Author Topic: EM Drive Developments Thread 1  (Read 782399 times)

Online aero

  • Senior Member
  • *****
  • Posts: 2778
  • 92129
  • Liked: 723
  • Likes Given: 246
Re: EM Drive Developments
« Reply #2300 on: 10/19/2014 08:16 PM »
Still toying with numbers. I divided the COP I posted previously by Q and got this:

                Experiment      COP          COP/Q
    "Shawyer (2008) a"       5,643     0.96
    "Shawyer (2008) b"      64,156     1.43
    "Juan (2012) TE011"    64,156     2.00
    "Juan (2012) TE012"    94,435     1.89
    "Brady et al. (2014) a"    1,618     0.22
    "Brady et al. (2014) b"    899         0.05
  "Brady et al. (2014) c"    6,388     0.29

I don't know that it has any meaning but now at least the number values are near the ideal photon rocket thrust. That is, the thrust of an ideal photon rocket using the stored energy of the cavity/second comes close to the experimentally derived thrusts.
Retired, working interesting problems

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6043
  • Likes Given: 5317
Re: EM Drive Developments
« Reply #2301 on: 10/19/2014 08:32 PM »
Still toying with numbers. I divided the COP I posted previously by Q and got this:

                Experiment      COP          COP/Q
    "Shawyer (2008) a"       5,643     0.96
    "Shawyer (2008) b"      64,156     1.43
    "Juan (2012) TE011"    64,156     2.00
    "Juan (2012) TE012"    94,435     1.89
    "Brady et al. (2014) a"    1,618     0.22
    "Brady et al. (2014) b"    899         0.05
  "Brady et al. (2014) c"    6,388     0.29

I don't know that it has any meaning but now at least the number values are near the ideal photon rocket thrust. That is, the thrust of an ideal photon rocket using the stored energy of the cavity/second comes close to the experimentally derived thrusts.

Taking into account the frequency drift and bandwidth issues that the researchers have tuning the device under resonance, with concomitant drift in Q (which therefore cannot be a constant during the measurements) these results are quite interesting!

Perhaps we need some more frobnicating , including  re-running the formulas with the updated geometry (see http://physicsfromtheedge.blogspot.it/2014/10/emdrive-mihsc-dream-of-horizon-physics.html) and allowing for Sqrt[ 1/big^2 + 1/L^2] and perhaps all the other square roots as well.   Also using the new formulas taking into account wavelength:  http://forum.nasaspaceflight.com/index.php?topic=29276.msg1272925#msg1272925


Also frobnicating with the photon rocket analogy...
« Last Edit: 10/19/2014 08:37 PM by Rodal »

Offline Notsosureofit

  • Full Member
  • ****
  • Posts: 656
  • Liked: 704
  • Likes Given: 1367
Re: EM Drive Developments
« Reply #2302 on: 10/19/2014 09:43 PM »
Couple things:
Won't get back to the lab til monday for measurements from picture
Both end plates have something else going on.  Looks likr the top one has a tuning plate(?) of some kind, not motorized
Why the extra plate below the bottom, can't see in pic

Found another paper that was bothering me in the piles behind the desk
"Ionization instabilities and resonant acoustic modes", Physics of Plasmas, V8, N0.11, p.5018

It was concerned w/ the coupling of ions w/ dust particles.  Reminds me of RF w/ axions  (Ya, I'm still chasing the axion connection) no cavities involved, but "It is found that an unstable dust-acoustic mode of nonzero real frequency can be generated via a resonance phenomenon."  ... "As the charge on dust particles exceeds a threshold, multiple low-frequency modes with large growth rates are excited suddenly."

I just had to throw that in cuz I finally found the d**m thing !  (You youngsters can play w/ it for now)

Ok, so in practical terms it means you can (theoretically)induce feedback into the coupling constant if you can set up the dispersion relations properly.

Got a couple minutes to myself, so I thought I'd throw out a bit of rationalization as to why I wanted to find the above mentioned paper (there is probably much better stuff out there for them what's got access)

Anywho, those trying to detect axions look at the interaction as a 3 legged Feynman loop in the presence of a magnetic field, etc.

I think of an interaction between the energy of the cavity photons and momentum transfer to some barely interacting fluid (?, dark matter, ether, etc) as a 4 legged Feynman loop.  It occurred to me that I had seen something analogous to that before, hence the referenced paper (once I found it)

An analogy only, of course, but expressed as a resonant interaction between weakly interacting fields.

