Author Topic: EM Drive Developments - related to space flight applications - Thread 3  (Read 1797579 times)

Online SeeShells

  • Senior Member
  • *****
  • Posts: 2325
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2953
  • Likes Given: 2587
So we have 7 drawings from the author's in the author's previously ballyhooed "peer-reviewed papers".

The smallest angle in any of these drawings is practically TWICE as much as 6 degrees, the median is 15 degrees and the max is 25 degrees.  Yet we are going to ignore all these drawings in peer-reviewed journals?

The drawings are part of a big master conspiracy?

The peer-reviewers are part of a master conspiracy or they just did not realize that the drawings did not match the text?

Since we are at it, why not just say that Yang used a cylinder for her experiments?
You know for sure what she used? She could have used a box for all we know. Peer reviewed or not. A real picture would be better not by much.

I find this interesting as well. Every open test that could be verified with real time video and EW as well has showed very low thrusts and here we have the Chinese and RS claiming outrageous thrusts. Why is that? Wouldn't that fact alone lead you to question not only thrusts but configurations and other important data... peer reviewed hmmm.

Of course it leads me to question everything that Yang reported.  But the alternatives are either to disregard what she reported or to take into consideration all her drawings,  but not to throw all her drawings to the garbage and substitute for her drawings pre-conceived notions.  Either I believe nothing she reported or I take into consideration her drawings.  It does not make any sense whatsoever to disregard all her drawings and substitute them with an estimate of 6 degrees based on a single paper she published in 2010 dealing mainly with her Finite Element formulation prior to her large thrust measurement papers

Not only her's but  we need question with a jaundiced eye all unverifiable reports extraordinary in nature. That said I believe they are gems of information buried in many of these extraordinary reports. Like I've said many times there is no bad data, the problem is distinguishing the good data.

I'm tired of beating this 6 degree angle with a brick. It's done, so called water under the bridge. I'll test it and throw a few variables at it and when I'm done it can become another brick in the wall.

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5901
  • Likes Given: 5249
...The only way to obtain the correct angle is either by calculus or by engineering drawings. Performing statistical distribution diagrams , medians , etc is all pointless in this case....
Thanks, we look forward to your use of Calculus to elucidate this problem and further teaching us about statistical methods
« Last Edit: 08/03/2015 01:05 PM by Rodal »

Online SeeShells

  • Senior Member
  • *****
  • Posts: 2325
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2953
  • Likes Given: 2587
...
What I find interesting is it seems the stress values are located big end or small end and little shows up in the sidewalls which is what the current theory of Yang's proposes as to thrust....
Shell
You are completely wrong in the above statement.  The calculation of stress shows that the stresses on the sidewalls are most significant.   I discussed in previous posts, and Todd understood it, as he actually said that's what he expected.
Are those the images you showed for the large end around the outer diameter? I have a point here to follow through with and got your attention.

You showed a very strong stress on the center and outside flange of the large end with a very sharp cutoff where the flange meets the sidewalls. I found that very hard to see that that stress didn't show any propagation to the sidewalls. Did you just calculate for the endplates? Why?

Shell

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5901
  • Likes Given: 5249
...
What I find interesting is it seems the stress values are located big end or small end and little shows up in the sidewalls which is what the current theory of Yang's proposes as to thrust....
Shell
You are completely wrong in the above statement.  The calculation of stress shows that the stresses on the sidewalls are most significant.   I discussed in previous posts, and Todd understood it, as he actually said that's what he expected.
Are those the images you showed for the large end around the outer diameter? I have a point here to follow through with and got your attention.

You showed a very strong stress on the center and outside flange of the large end with a very sharp cutoff where the flange meets the sidewalls. I found that very hard to see that that stress didn't show any propagation to the sidewalls. Did you just calculate for the endplates? Why?

Shell

Wrong. Please refer to the prior posts, where it is stated a number of times that that is post-processing of Meep data using Wolfram Mathematica. As aero understands, I can only post-process data that aero produces.  He did not send me the over 200 files necessary to calculate the stresses on the side conical walls, only the data for the big and small ends.

I have calculated the stresses on the side walls with a different program I wrote in Mathematica and I have verified that the stresses on the side walls are most significant.  I have pointed out how Shawyer is wrong in neglecting the stresses on the side walls.

Aero sends me the information for the big and small ends, and that's what I post-process.
If aero sends me no data I cannot post-process non-existing data.

The statement <<  stress didn't show any propagation to the sidewalls.>> is incorrect.  High stresses at boundaries are shown and they satisfy the mode shape boundary conditions, for example:

« Last Edit: 08/03/2015 01:17 PM by Rodal »

Offline TheTraveller

The closest we have to 6 frustum verified data points is when Shawyer took my 3 Flight Thruster internal dimensions estimations, put them through the SPR in-house software and produced the Df and resonant frequency at TE013 being:

Frustum big diameter   m: 0.2314m (my data)
Frustum small diameter: 0.1257m (my data)
Frustum centre length: 0.1386m (my data)
Mode: TE013 (my data)

Resonance: 3.9003 GHz (SPR data based on the above)
Df: 0.634 (SPR data based on the above)

With Shawyer we have multiple photographs of 4 EMDrives and the test rigs, a video of the Demonstrator EMDrive accelerating on a rotary test table, willing to corro via email, continual updates on SPR progress, a web site with experimental data and many papers with the last being peer reviewed.

Yet most here reject him like he has the plague. Go figure?
« Last Edit: 08/03/2015 01:09 PM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.
Herman Melville, Moby Dick

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5901
  • Likes Given: 5249
The closest we have to 6 frustum verified data points is when Shawyer took my 3 Flight Thruster internal dimensions estimations, put them through the SPR in-house software and produced the Df and resonant frequency at TE013 being:

Frustum big diameter   m: 0.2314m
Frustum small diameter: 0.1257m
Frustum centre length: 0.1386m
Resonance: 3.9003 GHz
Mode: TE013
Df: 0.634

With Shawyer we have multiple photographs of 4 EMDrives and the test rigs, a video of the Demonstrator EMDrive accelerating on a rotary test table, willing to corro via email, continual updates on SPR progress, a web site with experimental data and many papers with the last being peer reviewed.

Yet most here reject him like he has the plague. Go figure?
Incorrect.  In a prior post I had given the advice that if I were to select another geometry to compare with NASA's (used by Rfmwguy and Iulian Berca and Mulletron in their experiments) I strongly advised to use Shawyer's EM Drives as we know their geometry much better than Yang's, I repeat, emphatically, that advise again in this message

This is actually what I have been arguing for in the last dozen of pages.
And to repeat once again: I advice against using 6 degree cone half-angles that go against everything that Shawyer and NASA have been testing.  And that go against all the published formulas: Shawyer, McCulloch and Notsosureofit.
« Last Edit: 08/03/2015 01:19 PM by Rodal »

Offline rfmwguy

  • EmDrive Builder (retired)
  • Senior Member
  • *****
  • Posts: 2165
  • Liked: 2681
  • Likes Given: 1124
Busy with my build, but have reviewed many pages of posts, languishing over Yang's dimensions. These peeps are even more dark than EW has been. Face it, knowing how CN values other's IP, there's no way a military college will disclose anything. Best to move on IMHO. Take your best shot and not ruminate.

Have a question for NSFers here, since it looks like NSF-1701 will be OK at 100% power cycle for 1 minute, I now need to plan on the first real test run. Here's my thought for fulcrum force measurement (Ds down).

30% power cycle (setting) for 5 minutes.

I am guessing this will maintain adequate thermal management plus cycle ON for a decent repetitive rate. With an oil dampener, I will need to allow for resettling time before the next pulse occurs. Comments welcomed.

Offline TheTraveller

This is actually what I have been arguing for in the last dozen of pages.

BTW what resonance and Q do you get for my Flight Thruster estimated dimensions?

Frustum big diameter   m: 0.2314m
Frustum small diameter: 0.1257m
Frustum centre length: 0.1386m
Mode: TE013
« Last Edit: 08/03/2015 01:14 PM by TheTraveller »
"As for me, I am tormented with an everlasting itch for things remote. I love to sail forbidden seas.
Herman Melville, Moby Dick

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5901
  • Likes Given: 5249
This is actually what I have been arguing for in the last dozen of pages.

BTW what resonance and Q do you get for those dimensions?

Frustum big diameter   m: 0.2314m
Frustum small diameter: 0.1257m
Frustum centre length: 0.1386m
Mode: TE013

I have 3 computers busy running $$$ paying statistical work.  Won't be able to calculate that today, perhaps tonight?. (I only have 2 Mathematica licenses and both are running several hour long programs at the moment).  Mathematica does not allow you to run more simultaneous processes than the number of Mathematica licenses you have.
« Last Edit: 08/03/2015 01:23 PM by Rodal »

Online SeeShells

  • Senior Member
  • *****
  • Posts: 2325
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2953
  • Likes Given: 2587
...
What I find interesting is it seems the stress values are located big end or small end and little shows up in the sidewalls which is what the current theory of Yang's proposes as to thrust....
Shell
You are completely wrong in the above statement.  The calculation of stress shows that the stresses on the sidewalls are most significant.   I discussed in previous posts, and Todd understood it, as he actually said that's what he expected.
Are those the images you showed for the large end around the outer diameter? I have a point here to follow through with and got your attention.

You showed a very strong stress on the center and outside flange of the large end with a very sharp cutoff where the flange meets the sidewalls. I found that very hard to see that that stress didn't show any propagation to the sidewalls. Did you just calculate for the endplates? Why?

Shell

Wrong. Please refer to the prior posts, where it is stated a number of times that that is post-processing of Meep data using Wolfram Mathematica. As aero understands, I can only post-process data that aero produces.  He did not send me the over 200 files necessary to calculate the stresses on the side conical walls, only the data for the big and small ends.

I have calculated the stresses on the side walls with a different program I wrote in Mathematica and I have verified that the stresses on the side walls are most significant.  I have pointed out how Shawyer is wrong in neglecting the stresses on the side walls.

Aero sends me the information for the big and small ends, and that's what I post-process.
If aero sends me no data I cannot post-process non-existing data.

The statement <<  stress didn't show any propagation to the sidewalls.>> is incorrect.  High stresses at boundaries are shown and they satisfy the mode shape boundary conditions, for example:


Not seeing the sidewall angle is because it's just a thin slice through the frustum showing stress in XY

Here is where I'm going.

How could you show the stress, big end to small end by taking a point on the outside wall where the large plate connects to the side wall and project along that wall (z axis) to a point on the small plate.

Drew a quick pic to show you what I mean. What I'm looking for here is a traveling stress value over a X amount of time to see how the Stress propagates down the sidewalls.

Can you do this?

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5901
  • Likes Given: 5249
...
Not seeing the sidewall angle is because it's just a thin slice through the frustum showing stress in XY

Here is where I'm going.

How could you show the stress, big end to small end by taking a point on the outside wall where the large plate connects to the side wall and project along that wall (z axis) to a point on the small plate.

Drew a quick pic to show you what I mean. What I'm looking for here is a traveling stress value over a X amount of time to see how the Stress propagates down the sidewalls.

Can you do this?
Yes, I have done this with my program.  Cannot be done with Meep at the moment, if you want to know why PM aero, he will tell ya.  You are tired of the 6 degree? I'm tired of explaining why. It has to do with Meep not with me. 
If you want to have an idea of what the force looks like on the lateral walls take a gander at Greg Egan:

(see the difference between TE and TM modes ;)  -what difference do you see between TE and TM ? - )

Your drawing of side forces on conical walls is very wrong for TE modes

your using of z axis for the longitudinal is in disagreement with the aero-chosen Meep axis which is the x axis, this is important in discussing Meep results Use of different axes conventions leads to confusion.

http://gregegan.customer.netspace.net.au/SCIENCE/Cavity/Cavity.html





« Last Edit: 08/03/2015 02:25 PM by Rodal »

Online SeeShells

  • Senior Member
  • *****
  • Posts: 2325
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2953
  • Likes Given: 2587
...
Not seeing the sidewall angle is because it's just a thin slice through the frustum showing stress in XY

Here is where I'm going.

How could you show the stress, big end to small end by taking a point on the outside wall where the large plate connects to the side wall and project along that wall (z axis) to a point on the small plate.

Drew a quick pic to show you what I mean. What I'm looking for here is a traveling stress value over a X amount of time to see how the Stress propagates down the sidewalls.

Can you do this?
Yes, I have done this with my program.  Cannot be done with Meep at the moment, if you want to know why PM aero, he will tell ya.  You are tired of the 6 degree? I'm tired of explaining why. It has to do with Meep not with me. 
If you want to have an idea of what the force looks like on the lateral walls take a gander at Greg Egan:

(see the difference between TE and TM modes ;)  -what difference do you see between TE and TM ? - )

Your drawing of side forces on conical walls is very wrong for TE modes

http://gregegan.customer.netspace.net.au/SCIENCE/Cavity/Cavity.html




The TE modes are a better way to add stress into the frustum walls and one of the reasons I picked a TE mode.

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5901
  • Likes Given: 5249
...
The TE modes are a better way to add stress into the frustum walls and one of the reasons I picked a TE mode.
This bears emphasizing: there has NOT been a single Meep computer run (csv file for sure, and I don't recall seeing any images for TE mode excitation either) for Yang/Shell exciting the frustum in a TE  (transverse electric) mode.
« Last Edit: 08/03/2015 03:14 PM by Rodal »

Online Flyby

  • Full Member
  • ***
  • Posts: 382
  • Belgium
  • Liked: 447
  • Likes Given: 46
...The only way to obtain the correct angle is either by calculus or by engineering drawings. Performing statistical distribution diagrams , medians , etc is all pointless in this case....
Thanks, we look forward to your use of Calculus to elucidate this problem and further teaching us about statistical methods
No need to be edgy about it, dr Rodal...
My English is good enough to recognize sarcasm...

As research and maths are part of your professional environment, you obviously are considerably more skilled in those things then me. It is not my world, it's yours.
That is why I also do not engage in your - be slightly snidy perhaps - requests to fill in the data and provide mathematical/statistical method.

However, handling architectural and engineering drawings, making illustrations and communicating with clients and ponder about perception, that is my world, since well over 30years.
Visual communication is part of what i do on a daily basis, just as - I assume? - juggling with formula's and data analysis is what you do on a daily basis.

Consequently, my contribution to this forum is limited to visual information, so if there is data to be obtained from the provided photographs or drawings, I'll gladly assist in analyzing/reconstructing.

I have a great respect for your abilities and the lead you're taking in this topic(s), but I would appreciate you stop belittling some of us.

I do not comment on how correct or incorrect your formula's are or the mathematical methods you use to obtain your results. I do not challenge your math abilities and insights and do not venture on terrains i know little of.
I'm actually learning a lot from you and others like Todd.

But neither do I not ask you to make engineering drawings and 3d models, so why do you insist that I "provide calculus to elucidate the problem". You know damn well i can't...
I'm too old for these type of games...

But as far as visual perception, drawings, design intend and communication goes... that's my playground.
You do with your results what you want, but my 30 year experience with drawings tells me you went overboard with the average angle calculation....

With this from my chest and as I'm not a vindictive person,  I'll gladly continue to provide assistance if you need it.... Let's wipe the board clean and start over... ;)
« Last Edit: 08/03/2015 02:09 PM by Flyby »

Online Rodal

  • Senior Member
  • *****
  • Posts: 5833
  • USA
  • Liked: 5901
  • Likes Given: 5249
... Let's wipe the board clean and start over... ;)

OK,  :)  same here

« Last Edit: 08/03/2015 02:12 PM by Rodal »

Online SeeShells

  • Senior Member
  • *****
  • Posts: 2325
  • Every action there's a reaction we try to grasp.
  • United States
  • Liked: 2953
  • Likes Given: 2587
...
The TE modes are a better way to add stress into the frustum walls and one of the reasons I picked a TE mode.
This bears emphasizing: there has NOT been a single Meep computer run (csv file for sure, and I don't recall seeing any images for TE mode excitation either) for Yang/Shell exciting the frustum in a TE  (transverse electric) mode.  Not even one. Nada. Zilch.
I'm not sure what to do with this Jose. What do you expect me to do? I pretty much know what I'm going to see with this 6 degree frustum and like I said a few pages ago it is a start.

So if your intent is to get me so upset that I leave this forum by beating this dead horse into the ground, continue on this thread and I'll oblige you.

I have the other side of the build to finish so I'm off to the shop.

Shell




Online Flyby

  • Full Member
  • ***
  • Posts: 382
  • Belgium
  • Liked: 447
  • Likes Given: 46
straight from the press :
The baby-EMdrive team is going for a new build, this time a 3dprinted adjustable cavity...

https://hackaday.io/project/5596-em-drive/log/22027-3d-printed-tuneable-cavity



As I have a 3dprinting business myself, I do foresee problems with the surface finish due to the layering.

From what I know, the (european) machine park of Shapeways does not include a combined 3dprint/cnc machine, known for their nice surface finish. They'll need to do something about that or they'll have an extreme low Q.

I wish them all the best with the second experiment, but I fear that they making it hard on themselves : by making it 10x smaller, they also make it 10x more vulnerable to material imperfections. Hopefully they can compensate a bit with the adjustable small end.

also, any1 has an idea to why they added a longitudinal slit/opening ? a position adjustable Rf feed?
« Last Edit: 08/03/2015 03:38 PM by Flyby »

Offline D_Dom

  • Global Moderator
  • Full Member
  • *****
  • Posts: 456
  • Liked: 154
  • Likes Given: 105
I am guessing the media will be non-conductive which seems to indicate the cavity will be electroplated after build/polish? Considering your business Flyby what would you charge for this part?

From the link; https://hackaday.io/project/5596-em-drive/log/22027-3d-printed-tuneable-cavity

"a sliding rail for the custom RF board. It will make it possible to position the antenna freely inside the cavity, and it will make it easier replicable"

It seems the opening is an adjustable RF feed slot, not sure if the dimensions are similar to their previous build?!?
« Last Edit: 08/03/2015 03:44 PM by D_Dom »
Space is not merely a matter of life or death, it is considerably more important than that!

Online Flyby

  • Full Member
  • ***
  • Posts: 382
  • Belgium
  • Liked: 447
  • Likes Given: 46
I doubt it will be in a non-metal material. All those materials, except ceramics, have very low melting points.

so it has to be in a metal...
 
Titanium would be excluded, as I know from colleagues that it prices around 50euro/cm....expensive to build solid welled objects in... ???

I don't think Alumide would be a good idea. Although it contains a high amount of alu particles, it still is a plastic, sensitive to temperature.

My best guess would be they'll be using 3dprinted brass or steel. These are still affordable materials...

but this closeup shows to why I have questions about the surface finish:

notice the layering on the surface, even after polishing...

You need a very specialized (read expensive) metal 3dprinter to achieve smooth surfaces that can compare to cnc milling. In fact , the best industry results are obtained with a combo machine that first 3dprints and does a second cnc milling pass, before printing a new layer.

added.
come to think about it, the baby EMdrive team uses a very low powered device. So, yes, maybe a metal plated plastic model might work as very little heat will be produced (unlike the 1kW microwave generators)...
hmmm... to reduce the layering, they should use Shapeways Objet 3dprinter, those have layer accuracies of 16m, barely visible with the eye...
I'm not sure if these type of plastic materials can be copper plated...
To smooth out the surface it should be really heavy plated.

I know certain stereo-lithography resins can be easily copper plated, but those need to be hand polished first before they're send off to the copper plating. Some of the highly specialized dental or jewelry stereo-lithograph go down to even 5m layer thickness, but the downsize is their small printable size.
As you see.. many options... :)
« Last Edit: 08/03/2015 04:18 PM by Flyby »

Offline Donosauro

  • Full Member
  • *
  • Posts: 131
  • Liked: 40
  • Likes Given: 0
I doubt it will be in a non-metal material. All those materials, except ceramics, have very low melting points.

so it has to be in a metal...
 
Titanium would be excluded, as I know from colleagues that it prices around 50euro/cm....expensive to build solid welled objects in... ???

http://www.amazon.com/One-Pound-999-Fine-Titanium/dp/B007RFTC7I shows a one-pound bar of .999
fine titanium selling for less than $50, which is much less than 50 euros per cubic centimeter. Did you mean platinum?

Tags: