-
#3960
by
X_RaY
on 19 Jul, 2016 18:19
-
Still not quite clear about how folks are estimating Q: My scaled down frustum has a Q of +/- 19,000 mode TE012 @ 2.450161 GHz. Dimensions SD = 12.183 cm, L = 17.544 cm and LD = 21.443 cm. This is indeed quite small when compared to Monomophics's wedge and TM311 frustum. Is this because of the different mode/frequency being stimulated? Or? With bigger cavity shouldn't Q increase? Thanks for the help. , FL
The Q of such a cavity depends on a number of variables, one of them is the mode shape/ field pattern. Size or better volume to surface ratio, frequency and surface resistivity have also an impact on the resulting Q. The pic below shows the different equations for TE (4.202) and TM (4.203)modes for the case of a cylindrical cavity resonator.
EDIT:
Please note that "μ" in the term Rs=√ωμ/2σ means μ
r≈1(for copper) not µ
0 ,
the value for σ is 5,80E+007S in the case of copper.
-
#3961
by
Monomorphic
on 19 Jul, 2016 18:45
-
Super easy to grind the tops of soft brass screws away and replace the old ones. Waveguide connection is much cleaner now.
-
#3962
by
FattyLumpkin
on 19 Jul, 2016 19:25
-
X_Ray, thanks for the education re Q! One can see the dramatic differences in Q values in NASA's frustum when tested in different modes and frequencies, and all without changing the cavity dimensions. Another interesting fact that NASA had difficulties getting more power into their frustum in TE012...Do you think this might have been due to antenna location? (attached) FL
-
#3963
by
X_RaY
on 19 Jul, 2016 19:45
-
X_Ray, thanks for the education re Q! One can see the dramatic differences in Q values in NASA's frustum when tested in different modes and frequencies ,and all without changing the cavity dimensions. Another interesting fact that NASA had difficulties getting more power into their frustum in TE012...Do you think this might have been due to antenna location? (attached) FL
Not really, in the related paper they stated that other modes are very close to the TE012 eigenmode, that means the mode was degenerated/mixed by another mode. Therefore a huge part of the input energy was transfered into the parasitic mode rather than in the mode of interest.
The antenna position and shape seems a good choice and it was backed by Comsol FEA as shown in the pic you have posted.
On the other hand I don't know really they wasn't able to put more power into the mode. If the antenna impedance don't match the input impedance of the cavity, a part of power would be reflected back into the load of the used isolator. Also the forward power was tunable, I dont know what amplification they had used during this TE012 experiment.
-
#3964
by
FattyLumpkin
on 19 Jul, 2016 20:15
-
Yes, they were only able to get 2.6 Watts into the frustum in TE012. I consider this of vital interest to find out why, since this mode generated the highest force/Watt measured in their experiments. I know PLL would help with this but is very pricey. A while back (IMS) I believe some here at NSF pointed to antenna design and I believe others mentioned position.
During his lecture Sonny said "couldn't get more than 2.6 Watts into (the frustum) in it's current configuration" Of course, he did not indicate what it was about the configuration of their setup that might need to be changed in order to get more power into the frustum in this mode. Other than the use of a PLL mechanism, do you have any ideas about how one might go about solving this problem "in simulation"? FL
-
#3965
by
X_RaY
on 19 Jul, 2016 20:25
-
Yes, they were only able to get 2.6 Watts into the frustum in TE012. I consider this of vital interest to find out why, since this mode generated the highest force/Watt measured in their experiments. I know PLL would help with this but is very pricey. A while back (IMS) I believe some here at NSF pointed to antenna design and I believe others mentioned position.
During his lecture Sonny said "couldn't get more than 2.6 Watts into (the frustum) in it's current configuration" Of course, he did not indicate what it was about the configuration of their setup that might need to be changed in order to get more power into the frustum in this mode. Other than the use of a PLL mechanism, do you have any ideas about how one might go about solving this problem "in simulation"? FL
There was no problems with the simulation as far as I know. To solve the "problem" it would be needed to choose the cavity dimensions in a way to be sure all other eigenmodes are as far away (in frequency) as possible.
This means weeks full of simulation and several test cavities
but yes this goal is reachable**.
**I say this due to some experience with problems like that related to microwave sensor applications.
-
#3966
by
SeeShells
on 19 Jul, 2016 20:39
-
-
#3967
by
X_RaY
on 19 Jul, 2016 20:45
-
FYI
Shell
Yes this is for the empty 1m x1m x 0.5m cavity. Later it was scaled to the one they used for the Brady cone report, therefore the relations are the same*
*The later cavity involved the dielectric isert, maybe it was another parasitic mode in this case?
-
#3968
by
SeeShells
on 19 Jul, 2016 21:10
-
FYI
Shell
Yes this is for the empty 1m x1m x 0.5m cavity. Later it was scaled to the one they used for the Brady cone report, therefore the relations are the same*
*The later cavity involved the dielectric isert, maybe it was another parasitic mode in this case?
This was one of the main reasons I focused on stabilizing the thermal issues with the cavity by creating captured endplates, therefor as the cavity expanded from heat the distance between the plates wouldn't shift along with the growth of the cavity. The cavity would just "slide past" the small endplate.
The next thing was to stabilize the magnetron output which I was finally able to do with a clean HV DC, current controlled DC heater and a water jacket of copper tubing for the magnetron. This gives me <1MHz stable bandwidth of variable high power out of the magnetron to use TE012.
I know it was a long way to get there but it's worth it.
My Best...
Shell
-
#3969
by
FattyLumpkin
on 19 Jul, 2016 21:31
-
Shell, how much heating would one expect say within an eleven second period, given 100 Watts "injected" into a cavity? Assuming the clean type of RF you have described from maggie or from dielectric antenna. Would thermal expansion of the frustum be to the extent to where the resonant characteristics of the cavity would be altered? (in such a short time-frame) K
-
#3970
by
Monomorphic
on 19 Jul, 2016 21:40
-
Sim of Fatty's frustum. You can see the mode interaction. I get TE012 at 2.44562Ghz (for just an instant). Can anyone recognize the mode that looks similar?
-
#3971
by
dustinthewind
on 19 Jul, 2016 22:01
-
Yes, they were only able to get 2.6 Watts into the frustum in TE012. I consider this of vital interest to find out why, since this mode generated the highest force/Watt measured in their experiments. I know PLL would help with this but is very pricey. A while back (IMS) I believe some here at NSF pointed to antenna design and I believe others mentioned position.
During his lecture Sonny said "couldn't get more than 2.6 Watts into (the frustum) in it's current configuration" Of course, he did not indicate what it was about the configuration of their setup that might need to be changed in order to get more power into the frustum in this mode. Other than the use of a PLL mechanism, do you have any ideas about how one might go about solving this problem "in simulation"? FL
This may not be true, but I wonder if they were not able to store a lot of power in this mode TE012 because it was efficiently transferring the energy into thrust. If one is doing work that is, then the power should be going some where. Unless it was just that the power was being stored in some other mode but I would think they could measure that.
Some part of me wonders if the mixing of modes at similar frequencies may be necessary.
-
#3972
by
SeeShells
on 19 Jul, 2016 22:02
-
Shell, how much heating would one expect say within an eleven second period, given 100 Watts "injected" into a cavity? Assuming the clean type of RF you have described from maggie or from dielectric antenna. Would thermal expansion of the frustum be to the extent to where the resonant characteristics of the cavity would be altered? (in such a short time-frame) K
There is much more than just the thermal expansion physics of slower growth. There are parts of the cavity that will grow (also warp) faster than others dependent of the mode selected and an small area warpage can be felt at the physical deforming speed in other parts of the frustum... which is the speed of sound in copper
http://hyperphysics.phy-astr.gsu.edu/hbase/tables/soundv.htmlWhich works out to 3560 m/s.
In other words a very small spot of thermal heating can effect other areas on the frustum very quickly < 9.363295880149813e-5 seconds
Shell
As monomorphic pointed out in a deforming frustum simulation in FEKO (great work btw) that the two factors that detune a cavity from generation of a mode are endplate spacing and Big End warpage.
-
#3973
by
dustinthewind
on 19 Jul, 2016 22:24
-
Sim of Fatty's frustum. You can see the mode interaction. I get TE012 at 2.44562Ghz (for just an instant). Can anyone recognize the mode that looks similar?
The frustum looks kind of blocky on the sides. Is that supposed to be that way? So your trying to subtract out a known mode and observe another mode hidden in this mode with less power?
If so subtracting out the known mode at some known power might make it more obvious.
-
#3974
by
Monomorphic
on 19 Jul, 2016 23:14
-
Speaking of distorting the frustum walls, it turned out tuning the wedge is really that simple. By building in the ability to deform the walls of the wedge, I have been able to not only tune the emdrive, but increase Q-factor.
Image of of the prototype system. Final system will be more robust and include suction cups to push and pull the walls both ways. Plan is to have the ability to deform all walls, except for the wall the magnetron is attached to as it is reinforced. This reminds me of adaptive optics in astronomy!
With just distorting the one wall seen in the image, I was able to increase RL to -31.60 and move the center freq several Mhz. Measured Q-factor was 7,921.
-
#3975
by
rfmwguy
on 19 Jul, 2016 23:18
-
-
#3976
by
Monomorphic
on 19 Jul, 2016 23:23
-
I also performed the "just push it" experiment proposed by WarpTech. When the wedge is moving I saw a slight decrease in RL for the three tests I did. In the chart, Light blue is stationary (RL-29.54dB), while dark blue is when the wedge is moving (RL-29.10dB).
-
#3977
by
dustinthewind
on 19 Jul, 2016 23:28
-
I also performed the "just push it" experiment proposed by WarpTech. When the wedge is moving I saw a slight decrease in RL for the three tests I did. In the chart, Light blue is stationary (RL-29.54), while dark blue is when the wedge is moving (RL-29.10).
Something else you might do is take the data-points set 1 and subtract them from data-points set 2. The result will be the difference between the two, which will amplify any differences. Was it moving in only one direction or osculating back and forth? Was the frustum experiencing torque on the side walls (change in dimension) or was it just the beam moving?
-
#3978
by
Monomorphic
on 19 Jul, 2016 23:38
-
I also performed the "just push it" experiment proposed by WarpTech. When the wedge is moving I saw a slight decrease in RL for the three tests I did. In the chart, Light blue is stationary (RL-29.54), while dark blue is when the wedge is moving (RL-29.10).
Something else you might do is take the data-points set 1 and subtract them from data-points set 2. The result will be the difference between the two, which will amplify any differences. Was it moving in only one direction or osculating back and forth? Was the frustum experiencing torque on the side walls (change in dimension) or was it just the beam moving?
I pushed the oil dampening reservoir so the motion would be fairly consistent. So the beam and emdrive were moving in one direction (forward), not oscillating. Total displacement was about 30cm over about 5 seconds. I think it is worth pointing out Shawyer noted a decrease in Q due to doppler effect.
-
#3979
by
Bob Woods
on 19 Jul, 2016 23:41
-
Speaking of distorting the frustum walls, it turned out tuning the wedge is really that simple. By building in the ability to deform the walls of the wedge, I have been able to not only tune the emdrive, but increase Q-factor.
Image of of the prototype system. Final system will be more robust and include suction cups to push and pull the walls both ways. Plan is to have the ability to deform all walls, except for the wall the magnetron is attached to as it is reinforced. This reminds me of adaptive optics in astronomy!
With just distorting the one wall seen in the image, I was able to increase RL to -31.60 and move the center freq several Mhz. Measured Q-factor was 7,921.
Impressive engineering approach.
but yes this goal is reachable**.
