Cannae/EmDrive

[PB666]

Yes. It could be useful to know why it works. While no variation on QT can beat photon drive in absolute vacuum, this stuff might be interacting with atmosphere in interesting and very indirect ways. And even if it's simply a leak of some sort, we're still learning a lot about making fine measurements.

Caveat, I see absolutely no reason to test it as an orbital unit. There is nothing we can't test better and cheaper in the lab.

I absolutely want to see this tested as an orbital unit, particularly for the reasons you stated. Given the fact that the increased over photon drive thrust is undetected, I want to see if it will work in some orbital circumstances and fail in others. I think that Higgs fermion interaction may be accounting for the momentum. We need to get this sucker into a circumstance were local inertia is at a minimum and see if the force exerted changes. Have 6 identical units supplied with identical pointing each direction U,D,E,W,N,S in space use a second craft as a reference and see if the craft deflects in one direction.

- - - Updated - - -

Again. That's just theorising. Theorising helps no one. Observations do.

We observe a force, currently unexplained.

Cold Fusion! Doh!

- - - Updated - - -

That's correct because real-life lazors work exactly like they do in Star Wars.

Actually, using lasers to slowly ablate the surface of an asteroid can in fact work... but like most deflection schemes, it just takes a long time.

Only if the asteroids surface is partially metalic and not composed of pits and cavities. If the laser manages to hit a cavity and begin heating mater close to absolute zero, it will produce volatiles with unpredictable consequences.

[Rakaydos]

Only if the asteroids surface is partially metalic and not composed of pits and cavities. If the laser manages to hit a cavity and begin heating mater close to absolute zero, it will produce volatiles with unpredictable consequences.

An earthcrossing asteroid is unlikely to be near absolute zero. As the asteroid spins, it spends half of it's "day" bathed in unfiltered sunlight, quite often within the sun's Goldilox zone. At "night" it will radiate some of that away, but the night would have to be exceptionally long- Even the moon, with a 2 week long "night" only gets down to -173c. And that's ignorig thermal conduction of the rock, from day to core to night, something that may still matter on an object a few KM across.

[Technical Ben]

"Only if the asteroids surface is partially metalic and not composed of pits and cavities. If the laser manages to hit a cavity and begin heating mater close to absolute zero, it will produce volatiles with unpredictable consequences. "

No. That's again more dreaming.

[PB666]

An earthcrossing asteroid is unlikely to be near absolute zero. As the asteroid spins, it spends half of it's "day" bathed in unfiltered sunlight, quite often within the sun's Goldilox zone. At "night" it will radiate some of that away, but the night would have to be exceptionally long- Even the moon, with a 2 week long "night" only gets down to -173c. And that's ignorig thermal conduction of the rock, from day to core to night, something that may still matter on an object a few KM across.

http://en.wikipedia.org/wiki/Black_swan_theory <----never wrong, lol. I win.

[0111narwhalz]

If/when this drive is explained and put into space, I hypothesize a stray Kraken will drift by and casually nab it.

No, but seriously. This sounds like a kraken drive from an initial explnation. After closer explantion, it still looks like a damn kraken drive. Now, I've dabbled in physics just enough to look like I know what I'm talking about (but not much more), and this EM drive is vaguely terrifying in its implications. Either there's instrument error or other artifacts, or there's something to this.

Also, was it this thred with the scifi-predicting future tech? Anyway, Verne was quite successful, especially for a non-scientist. Moon expeditions, rigid, powered submarines, heavier-than-air flight, etc.

[John FX]

Also gene Roddenberry predicted the future.

Star Trek communicators were very similar to mobile phones.

The hand held work pads on star trek TNG are very similar to our tablets.

The personal computer everyone had in there quarters in tng are pretty much laptops.

Phaser which are technically lasers, the USAF are experimenting with lasers on a 747 to shoot down nukes.

Antimatter, back in the late 60s we didn't even know it existed.

Fusion reactors from TNG again we didn't think it was possible then but now a company, I beleive it's Lockheed Martin have created a small one.

The shuttle, a reuseable vehicle, ok it wasn't exactly safe. This is why the test bed is called STS Enterprise.

Transporters, well now we can make stuff dematerialize thing but not rematerialize them, good luck finding much info on this I read about it a few years ago.

These are just to name a few.

The phaser I heard about was quite cool. It fired a targeting laser which created an ionised channel in the air toward the target. A high voltage electric current was then passed down the ionised channel which could stun or kill someone presumably.

