by the Q of the cavity, so maybe a few hundred cycles before it becomes half the intensity.
     
  5. Nov 11, 2017 #4

    Tom.G

    User Avatar
    Science Advisor

    Sorry, other than free space, I haven't heard of a superconductor for light, yet. Sometime in the future it may be worth a Noble Prize for someone.
     
  6. Nov 12, 2017 #5

    mfb

    User Avatar
    2016 Award

    Staff: Mentor

    LIGO's mirrors only absorb 1 out of 3.3 million photons - but require the light to be precisely perpendicular to achieve that, so a sphere is not the best shape. Take the 4 km LIGO tunnel, put these mirrors on both sides and you get a lifetime of 4km*3.3 million/c = 44 seconds. You can inject light, wait a minute, get it out and you still see it.
     
  7. Nov 12, 2017 #6

    ZapperZ

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2016 Award

    Do you know what a typical mirror is made of?

    Zz.
     
  8. Nov 12, 2017 #7
    Are you asking me? I actually did not know what a 'typical' mirror uses. I know you can use almost any type of metal, and they have different reflective properties.


    Mainly I was thinking of ways to collect laser light to distribute it over pv.
     
  9. Nov 12, 2017 #8

    ZapperZ

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2016 Award

    But this is the central "physics" of your question. You need to know what exactly is a "mirror".

    A typical mirror is a thin film of metal, usually aluminum. And that's the key here, i.e. it is a metallic surface. The presence of conduction electrons is essential in what makes a metal "reflective" in the visible range (it isn't very reflective in the UV range, for example). Even within the visible range, a particular metallic surface may have differing reflectivity for different color.

    So already this is more complicated than a first glance. The conduction electrons respond to the incoming light, and there is always a non-zero probability that the incoming photons never get reflected but instead are absorbed by the metal as heat. One can naively imagine that the conduction electrons that are excited by the photons undergo scattering with the lattice ions before they have a chance to transmit the reflected light.

    Real material have real properties, not idealized ones.

    Zz.
     
  10. Nov 12, 2017 #9

    mfb

    User Avatar
    2016 Award

    Staff: Mentor

    Sunlight has an intensity of about 1 kW/m2, that corresponds to an energy density of 3.3 µJ/m3 (divide by the speed of light). Even a cubic kilometer of sunlight can just store 3.3 kJ, about the amount of power a hair dryer needs in 3 seconds. With lasers you might get a bit more but then you have the sunlight -> electricity -> laser conversion where you lose at least 90%. Even if we could store the light for hours instead of seconds (and even these seconds are achievable only for very specific wavelengths), it wouldn't help.
     
  11. Nov 13, 2017 #10

    Yes-in regards to heat and mirrors-yeah.. if we wanted to have sever MJ or even a few GJ laser-there would be no way to be able to reflect that high of a density of photons. I have instead though of centering a laser with convex lenses onto something that would break the laser up and distribute the light onto say GaAs pv cells. SO Would there be a lens that would be light enough and strong enough to "center" the laser. You would need at least one lense to be able to adjust rapidly. The laser could be centered on a spherical diamond-which would then be distributed on GaAs. With this design (assuming you could have servos accurate enough) you could preserve way more than 10% of that photonic energy.
     
  12. Nov 13, 2017 #11
    Also, we need a laser that powerful (>10 MJ) to be generated by single-pass fiber pumping.. idk how to do that
     
    Last edited: Nov 13, 2017
  13. Nov 13, 2017 #12

    Baluncore

    User Avatar
    Science Advisor

    Any solid or liquid lens, or gaseous chamber filling will rapidly scatter and absorb energy from the stored light.

    Light is an EM wave. The incident magnetic field induces a perpendicular current of free electrons in the mirror surface. That in turn generates a perpendicular magnetic field, that is then equal and opposite to the incident wave, hence cancellation of the EM wave into the conductive mirror surface, that becomes reflection back from the surface.

    An efficient mirror would need to have a highly conductive front face, I would expect a dense polished coating of silver. Oxidation of the metal surface will need to be prevented by an inert gas until the mirror is transferred to the vacuum of the light storage chamber.

    Colder mirrors will probably work better than warm ones because cold metals are less resistive, but I don't believe a superconducting metal mirror has an advantage at optical wavelengths, since the photons have higher energy than the superconducting metal atoms in the mirror.
     
  14. Nov 13, 2017 #13

    ZapperZ

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2016 Award

    Remember that while DC resistivity of a superconductor is zero, the AC resistivity is NOT zero. So superconducting mirror will still present a non-zero resistance to the oscillating conducting electrons.

    Zz.
     