Now, the dispersion relation for the axions (?, etc) is unknown(?) but the relation for cavity photons is at least suseptable to some control, for instance, by changing the shape of the cavity, I think.

One thing, the hint is there that the cavity might be exhibiting parametric conversion of some of the photon energy under these circumstances, but there was no effort to look for the presence of other frequencies as far as I know.

Still a long way from numerical calculations.  Thanks for your patience.


Online aero

  • Senior Member
  • *****
  • Posts: 2778
  • 92129
  • Liked: 723
  • Likes Given: 246
Re: EM Drive Developments
« Reply #2303 on: 10/19/2014 09:53 PM »
OK - I took out all possible sources of loss and inefficiency by simply replacing the actual measured trust with Prof. M's new formula predicted values.

                                                            Formula thrust
    Experiment                  COP    COP/Q   COP/Q
  "Shawyer (2008) a"    5,643     0.96     0.44
  "Shawyer (2008) b"    64,156     1.43     0.82
  "Juan (2012) TE011"    64,156     2.00     0.82
  "Juan (2012) TE012"    94,435     1.89     0.82
  "Brady et al. (2014) a"    1,618     0.22     0.46
  "Brady et al. (2014) b"    899     0.05     0.46
  "Brady et al. (2014) c"    6,388     0.29     0.46

Now the values are all geometry because Q is multiplied the formula and divided out.  The values though are getting into the range of what a COP based on an ideal photon rocket should be. I could factor cosine losses into the thrust formula but that's a lot of trouble for little benefit.

This formula says to me that what is wanted is a big cavity with the ratio (w_big/w_small) large, and height/length, (s) large. Height/length, (s) large is the new factor in Prof. M's latest formula and s is the variable name he uses.

 It seems that current tech could deal with that but of course the larger the cavity is, the heavier it is so the T/W ratio suffers at some point.
Retired, working interesting problems

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6043
  • Likes Given: 5317
Re: EM Drive Developments
« Reply #2304 on: 10/19/2014 09:57 PM »
OK - I took out all possible sources of loss and inefficiency by simply replacing the actual measured trust with Prof. M's new formula predicted values.

                                                            Formula thrust
    Experiment                  COP    COP/Q   COP/Q
  "Shawyer (2008) a"    5,643     0.96     0.44
  "Shawyer (2008) b"    64,156     1.43     0.82
  "Juan (2012) TE011"    64,156     2.00     0.82
  "Juan (2012) TE012"    94,435     1.89     0.82
  "Brady et al. (2014) a"    1,618     0.22     0.46
  "Brady et al. (2014) b"    899     0.05     0.46
  "Brady et al. (2014) c"    6,388     0.29     0.46

Now the values are all geometry because Q is multiplied the formula and divided out.  The values though are getting into the range of what a COP based on an ideal photon rocket should be. I could factor cosine losses into the thrust formula but that's a lot of trouble for little benefit.

This formula says to me that what is wanted is a big cavity with the ratio (w_big/w_small) large, and height/length, (s) large. Height/length, (s) large is the new factor in Prof. M's latest formula and s is the variable name he uses.

 It seems that current tech could deal with that but of course the larger the cavity is, the heavier it is so the T/W ratio suffers at some point.

Very interesting.   On this topic, have people commented on this paper

Photonic Laser Propulsion (PLP): Photon Propulsion Using an Active Resonant Optical Cavity ?

http://arc.aiaa.org/doi/abs/10.2514/6.2007-6131

He claims to have measured 35 microNewtons thrust at 1.7 watts with propellant less photon rocket amplification in an optical cavity

This is a much higher thrust/power input than normal photon rockets and even superior to NASA Eagleworks truncated cone experiments



« Last Edit: 10/19/2014 10:00 PM by Rodal »

Offline frobnicat

  • Full Member
  • ****
  • Posts: 518
  • Liked: 500
  • Likes Given: 151
Re: EM Drive Developments
« Reply #2305 on: 10/19/2014 10:10 PM »
Still toying with numbers. I divided the COP I posted previously by Q and got this:

                Experiment      COP          COP/Q
    "Shawyer (2008) a"       5,643     0.96
    "Shawyer (2008) b"      64,156     1.43
    "Juan (2012) TE011"    64,156     2.00
    "Juan (2012) TE012"    94,435     1.89
    "Brady et al. (2014) a"    1,618     0.22
    "Brady et al. (2014) b"    899         0.05
  "Brady et al. (2014) c"    6,388     0.29

I don't know that it has any meaning but now at least the number values are near the ideal photon rocket thrust. That is, the thrust of an ideal photon rocket using the stored energy of the cavity/second comes close to the experimentally derived thrusts.