[PB666]

If/when this drive is explained and put into space, I hypothesize a stray Kraken will drift by and casually nab it.

No, but seriously. This sounds like a kraken drive from an initial explnation. After closer explantion, it still looks like a damn kraken drive. Now, I've dabbled in physics just enough to look like I know what I'm talking about (but not much more), and this EM drive is vaguely terrifying in its implications. Either there's instrument error or other artifacts, or there's something to this.

Also, was it this thred with the scifi-predicting future tech? Anyway, Verne was quite successful, especially for a non-scientist. Moon expeditions, rigid, powered submarines, heavier-than-air flight, etc.

I think they are pushing off real matter, not quantum fluctuations. However I think that expulsion mass nature is immaterial, as long as the expulsion mass (be it quantum fluctuations or gas or instrumentation) was not initially part of the ship, then I think we can move on to the next step of questioning..........energy efficiency.

kw of power givens micronewton/kg thrust???

[Accelerando]

If/when this drive is explained and put into space, I hypothesize a stray Kraken will drift by and casually nab it.

No, but seriously. This sounds like a kraken drive from an initial explnation. After closer explantion, it still looks like a damn kraken drive. Now, I've dabbled in physics just enough to look like I know what I'm talking about (but not much more), and this EM drive is vaguely terrifying in its implications. Either there's instrument error or other artifacts, or there's something to this.

Also, was it this thred with the scifi-predicting future tech? Anyway, Verne was quite successful, especially for a non-scientist. Moon expeditions, rigid, powered submarines, heavier-than-air flight, etc.

Indeed, he did so well that the Columbia capsule was named after the giant cannon from the story. Moon expeditions go even further back, too - Kepler outlined a very well thought-out one for his time, in his novel, Somnium, and Lucian of Samosata much further back, in the 2nd century, with his True History.

Edited February 16, 2015 by Accelerando

[magnemoe]

The phaser I heard about was quite cool. It fired a targeting laser which created an ionised channel in the air toward the target. A high voltage electric current was then passed down the ionised channel which could stun or kill someone presumably.

Read about it however then it was as an weapon to stop cars, it works like an taser except it don't need an projectile.

The stun gun is an old sci-fi idea and it has been lots of tries to make it in real world as it would be very useful.

Makes me wonder why no stun grenade? It would be just as useful in that you could hit someone in hiding.

The star trek phaser has no real life equivalent and its pretty magical, not only can it stun harmlessly but its also an heavy weapon, its also able to make stuff disappear without harming the surroundings.

[Northstar1989]

OK, so there's been a further set of studies that confirm the results in vacuum, and even a theory based on the vacuum plasma that explains and even PREDICTS its behavior it WITHOUT breaking Conservation of Momentum (essentially the device pushes against transient particles that normally briefly and disappear all around the universe...)

Time for some additional discussion!

Regards,

Northstar

[Streetwind]

Eighteen pages of additional discussion have already happened here over the past weeks: http://forum.kerbalspaceprogram.com/threads/109851-EMDrive-vacuum-testing-give-positive-results

[N_las]

OK, so there's been a further set of studies that confirm the results in vacuum, and even a theory based on the vacuum plasma that explains and even PREDICTS its behavior it WITHOUT breaking Conservation of Momentum (essentially the device pushes against transient particles that normally briefly and disappear all around the universe...)

Someone who claims that "vacuum plasma" explanation wouldn't break conservation of momentum just doesn't understand a thing, or tries to lie to make the explanation appear more plausible than it is. If a pair of particles with net momentum can simply disapper, than conservation of momentum is broken, easy as that.

The only reason a virtual particle pair can "appear and disappear" in vaccum is because the pair has no net momentum and energy. If you "force" some momentum on such a particle, it would "become" a real particle, so it wouldn't disapear anymore. That would just be a version of creating your own reaction mass, thus a photon drive.

The "theory" that "explains and predicts" the drives behaviour is a piece of code were you input the data about the electromagnetic field, and the output is that supposed thrust. There is no reasonable explanation how that electromagnetic field would even influence the vacuum. Also, that piece of code was written after the first measurments, so that "prediction" could just be a tinkering with code until it shows the measured behaviour.

[Rakaydos]

So if imparting momentum on the virutal particals makes them real, what's the problem? You've spent energy making your ship move using reaction mass you didnt bring with you. The energy to propel the particle IS the energy to create it, because it already has a statistical chance to exist, and you're just ruining the balance of the universe by pushing off it.