  15. Nov 14, 2017 #14

    mfb

    User Avatar
    2016 Award

    Staff: Mentor

    The 10% are optimistic without any optical elements. Adding more points where energy gets lost makes it worse. The combination of extremely efficient solar cells (30%) and extremely efficient lasers (30%) gives you a 10% light->light conversion - into the infrared as you probably need a CO2 laser to get 30% wall-plug efficiency.
     
  16. Nov 18, 2017 #15
    Well sir, let's neglect the efficiency of generating the laser. Experimentally you are right- 30% of laser light we do hit our target with would be remarkable-let's compare that to any other type of space propulsion however-if we can land 1 MW on target we could get a sizeable mass moving pretty quickly-pretty quickly.
     
  17. Nov 18, 2017 #16

    mfb

    User Avatar
    2016 Award

    Staff: Mentor

    You can’t neglect efficiencies of (non-mobile) storage systems. Their whole point is to store energy efficiently over some relevant timescale.
     
  18. Nov 21, 2017 #17
    My point is to compete with chemical rocket fuel for large payload spacecraft
     
  19. Nov 22, 2017 #18

    mfb

    User Avatar
    2016 Award

    Staff: Mentor

    See the tiny energy density discussed above. You can’t increase that to relevant values without burning through all the mirrors (destroying the storage system). Laser beams from the ground (e.g. to evaporate propellant) have been proposed, but never tested on a relevant scale.
     
  20. Nov 22, 2017 #19

    Baluncore

    User Avatar
    Science Advisor

    Are you really advocating releasing all that energy under a spacecraft in order to launch it? Rockets fuelled with liquid chemicals use the propellants to cool the combustion chambers. How will you prevent the inefficient mirror component of your incident light from melting the bottom of your spacecraft? Will you carry a liquid coolant such as liquid oxygen and hydrogen that can then be used as fuel to correct orientation during launch. Will the resulting cloud of condensate scatter the laser light? How much laser light will be scattered by the atmospheric wake turbulence that follows a supersonic vehicle? Will the atmospheric density behind the vehicle not form a lens that diverges the light beam to miss the spacecraft being launched?
     
  21. Nov 22, 2017 #20

    mfb

    User Avatar
    2016 Award

    Staff: Mentor

    Let's make a rough estimate: Concentrator solar cells and solar thermal towers can cool away ~1000 times the solar irradiation, or about 1 MW/m2. Fission reactors have a bit more, for fusion people look at 10-25 MW/m2 (example 1, example 2).

    Let's be super optimistic and combine 50 MW/m2 with LIGO-quality mirrors (1 photon out of 3.3 million is absorbed). They don't work at temperatures where you can cool this heat flow but this gives a conservative upper limit. That means we can provide 3.3 million times 50 MW per square meter, or 165 TW/m2. To get the corresponding pressure divide it by the speed of light and multiply by 2 (arriving+leaving light). The result? 1.1 MPa, about 11 times the atmospheric pressure. That is in principle sufficient to launch a rocket.

    Caveats: 165 TW (a single square meter in the calculation above) exceeds the global power consumption by a factor 10 and the electricity production by a factor 100. That is not really feasible. If you can dedicate a full large power plant block (1 GW) to the rocket launch and assume a super optimistic 50% laser efficiency, the force you get is only 1 GW/c = 3.4 N, not even sufficient to launch half a kg. With 10 power plant blocks you can launch two kilograms, still not reasonable as the mirror and other infrastructure will probably have a larger mass already.

    What about storing the energy in advance?
    Accelerating a payload of mass m to 8 km/s (roughly orbital speed, ignoring drag and gravity losses) at constant acceleration needs a force of mx over 8 km/s / x time. A force of mx needs mxc/2 laser power, or mxc electricity. Multiplied by the time we get an energy demand of 8 km/s * c = 2.4 TJ/kg. This energy doesn't depend on the acceleration profile, I just chose the easiest one here.
    Superconducting coils are probably the cheapest technology that can store several TJ and deliver them back within minutes. A typical investment cost is about $100/kWh, so we need $100 millions per kg of lauch mass. A very small spacecraft of 100 kg would need $10 billion investment just for the energy storage. A one tonne payload would need $100 billions.

    You could make a cavity with a stationary mirror and the rocket as other side. If we handwave away all focusing issues, that approach recycles the power efficiently, but unfortunately the moving rocket will shift the light frequency - and that means your mirrors get a much worse efficiency. Oh, and the focusing won't work.
     
    Last edited: Nov 22, 2017
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted



Similar Discussions: Methods of Trapping Laser Light
  1. Laser Light (Replies: 14)

Loading...