Author Topic: EM Drive Developments - related to space flight applications - Thread 3  (Read 3131023 times)

Offline TheTraveller

http://physics.stackexchange.com/questions/190018/does-radiation-force-depend-on-group-velocity-or-on-phase-velocity

Is that all you have to prove momentum moves at Phase Velocity inside a waveguide?

The Phase Velocity in a waveguide is greater than c and nothing moves at greater than c.

Here is what a microwave engineers web site has to say about Phase Velocity inside a waveguide.
http://www.microwaves101.com/encyclopedias/waveguide-mathematics

Quote
Phase velocity is an almost useless piece of information you'll find in waveguide mathematics; here you multiply frequency times guide wavelength, and come up with a number that exceeds the speed of light!

Be assured that the energy in your wave is not exceeding the speed of light, because it travels at what is called the group velocity of the waveguide:


Cullen, Shawyer and Prof Yang are correct that the Force generated on the end plates as the EM wave bounces off them is from momentum that travers, inside a waveguide, at Group Velocity. The Group Velocity and thus the Force generated from the EM waves bounce at each end varies as to the EM waves Group Velocity at each end varies.
« Last Edit: 07/14/2015 01:14 pm by TheTraveller »
It Is Time For The EmDrive To Come Out Of The Shadows

Offline rfmwguy

  • EmDrive Builder (retired)
  • Senior Member
  • *****
  • Posts: 2205
  • Liked: 2713
  • Likes Given: 1134
Congrats New Horizons team, very impressed...the best media question I heard was "when are we going back?" The answer was interesting, there are designs being worked on.

IMHO, the key for planetary science and space exploration success is becoming clear. It is not more sensitive  instrumentation, it is not lower cost hardware...it is faster propulsion. This is your challenge scientific community...to shrink the fabric of space and time with modern propulsion research.

To the propulsion industry...think outside the box. Your equations have shown us that propellants will never get us to where we want to go...faster and farther. Perhaps in this forum, one of us may help show you the way.

Offline TheTraveller

Congrats New Horizons team, very impressed...the best media question I heard was "when are we going back?" The answer was interesting, there are designs being worked on.

IMHO, the key for planetary science and space exploration success is becoming clear. It is not more sensitive  instrumentation, it is not lower cost hardware...it is faster propulsion. This is your challenge scientific community...to shrink the fabric of space and time with modern propulsion research.

To the propulsion industry...think outside the box. Your equations have shown us that propellants will never get us to where we want to go...faster and farther. Perhaps in this forum, one of us may help show you the way.

The 20 x 4N/kW EMDrive powered IXS Clark could have a 90t crewed spacecraft in orbit around Pluto in 167 days:
http://emdrive.wiki/Potential_EMDrive_solar_system_explorer_ship
« Last Edit: 07/14/2015 01:15 pm by TheTraveller »
It Is Time For The EmDrive To Come Out Of The Shadows

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
« Last Edit: 07/14/2015 01:26 pm by Rodal »

Offline TheTraveller

Interesting.

Just found a 1kW 2.4-2.5GHz Rf amp. Don't think I'll need it for my tests but nice to know it is available.

With that Rf amp could maybe generate 1N of Force.

Wow. Do you have a web link for that beast?

No link. Was told it was possible on a 8-10 week lead time. Price $2.5k. Min order 10.
It Is Time For The EmDrive To Come Out Of The Shadows

Offline leomillert

  • Member
  • Posts: 34
  • Liked: 21
  • Likes Given: 12
Doctor Rodal, I found an interesting discussion about validating an FDTD model by calculating the Poynting vector field.
Maybe it's old news for you, but I'm posting it here for others trying to independently study the aero's MEEP model. It provides some tips.
https://scicomp.stackexchange.com/questions/10692/fdtd-poynting-vector

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
Doctor Rodal, I found an interesting discussion about validating an FDTD model by calculating the Poynting vector field.
Maybe it's old news for you, but I'm posting it here for others trying to independently study the aero's MEEP model. It provides some tips.
https://scicomp.stackexchange.com/questions/10692/fdtd-poynting-vector
Yes he made a number of mistakes, most importantly not understanding that:

<< the source is a sinusoid, the magnitude (and Poynting vector) will, of course, oscillate.>>

He initially plotted the magnitude of the Poynting vector and disregarded its time variation.

To compound the problem, he was performing calculations in single precision ( !!! ) instead of double precision

<<This is running on consumer-grade GPUs, most of which don't support double.>> ???

Please notice:

1) I have plotted the Poynting vector as a vector field and not its absolute magnitude
2) I have taken into account its time variation
3) I have calculated the stress tensor separately, without any involvement whatsoever of the Poynting vector.

Others have discussed in this thread obtaining forces from the Poynting vector.  This is not the best way to go (as it would involve numerical differentiation, which is always something to be avoided in numerical schemes) as well known in fluid mechanics: the best thing is to calculate the stress tensor directly and not from the momentum equilibrium equation.
« Last Edit: 07/14/2015 02:13 pm by Rodal »

Offline wallofwolfstreet

  • Full Member
  • *
  • Posts: 165
  • Liked: 169
  • Likes Given: 436
...

Thanks for the reply.

Before I address you're comments, maybe one way for me and people like me to get our minds around what you are doing is for you to use actual numbers in place of variables within your equations, and in so doing demonstrate COE.  Actual numbers have a way of cutting through the abstractions. 

Ex:  We have an emdrive that gets 10 N/W, less than Shawyer's claimed superconducting specific thrust.  We run it for 1 s with an electrical power input of 1 W when it is attached to a 1 kg spaceship.

At t=1, with Newtonian approximation:

Ein=Pin*t
     =(1 W)*(1 s)
     =1 J

Eout=0.5*m*v^2
      =0.5*m*(int(F/m,dt))^2
      =0.5*m*(k*Pin*t/m)^2
      =0.5*1*(10*1*1/1)^2
      =50 J   

The mass change, delM, of the battery is calculated as:

delM=Ein/c^2
       =1/300,000,000^2
       =1.1*(10^-17) kg

So with all of those numbers in mind, can you show me, preferably for all t in [0,1] but just for t=1 is fine, how your math results in Ein=Eout?  Showing as many steps as possible would also be appreciated.
« Last Edit: 07/14/2015 02:51 pm by wallofwolfstreet »

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
...

Thanks for the reply.

Before I address you're comments, maybe one way for me and people like me to get our minds around what you are doing is for you to use actual numbers in place of variables within your equations, and in so doing demonstrate COE.  Real numbers have a way of cutting through the abstractions. 

Ex:  We have an emdrive that gets 10 N/W, less than Shawyer's claimed superconducting specific thrust.  ....
10 N/W cannot be considered to be a "real" number: there is no experimental data justifying this number for an EM Drive.   It appears to be an extrapolation orders of magnitude greater than present experimental data.

Bringing this back to reality, the highest force/InputPower measured in vacuum has been by NASA:

1.1*10^(-6)  N/W

This actual measured force is ten million times less than the number assumed above

http://emdrive.wiki/Experimental_Results

I agree with your statement quoted below, running numbers brings things back to reality.  Numbers that are 7 orders of magnitude less than 10 N/W .

Quote from: WallOfWolfStreet
  Real numbers have a way of cutting through the abstractions.   

To further bring this back to reality, Prof. Tajmar from TU Dresden University in Germany will be the second prestigious institution reporting on scientific tests performed in vacuum (instead of fictional numbers).  Tajmar's numbers are substantially lower than 1.1*10^(-6)  N/W

SUGGESTION: Use of real numbers, like 1.1*10^(-6)  N/W would bring this discussion back to reality.

Even if one were to use the highest number reported by Yang (not performed in vacuum) (which is slightly higher than Cannae LLC, G. Fetta, Superconducting test )

0.001070 N/W

is 10,000 times less than the assumed value of 10 N/W
« Last Edit: 07/14/2015 02:46 pm by Rodal »

Offline SeeShells

  • Senior Member
  • *****
  • Posts: 2442
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 3186
  • Likes Given: 2708
Interesting.

Just found a 1kW 2.4-2.5GHz Rf amp. Don't think I'll need it for my tests but nice to know it is available.

With that Rf amp could maybe generate 1N of Force.

Wow. Do you have a web link for that beast?

No link. Was told it was possible on a 8-10 week lead time. Price $2.5k. Min order 10.
I'm waiting for a reply from L3 communications I'll update when I do.
http://www.l-3com.com/

Offline wallofwolfstreet

  • Full Member
  • *
  • Posts: 165
  • Liked: 169
  • Likes Given: 436
......

I will echo what @deltamass has said, in that I would strongly advise against including any reference to "misunderstandings" or incorrect teaching methods if/when you send this off for peer review.  I can't imagine any peer reviewer will look favourably on that, and they will definitely request it's removal.

Quote
Do you believe that Special Relativity is an accurate and correct theory, or are you one of "those" people who thinks it is wrong?

uhhh, No?  About as orthodox a student as you can get.

Quote
This is the relativistic energy equation for a rocket, or any vehicle where mass-energy is being expended to produce propulsion.

I'm quite certain it's not.  If it is, it is at the very least missing some terms.  As I said in my original post, there must be some dependence on KE within the Eout expression.  How can your Eout expression possibly be divorced from KE?   

Quote
You may connect an external power source. Simply change the sign of Pin*t from negative, to positive. Very simple. But now you have an unlimited external supply of energy to start with. You don't need an over-unity machine. :)

The power supply need not be "unlimited".  It could be a basic car battery.  Simply removing it from the "spaceship" breaks any notion of "relativistic rocket equation", so your equations no longer apply. 

Quote
You will need an accelerometer AND two clocks, (one left behind) not a velocimeter, to determine the state of the vehicle. What I am claiming, and "correctly" showing is that it obeys Special and General Relativity.

Nope.  No need for a clock.  Use the accelerometer and the known mass of you ship to solve for force.  Use your multimeter to calculate the energy you're feeding the emdrive.  Use the equations you have presented in your paper to uniquely solve for velocity.

Quote
Tell me, do you remember the resolution of the Twin Paradox? Because it is the same paradox, just offered under a different scenario. If you do not remember how to solve it. Please go back and review it as I have done. It may jog your memory.

No idea what the twin paradox has to do with this.  While you brought it up, I don't agree with you're resolution to the twin paradox anyway and it is not the way I ever learned it.  The wikipedia article even lays it out:
Quote
Although some texts assign a crucial role to the acceleration of the travelling twin at the time of the turnaround,[7][8][9][10] others note that the effect also arises if one imagines separate outward-going and inward-coming travellers, who pass each other and synchronize their clocks at the point corresponding to "turnaround" of a single traveller. In this version, acceleration plays no direct role

Quote
You must prove over-unity using a finite supply of energy, not an external source

A thing can't be external and finite?  And within your equations, I have worked out overunity using a finite supply of energy.  As I said above, I'm not following this line of reasoning. 
« Last Edit: 07/14/2015 02:46 pm by wallofwolfstreet »

Offline wallofwolfstreet

  • Full Member
  • *
  • Posts: 165
  • Liked: 169
  • Likes Given: 436
10 N/W cannot be considered to be a real number.  There is no experimental data justifying this number for an EM Drive.

You misunderstand my use of "real".  I don't mean real as in experimentally measured, I mean real as in an element of R

In @warptech's math, k is a variable, completely unbounded.  It could by anything. 0.000001 N/kW.  1 N/kW  10000 N/W.  If the math is right, the value of k is irrelevant, because the math must work for all k in R.

Talking about whether or not the k is experimentally real is a totally different issue completely outside of my suggestion.

A better word to have used on my part is "actual", since of course the variable's themselves can be real numbers.
   
« Last Edit: 07/14/2015 02:51 pm by wallofwolfstreet »

Offline TheTraveller

Direction and Strength of Generated Force

With all the theory walk, it should be noted that using a dielectric in the cavity generates a weaker externally generated Force pushing in the opposite direction than if there was no dielectric in the cavity.

1) Dielectric attached to small end: Generated Force is weakly pushing the Small end to the Big end.

2) No dielectric in cavity: Generated Force is strongly pushing the Big end to the Small end.

If an EMDrive theory can't handle both conditions and predict the correct Force direction / strength, it may need to be worked on.
« Last Edit: 07/14/2015 02:55 pm by TheTraveller »
It Is Time For The EmDrive To Come Out Of The Shadows

Online aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1145
  • Likes Given: 360
I have just installed MEEP and NSF-1701.ctl is sitting here on my hard drive.
How can I help?

Good! Run it and see how the data compares to posted data. In particular, generate some csv files, then "diff" them with the "identical" data on Google drive. If you haven't changed any settings, the data should be there. If you change anything at all, chances are your data will be different. If that is the case post a copy of your file to me and I'll run it as is then post a some csv files back to you.

I think that is probably the most straight forward way to check installations. Of course your computer word length may be different so that might not work either.

aero

I ran "meep NSF-1701.ctl" and got the following files:

      Name                        Size                                          MD5
--------------------------    ------------     ---------------------------------------------------------
eps-000000000.h5    127.3M         7028e98b42305c0fc5bd285c70766e17
ex.h5                             1.7G              f515221aee0ff2b82a32c69ee4c4a54b
ey.h5                             1.7G              069ec826d6803e2e67f0619617c641b7
ez-000000003.h5      127.3M         4865721e49192853eb998efe0fa76b0f
ez-000000021.h5      127.3M         8675ee3158b64285aa1fb4e36833217b
ez.h5                             1.7G             920f598802f7eee417465c0f2703872c
hx.h5                             1.7G             c4c14a37208eed92f8dff0c1aef7762c
hy.h5                             1.7G             01ff935ad90d58e0d31f93c84244ee95
hz.h5                             1.7G             f9ee7c2e08c895bd57ac7b7cbc200913


Should I use h5totxt to convert each file, individually?
What is the expected output? The Google Drive has a lot of .csv files, I'm not sure which ones I'm supposed to independently replicate and compare to.

If you have HDFview installed, look at the images of the eps file, then the two ez-000000003.h5 and ez-000000021.h5 just to verify that the antenna is there and in the correct location near the small end of the cavity. If you don't have or don't know how to use HDFview then just use h5topng to generate images of those files.

Make one or more csv files of any of the other .h5 files. Just pick one, ex.h5 is the first. Make a csv file of the final image, image 13, then we'll help you to find it on Google drive. It will be in the NSF-1701 folder under csv files but I need to know which one to look for.
Retired, working interesting problems

Offline Rodal

  • Senior Member
  • *****
  • Posts: 5911
  • USA
  • Liked: 6124
  • Likes Given: 5564
McCulloch has been writing more on his theory:

MiHsC and EmDrive: Clarification (dated today July 14 2015)
http://physicsfromtheedge.blogspot.com/2015/07/mihsc-and-emdrive-clarification.html

McCulloch<<It seems, after several comments I've received, and from reviewers too, that the way MiHsC works on the emdrive is not quite clear, so here is an attempt to clarify why I think you get a push consistently towards the narrow end from MiHsC and why, although new physics, it is at least perfectly self-consistent.>>

and has also written about Shawyer's theory:

Critique of Shawyer's emdrive theory
http://physicsfromtheedge.blogspot.com/2015_06_01_archive.html

McCulloch<< I'd like here to criticise Shawyer's theory of it, which I believe is confused. >>

____________

I don't recall whether McCulloch's theory has anything to say (and if so what does it have to say) about the effect of the dielectric insert used by NASA.   Does it?
« Last Edit: 07/14/2015 03:05 pm by Rodal »

Offline leomillert

  • Member
  • Posts: 34
  • Liked: 21
  • Likes Given: 12
I have just installed MEEP and NSF-1701.ctl is sitting here on my hard drive.
How can I help?

Good! Run it and see how the data compares to posted data. In particular, generate some csv files, then "diff" them with the "identical" data on Google drive. If you haven't changed any settings, the data should be there. If you change anything at all, chances are your data will be different. If that is the case post a copy of your file to me and I'll run it as is then post a some csv files back to you.

I think that is probably the most straight forward way to check installations. Of course your computer word length may be different so that might not work either.

aero

I ran "meep NSF-1701.ctl" and got the following files:

      Name                        Size                                          MD5
--------------------------    ------------     ---------------------------------------------------------
eps-000000000.h5    127.3M         7028e98b42305c0fc5bd285c70766e17
ex.h5                             1.7G              f515221aee0ff2b82a32c69ee4c4a54b
ey.h5                             1.7G              069ec826d6803e2e67f0619617c641b7
ez-000000003.h5      127.3M         4865721e49192853eb998efe0fa76b0f
ez-000000021.h5      127.3M         8675ee3158b64285aa1fb4e36833217b
ez.h5                             1.7G             920f598802f7eee417465c0f2703872c
hx.h5                             1.7G             c4c14a37208eed92f8dff0c1aef7762c
hy.h5                             1.7G             01ff935ad90d58e0d31f93c84244ee95
hz.h5                             1.7G             f9ee7c2e08c895bd57ac7b7cbc200913


Should I use h5totxt to convert each file, individually?
What is the expected output? The Google Drive has a lot of .csv files, I'm not sure which ones I'm supposed to independently replicate and compare to.

If you have HDFview installed, look at the images of the eps file, then the two ez-000000003.h5 and ez-000000021.h5 just to verify that the antenna is there and in the correct location near the small end of the cavity. If you don't have or don't know how to use HDFview then just use h5topng to generate images of those files.

Make one or more csv files of any of the other .h5 files. Just pick one, ex.h5 is the first. Make a csv file of the final image, image 13, then we'll help you to find it on Google drive. It will be in the NSF-1701 folder under csv files but I need to know which one to look for.

h5topng asks me to specify a slice.
Which one should I use?
Anyway, if this is just to verify our files is the same, you could do a simple "md5sum" on your .h5 files and confirm the hash values. If they match, our files are the same.

Offline TheTraveller

McCulloch has been writing more on his theory:

MiHsC and EmDrive: Clarification (dated today July 14 2015
http://physicsfromtheedge.blogspot.com/2015/07/mihsc-and-emdrive-clarification.html

McCulloch<<It seems, after several comments I've received, and from reviewers too, that the way MiHsC works on the emdrive is not quite clear, so here is an attempt to clarify why I think you get a push consistently towards the narrow end from MiHsC and why, although new physics, it is at least perfectly self-consistent.>>


Critique of Shawyer's emdrive theory
http://physicsfromtheedge.blogspot.com/2015_06_01_archive.html

McCulloch<< I'd like here to criticise Shawyer's theory of it, which I believe is confused. >>

Mike doesn't understand that an EM wave in a waveguide is velocity restricted below c, moves at Group Velocity and not at c.

Thus his theory, based on constant EM wave velocity at c, inside a waveguide, can't be valid as it is against all known microwave waveguide physics, as I posted on his Blog:
http://physicsfromtheedge.blogspot.com/2015/07/mihsc-and-emdrive-clarification.html?showComment=1436887078021#c266729468057178512
« Last Edit: 07/14/2015 03:23 pm by TheTraveller »
It Is Time For The EmDrive To Come Out Of The Shadows

Offline rfmwguy

  • EmDrive Builder (retired)
  • Senior Member
  • *****
  • Posts: 2205
  • Liked: 2713
  • Likes Given: 1134
NSF-1701 will be using:

Item No.: 2M218J Magnetron tube
Peak anode voltage (ebm): 4.0+/-0.2KV
Average anode current (ib): 300mAdc
Average output power (Po): 900+/-40W
Frequency (Fo): 2458+/-10MHz
Filament voltage(Ef): 3.3V

Online aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1145
  • Likes Given: 360
Don't have the software to do the integration on the images but I did the large end.

I hope you have good solder joints people! The maximum stress seems to ride the seem of the large end. Hence, the use of bolts, not solder to hold it together. Hmmm...

It makes sense though, that when the mode energy reaches the big end, it has already expanded. So after that, most of the reflections must be concentrated around the perimeter of the big end. That's why I want to see a simulation of it turned off. I want to see how far back up the frustum the wave goes after it is reflected from the big end, and for how long.
Todd

It is a location NEAR the big end, but NOT at the big end.  I still have to plot the stresses at the big and at the small ends

In order to do that, don't you need csv files at the big and small ends? If so, tell me where they are. That is, what rows of the xz or xy csv files that you have correspond to the big and small ends? Then I can cut x slices at those rows and make the right csv files.

I thought you had determined that these csv files were one of the circular cross-sections bases

\\ts03\\s3-exx.csv
\\ts03\\s3-exx.csv
\\ts03\\s3-eyx.csv
\\ts03\\s3-ezx.csv
\\ts03\\s3-hxx.csv
\\ts03\\s3-hyx.csv
\\ts03\\s3-hzx.csv

etc. up to ts13

dated 07/03/2015 in the folder "time slices 3-13"

Please let me know whether these cross-sections are of one of the bases (if so what base) or if this cross section is not located at a base at what column is this cross-section located (the files are not labeled as to such location, all they state is exx.csv)

Reminder:  remember that your original csv file had a cross-section that contained noise, and you traced it back to the fact that the cross-section was outside the EM Drive in nowhere land.  Then you determined where the metal base was located.

If that is what I said, then those files are likely the best I can come up with for the inside of the big end. Pardon my poor memory.
Retired, working interesting problems

Online aero

  • Senior Member
  • *****
  • Posts: 3629
  • 92129
  • Liked: 1145
  • Likes Given: 360
Congrats New Horizons team, very impressed...the best media question I heard was "when are we going back?" The answer was interesting, there are designs being worked on.

IMHO, the key for planetary science and space exploration success is becoming clear. It is not more sensitive  instrumentation, it is not lower cost hardware...it is faster propulsion. This is your challenge scientific community...to shrink the fabric of space and time with modern propulsion research.

To the propulsion industry...think outside the box. Your equations have shown us that propellants will never get us to where we want to go...faster and farther. Perhaps in this forum, one of us may help show you the way.

In case you are wondering, Space is big. Here is a "to scale" map of the solar system with the moon diameter as the scale length of one pixel. http://joshworth.com/dev/pixelspace/pixelspace_solarsystem.html No wonder it takes so much time to get anywhere.
Retired, working interesting problems

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1