[Mazon Del]

The issue Rakaydos that N_Las and others point out, is that if the engines DO work by converting virtual particles into real, current physics still says that there is a maximum efficiency that this can operate at which is being broken by the input energy to thrust of this drive.

I of course am hopeful that maybe the drive system is doing something new or unrealized, but only time will tell. Within the next couple of months they will likely have shipped the test article off to at least one alternate lab for verification purposes, and probably 2-6 months after that we should have their report.

[SuperFastJellyfish]

In the spirit of reposting, Paul March from the Eagleworks team has been corresponding with a few people here. Lots of over my head math, but it has been fun and informative lurking on that thread.

[N_las]

So if imparting momentum on the virutal particals makes them real, what's the problem? You've spent energy making your ship move using reaction mass you didnt bring with you. The energy to propel the particle IS the energy to create it, because it already has a statistical chance to exist, and you're just ruining the balance of the universe by pushing off it.

In a certain way, you did bring your reaction mass with you. Imagine the following: You have a LED flashligh in space. It needs 5 Watt and produces light with 100% efficency. It has a battery which contains 5 Wh of energy. You are creating your own reaction mass in form of photons out of the energy in the battery. The Energie of a photon is E = h c/wavelength and the momentum of a photon is p = h/wavelength. If you do the calculation you get thrust force from the flashlight of Force = Power/c. That is the ultimative limit for efficentcy if you "produce" your own reaction mass.

After 1 hour, the battery will be empty. The total energy content of your "ship" is now 5 Wh smaller than before. This has a direct effect on the restmass of the flashlight. With E=mc^2 you can calculate, that the flashlight has lost 2*10^-13 kg of mass. That is the mass of about 1.2*10^14 hydrogen atoms. You could say, that this was the reaction mass you brought with you. It ultimately doesn't matter if you bring energy or mass along, because they are kinda the same.

If you don't bring a battery, but use solar power: congratulation! You could just as well have brought a mirror with you! Using solar cells to collect the energy of photons, converting this energy in photons again an "shining" them out trough the flashlight is exactly the same as just letting them bounce of a mirror.

[Rakaydos]

But that's not what this drive is doing.

Virtual particle pairs can be assumed to already exist, because that's what they do- start and stop existing. Therefore you do not need to create them out of energy, only to push off them. Their net energy is no longer zero, but it's not their mass energy either- it's the equal and opposite of the energy gained by the ship pushing off them.

That's what the articles mean by "quasi-virtual particles"- they're not "real" yet, but their energy isnt zero, either. as these quasi-virtual particles intereact, they will eventually sort themselves into those tho have passed on their momentum to other particles and can resume not-existing, or those who have gained enough energy to become real.

[Technical Ben]

There is also the effect of wheels. These do not expel reaction mass, but do drive along a road. Is it possible to use a similar effect with the "fields" in the likes of gravity or the virtual particle field? I'd doubt so, but am not able to explain why that would be.

[N_las]

But that's not what this drive is doing.

Virtual particle pairs can be assumed to already exist, because that's what they do- start and stop existing. Therefore you do not need to create them out of energy, only to push off them. Their net energy is no longer zero, but it's not their mass energy either- it's the equal and opposite of the energy gained by the ship pushing off them.

But that explanation is total nonsense. The only reason virtual particles can "start and stop existing" is because they "pay back their debt" and the net energy and momentum is conserved. If you wan't to push off virtual particles, you have to invest energy into them. You have to create real particles by spending energy. You are creating your own reaction mass.

That's what the articles mean by "quasi-virtual particles"- they're not "real" yet, but their energy isnt zero, either. as these quasi-virtual particles intereact, they will eventually sort themselves into those tho have passed on their momentum to other particles and can resume not-existing, or those who have gained enough energy to become real.

So they eventually sort themselfs into those who can resume not-existing and those wha have gained enough energy to become real. But were did this energy came from? From the drive. If there are real particles left, you had to create them (make them real) with the energy of your ship. So even according to that perspective you are creating your own reaction mass. Even if that explanation with virtual particles would work, it couldn't show a greater thrust to power ratio than a photon drive.

Edited February 17, 2015 by N_las

[B787_300]

Please note this is now one larger thread because someone posted a new topic. It is worth seeing what others have said before replying to this one.