Yes, from my automated search most formulas that fit data without big constant magnitude fudge factor show  F/P = 1/c * Q * geometric_factor. In the geometric factor there is the wavelength (of the driving frequency), the small  end and big end size. All those sizes are in the same ballpark and combined in the geometric_factor which is dimensionless => geometric_factor is around one unit. That is geometric_factor=COP/Q, as what you define as COP=(claimed thrust/power)/(perfect photon rocket/power) = (1/c*Q*geometric_factor)/(1/c) = Q*geometric_factor.

If we discard the outlier Brady b, and ignore the geometric_factor for the moment (while data shows a statistically strong linear relation F/P = 1/c Q times something, this looks more fragile when it comes to pinning down the exact expression of this something in function of remaining parameters)... well, so , if we consider as a simplification geometric_factor=1, that is something that pushes simply as F/P = 1/c * Q  then it is like the force between two plates defining a cavity where photons bounce back and forth an average Q number of times before being lost or absorbed.

Note that it would be the force between two plates, that is to get useful work (not more than what was invested for photons, but better than for a photon rocket) from this "photon pressure buildup" you have to get a fixed plate and a moving plate. If the two plates are comoving because they are part of the same rigid object then it is useless. It's like pressure : more pressure gives more acceleration to the two parts of a piston, but more pressure in a rigid device (no parts moving relative to each others) gives nothing useful in term of displacements. If there is a leak in the walls of a rigid device then more pressure means more thrust, but this is simply letting some more momentum escape, more action => more reaction : this is a classical rocket open system, and the efficiency is given by the speed of ejection. If photons escape at a certain rate, then we are losing energy at a certain rate, that is we have to "lose" a certain power, and the thrust/power is again 1/c, regardless of photons coming from a highly resonant cavity or from a mere tungsten filament.

For the variable length high Q "cavity" space applications : some musings by me.

More later as per Rodal answer to this nifty toy table : we must seriously address the question of the role of resonance and Q in a cavity, and first I feel an urgent need to clarify my (and maybe the general audience's) understanding for those mundane mechanisms in usual classical setting, before tuning for more exotic storylines.

Offline JohnFornaro

  • Not an expert
  • Senior Member
  • *****
  • Posts: 9162
  • Delta-t is the salient metric.
  • Planet Eaarth
    • Design / Program Associates
  • Liked: 610
  • Likes Given: 314
Re: EM Drive Developments
« Reply #2306 on: 10/19/2014 10:17 PM »
Quote from: Steve Kelsey
Please forgive an idle speculation on a Sunday evening from the peanut gallery.
Quote

Thanks.  I was really gettin'  kinda full, all by myself in that there gallery.
Sometimes I just flat out don't get it.

Online aero

  • Senior Member
  • *****
  • Posts: 2778
  • 92129
  • Liked: 723
  • Likes Given: 246
Re: EM Drive Developments
« Reply #2307 on: 10/19/2014 10:22 PM »
@frobnicat

Note that Prof. M's revised thrust formula,

F = PQs/c * ((1/w_big)-(1/w_small))     (2),

he has replaced the frequency term with c/s, where he is using the length of the cavity as s. That of course predicts a higher value for thrust. But with the revised cavity dimensions, he needed a larger predicted value to come close to the experimental results.

And we do need better estimates of Brady's cavity dimensions as a minimum.

aero.
Retired, working interesting problems

Offline IslandPlaya

  • Full Member
  • ****
  • Posts: 582
  • Outer Hebrides
  • Liked: 163
  • Likes Given: 166
Re: EM Drive Developments
« Reply #2308 on: 10/19/2014 10:23 PM »
Quote
we must seriously address the question of the role of resonance and Q in a cavity
Indeed!
We need microwave sources that are stable < 0.1Hz
and tuneable.

Online aero

  • Senior Member
  • *****
  • Posts: 2778
  • 92129
  • Liked: 723
  • Likes Given: 246
Re: EM Drive Developments
« Reply #2309 on: 10/19/2014 10:43 PM »
OK - I took out all possible sources of loss and inefficiency by simply replacing the actual measured trust with Prof. M's new formula predicted values.

