This is interesting (and possibly dismaying for solar sail fans). Thomas Gold says that they won't work, at least not the way people have had in mind.

Basically, he contends that none of the analyses of their performance has taken into account thermodynamics--that they only consider the momentum change of the photon. The sails are built to be as reflective as possible to minimize heating of the fragile sail fabric. But according to Gold, if no heat is absorbed (i.e., no temperature change in the photon), the Carnot's Rule says that no energy can be extracted.

It's been too long since my thermo courses to know if he's right or not, but he's not obviously wrong. Considering that people have been thinking about this for decades, I find it a little surprising that the physics remains unsettled. We'll find out when the first demo sails are flown shortly, and we'll see if the performance matches predictions.

[Update at 11:26 AM PDT]

There's been a lot of discussion in the comments section, but considering the source (Henry Spencer, over at sci.space.*), I'll consider this the last word for now.

A *moving* perfect mirror *does* reduce the "temperature" of photons reflected from it -- by Doppler shift! Where does the energy lost in Doppler shift go? Into added kinetic energy of the mirror.

(If the mirror is held stationary -- relative to the observer who is measuring the details -- by some means, then there can be no Doppler shift. But there is also no work done on the mirror, since work is thrust times *distance*, and hence there is no added kinetic energy.)

Yes, Doppler shift at ordinary velocities is pretty damn small. But so is the acceleration produced by light pressure.

Gold appears to be unaware that the physics of light pressure are well understood and have been demonstrated many times -- in the laboratory, in precision tracking of spacecraft, and in attitude control of spacecraft. A particularly glaring example is Radarsat 1, which is in a dawn-dusk sun-synchronous orbit (i.e. essentially continuous sunlight) and flies with an essentially constant attitude. Its designers overlooked solar-sail effects on its big solar arrays and radar antenna, which are slightly tilted with respect to the Sun for engineering reasons. Turns out that nearly 2/3 of Radarsat's stationkeeping fuel goes to fight light-pressure drag -- it's trying to sail down into the atmosphere. See "Radarsat Time Rate of Mean Semi-Major Axis Due to Drag", by Said R. Marandi, in the AAS/GSFC 13th International Symposium on Space-Flight Dynamics, 1998.

Note that the experts consulted for the article were a thermodynamicist and an astronomer, neither of them a physicist. (Citing the Crookes radiometer is just plain embarrassing -- it turns by thermal effects, not by light pressure.)

Well, certainly in theory both thermodynamicists and astronomers are supposed to have a good grounding in physics (and arguably, the former is a specialized form of physicist), but other than that bit of ad hominem, I've nothing to dispute.

[One more a few minutes later]

OK, Geoff Landis has also weighed in.

Unfortunately, Gold has apparently forgotten to account for a well-known physical effect: the Doppler shift.

It's worth saying that the photon pressure on a spacecraft is not theoretical; its effect on spacecraft is measurable, and it has been observed and measured to great precision routinely in space. Photon pressure-- the solar sail effect-- has already been used for an operational space mission; it was for spacecraft attitude control on the Pioneer Venus-Mercury mission.

The Crookes radiometer does not operate on photon pressure, and the explanation for how it operates has been known for over a century.

The energy transfer to a solar sail can be accounted for from the Doppler shift of reflected photons; even when the reflectivity is 100%, a photon looses [sic] energy when reflecting from a moving sail. This effect exactly corresponds to the energy increase of the sail. No sophisticated physics is needed to analyze this effect, it is a problem suitable for a homework assignment for a college undergraduate.

When the sail is moving, then the reflected photons are Doppler shifted, and leave the sail with lower energy than they arrived. This loss of energy exactly equals the energy imparted to the sail, a fact which can be trivially verified by using Newton's laws, the Doppler formula, and the Einstein equation for photon momentum p=E/c.

If the sail is not moving, there is no Doppler shift. However, note that since energy is proportional to momentum squared, the derivative of energy with respect to momentum is zero for a non-moving sail. Thus, when the sail is stationary, it can reflect photons with perfect efficiency and still gain momentum at no energy cost.

For completeness, note that if the sail is moving *toward* the light source, then the phtons [sic] are Doppler shifted to *higher* energy by the reflection. This implies that the sail must lose energy-- which is correct; when the sail moves toward the light source, it slows down.

The photon suffers a loss of energy, and is thus reduced in frequency (red shifted). In effect it is "cooled". Thermodynamics and Conservation of Momentum are both obeyed.
Gold has lost it.

Yes, that's what I'd always thought, except that Planck's Law was a frequency shift attributed to a change in momentum, not energy. I'd never actually considered the energy, per se, just assuming that it all came out in the wash.