[PB666]

But that explanation is total nonsense. The only reason virtual particles can "start and stop existing" is because they "pay back their debt" and the net energy and momentum is conserved. If you wan't to push off virtual particles, you have to invest energy into them. You have to create real particles by spending energy. You are creating your own reaction mass.

So they eventually sort themselfs into those who can resume not-existing and those wha have gained enough energy to become real. But were did this energy came from? From the drive. If there are real particles left, you had to create them (make them real) with the energy of your ship. So even according to that perspective you are creating your own reaction mass. Even if that explanation with virtual particles would work, it couldn't show a greater thrust to power ratio than a photon drive.

If I am not mistaken virtual particles do not need to perfectly conserve energy and momentum.

There is two things I could see here.

-If violations were common then heat insulation should not work, because virtual momentum and energy could be transferred across space indifferent to the intervening matter.

-Violations could be common but in 'local' space generally cancel out leaving no net momentum that can be transferred.

-Directional violations can occur, momentum and energy are transferred to somewhere else in virtual space.

Thing that only exist for Planck's time are never individually observable.

Thing that only travel for Planck's length can never demonstrate a displacement.

Therefore something may be accelerated for Planck's time at C/Planck's time and momentum and energy not detected individually.

The question is if these things are done directionally on mass scale would the net effects in our time and space be detectable.

For this to work there would need to be virtual cohesion (or alignment, or virtual creation resonance) between the particles that allow them to impart their forces collectively. In theory such a resonance could also create a collective virtual space where violations of momentum and energy return and are dealt with. Alternatively the resonance may already exist and the EM drive is synchronizing its resonance to exploit it.

Theoretically, at least, this could work, I would be dubious if it was infinitely scalable.

[Northstar1989]

In a certain way, you did bring your reaction mass with you. Imagine the following: You have a LED flashligh in space. It needs 5 Watt and produces light with 100% efficency. It has a battery which contains 5 Wh of energy. You are creating your own reaction mass in form of photons out of the energy in the battery. The Energie of a photon is E = h c/wavelength and the momentum of a photon is p = h/wavelength. If you do the calculation you get thrust force from the flashlight of Force = Power/c. That is the ultimative limit for efficentcy if you "produce" your own reaction mass.

Glad to see you're thinking about this carefully N_las, but a photon drive is NOT the ultimate in efficiency for creating your own reaction mass (out of energy).

Thrust = Mass Flow Rate * Exhaust Velocity

*AND*

E = 1/2 * m * v^2

Take a moment to think about these two equations. You'll *QUICKLY* realize that the higher your Exhaust Velocity, the lower your Thrust for the same Energy requirement. Since a photon has the HIGHEST POSSIBLE EXHAUST VELOCITY, it also has the LOWEST possible Thrust/MW.

A Cannae/EmDrive is not postulated to actually CREATE the virtual particles (they normally exist as part of the ebb and flow of the universe), but to give energy (momentum) to them. Even if they promptly disappeared again, if they existed long enough to exert a reaction force back on the drive, then it's essentially the same as having infinite reaction mass, and imparting energy to it to generate Thrust.

An interesting caveat follows- if the amount of power you pump into the Cannae/EmDrive did not lead to more virtual particles appearing (in actuality, it's theorized to do precisely* that), then you would get exponentially declining Thrust/MW due to the equations above. Luckily, pumping more energy into a space *IS* theorized to increase the generation of virtual particles- though once again unless using twice as much power led to twice as many virtual particles, I would expect Thrust/MW to decline (reaching an ultimate limit at the efficiency of a photon-drive) the higher the power-level you put into the system...

The efficiency of a Cannae/EmDrive should indeed reach a limit of that of a photon-drive, as some people have stated: but they are incorrect about the relationship- a Cannae/EmDrive should reach a MINIMUM efficiency equivalent to that of a photon-drive when it is accelerating the virtual particles to the speed of light. Any slower Exhaust Velocity than the speed of light, and you will get a higher Thrust/MW...

After 1 hour, the battery will be empty. The total energy content of your "ship" is now 5 Wh smaller than before. This has a direct effect on the restmass of the flashlight. With E=mc^2 you can calculate, that the flashlight has lost 2*10^-13 kg of mass. That is the mass of about 1.2*10^14 hydrogen atoms. You could say, that this was the reaction mass you brought with you. It ultimately doesn't matter if you bring energy or mass along, because they are kinda the same.

Indeed. E = m * c^2. Ultimately, with unlimited energy you have unlimited Thrust no matter how you look at it. Of course, you DON'T have unlimited energy...

If you don't bring a battery, but use solar power: congratulation! You could just as well have brought a mirror with you! Using solar cells to collect the energy of photons, converting this energy in photons again an "shining" them out trough the flashlight is exactly the same as just letting them bounce of a mirror.

You see, though, you can generate changes in momentum with mirrors in space. There are these things called Solar Sails- have you ever heard of them? Essentially, they work off changing the direction/energy of the Sun's light int order to change the direction/energy of the spacecraft. The momentum-change they impart is VERY slow, though...

A solar panel/ photon-drive combo may work better than a Solar Sail, though, because ultimately it's easier to control the direction you change your momentum in... You can point a photon-drive any direction you'd like, but a Solar Sail can only accelerate you in a combination of six directions: prograde, retrograde, normal, anti-normal, or radially outwards from the Sun (but NOT radially inwards)- all assuming a circular orbit around the Sun...

A photon-drive, on the other hand, could angle the solar panels such as to produce a prograde thrust from the radiation pressure of the Sun (by pointing the solar panels at a 45-degree angle to the Sun), then use energy collected and stored from the panels to exert a Thrust in the radially-inward direction. So, you have a bit more control over your trajectory-changes with a solar panel/ photon-drive combination than just using Solar Sails...

You could always use a nuclear reactor with a photon-drive (or better yet- a Cannae/EmDrive!), though, to operate without reflecting large amounts of light.

Regards,

Northstar

[N_las]

Glad to see you're thinking about this carefully N_las, but a photon drive is NOT the ultimate in efficiency for creating your own reaction mass (out of energy).

Thrust = Mass Flow Rate * Exhaust Velocity

*AND*

E = 1/2 * m * v^2

Take a moment to think about these two equations. You'll *QUICKLY* realize that the higher your Exhaust Velocity, the lower your Thrust for the same Energy requirement. Since a photon has the HIGHEST POSSIBLE EXHAUST VELOCITY, it also has the LOWEST possible Thrust/MW.

You haven't thought that trough. If you use a lower exhaust velocity (than c), you have to put more energy into it because of all the invariant mass you have to create.

A Cannae/EmDrive is not postulated to actually CREATE the virtual particles (they normally exist as part of the ebb and flow of the universe), but to give energy (momentum) to them.

To give momentum to those virtual particles you have to invest so much energy into them, that it is identical to you creating real particles out of energy. When I say "creating", I don't mean "creating virtual particles" but just "creating particles".

Think about hawking radiation: it is commonly explained with virtual particles in the proximity of the even horizont. You have a pair of virtual particles. One will fall into the event horizont, the other will become hawking radiation. But why? Why can't both particles just be hawking radiation. Because if you wan't virtual particles to actually do anything useful (like being hawking radiation), you have to invest energy into it. The energy-invest-mechanism at the event horizont is the virtual particle falling into the event horizont. The black hole loses the energy in the process in the exact same quantity as if it just would have created the hawking radiation out of energy.

The same is true for any other case: To actually interact with the virtual particle "in a meaningful way" (in this example to impart net momentum onto them) you will have to invest a big amout of energy, exactly as much as if you would create matter out of energy. Virtual particles "poping in and out of existens" doesn't mean you don't have to pay your bill at the end of the night.

The efficiency of a Cannae/EmDrive should indeed reach a limit of that of a photon-drive, as some people have stated: but they are incorrect about the relationship- a Cannae/EmDrive should reach a MINIMUM efficiency equivalent to that of a photon-drive when it is accelerating the virtual particles to the speed of light. Any slower Exhaust Velocity than the speed of light, and you will get a higher Thrust/MW...

No, if you do the actual math, the poton drive will give the upper efficency:

Energy you have to invest (into virtual particles) to give momentum to them: E = sqrt((m_rest c^2)^2 + (pc)^2)

To get the maximum momentum per invested energy you wan't to use particles without rest mass, hence photons.

Momentum per invested Energy is exactly the same as Force per invested Power.

Or, you could just use real particles in the first place. Since you don't have to pay the heavy bill of their rest mass, you will get amazing amounts of momentum per invested energy (hence the usefulness of ion drives in comparison to photon drives)

You see, though, you can generate changes in momentum with mirrors in space.

Yes, thats why I brought up the example. If you do the actual math you will see that the standart solar sail will be incredibly more powerful than the "solar power photon drive", because photovoltaics are so much heavier per area than a thin sail foil.

You could always use a nuclear reactor with a photon-drive (or better yet- a Cannae/EmDrive!), though, to operate without reflecting large amounts of light.

Yes, this is just another version of the battery example I stated earlier. So the conversation went from: "You could use an energy source on the ship" over "but an alternative would be to use solar power" back to "another alternative could be to use an energy source on the ship".

Edited February 18, 2015 by N_las

[Rakaydos]

When I say "creating", I don't mean "creating virtual particles" but just "creating particles".

Then please stop. Nobody's talking about creating particles. we're talking about using the ones that are already statistically there.

Think about hawking radiation: it is commonly explained with virtual particles in the proximity of the even horizont. You have a pair of virtual particles. One will fall into the event horizont, the other will become hawking radiation. But why? Why can't both particles just be hawking radiation. Because if you wan't virtual particles to actually do anything useful (like being hawking radiation), you have to invest energy into it. The energy-invest-mechanism at the event horizont is the virtual particle falling into the event horizont. The black hole loses the energy in the process in the exact same quantity as if it just would have created the hawking radiation out of energy.

There's not actually an energy transfer in the black hole. 99+% of the time, both particles fall in or both particles escape. But if the aintparticle falls in first, the black hole is some immesurably tiny amount smaller and the particle escapes easier- conversely, if the particle falls in first, the black hole is more massive and more likely to absorb the antiparticle, canceling out the virtual particle growth. This tiny imbalance in favor of particles escaping over antiparticles is what causes hawking radiation, not "energy transfer" past an event horizon.

[PB666]

An interesting caveat follows- if the amount of power you pump into the Cannae/EmDrive did not lead to more virtual particles appearing (in actuality, it's theorized to do precisely* that), then you would get exponentially declining Thrust/MW due to the equations above. Luckily, pumping more energy into a space *IS* theorized to increase the generation of virtual particles- though once again unless using twice as much power led to twice as many virtual particles, I would expect Thrust/MW to decline (reaching an ultimate limit at the efficiency of a photon-drive) the higher the power-level you put into the system...

The efficiency of a Cannae/EmDrive should indeed reach a limit of that of a photon-drive, as some people have stated: but they are incorrect about the relationship- a Cannae/EmDrive should reach a MINIMUM efficiency equivalent to that of a photon-drive when it is accelerating the virtual particles to the speed of light. Any slower Exhaust Velocity than the speed of light, and you will get a higher Thrust/MW...

First, I don't think you can induce an infinite number of virtual particles into spaces that will resonate with the system. I think the system becomes more efficient at corralling and accelerating particles that it can resonate with, that's all. Virtual particles according to uncertainty should come and go out of existence in complete spontaneity. Or lets just put it another way a 16W resonator is not likely to put enough strain on space time to alter the appearance of virtual particles, a quantum singularity that becomes unstable might

Second, the accelerations are not via a nozzle, the particles are being accelerated by some sort of inductive resonator but if they approach the speed of light, they will gain virtual mass and slow down, which can also be push against since the push is EM in nature the reaction force can act over space.

You see, though, you can generate changes in momentum with mirrors in space. There are these things called Solar Sails- have you ever heard of them? Essentially, they work off changing the direction/energy of the Sun's light int order to change the direction/energy of the spacecraft. The momentum-change they impart is VERY slow, though...

This is straying off the topic, bouncing photon's conserve most of their energy, although I did read somewhere that this does affect orbits and NASA takes this into consideration for plotting courses of interplanetary missions. The sun also produces a wind, which itself has momentum and charge this also can be used to accelerate.

A solar panel/ photon-drive combo may work better than a Solar Sail, though, because ultimately it's easier to control the direction you change your momentum in... You can point a photon-drive any direction you'd like, but a Solar Sail can only accelerate you in a combination of six directions: prograde, retrograde, normal, anti-normal, or radially outwards from the Sun (but NOT radially inwards)- all assuming a circular orbit around the Sun...

A solar panel takes ~40% of a photons energy and delivers it for whatever purpose, a mirror captures less than 5%. If a solar panel uses its energy efficiently in could convert up to 50% of the energy gained into acceleration of the space-craft. This means ~20% could end up in acceleration. As the other posters have stated, this is probably going to involve accelerating some mass (e.g. xenon).