                                                            Formula thrust
    Experiment                  COP    COP/Q   COP/Q
  "Shawyer (2008) a"    5,643     0.96     0.44
  "Shawyer (2008) b"    64,156     1.43     0.82
  "Juan (2012) TE011"    64,156     2.00     0.82
  "Juan (2012) TE012"    94,435     1.89     0.82
  "Brady et al. (2014) a"    1,618     0.22     0.46
  "Brady et al. (2014) b"    899     0.05     0.46
  "Brady et al. (2014) c"    6,388     0.29     0.46

Now the values are all geometry because Q is multiplied the formula and divided out.  The values though are getting into the range of what a COP based on an ideal photon rocket should be. I could factor cosine losses into the thrust formula but that's a lot of trouble for little benefit.

This formula says to me that what is wanted is a big cavity with the ratio (w_big/w_small) large, and height/length, (s) large. Height/length, (s) large is the new factor in Prof. M's latest formula and s is the variable name he uses.

 It seems that current tech could deal with that but of course the larger the cavity is, the heavier it is so the T/W ratio suffers at some point.

Very interesting.   On this topic, have people commented on this paper

Photonic Laser Propulsion (PLP): Photon Propulsion Using an Active Resonant Optical Cavity ?

http://arc.aiaa.org/doi/abs/10.2514/6.2007-6131

He claims to have measured 35 microNewtons thrust at 1.7 watts with propellant less photon rocket amplification in an optical cavity

This is a much higher thrust/power input than normal photon rockets and even superior to NASA Eagleworks truncated cone experiments

Does anyone have access to this full paper?

From the abstract, It looks like his resonant cavity has a Q of 3000. At 35 muN thrust, 1.7 watts, and frobnicat's simplified rule of thumb,  F/P = 1/c * Q predicts F = 17. muN . Only missing a factor of 2.

I'd say that's close enough to add a constant factor of 2 as a possible choice in your formula search. Have we exceeded a million equation search yet?
Retired, working interesting problems

Offline IslandPlaya

  • Full Member
  • ****
  • Posts: 582
  • Outer Hebrides
  • Liked: 163
  • Likes Given: 166
Re: EM Drive Developments
« Reply #2310 on: 10/19/2014 10:48 PM »
Respect to all the theory.
Shirley you guys can come up with an experiment to determine reality?

Offline IslandPlaya

  • Full Member
  • ****
  • Posts: 582
  • Outer Hebrides
  • Liked: 163
  • Likes Given: 166
Re: EM Drive Developments
« Reply #2311 on: 10/19/2014 10:50 PM »
I have my sheets of copper and a dismantled microwave oven at the ready!

Offline frobnicat

  • Full Member
  • ****
  • Posts: 518
  • Liked: 500
  • Likes Given: 151
Re: EM Drive Developments
« Reply #2312 on: 10/19/2014 10:53 PM »
Taking into account the frequency drift and bandwidth issues that the researchers have tuning the device under resonance, with concomitant drift in Q (which therefore cannot be a constant during the measurements) these results are quite interesting!
More on that later please, as relation of resonance with Q not clear to me (higher Q => narrower bandwidth not the point, that's perfectly clear). Needs time to think of a clear enough way to express what's not clear, that's difficult.

Quote
Perhaps we need some more frobnicating , including  re-running the formulas with the updated geometry (see http://physicsfromtheedge.blogspot.it/2014/10/emdrive-mihsc-dream-of-horizon-physics.html) and allowing for Sqrt[ 1/big^2 + 1/L^2] and perhaps all the other square roots as well.   Also using the new formulas taking into account wavelength:  http://forum.nasaspaceflight.com/index.php?topic=29276.msg1272925#msg1272925


Also frobnicating with the photon rocket analogy...

You know the definition of what it means to frobnicate, don't you ? Been warned : don't expect too much of frobnicating, or of frobnicat for that matter.

I can release source code for anyone with a standard C compiler at hand, but there also be warned than in its present version it is risky (risky like dancing with a werewolf atop a cliff) to make modifications on it, unless you are me (are you ?) or a veteran programmer with quite a few hours to kill. Why I haven't proposed that before. Time permitting, a refactoring could help make it more useable.