If solar sails can't work, I wonder how radiometers work? http://www.physics.purdue.edu/demo/3D/radiometer.html

Simple calculations show that the photon loses a small amount of energy to the sail, leading to a decrease in frequency of the photon. This also results in a lower total momentum for the photon after the interaction. The decrease in frequency exactly compensates for both the energy and momentum transferred to the sail. This is a very simple scattering problem, and it's frankly amazing that a physicist of Gold's caliber could make such a simple error.

Andy, the article points out that the direction a radiometer spins supports Gold.

According to the interview, Gold only has an issue with a sail that acts as a 'perfect mirror'. He readily admits that a comet's tail shows a momentum change from the photons, as will other non-perfect reflecting surfaces.

If the Sail is indeed a 'perfect mirror', then there should be no frequency shift of the photons. His point is still valid.

-S

Even with a perfect reflection, there will still be a frequency shift due to momentum change (changing frequency a la Planck is the only way to change the momentum of a photon, since its speed can't change, and it has no rest mass).

JPL has done the test, at least for their material of choice. It worked.

I don't follow Gold's argument; how can lack of temperature change invalidate conservation of momentum and energy?

Troy, as I read that test, it indicated that they could indeed put pressure on the sail, but it wasn't clear from the story whether or not it matched up quantitatively to theoretical predictions.

One would presume so, though, because they would have had to figure out why if it hadn't.

I must really still be back in the buggy-whip days on this technology (what else is new?), because I thought the solar wind also includes heavier particles that can impart momentum in a way that even us buggy-whippers can wrap our minds around.

I thought that would be how a solar sail would really work; didn't know the bulk of the energy (pardon the expression) was supposed to come from photons.

No, the solar wind is almost irrelevant to a light sail, and in fact degrades its performance over the long term by putting holes in it. It's propelled by photons.

That was my take, Rand -- if the result was wildly out-of-whack with predictions, they'd not be likely to stick their necks out by announcing a successful test. Of course, I haven't seen the results in a peer-reviewed journal, but then I don't get all the relevant journals.

I recall the Echo satellites (huge aluminized Mylar balloons) were pushed around quite a bit by solar radiation; I was under the impression that this sort of change was routinely predicted for lightweight satellites these days -- I've seen published notices that such-and-such was expected to move due to solar activity, even. Solar wind forces on these objects are orders of magnitude lower than solar radiation forces, so what else could be causing it?

Upon reflection, I believe solar sails will work.

Carnot law does not say that the temperature must go down..it says that the entropy of the universe must go up. For thermal systems, there is a temperature driven maximum efficiency. For non thermal systems, there is no such limitation, hence fuel cells, which use gibbs free energy to calculate efficiency versus carnot efficiency, have very high efficiency rates. Photons reflected from a solar cell (due to compton scattering) are at a lower energy ... Energy of photons e=hv (where h is planks constant and v is frequency) and E=mc2 (where E is energy, c is speed of light and m is mass, photons have relativistic mass = hv/c2 since momentum p is = mass x speed , thus hv=pc..h and c are constants, so if p goes down (loss of momentum, and momentum is always conserved) v (frequency) must go down, and energy was hv, thus energy went down. Since energy was transfered, the future usefulness of the reflectant photons goes down, and thus entropy of the universe goes up. The photons cannot possibly transfer more momentum than the differnce between current momentum (which was reduced by the solar cell) and the momentum due to brownian motion of the interstellar hydrogen (about 3K)with which they must be discharged. The crookes radiometer is consistent, since the mirrored surface encounters much less compton scattering it reflects most photons at a higher energy (thus on that side less momentum is transfered) than the black side which absorbs the high energy photons (visible light) and emmits lower energy photons at near infrared (it will heat up more). Thus the black side is absorbing more energy, and hence transfering more momentum, so the net moment arm pushes forward on the black side.

Actually, that's not right. The Crooke radiometer is a heat engine, and doesn't work directly on a photonic propulsion principle. If you totally evacuate it, in fact, it doesn't work. It's propelled by heated air on the black side.

"Thus, when the sail is stationary, it can reflect photons with perfect efficiency and still gain momentum at no energy cost."

Huh? How do you gain momentum while remaining stationary?

Is Thomas Gold running for something? They say any media is good media, but good heavens. If he wants name recognition he's got it now.

This is the most absurd thing I've seen since Bill Clinton talked about the shuttle Columbia on Larry King.

Sean Kiely: Upon reflection, I believe solar sails will work.

<groan!> (Never let it be said I didn't show proper appreciation for a good bad pun.

J. Bowen: "Thus, when the sail is stationary, it can reflect photons with perfect efficiency and still gain momentum at no energy cost."

Huh? How do you gain momentum while remaining stationary?

Doesn't say it "remains" stationary. Says it "is" stationary, which I read to mean, before it starts moving.

Here's what I'm missing: If, at V-0 the photon emparts no energy on the sail, How does the spacecraft accelerate from adead stop?

I still don't understand how a doppler shift would occur if the sail is at rest.