-
#1940 by TheTraveller on 23 Oct, 2016 15:59
-
Interesting rumour just arrived.
Seems the Chinese have tested a EmDrive on station but have no idea why it works. Maybe they should talk to Roger or Gilo Industries?
Sure hope the X-37B is testing a better EmDrive on station.
-
#1941 by Rodal on 23 Oct, 2016 16:26
-
Interesting rumour just arrived.
What would be the advantage(s) and therefore the need of testing an EM Drive inside a Space Station or inside the X-37B as compared to just testing it in Space as proposed by Cannae ?
Seems the Chinese have tested a EmDrive on station but have no idea why it works. Maybe they should talk to Roger or Gilo Industries?
Sure hope the X-37B is testing a better EmDrive on station.
http://cannae.com/cubesat-mission-clarification/
http://cannae.com/cannae-is-developing-a-cubesat-thruster/
-
#1942 by Monomorphic on 23 Oct, 2016 16:46
-
What would be the advantage(s) and therefore the need of testing an EM Drive inside a Space Station or inside the X-37B as compared to just testing it in Space as proposed by Cannae ?
The obvious advantage is having direct access to the test article. An astronaut can make physical adjustments, reorient/reset the experiment.
Second advantage is secrecy. Anybody can track a satellite. -
#1943 by Rodal on 23 Oct, 2016 16:48
-
What would be the advantage(s) and therefore the need of testing an EM Drive inside a Space Station or inside the X-37B as compared to just testing it in Space as proposed by Cannae ?
The obvious advantage is having direct access to the test article. An astronaut can make physical adjustments, reorient/reset the experiment.
Second advantage is secrecy. Anybody can track a satellite.
There is no astronaut inside the X-37B, so the first "advantage" is only for the Space Station, and only while there are astronauts inside the Space Station, and only if such adjustments are necessary and cannot be made by remote or automatic control.
Concerning the second point (avoiding enemies or others tracking a satellite), there are ways to get around that. -
#1944 by Monomorphic on 23 Oct, 2016 16:55
-
Concerning the second point <<Anybody can track a satellite>>, there are ways to get around that (USAirForce)
Wouldn't a small emdrive test satellite need to broadcast a signal to be tracked? -
#1945 by Rodal on 23 Oct, 2016 16:57
-
https://en.wikipedia.org/wiki/X-ray_pulsar-based_navigationConcerning the second point <<Anybody can track a satellite>>, there are ways to get around that (USAirForce)
Wouldn't a small emdrive test satellite need to broadcast a signal to be tracked? -
#1946 by TheTraveller on 23 Oct, 2016 16:58
-
...
Would suggest what Gilo Industries may release should eliminate doubt. -
#1947 by Monomorphic on 23 Oct, 2016 17:02
-
https://en.wikipedia.org/wiki/X-ray_pulsar-based_navigationConcerning the second point <<Anybody can track a satellite>>, there are ways to get around that (USAirForce)
Wouldn't a small emdrive test satellite need to broadcast a signal to be tracked?
I could see that being used in a mature emdrive spy satellite, but not on an experimental satellite the purpose of which is to validate the emdrive technology. The small Cannae satellite will need to broadcast a signal, at least intermittently, if they ever hope to track its location to prove it works. Or they could use radar to track it. -
#1948 by Rodal on 23 Oct, 2016 17:10
-
Please notice that Todd updated his spreadsheet with the copper Cannae data and I updated the Log Log plot in:
http://forum.nasaspaceflight.com/index.php?topic=40959.msg1602555#msg1602555
This is very interesting, because it shows that the Cannae cavity has its own regression line slope, which is parallel and different from the truncated cone geometry of Shawyer/NASA/Yang/Tajmar and others.
So Todd's theory shows a fundamental difference between the Cannae design and the truncated cone design of Shawyer, which is due to the Cannae geometry and not due to whether it is superconducting.
There are two Log-Log lines: one for the truncated cone geometry and another one for the Cannae pillbox/flying-saucer geometry.
I also now calculated the Regression Line only taking into account the Truncated Cone geometry pioneered by Shawyer. The data gives an excellent R2=0.99
-
#1949 by WarpTech on 23 Oct, 2016 17:36
-
Please notice that Todd updated his spreadsheet with the copper Cannae data and I updated the Log Log plot in:
http://forum.nasaspaceflight.com/index.php?topic=40959.msg1602555#msg1602555
This is very interesting, because it shows that the Cannae cavity has its own regression line slope, which is parallel and different from the truncated cone geometry of Shawyer/NASA/Yang/Tajmar and others.
So Todd's theory shows a fundamental difference between the Cannae design and the truncated cone design of Shawyer, which is due to the Cannae geometry and not due to whether it is superconducting.
There are two Log-Log lines: one for the truncated cone geometry and another one for the Cannae pillbox/flying-saucer geometry.
I also now calculated the Regression Line only taking into account the Truncated Cone geometry pioneered by Shawyer. The data gives an excellent R2=0.99
Thank you for doing the log-log plot & regression. That looks great!
I'm not 100% certain that air "inside" the frustum doesn't help the thrust by design. After all, in @Notsosureofit's model and in my model, there is an accelerated reference frame inside the frustum. The force acting on the frustum mass, is equal and opposite the force acting on the internal EM field equivalent mass. Having air inside, if it accelerates along with the EM field, adds a significant amount of mass to the equation, and therefore will add a significant amount of force to the EM drive. A compressed gas inside the frustum may be a necessary part of the design, to enhance the thrust, provided the kinetic energy of the gas' mass can be dissipated at the big end, before returning to the small end, or at least over many cycles. i.e., those data points may not be 100% anomalous data. There may be a real affect enhancement due to gas with high kinetic energy, inside the frustum. -
#1950 by as58 on 23 Oct, 2016 17:37
-
Have I completely misunderstood something? As the equations in the spreadsheet are written, log(R2/R1) is a directly proportional to thurst/power. So the amazingly good fit seems to me to be only a reflection of the fact that all devices are of similar size and operate at similar frequencies.
edit. thrust/power instead of thrust/weight as I'd originally written -
#1951 by as58 on 23 Oct, 2016 17:47
-
Have I completely misunderstood something? As the equations in the spreadsheet are written, log(R2/R1) is a directly proportional to thurst/power. So the amazingly good fit seems to me to be only a reflection of the fact that all devices are of similar size and operate at similar frequencies.
I had already pointed that out, explicitly, and discussed that here:
edit. thrust/power instead of thrust/weight as I'd originally written
http://forum.nasaspaceflight.com/index.php?topic=40959.msg1602526#msg1602526
and I already pointed out that the different Cannae geometry belongs in its own separate line here:
http://forum.nasaspaceflight.com/index.php?topic=40959.msg1602584#msg1602584
But what is the point of doing regression then?
Why not just divide thrust/power out? -
#1952 by as58 on 23 Oct, 2016 18:00
-
What is this particular 'statistical analysis' trying to achieve? It certainly isn't useful in predicting thrust/power, because the explanatory variable itself includes thrust/power. However, for a casual observer who doesn't know what R2/R1 means the plots definitely give that impression.
-
#1953 by Rodal on 23 Oct, 2016 18:09
-
What is this particular 'statistical analysis' trying to achieve? It certainly isn't useful in predicting thrust/power, because the explanatory variable itself includes thrust/power. However, for a casual observer who doesn't know what R2/R1 means the plots definitely give that impression.
I first made that very same point, and I made it explicitly, rather than using words.
Todd chose those variables. For me, formulas, plots and analysis speak for themselves more clearly than using words.
You are welcome to use words, or nothing (no discussion of the variables and the formula, no plots and no statistical analysis), if you prefer, to describe this.
I already pointed out a useful thing that came from the analysis (All of Yang results fall into line with the other other truncated cone results ). Here is another one: obviously the EM Drive data is not random or faked data, because it plots the way it is supposed to behave, when regressed. The different L, Cos[theta] and omega for Cannae fall on their own line, and as an outlier frm the truncated cone geometry. That does not mean that the EM Drive can necessarily be used for Space Propulsion, it just shows that the EM Drive data is not randomly generated.
The different geometries, were not always provided by the experimenters. As I recall only Paul March (NASA) and Tajmar (TU Dresden) have provided their actual geometry. The other geometries had to be painstakingly ascertained by the NSF team from a number of arguments. Yang's geometry was very controversial, and very difficult to ascertain. It is comforting to see that the regression confirms that the geometry calculated by the NSF team gives reasonable results. -
#1954 by demofsky on 23 Oct, 2016 18:22
-
Please notice that Todd updated his spreadsheet with the copper Cannae data and I updated the Log Log plot in:
http://forum.nasaspaceflight.com/index.php?topic=40959.msg1602555#msg1602555
This is very interesting, because it shows that the Cannae cavity has its own regression line slope, which is parallel and different from the truncated cone geometry of Shawyer/NASA/Yang/Tajmar and others.
So Todd's theory shows a fundamental difference between the Cannae design and the truncated cone design of Shawyer, which is due to the Cannae geometry and not due to whether it is superconducting.
There are two Log-Log lines: one for the truncated cone geometry and another one for the Cannae pillbox/flying-saucer geometry.
I also now calculated the Regression Line only taking into account the Truncated Cone geometry pioneered by Shawyer. The data gives an excellent R2=0.99
Thank you for doing the log-log plot & regression. That looks great!
I'm not 100% certain that air "inside" the frustum doesn't help the thrust by design. After all, in @Notsosureofit's model and in my model, there is an accelerated reference frame inside the frustum. The force acting on the frustum mass, is equal and opposite the force acting on the internal EM field equivalent mass. Having air inside, if it accelerates along with the EM field, adds a significant amount of mass to the equation, and therefore will add a significant amount of force to the EM drive. A compressed gas inside the frustum may be a necessary part of the design, to enhance the thrust, provided the kinetic energy of the gas' mass can be dissipated at the big end, before returning to the small end, or at least over many cycles. i.e., those data points may not be 100% anomalous data. There may be a real affect enhancement due to gas with high kinetic energy, inside the frustum.
First of all these plots are breathtaking!! This is truly a case of the data speaking for itself. Even Yang 3! (Now why Yang 3 should show up there, so nicely correlated, is quite another story... Mutter, mutter...)
However, if having a gas present is important to improve the mass it might be tricky to choose the right one. For instance, one could consider a heavy gas like Sulphur Hexafluoride but how would it affect Q and other EM characteristics? Or, as Rodal pointed out some time ago, one could aim for a maser effect using a gas like ammonia as they did in the early masers.
There are a lot of factors to use to tune this phenomena. Also, the latest Sawyer "bell" fustrum appears to create very intense conditions so that is another consideration as well! -
#1955 by as58 on 23 Oct, 2016 18:27
-
What is this particular 'statistical analysis' trying to achieve? It certainly isn't useful in predicting thrust/power, because the explanatory variable itself includes thrust/power. However, for a casual observer who doesn't know what R2/R1 means the plots definitely give that impression.
I first made that very same point, and I made it explicitly, rather than using words.
Todd chose those variables. For me, formulas, plots and analysis speak for themselves more clearly than using words.
You are welcome to use words, or nothing (no discussion of the variables and the formula, no plots and no statistical analysis), if you prefer.
I already pointed out a useful thing that came from the analysis (Yang 3).
Here is another one: obviously the EM Drive data is not random or faked data, because it plots the way it is supposed to behave, when regressed. The different L, Cos[theta] and omega for Cannae fall on their own line, and as an outlier form the truncated cone geometry. That does not mean that the EM Drive can necessarily be used for Space Propulsion, it just shows that the data is not random.
How could Yang 3 not be on the same line? It's the same geometry and frequency as the other two Yang's tests. And I don't get your point about not faked/random data: it shows that device dimensions and frequency used are not random, but nothing at all about thrust/power. -
#1956 by Rodal on 23 Oct, 2016 18:29
-
Yes, you are correct that all of the Yang results necessarily should fall into their own same line if they have the same length L, cosine of cone angle Cos[theta], and natural frequency.What is this particular 'statistical analysis' trying to achieve? It certainly isn't useful in predicting thrust/power, because the explanatory variable itself includes thrust/power. However, for a casual observer who doesn't know what R2/R1 means the plots definitely give that impression.
I first made that very same point, and I made it explicitly, rather than using words.
Todd chose those variables. For me, formulas, plots and analysis speak for themselves more clearly than using words.
You are welcome to use words, or nothing (no discussion of the variables and the formula, no plots and no statistical analysis), if you prefer.
I already pointed out a useful thing that came from the analysis (Yang 3).
Here is another one: obviously the EM Drive data is not random or faked data, because it plots the way it is supposed to behave, when regressed. The different L, Cos[theta] and omega for Cannae fall on their own line, and as an outlier form the truncated cone geometry. That does not mean that the EM Drive can necessarily be used for Space Propulsion, it just shows that the data is not random.
How could Yang 3 not be on the same line? It's the same geometry and frequency as the other two Yang's tests. And I don't get your point about not faked/random data: it shows that device dimensions and frequency used are not random, but nothing at all about thrust/power.
This follows from what I pointed out here:QuoteSince Log[R2/R1] is a linear function of Thrust/Power, times other factors (2L/Cos[theta]), it is not surprising that it shows a linear relationship with Thrust/Power:
-Log[R2/R1]= (Thrust/Power)* (2 omega * L/Cos[theta])
when the factor (2 omega * L/Cos[theta]) does not change much between different experiments
However, they will not necessarily fall into the same line as other results having different L, Cos[theta] and frequency. For example, the Cannae (both copper and superconducting) results fall into a different line. This makes sense. It appears as a trivial result if all the geometries would be well known and provided by the researchers, but unfortunately that was not the case.
The only experimenters that fully provided their geometry were NASA and TU Dresden (Tajmar). Tajmar's initially described geometry was in error, as pointed out by a NSF member, and later corrected by Tajmar as a result of the NSF member's communication.
The geometries (other than NASA's) had to be painstakingly ascertained by the NSF team from a number of arguments. Yang did not provide the complete geometry. Instead of making a statement about Yang 3, I should have made a statement about all the Yang results as a group: Yang 1, Yang 2 and Yang 3. Yang's geometry, particularly the Length (L) and the cone angle (theta), were very controversial, and very difficult to ascertain and to reach a consensus.
To refresh memories, a geometry for Yang was initially proposed that had a very different Length L and cosine angle. SeeShell even made a prototype for this initial geometry, and we had lots of discussion with SeeShells, Flyby and others concerning this controversy around Yang's geometry.
It is comforting to see that the regression confirms that the geometry calculated by the NSF team, for all the truncated cones, including Yang's results, gives reasonable results, that fall into approximately the same line as the other truncated cone geometries. And that the fundamentally different geometry for the Cannae drives falls into a different line.... it shows that device dimensions and frequency used are not random, but nothing at all about thrust/power.
Yes, your statement is correct and is in agreement with my above remark << Since Log[R2/R1] is a linear function of Thrust/Power, times other factors (2L/Cos[theta]), it is not surprising that it shows a linear relationship with Thrust/Power>> and thatQuoteThat does not mean that the EM Drive can necessarily be used for Space Propulsion, it just shows that the data is not random.
-
#1957 by RotoSequence on 23 Oct, 2016 18:31
-
What does the Log-Log R1 R2 axis measure?
-
#1958 by demofsky on 23 Oct, 2016 18:33
-
Have a little question about miniaturization. Idea is to use small thin copper foil frustum with very small mass.
Here is everything in picture.
Sorry if its...
The problem with this approach is that the small geometry necessitates the use of a high frequency like your proposed 24GHz. As a practical matter, microwave tech at these frequencies is lossy and the fustrum will also have a low Q. The issue is you may get thrust but it is so low (low Q, low energy due to losses) that it gets lost in the noise. Small vacuum chambers have, in my opinion, been the bane EM drive research. Also, having a gas present may be important to higher thrust.
-
#1959 by Rodal on 23 Oct, 2016 18:35
-
What does the Log-Log R1 R2 axis measure?
As chosen by Todd:
-Log[R2/R1]= (Thrust/Power)* (2 omega * L/Cos[theta])
I leave it up to Todd to explain/defend as a variable, but I think he was trying to find empirically what is the (1/Q)dQ/dr=d[ln Q]/dr functional relationship (due to geometry) in his equation (1/Q)dQ/dr
This relationship of Thrust/Power vs Ln[R2/R1] implies, besides geometry, the quality of resonance factor Q, which varies by several orders of magnitude between Tajmar's EM Drive and Shawyer's and Yang's, and does not appear as an explicit variable in the regression plot.
Q is a function of geometry, material, and mode shapes.
