-
#1140 by VAXHeadroom on 02 Jan, 2016 16:50
-
Well...here's my first shot at computing an endplate 'energy' picture from the meep data.
What model are you running?
I hesitate to call this the thermal signature because I'm guessing at the equation.
This picture was generated with the following algorithm:
(where each frame element is the <x,y,z> H field 3 vector)
for each frame 1..112
for each row
for each col
sum[row][col] += abs( VCos_Angle(data[frame][row][col] , z) ) * vlength(data[frame][row][col]);
This sum is then output as the height of a cylinder and colored as in the picture. Elements whose resulting value < 1e-6 are ignored. The "MAX" shown is automatically computed as the highest value of the sum array. The MIN is 0 to 3 decimal points as shown. Values are in 'meep units'.
This was my best guess at a reasonable way to compute this. If a different algorithm is desired, let me know.
This meep model file is from aero modeling SeeShells' device. It is the same data as used to generate the H field animations I've been posting.
It looks pretty darn close to both TM12 and TE12!
Doing this for the small end is pretty trivial. Doing it for the conical section is a son-of-a-gun, but I'm working on it
QuoteThis meep model file is from aero modeling SeeShells' device. It is the same data as used to generate the H field animations I've been posting.
(I thought that Shell and aero were expecting TE01 instead of TE12)(completely different mode !!! )
Actually the mode shown by the Meep model does not look like any of these modes, really...
That's the reason why I had suggested making this verification comparison
All Meep runs need verification to compare with reality The problem is not with Meep, the issue is with the particular models
pardon the word "garbage" which is not meant to be offensive, it is enshrined in computer science:
https://en.wikipedia.org/wiki/Garbage_in,_garbage_out
I'm a programmer, I totally understand GIGO
There are a LOT of steps in this chain, and some guesswork on my part, so each of those requires scrutiny.
FIRST and foremost in my mind: is the algorithm (posted above) correct?
Following right behind that is: did I implement it correctly? i.e. does this output accurately reflect what the meep model generates? If I have that right then we work back up the chain to the meep model itself.
One note: This looks to me to match the data I see in the animations. Of course part of the same tool chain is used (conversion of the CSV files output from the h5totxt to the POV-Ray include files using C++ code I wrote), so the errors could be common to that. I have manually spot-checked data values and found them to have been copied correctly from input to output, but peer review is a 'very good thing'.
I was never comfortable with the fact that the above graphic didn't show the square waveguides. And I was right to question that - this is the wrong Z slice! ZERO is the middle of the simulation grid (the way I used h5totxt) and therefore the 'big end' is -112 and the 'small end' is 58. Verified by running this against -113 (which shows basically nothing as all the data is <1e-6) and 59 also shows nothing. The above graphic was 34 so almost at the small end, but not quite..
Attached is the big end and small end graphics with the correct Z slices.
-
#1141 by Fugudaddy on 02 Jan, 2016 16:53
-
With all due respect to the work that @TheTraveller is doing, I think there's a problem relying on spreadsheets to count on for accurate data.
The problem isn't with the input, but rather that spreadsheets have several inherent issues in regards to accuracy. Unless one knows the underlying code, one is 'at the mercy' of the people who programmed that tool. That's true, of course, of any piece of software (including MEEP, et al). However with spreadsheets, the problems are magnified.
Spreadsheets are great for providing overviews and oversight, but fail at finer levels of accuracy. There's 'size/placement of application and data in memory', 'floating point processing', hell, how the programmers implemented certain stacks and other issues that get amplified the more complex the calculations become. There was a FP error in Excel that existed for years and years before anybody tripped across it. The input is simply 'given over' to the code to process.
Mathematical processing software has the advantage that the user inputting the data is also responsible for making sure things are processed in the right order. This means that errors in calculating can be discerned, code corrected, and data processed again. It is a heck of a lot more tedious and its biggest problem is that one has to really understand the maths and get them all right in order to ensure accuracy.
Don't get me wrong- modern spreadsheets are really good, but they have some serious limitations especially in regards to complex processing. They're not designed for that however; they're designed as general business tools. For general engineering, they're fine. Good enough, IMO, to produce accurate enough data that a proof of direction (ie. there is a significant force from an EM drive that it goes past margin of error) is possible.
To use them to try and figure out exactly what is happening from a physics vs. engineering POV? Yeah nah, mate. Best to rely on the raw math tools for that work.
This debate on the maths is more religious than necessary. We'll wait and see what data comes from the DIY experiments, where there continues to be hints and smells, but nothing of 'proof' yet. We'll also wait and see what data comes from the math/physics folks, as they've been pushing new boundaries of understanding as to WTF is going on with this thing. Like the DIY, that there continues to be hints and smells of something interesting, but again still no proof.
As different theory have been proposed, new math and new designs are worked and new data is produced and the maths and designs are tightened even further, and we're all coming one step closer to learning if this is a real effect or, if you'll forgive me, just a bunch of hot air.
Peace and data in the new year.
Ronald
-
#1142 by Rodal on 02 Jan, 2016 17:38
-
...
Interesting !
I was never comfortable with the fact that the above graphic didn't show the square waveguides. And I was right to question that - this is the wrong Z slice! ZERO is the middle of the simulation grid (the way I used h5totxt) and therefore the 'big end' is -112 and the 'small end' is 58. Verified by running this against -113 (which shows basically nothing as all the data is <1e-6) and 59 also shows nothing. The above graphic was 34 so almost at the small end, but not quite..
Attached is the big end and small end graphics with the correct Z slices.
So it looks like Shell's experiment should not give a TE013 mode, according to this Meep model. It looks like m=1 rather m=0 as you previously stated...
As you (and others) previously stated it looks like the circumferential symmetry may be broken because of the dual waveguides used by Shell, at least at the early time modeled by Meep (is it 0.01 microseconds ?)
Wish I had the time to check this with Mathematica, which I don't at present. Meanwhile I can only thank you, VaxHeadRoom, for devoting your time to this
-
#1143 by rfmwguy on 02 Jan, 2016 17:49
-
General cleanup of some recent posts, friendly reminders apparently not enough.
Escalation of 1-1 should be PM only
-
#1144 by VAXHeadroom on 02 Jan, 2016 18:11
-
...
Interesting !
I was never comfortable with the fact that the above graphic didn't show the square waveguides. And I was right to question that - this is the wrong Z slice! ZERO is the middle of the simulation grid (the way I used h5totxt) and therefore the 'big end' is -112 and the 'small end' is 58. Verified by running this against -113 (which shows basically nothing as all the data is <1e-6) and 59 also shows nothing. The above graphic was 34 so almost at the small end, but not quite..
Attached is the big end and small end graphics with the correct Z slices.
So it looks like Shell's experiment should not give a TE013 mode, according to this Meep model. It looks like m=1 rather m=0 as you previously stated...
As you (and others) previously stated it looks like the circumferential symmetry may be broken because of the dual waveguides used by Shell, at least at the early time modeled by Meep (is it 0.01 microseconds ?)
Wish I had the time to check this with Mathematica, which I don't at present. Meanwhile I can only thank you, VaxHeadRoom, for devoting your time to this
You're quite welcome. FYI, the value in the middle of the big-end grid is about 0.8, it only goes above 1.0 around the edges and then only in isolated places. The large values at the corner of the waveguide is expected, but somewhat unfortunate in this instance as it swamps the scale.
The small end is where it gets interesting as the values are near the max value for quite a large area of the plate... -
#1145 by rfmwguy on 02 Jan, 2016 21:46
-
additional thread cleanup completed by nsf staff. I could not restore posts. Any new, specific data, not appearing in T1-T6 should be restated
-
#1146 by bprager on 02 Jan, 2016 22:07
-
...
Has it ever been discussed, that we might have accidently stumbled over some kind of a "Maxwell Daemon"?: http://physicsworld.com/cws/article/news/2010/nov/19/information-converted-to-energy
Lately I've been focused on just the type 3 cavity and really thinking hard about all of the properties of the air and why it seems important to thrust (Shawyer did away with the dielectric). It must be the related to the collective natural motion of the molecules in the cavity, but perturbed by some still unknown method. No doubt, that if the reports of thrust are not due to some artifact, there must be an explanation; maybe even one of the above. I don't know what could be behind the thrust with any certainty. What I'm proposing with the sound is (in my view) agnostic to the underlying thrust mechanism. It is simply adding energy to the system. If an anisotropy is present, I believe this may be a way to expose it by bringing forces up to a measurable level.
-
#1147 by SeeShells on 02 Jan, 2016 22:10
-
Great minds...I agree shell. I really think people are attracted here for something new. New tests, theories, data and pics (hahaha)I read the deleted posts and I can't help but don't see any reason for this "cleaning" campaign at all
I'm not taking sides although I think being open to see both sides is important.
I know moderators read and even comment on other postings in other threads and I'm sure due to the nature of this thread you can be assured this one is as well.
Another note...
We all are waiting for good solid data to hit, (me probably more than anyone here). I wonder how many visit more than once a day hoping to read about some test, some computational results, so we don't loose touch on this hot topic.
The stakes could be enormous if you consider what Paul March posted in his bombshell "And yet the anomalous thrust signals remain...".
As I said before, you OWE us and we want data...actually, you owe us nothing and we are grateful you are out there tearing apart your home to make room for an experiment just to give us data. For that, I commend you. Can't wait to see it...and yes, I didn't spill the beans
I have some equipment to order this week. Getting a set of digital logging scales because this last time convinced me that there was no way I could monitor the display by videoing, the blurring the digital scale display made it almost unreadable. Still plan of videoing the new digital scales.
I have a few other mods I'm going to be doing, but working on delivery times and costs. When I have those figures down better I'll let everyone know what it is I'm doing and how I'm doing it. -
#1148 by Andy USA on 02 Jan, 2016 22:53
-
This thread is only required for EM Drive discussion. Posts not discussing EM Drive will be removed.
-
#1149 by rfmwguy on 02 Jan, 2016 23:00
-
Nice work shell. Before you do the digital scale thing, did you consider beam deflection monitoring via laser displacement sensor?
Great minds...I agree shell. I really think people are attracted here for something new. New tests, theories, data and pics (hahaha)I read the deleted posts and I can't help but don't see any reason for this "cleaning" campaign at all
I'm not taking sides although I think being open to see both sides is important.
I know moderators read and even comment on other postings in other threads and I'm sure due to the nature of this thread you can be assured this one is as well.
Another note...
We all are waiting for good solid data to hit, (me probably more than anyone here). I wonder how many visit more than once a day hoping to read about some test, some computational results, so we don't loose touch on this hot topic.
The stakes could be enormous if you consider what Paul March posted in his bombshell "And yet the anomalous thrust signals remain...".
As I said before, you OWE us and we want data...actually, you owe us nothing and we are grateful you are out there tearing apart your home to make room for an experiment just to give us data. For that, I commend you. Can't wait to see it...and yes, I didn't spill the beans
I have some equipment to order this week. Getting a set of digital logging scales because this last time convinced me that there was no way I could monitor the display by videoing, the blurring the digital scale display made it almost unreadable. Still plan of videoing the new digital scales.
I have a few other mods I'm going to be doing, but working on delivery times and costs. When I have those figures down better I'll let everyone know what it is I'm doing and how I'm doing it.
I found this simple and easy to integrate and rebound or resistance forces are nil since its a laser spot only. I believe a scale will add some degree of attenuation to movement once your emdrive fires up...may not much, but some.
Either way your data should be great. Lds is just a non-contact method. -
#1150 by SeeShells on 02 Jan, 2016 23:45
-
The scales and the laser deflection system are still the same. Both will operate independently of each other, data from pressure recorded by the digital scales. And on acceleration via the laser.
Nice work shell. Before you do the digital scale thing, did you consider beam deflection monitoring via laser displacement sensor?
Great minds...I agree shell. I really think people are attracted here for something new. New tests, theories, data and pics (hahaha)I read the deleted posts and I can't help but don't see any reason for this "cleaning" campaign at all
I'm not taking sides although I think being open to see both sides is important.
I know moderators read and even comment on other postings in other threads and I'm sure due to the nature of this thread you can be assured this one is as well.
Another note...
We all are waiting for good solid data to hit, (me probably more than anyone here). I wonder how many visit more than once a day hoping to read about some test, some computational results, so we don't loose touch on this hot topic.
The stakes could be enormous if you consider what Paul March posted in his bombshell "And yet the anomalous thrust signals remain...".
As I said before, you OWE us and we want data...actually, you owe us nothing and we are grateful you are out there tearing apart your home to make room for an experiment just to give us data. For that, I commend you. Can't wait to see it...and yes, I didn't spill the beans
I have some equipment to order this week. Getting a set of digital logging scales because this last time convinced me that there was no way I could monitor the display by videoing, the blurring the digital scale display made it almost unreadable. Still plan of videoing the new digital scales.
I have a few other mods I'm going to be doing, but working on delivery times and costs. When I have those figures down better I'll let everyone know what it is I'm doing and how I'm doing it.
I found this simple and easy to integrate and rebound or resistance forces are nil since its a laser spot only. I believe a scale will add some degree of attenuation to movement once your emdrive fires up...may not much, but some.
Either way your data should be great. Lds is just a non-contact method.
There shouldn't be any difference in the data from pressures or thrust vs acceleration but it is something that needs to be tested for.
Shell -
#1151 by birchoff on 03 Jan, 2016 00:02
-
...
I guess I can't see why, when you have just developed a nifty propulsion system that needs no propellant, you would power it with a power system that needs to be supplied with fresh reactants after a few weeks of use.
To my knowledge, there are no current spacecraft which use fuel cells. The Space Shuttle did (as did the Apollo and Gemini spacecraft), but this was because the mission duration was short, and the by-product of the FCs was potable water for the crew.
The reason for using a Fuel Cell is not because you intend to always use one. But because it would immediately allow you to prove out different mission types that were previously not possible before in the chemical rocket regime. Of all the information I have seen about RTG's none of them are rated at 1KWe. I think their was an experimental proposal using a sterling engine instead of a thermoelectric generator that was in the 1KWe+ range but not sure.
But the point I was trying to make was we have the ability to deploy multi kilowatt power systems in space using fuel cells and solar cells. If someone delivers a working EmDrive. the last thing we will be dependent on is power. Even if it means powering the thing with fossil fuels. A Working EmDrive gets rid of a whole slew of problems we have with getting things into space. Assuming its initial usage is not limited to only government use cases. I dont see anyone waiting around for LENR much less thinking about it as being a requirement for usage. -
#1152 by SeeShells on 03 Jan, 2016 00:30
-
Nice work shell. Before you do the digital scale thing, did you consider beam deflection monitoring via laser displacement sensor?
Great minds...I agree shell. I really think people are attracted here for something new. New tests, theories, data and pics (hahaha)I read the deleted posts and I can't help but don't see any reason for this "cleaning" campaign at all
I'm not taking sides although I think being open to see both sides is important.
I know moderators read and even comment on other postings in other threads and I'm sure due to the nature of this thread you can be assured this one is as well.
Another note...
We all are waiting for good solid data to hit, (me probably more than anyone here). I wonder how many visit more than once a day hoping to read about some test, some computational results, so we don't loose touch on this hot topic.
The stakes could be enormous if you consider what Paul March posted in his bombshell "And yet the anomalous thrust signals remain...".
As I said before, you OWE us and we want data...actually, you owe us nothing and we are grateful you are out there tearing apart your home to make room for an experiment just to give us data. For that, I commend you. Can't wait to see it...and yes, I didn't spill the beans
I have some equipment to order this week. Getting a set of digital logging scales because this last time convinced me that there was no way I could monitor the display by videoing, the blurring the digital scale display made it almost unreadable. Still plan of videoing the new digital scales.
I have a few other mods I'm going to be doing, but working on delivery times and costs. When I have those figures down better I'll let everyone know what it is I'm doing and how I'm doing it.
I found this simple and easy to integrate and rebound or resistance forces are nil since its a laser spot only. I believe a scale will add some degree of attenuation to movement once your emdrive fires up...may not much, but some.
Either way your data should be great. Lds is just a non-contact method.
Send me the info, I'll look at it again. Still trying to resurrect and file all the data again. -
#1153 by rfmwguy on 03 Jan, 2016 01:07
-
Got to admit, the subscriber or L2 section of nsf would make your eyes light up regarding future designs in the works not only by nasa but subcontractors like bigelow.. Some of it is nsf exclusive info: http://forum.nasaspaceflight.com/index.php?board=29.0...
I guess I can't see why, when you have just developed a nifty propulsion system that needs no propellant, you would power it with a power system that needs to be supplied with fresh reactants after a few weeks of use.
To my knowledge, there are no current spacecraft which use fuel cells. The Space Shuttle did (as did the Apollo and Gemini spacecraft), but this was because the mission duration was short, and the by-product of the FCs was potable water for the crew.
The reason for using a Fuel Cell is not because you intend to always use one. But because it would immediately allow you to prove out different mission types that were previously not possible before in the chemical rocket regime. Of all the information I have seen about RTG's none of them are rated at 1KWe. I think their was an experimental proposal using a sterling engine instead of a thermoelectric generator that was in the 1KWe+ range but not sure.
But the point I was trying to make was we have the ability to deploy multi kilowatt power systems in space using fuel cells and solar cells. If someone delivers a working EmDrive. the last thing we will be dependent on is power. Even if it means powering the thing with fossil fuels. A Working EmDrive gets rid of a whole slew of problems we have with getting things into space. Assuming its initial usage is not limited to only government use cases. I dont see anyone waiting around for LENR much less thinking about it as being a requirement for usage.
This section has all of emerging tech including spacecraft power management. Think there's a trial subscription available to check it out. If emdrive turns into a workable tool, somethings got to fire it up! -
#1154 by TheTraveller on 03 Jan, 2016 04:17
-
This is a very custom design job. A very rough sketch of my design is attached.
This design needs no circulator and Rf dummy load as the real time Rf environment is monitored at least 100 times a second and will never allow the Rf amp to be over stressed. Rf power is programmed from approx 200mW to 100W with real time Forward and Reflected power feedback. If reflected power exceeds programmed limits, power output levels are dropped back in real time.
I am envy of your RF amp. What is it? Who makes them?
Also, what are you using the 3 way splitter for? Just to excite 3 separate loops at the same time?
@ TT - I second the question re: your RF amp.
http://m.alibaba.com/product/60125990298/100w-High-Power-Broadband-Power-amplifier.html
In bed & out of action for a few days. Seems Christmas and New Years celebrations and my radiation treated gut do not agree. -
#1155 by TheTraveller on 03 Jan, 2016 04:33
-
With all due respect to the work that @TheTraveller is doing, I think there's a problem relying on spreadsheets to count on for accurate data.
The problem isn't with the input, but rather that spreadsheets have several inherent issues in regards to accuracy. Unless one knows the underlying code, one is 'at the mercy' of the people who programmed that tool. That's true, of course, of any piece of software (including MEEP, et al). However with spreadsheets, the problems are magnified.
Spreadsheets are great for providing overviews and oversight, but fail at finer levels of accuracy. There's 'size/placement of application and data in memory', 'floating point processing', hell, how the programmers implemented certain stacks and other issues that get amplified the more complex the calculations become. There was a FP error in Excel that existed for years and years before anybody tripped across it. The input is simply 'given over' to the code to process.
Mathematical processing software has the advantage that the user inputting the data is also responsible for making sure things are processed in the right order. This means that errors in calculating can be discerned, code corrected, and data processed again. It is a heck of a lot more tedious and its biggest problem is that one has to really understand the maths and get them all right in order to ensure accuracy.
Don't get me wrong- modern spreadsheets are really good, but they have some serious limitations especially in regards to complex processing. They're not designed for that however; they're designed as general business tools. For general engineering, they're fine. Good enough, IMO, to produce accurate enough data that a proof of direction (ie. there is a significant force from an EM drive that it goes past margin of error) is possible.
To use them to try and figure out exactly what is happening from a physics vs. engineering POV? Yeah nah, mate. Best to rely on the raw math tools for that work.
This debate on the maths is more religious than necessary. We'll wait and see what data comes from the DIY experiments, where there continues to be hints and smells, but nothing of 'proof' yet. We'll also wait and see what data comes from the math/physics folks, as they've been pushing new boundaries of understanding as to WTF is going on with this thing. Like the DIY, that there continues to be hints and smells of something interesting, but again still no proof.
As different theory have been proposed, new math and new designs are worked and new data is produced and the maths and designs are tightened even further, and we're all coming one step closer to learning if this is a real effect or, if you'll forgive me, just a bunch of hot air.
Peace and data in the new year.
Ronald
Frustum TE013 resonances predicted by my spreadsheet have been checked against SPR's inhouse system for 3 frustum designs and found to agree to around +-0.3%.
Additionally, resonance agreement have been shown as against VNA resonance scan data versus frustum dimensions for all known DIY builds that I have the scans and dimensions for.
Should add that Roger advised either the freq or the frustum length or both need to be adjustable to obtain the best Force generation result as manufacturing errors will always occur. -
#1156 by TheTraveller on 03 Jan, 2016 06:20
-
"1/2 Loop Antenna"
During my conversations with Roger, I did learn the Flight Thruster used a loop antenna. Stub antennas (as in attachment 5) can get burnt as Shell and EWs discovered.
My design is like the 1st attachment and not like the 2nd or 3rd.
This 1/2 loop also fits the design of the multi loop coupler I plan to test as in the 4th attachment. That design may grow to 4 - 8 individually driven loops, equally spaced and installed at the same radius from the big end centre. Or maybe the 2 or 3 or 4 loop coupler will prove to do the job well enough. Nice feature about the internally connected 2 or 3 or 4 x 1/2 loops is there is no need for an external coax splitter and multiple install sites in the big end.
Ideally only the 1/2 loop antenna should be inside the frustum, so no part of the N type connector outer ground fitting should protrude inside the frustum, nor should any connector mounting bolts, nuts or screws protrude inside the frustum.
-
#1157 by TheUberOverLord on 03 Jan, 2016 10:50
-
I think I may have figured out a way to increase thrust. It involves resonating sound waves within the cavity simultaneously with the rf.
I've been thinking about the matter inside the cavity for a while now and about how thrust seems to disappear in vacuum or when the dielectric insert was removed as reported by EW. Lots of microscopic effects have been considered, like Casimir momentum, magnetoelectric anisotropies, and lately gravitomagnetism. None appear capable of generating or explaining macroscopic forces.
Lately I've been focused on just the type 3 cavity and really thinking hard about all of the properties of the air and why it seems important to thrust (Shawyer did away with the dielectric). It must be the related to the collective natural motion of the molecules in the cavity, but perturbed by some still unknown method. No doubt, that if the reports of thrust are not due to some artifact, there must be an explanation; maybe even one of the above. I don't know what could be behind the thrust with any certainty. What I'm proposing with the sound is (in my view) agnostic to the underlying thrust mechanism. It is simply adding energy to the system. If an anisotropy is present, I believe this may be a way to expose it by bringing forces up to a measurable level.
Would it not be interesting to see the cavity pressurized at normal atmospheric levels and then times x atmospheric levels. Using our typical atmosphere components as well as trying other gases in both normal and vacuum testbed environments? Down the road, but soon?
After all. Since the contents of the cavity may have already shown decreases and increases in measured thrust output. Might it not be possible that the contents of the cavity could increase or decrease the magnitude of thrust many times over? Depending on what the contents of the cavity contains. In both a vacuum and at normal atmospheric pressures.
Since it might be possible that depending on the contents of the cavity. More thrust in a vacuum might be created than at normal atmospheric pressures.
If particles in the cavity are somehow involved in this thrust process. Would it not be worthwhile to be thinking about this for future tests?
Note: One thing that always perplexed me about the vacuum tests so far. Was that there was no test data on if the cavity had normal atmospheric contents and pressure in the vacuum. Would the thrust results have remained as low as they were?
Would love to see a test with a cavity using normal atmospheric pressure but the cavity contents filled with Hydrogen and then compare the results using the same EM Drive setup with the cavity contents replaced with normal atmospheric contents at the same atmospheric pressure.
Instead of water or air, think cavity contents. If thrust has anything to do with cavity contents, pressures and/or the resonance of the particles in the cavity contents:
Water tube resonance experiment Closed Cylinder Air Column
Water: http://hyperphysics.phy-astr.gsu.edu/hbase/waves/clocol2.html#c2
Air: http://hyperphysics.phy-astr.gsu.edu/hbase/waves/clocol.html
Don -
#1158 by rfmwguy on 03 Jan, 2016 11:47
-
Well...a cavity filled with nitrogen rather than hydrogen might be a better idea
-
#1159 by TheUberOverLord on 03 Jan, 2016 11:52
-
Well...a cavity filled with nitrogen rather than hydrogen might be a better idea
Yes, but I was thinking along the lines to choose the smallest atom to fill the cavity contents. Which is why I suggested Hydrogen. But any gas and pressures used. Would be an interesting additional test using the same EM Drive testbed. In very safe and controlled tests. If that's reasonably possible.
Whereas, nitrogen and oxygen are very close in atomic weight. Hydrogen has the smallest atomic weight. I'm thinking along particle involvement and or/resonance and trying to see if any difference in thrust can be shown and demonstrated, by changing the sizes of the atoms involved in EM Drive cavity contents using different pressures.
Don