Easiest way would be someone fills in data in format like that

#define Nrec 7
t_data data_in[Nrec] =
{
    //                                 w_big  w_small  lambda      Q     power   force
    {"Shawyer (2008) a",        1.0 ,  16    ,  8    , C/2.45  ,  5900 ,  850   , 16       },
    {"Shawyer (2008) b",        1.0 ,  28    ,  4    , C/2.45  , 45000 , 1000   , 214      },
    {"Juan (2012) TE011",       1.0 ,  28    ,  4    , C/2.5   , 32000 , 1000   , 214      },
    {"Juan (2012) TE012",       1.0 ,  28    ,  4    , C/2.45  , 50000 , 1000   , 315      },
    {"Brady et al. (2014) a",   1.0 ,  24.75 , 16.5  , C/1.933 ,  7320 ,   16.9 ,   0.0912 },
    {"Brady et al. (2014) b",   1.0 ,  24.75 , 16.5  , C/1.937 , 18100 ,   16.7 ,   0.0501 },
    {"Brady et al. (2014) c",   1.0 ,  24.75 , 16.5  , C/1.88  , 22000 ,    2.6 ,   0.0554 },
};


and specify the units of other fields (say, introducing a new column length => that's centimetres  :D)
The column full of ones are weights for the least square sum, but it is not actually used yet (could be in the future)
Oh, and remember I'm French, I don't know what is an inch or a foot.
« Last Edit: 10/19/2014 10:55 PM by frobnicat »

Offline IslandPlaya

  • Full Member
  • ****
  • Posts: 582
  • Outer Hebrides
  • Liked: 163
  • Likes Given: 166
Re: EM Drive Developments
« Reply #2313 on: 10/19/2014 10:55 PM »
Quick run with the magnetron:
Need to take the cat to the vet.
No anomalous force detected.
Minor fires extinguished.

Offline IslandPlaya

  • Full Member
  • ****
  • Posts: 582
  • Outer Hebrides
  • Liked: 163
  • Likes Given: 166
Re: EM Drive Developments
« Reply #2314 on: 10/19/2014 10:59 PM »
The bloody thing you get out of a microwave oven is a dangerous bloody thing.
Please, people don't fool around with these things.

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6043
  • Likes Given: 5317
Re: EM Drive Developments
« Reply #2315 on: 10/19/2014 11:13 PM »
I have my sheets of copper and a dismantled microwave oven at the ready!
All kidding aside, it is important that the inner cavity walls made out of a shiny highly conductive material like copper, in order to minimize losses and maximize Q.  The inner surface should be smooth copper like a mirror.
« Last Edit: 10/19/2014 11:21 PM by Rodal »

Offline frobnicat

  • Full Member
  • ****
  • Posts: 518
  • Liked: 500
  • Likes Given: 151
Re: EM Drive Developments
« Reply #2316 on: 10/19/2014 11:53 PM »

Very interesting.   On this topic, have people commented on this paper

Photonic Laser Propulsion (PLP): Photon Propulsion Using an Active Resonant Optical Cavity ?

http://arc.aiaa.org/doi/abs/10.2514/6.2007-6131

He claims to have measured 35 microNewtons thrust at 1.7 watts with propellant less photon rocket amplification in an optical cavity

This is a much higher thrust/power input than normal photon rockets and even superior to NASA Eagleworks truncated cone experiments

Does anyone have access to this full paper?

From the abstract, It looks like his resonant cavity has a Q of 3000. At 35 muN thrust, 1.7 watts, and frobnicat's simplified rule of thumb,  F/P = 1/c * Q predicts F = 17. muN . Only missing a factor of 2.

I'd say that's close enough to add a constant factor of 2 as a possible choice in your formula search. Have we exceeded a million equation search yet?

Yeah, I probably messed up with a factor of 2 in my previous post. This is just, you know, frobnicating. Seriously : a bounce gives twice the momentum of the photon so, here is the 2 factor probably. Now, if someone comes with another 2 factor (4 times what I told you so learnedly) I could still argue the subtle nuance of the chances being absorbed on one single pass or on a back and forth.

Anyway, this paper is, like, exactly what I said  ;D
From the summary (don't have access to full paper, would be interested) :
<<...trapping or bouncing photons between two high reflectance mirrors located separately in spacecraft platforms ... the laser cavity is directly formed between a pair of space platforms. >>

Sorry this is not exciting fringe science, this is action/reaction all the most classical. It is interesting, it deserves consideration in advanced concepts, but this is not about a box "self accelerating" in deep space. This is two spaceships pushing on each others at some distance, gaining more distance, and more relative speed. This is a cavity with two plates that are not part of one single rigid body. No fancy theory needed, no problem with momentum conservation, no problemo in using the same photon's momentum 3000 times, no problem with energy conservation. As the relative speed increases, the photons get more and more redshifted at each bounce until it gets significant. So expect a decreasing efficiency as approaching c/3000 = 100 km/s relative speeds. If the "amplification factor" of 3000 were to be raised to 3000000 then the efficiency would start to suffer at 100m/s relative speeds... (give or take a factor 2  ::)

This is not photon rocket or a "drive", this is a pusher beam. Could be applicable to beamed propulsion. Would love to know the details in the full paper : would work up to what distances ? Probably bounded by diffraction limit.

One relevant aspect though is that the cavity "becomes robust" because the "optical gain medium is within the cavity". This was advanced by someone here (don't recall who and when) that it would be a good idea to use the cavity itself as part of the microwave generator to avoid problems of fine tuning (for EMdrives designs).

<<the reason for this (resonance robustness) is that the optical gain medium is within the optical cavity, and the cavity is operated in multifrequency oscillation.>>
multifrequency oscillation : what's that ??

Offline frobnicat

  • Full Member
  • ****
  • Posts: 518
  • Liked: 500
  • Likes Given: 151
Re: EM Drive Developments
« Reply #2317 on: 10/20/2014 12:19 AM »
Have we exceeded a million equation search yet?

No but 21769 is already quite a lot. The data set is sparse and with some uncertainties : the risk is overfitting. Scanning on more than 15 bits (32768 combinations) worth of explanation could easily bring up more perfect formula for the specific available data but with less generalisation power : worse at predicting next data points to come. Need more data points before it's worth looking at much more formulas. Not a problem of computing power, not before reaching many billions of formulas.

At this stage with 7 data points, from a "phenomenological theoretically agnostic" point of view, simpler is better, and there is not that much simple equations.

Note : the number of combinations of exponents and added terms were 94 millions but of those only 21769 unique representations (discarding equivalents) made sense in dimensional analysis (kg m s).



Online aero

  • Senior Member
  • *****
  • Posts: 2778
  • 92129
  • Liked: 723
  • Likes Given: 246
Re: EM Drive Developments
« Reply #2318 on: 10/20/2014 12:43 AM »
Ok, that's good. Basically anything sensible we come up with can be computed then for this small data set.

Yes. I looked at some articles on PSP, its just another way of doing beamed propulsion. Since the photons within the resonance cavity (between the mirrors) actually are pushing, all of them count to give thrust. In the EM drive case, it seems that all of the photons are counting to give thrust but they are not leaving the cavity of the engine. I guess that is the trick and the problem. How can the momentum of the photons in the cavity be separated from the photons themselves? Momentum departs, photons remain.
Retired, working interesting problems

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5895
  • USA
  • Liked: 6043
  • Likes Given: 5317
Re: EM Drive Developments
« Reply #2319 on: 10/20/2014 01:12 AM »
Have we exceeded a million equation search yet?

No but 21769 is already quite a lot. The data set is sparse and with some uncertainties : the risk is overfitting. Scanning on more than 15 bits (32768 combinations) worth of explanation could easily bring up more perfect formula for the specific available data but with less generalisation power : worse at predicting next data points to come. Need more data points before it's worth looking at much more formulas. Not a problem of computing power, not before reaching many billions of formulas.

At this stage with 7 data points, from a "phenomenological theoretically agnostic" point of view, simpler is better, and there is not that much simple equations.

Note : the number of combinations of exponents and added terms were 94 millions but of those only 21769 unique representations (discarding equivalents) made sense in dimensional analysis (kg m s).

Not quite there yet.  The theoretical geometrical variable for a resonant cavity box is Sqrt[a^-2+L^-2]  and we didn't have square roots of additions allowed, is that right?

Moreover we had a number of formulas very close: [a^-2+L^-2]  and [a^-1+L^-1] 
So the data was telling us we need to allow Sqrt[a^-2+L^-2]

We had these square roots but they are missing the plus sign:

a^-2 b^-2 L^2  Q^1  P^1  F^-1 c^-1 sqrt(a^-2 L^-2)^-1   1.02   0.58
a^-1 b^-2 L^2  Q^1  P^1  F^-1 c^-1 sqrt(a^-1 L^-1)^-1   1.32   0.58
« Last Edit: 10/20/2014 01:18 AM by Rodal »

Tags: