-
#1700 by zellerium on 14 Jan, 2016 15:54
-
Yet, how does the Notsosureofit hypothesis address the conservation of energy issues and frame-indifference, as best articulated by Frobnicat in these threads?
We need another poem to continue the last one, as the conservation of energy and frame-indifference counter-arguments persist
Still slowly working on that.
The hypothesis, so far, treats only the static force on the cavity. I need to finish the dependence on acceleration next. (update soon) Then we can treat the above. Note that the conservation law in GR is NOT the same as in SR.
Did you notice that NSF user Zellerium, has listed in http://emdrive.wiki/Experimental_Results the results of his tests: California Polytechnic State Univ., San Luis Obispo, Zeller, Kraft, Echolstest at California Polytech University, claiming positive results for a cavity with constant cylindrical cross-section, using a dielectric insert ?
Length =0.180 m
Diameter = 0.1077 m
Force/Power = 2.22 mN/kW (this value, for a low Q=300, is 665 times better than a perfectly collimated photon rocket, similar to the values in NASA's report for the truncated cone with dielectric inserts)
Recall that NASA's report reported NO thrust without a dielectric insert (proof is found here: https://forum.nasaspaceflight.com/index.php?topic=39214.msg1469685#msg1469685 ), NASA's report only shows thrust using a dielectric insert.
Remember that we worked on a Notsosureofit formula for a cylindrical cavity with a dielectric insert?
EDIT:
1) I could not find an expression for a cavity with a cylindrical cross-section with asymmetric dielectric in the Notsosureofit hypothesis section on the EM Drive wiki: http://emdrive.wiki/@notsosureofit_Hypothesis. If there is such an expression, could you please include it in the Wiki?
2) Can you calculate what the Notsosureofit hypothesis would calculate for Zellerium's test at CalPoly for a cavity with constant cylindrical cross-section and compare with his results ?
Shawyer stopped using dielectric inserts, according to TheTraveller, because the dielectric insert lowers the quality factor of resonance Q.
Yet, the Q with a dielectric insert is easy to calculate, as the dissipation is just due to the tan delta of the material. The tan delta of HDPE used by NASA is very small (tan delta (HDPE) = 0.00031 @ 3 GHz). Hence the reduction in Q is not all that great. The low Q in CalPoly's test (much lower than the Q in NASA's test) may be due to other reasons, other than the HDPE dielectric. The main effect of the dielectric on Maxwell's equations is the reduction of the natural frequency.
It would be interesting to see what the Notsosureofit hypothesis predicts for the force/Power for the tests at CalPoly.
Please remember that the "thrust" observed in our experiment was not primarily in the axial direction. The pendulum twisted and did not swing forward very much.
However the deflections occured only when we predicted to be on resonance based on VNA analyses. No deflection occured when we moved off resonance.
The effects seemed to be too quick to be completely thermal. I would guess lorentz force, but have yet to calculate it.
We are about to resume testing this quarter. A microwave engineering professional has let us borrow a professionally made magnetron antenna and we are in the process of remaking the movable plate so that the dielectric is press fit onto the plate (no screws). We are also going to reduce the output power of the magnetron to completely remove arcing (which we were never able to do) and possibly reduce heating.
Dave: yes, I did.
-
#1701 by Rodal on 14 Jan, 2016 15:58
-
...
Please remember that the "thrust" observed in our experiment was not primarily in the axial direction. The pendulum twisted and did not swing forward very much.
However the deflections occured only when we predicted to be on resonance based on VNA analyses.
The effects seemed to be too quick to be completely thermal. I would guess lorentz force, but have yet to calculate it.
We are about to resume testing this quarter. A microwave engineering professional has let us borrow a professionally made magnetron antenna and we are in the process of remaking the movable plate so that the dielectric is press fit onto the plate (no screws). We are also going to reduce the output power of the magnetron to completely remove arcing (which we were never able to do) and possibly reduce heating.
Dave: yes, I did.
Would you agree that when and if you conclude that the 2 mN transverse force presently listed in http://emdrive.wiki/Experimental_Results for your test turns out to be explainable as a Lorentz force, that the data for your test in the Wiki should be changed to zero (0), since it is my understanding that the purpose of the table "force" entries in http://emdrive.wiki/Experimental_Results is to list only anomalous forces that cannot be explained as Lorentz forces or thermal forces or as other experimental artifacts unrelated to possible space propulsion of the kind proposed by EM Drive proponents.
If the force is a Lorentz force, the Lorentz force measurement of 2 mN in the transverse direction can be included in the Notes, for thoroughness, but the data in the table should be zero for the above mentioned reasons. -
#1702 by rfmwguy on 14 Jan, 2016 16:38
-
This would be the highest Lorentz force measurement I have seen, if true. Nothing Mr Li did even approached this level....
Please remember that the "thrust" observed in our experiment was not primarily in the axial direction. The pendulum twisted and did not swing forward very much.
However the deflections occured only when we predicted to be on resonance based on VNA analyses.
The effects seemed to be too quick to be completely thermal. I would guess lorentz force, but have yet to calculate it.
We are about to resume testing this quarter. A microwave engineering professional has let us borrow a professionally made magnetron antenna and we are in the process of remaking the movable plate so that the dielectric is press fit onto the plate (no screws). We are also going to reduce the output power of the magnetron to completely remove arcing (which we were never able to do) and possibly reduce heating.
Dave: yes, I did.
Would you agree that when and if you conclude that the 2 mN transverse force presently listed in http://emdrive.wiki/Experimental_Results for your test turns out to be explainable as a Lorentz force, that the data for your test in the Wiki should be changed to zero (0), since it is my understanding that the purpose of the table "force" entries in http://emdrive.wiki/Experimental_Results is to list only anomalous forces that cannot be explained as Lorentz forces or thermal forces or as other experimental artifacts unrelated to possible space propulsion of the kind proposed by EM Drive proponents.
If the force is a Lorentz force, the Lorentz force measurement of 2 mN in the transverse direction can be included in the Notes, for thoroughness, but the data in the table should be zero for the above mentioned reasons. -
#1703 by Rodal on 14 Jan, 2016 17:03
-
The important point is what is the purpose of the table in the http://emdrive.wiki/Experimental_Results :
This would be the highest Lorentz force measurement I have seen, if true. Nothing Mr Li did even approached this level....
Please remember that the "thrust" observed in our experiment was not primarily in the axial direction. The pendulum twisted and did not swing forward very much.
However the deflections occured only when we predicted to be on resonance based on VNA analyses.
The effects seemed to be too quick to be completely thermal. I would guess lorentz force, but have yet to calculate it.
We are about to resume testing this quarter. A microwave engineering professional has let us borrow a professionally made magnetron antenna and we are in the process of remaking the movable plate so that the dielectric is press fit onto the plate (no screws). We are also going to reduce the output power of the magnetron to completely remove arcing (which we were never able to do) and possibly reduce heating.
Dave: yes, I did.
Would you agree that when and if you conclude that the 2 mN transverse force presently listed in http://emdrive.wiki/Experimental_Results for your test turns out to be explainable as a Lorentz force, that the data for your test in the Wiki should be changed to zero (0), since it is my understanding that the purpose of the table "force" entries in http://emdrive.wiki/Experimental_Results is to list only anomalous forces that cannot be explained as Lorentz forces or thermal forces or as other experimental artifacts unrelated to possible space propulsion of the kind proposed by EM Drive proponents.
If the force is a Lorentz force, the Lorentz force measurement of 2 mN in the transverse direction can be included in the Notes, for thoroughness, but the data in the table should be zero for the above mentioned reasons.
1) Initially, this table was being curated by me, as I started this table in early EM Drive threads to bring some order to the confusion at that time, as to what forces had authors claimed.
2) The purpose of the force in the list was to list the "anomalous force" being claimed and not to list forces that can be explained as experimental artifacts.
3) The table http://emdrive.wiki/Experimental_Results went from being curated to being uncurated with the entry of rfmwguy's force. As I understand it rfmwguy relies on someone else's statistical analysis to be able to enter a numerical value for the anomalous force. My understanding is that the person doing the statistical analysis has agreed that the sample population is not large enough to arrive at a statistically strong conclusion (there is not even enough sample population to plot a well-formed histogram to strongly support the assumption of a particular statistical distribution). An informal agreement was reached at that point that in the case of DoItYourself experiments, that the author herself/himself would enter numerical data as they think best represents their beliefs of what they measured as "anomalous force".
4) With the reporting of a force in the transverse direction (instead of the axial direction) that the author himself suspects to be a Lorentz force instead of an anomalous force, the entries in the http://emdrive.wiki/Experimental_Results table acquire yet a new, different character. What is the purpose of dividing a transverse force/power, that is suspected to be due to a Lorentz artifact, by the force/power of a perfectly collimated photon rocket? What is the purpose of force entries in the table that the author himself suspects to be an experimental artifact?
It seems to me that we should agree, that if the author is going to be entering the numerical values, that those values should correspond to what herself/himself considers to be strictly due to anomalous forces and not to experimental artifacts.
What the author herself/himself suspects to be experimental artifacts like Lorentz forces, could be annotated, instead of being listed in the table itself. -
#1704 by zellerium on 14 Jan, 2016 18:14
-
...
...
Please remember that the "thrust" observed in our experiment was not primarily in the axial direction. The pendulum twisted and did not swing forward very much.
However the deflections occured only when we predicted to be on resonance based on VNA analyses.
The effects seemed to be too quick to be completely thermal. I would guess lorentz force, but have yet to calculate it.
...
If the force is a Lorentz force, the Lorentz force measurement of 2 mN in the transverse direction can be included in the Notes, for thoroughness, but the data in the table should be zero for the above mentioned reasons.
Yes I agree, and will change it as soon as I determine the source of this force.
-
#1705 by Rodal on 14 Jan, 2016 19:09
-
Thank you Mr. Li. There is only 1 small vertical drop of about 8 inches before it reaches the magnetron on the Frustum.
Hi rfmwguy, I watched your video. I think generally it is OK regarding Lorentz force. However, you'd better make the following changes, ..
Hi Mr.Li, we would appreciate your comments regarding the following:..Would you agree that when and if you conclude that the 2 mN transverse force presently listed in http://emdrive.wiki/Experimental_Results for your test turns out to be explainable as a Lorentz force, that the data for your test in the Wiki should be changed to zero (0), since it is my understanding that the purpose of the table "force" entries in http://emdrive.wiki/Experimental_Results is to list only anomalous forces that cannot be explained as Lorentz forces or thermal forces or as other experimental artifacts unrelated to possible space propulsion of the kind proposed by EM Drive proponents.
This would be the highest Lorentz force measurement I have seen, if true. Nothing Mr Li did even approached this level.
If the force is a Lorentz force, the Lorentz force measurement of 2 mN in the transverse direction can be included in the Notes, for thoroughness, but the data in the table should be zero for the above mentioned reasons.
Regarding your analysis of Lorentz forces, were your measurements carried out at (or did you analyze) power levels of 900 W as used by Zellerium ? (California Polytechnic State Univ., San Luis Obispo, Zeller, Kraft, Echols in http://emdrive.wiki/Experimental_Results )
Do you have any comments regarding whether the transverse force reported by Zellerium as 2 mN can be attributed to Lorentz force effects ?
For convenient reference, you can find all of Zellerium's papers, concerning his experiments, at the following link:Quote from: zellerium on 10/09/2015 05:45 PM
Hey everyone,
Papers have been published, the first details our investigation, the second outlines our newest proposal, and the third discusses the previous experiments. (The third hasn't changed much and could use more updating)
You can find them all on my linkedin: https://www.linkedin.com/in/kurtwadezeller
...
-Kurt -
#1706 by RFPlumber on 14 Jan, 2016 20:42
-
I have a question which I am sure have already been answered here before...
What is the force-over-time profile caused by thermal expansion? Say, we have a metal rod connected via an ideal force meter to a large mass (Fig 1). Now apply a high intensity pulse of inductive heating to a section of such rod. Will the force over time look more like Fig 2 or more like Fig 3? I honestly don't know and would likely pick Fig 2 if I had to choose. The rationale is that each small portion of energy taken in during heating (or lost during cooling) will force an expansion / compression, hence some amount of force, however small, will always be observed. Force magnitude will be proportional to change in the amount of stored thermal energy per unit time. Hence the force will be more during the period of intense heating, and them smaller and longer (and in the opposite direction) during subsequent cooling.
-
#1707 by RFPlumber on 14 Jan, 2016 21:36
-
With the oscillation in there swamping everything, I frankly couldn't confirm or deny that you have anything or nothing other than a noisy oscillator that responds well when you apply the high-voltage.
For the reddit trolls, this is neither a positive nor a null result. It's a characterization of a test environment showing measurement concerns prior to a test.
Btw, are we on the same page w.r.t. the expected magnitude of the abnormal force this rig was built to capture? To re-iterate, it was expected to be at least a few hundred uN. Are you saying there is so much noise that one will not be able to see such a change in force during the RF pulse? How come the ~300 uN high voltage pulse is visible then? Do you honestly think that if the timing of HV pulse is changed so the pulse happens somewhere in the middle of the RF pulse, then the electrostatic force will no longer be observable because of "noise"? It is hard to believe one may have such a prediction, but I can easily make this kind of run during the next serious of tests...
However, If one hopes to detect something like 30 uN, then I completely agree - this is not the right setup for the task. -
#1708 by rfmwguy on 14 Jan, 2016 21:39
-
A SLIGHT leak of recent EW information: http://www.qsconference.com/speakers-info.html
See Jeff Lee, one of the speakers at this conference scheduled early next month in Canada.
"From attending graduate school at NASA’s Johnson Space Center, and later returning there to lecture on interstellar travel – to advising the co-head of a NASA Blue Ribbon Panel investigating NASA’s Eagleworks’ Q-Thruster (EM Drive) claims – to being invited to confidential meetings at the U.S. Army Research Laboratory – Jeff Lee will take the audience on a fascinating, and yet equally improbable, journey through the experiences of being on the leading edge of Breakthrough Propulsion Physics, the quest for interstellar travel, and the search for a greater understanding of “extreme cosmological environments”."
Bold is my annotation -
#1709 by Rodal on 14 Jan, 2016 22:03
-
I have a question which I am sure have already been answered here before...
What is the force-over-time profile caused by thermal expansion? Say, we have a metal rod connected via an ideal force meter to a large mass (Fig 1). Now apply a high intensity pulse of inductive heating to a section of such rod. Will the force over time look more like Fig 2 or more like Fig 3? I honestly don't know and would likely pick Fig 2 if I had to choose. The rationale is that each small portion of energy taken in during heating (or lost during cooling) will force an expansion / compression, hence some amount of force, however small, will always be observed. Force magnitude will be proportional to change in the amount of stored thermal energy per unit time. Hence the force will be more during the period of intense heating, and them smaller and longer (and in the opposite direction) during subsequent cooling.
Can you please specify the boundary conditions on the left rod? Is it floating in space? Is it restrained somehow?
If the rod is floating in space, is its mass infinitesimal compared to the mass at the right? is the force measuring device essentially a spring rigidly attached to the rod and to the big mass? is there damping in the system? and if so what is the time scale of the heating period you are looking at compared to the period of oscillation of the rod's-mass-spring system ? -
#1710 by RFPlumber on 14 Jan, 2016 23:07
-
I have a question which I am sure have already been answered here before...
What is the force-over-time profile caused by thermal expansion? Say, we have a metal rod connected via an ideal force meter to a large mass (Fig 1). Now apply a high intensity pulse of inductive heating to a section of such rod. Will the force over time look more like Fig 2 or more like Fig 3? I honestly don't know and would likely pick Fig 2 if I had to choose. The rationale is that each small portion of energy taken in during heating (or lost during cooling) will force an expansion / compression, hence some amount of force, however small, will always be observed. Force magnitude will be proportional to change in the amount of stored thermal energy per unit time. Hence the force will be more during the period of intense heating, and them smaller and longer (and in the opposite direction) during subsequent cooling.
Can you please specify the boundary conditions on the left rod? Is it floating in space? Is it restrained somehow?
If the rod is floating in space, is its mass infinitesimal compared to the mass at the right? is the force measuring device essentially a spring rigidly attached to the rod and to the big mass? is there damping in the system? and if so what is the time scale of the heating period you are looking at compared to the period of oscillation of the rod's-mass-spring system ?
Well, this was supposed to be as ideal as possible just to simplify the case... I don't have a good feel for what particular constrains are important here. To-rephrase the same underlying question... Metal is being heated for some period of time during which its temperature keeps rising... So it is expanding... So its mass is moving in space... Does this produce force during the entire duration of heating or just at the on/off boundary?
...And thinking about it some more, I believe the right answer is that "this produces no force at all". Because at the end of the day, to move some mass from point A to point B (where starting velocity = end velocity = 0) requires zero average force over time (positive force to start moving it, then equal negative force to stop it). Applying this to thermal expansion being atoms pushing against each other to give them some more "room", it seems the average force on macro-scale timeline will be zero.
EDIT: Left side of the rod was meant free-floating. A linearly expanding metal constrained at both sides will definitely produce force
-
#1711 by frobnicat on 14 Jan, 2016 23:44
-
I have a question which I am sure have already been answered here before...
What is the force-over-time profile caused by thermal expansion? Say, we have a metal rod connected via an ideal force meter to a large mass (Fig 1). Now apply a high intensity pulse of inductive heating to a section of such rod. Will the force over time look more like Fig 2 or more like Fig 3? I honestly don't know and would likely pick Fig 2 if I had to choose. The rationale is that each small portion of energy taken in during heating (or lost during cooling) will force an expansion / compression, hence some amount of force, however small, will always be observed. Force magnitude will be proportional to change in the amount of stored thermal energy per unit time. Hence the force will be more during the period of intense heating, and them smaller and longer (and in the opposite direction) during subsequent cooling.
From similar discussion (In retrospect the signs and orientations conventions may not be the most obvious in the attached plots of this old post) :
Back to your Fig 1 above, with an "infinite" mass on the right and a small mass of m on the left, and a close to ideal force meter : If you are looking at recoil (reaction) forces from the mass on the left forced to accelerate leftward by the thermal expansion of connecting rod, then measured force F(t)=m×a(t).
To get anything even remotely close to a mesa as in you Fig 2 would mean (reversing all axis for convenience to manipulate positive values) F(t)>Fbase during the heating pulse, with Fbase being the lower magnitude recorded just before the end of heating pulse on your Fig 2
That would imply that to get the shape of Fig 2, that during all Heating pulse a(t)>=Fbase/m
=> v(t)>=(Fbase/m)×t starting at v(0)=0
=> L(t)>=0.5×(Fbase/m)×t˛+L(0)
With L(t) being the length of rod subject to heating at temperature T and α linear expansion coefficient
L(t)-L(0)=α×L(0)×(T(t)-T(0))
=> T(t)>=T(0)+((0.5×Fbase)/(m×α×L(0))×t˛
To get the kind of shape of Fig. 2 it would take something close to a very specific quadratic increase in temperature during heating pulse, followed by continued increase in temperature for quite some time as any discontinuity (or fast step) in the rate of T would derive to a short and strong opposite pulse of force (and not a smooth one as in your plot).
And working out the numbers is not favorable to an explanation (of experimental results) based on this kind of recoil. Just to give a rough idea, 20s stable above 20µN with m=4kg L(0)=0.3m α=17×10-6 (copper) gives a temperature increase of about 200°C at end of heating pulse, the rate is then about +20°C/s and can't have a short drop. Admittedly this is not plasma hot, but still hard to reconcile with known temperature increase of structural parts (as measured at EagleWorks for instance).
So, recoil induced force would probably look more like Fig 3., as only the change of rate of temperature increase induces forces (second derivative of temperature wrt. time). During a heating process at constant power input (and assuming little cooling for low ΔT initially) the rate of increase of temperature would be roughly constant. The length of rod expands at near constant speed, the pushed mass is "cruising" and there is no acceleration, hence no recoil force.
edit: unbalanced parenthesis
added:
A not ideal force measurement (not infinite stiffness of spring, time constants...) will smooth out the signal of recoil forces.
As specifically for EagleWorks balance, it is operated in a fashion that makes it record displacements (interpreted as forces) directly proportional to ΔL : F=cst×ΔL and I believe (from estimates) much higher (and in the ballpark of claimed results) than the recoil forces that are proportional to second derivative of ΔL. Unfortunately inconsistencies (yet unexplained) with the published data, and no specific test to address the question (controlled actuated small mass shift, 100$ budget max), makes it impossible to give a precise value to cst. (there is about an order of magnitude uncertainty). -
#1712 by Three1415 on 15 Jan, 2016 00:12
-
Greetings! Longtime lurker here. I noticed that, as of late, people have begun to despair over the persistence of nontrivial heating effects, and their invalidation, or at least reduction in quality, of much of the data we have on hand. So, I racked my brains for a possible way of alleviating the problem, and have come up with a fairly simple but hopefully effective method of dealing with convective effects.
Ideally, of course, we would be running these tests in a vacuum (no convection there...), but seeing as no one has a budget sufficient to accommodate a frustrum-sized vacuum chamber, what we must try to do instead is thermally isolate the exterior of the frustrum from the interior.
My solution was to have two nested frustrums, the inner one with RF energy being injected via waveguide (like Shell's build), and the outer one being separated from this by a thick layer of ice--see the attached diagram (I apologize in advance for my crude MS Paint drawings
). This could be done by simply placing the inner frustrum into the larger shell, then pouring water into the gap and letting it sit in a freezer overnight; no expensive components required.
Because water's heat of fusion is quite high, heat being dissipated by the frustrum should take a long time to melt completely through the ice, and while that is happening, the temperature of the outer surface should remain quite constant, perhaps changing slightly in response to the ambient temperature of the room. However, this does not matter, because at least it will be thermally isolated from the interior, where the less predictable things are happening, and so valid control runs should be possible with this setup.
The only reason I could think of that this build scheme would not work is if eddy currents were able to penetrate through the ice and reach the outer surface of the frustrum, heating it regardless of the separation, but given that magnetic fields rarely like to stray far from their sources (and particularly not through diamagnetic materials like water), I consider this unlikely.
Of course, I may be missing something obvious to those with more experience that would completely invalidate this procedure, but I feel that it is at least worth proposing, because it could eliminate one of the most irritating sources of error we have using only water and a freezer.
Thanks,
Three1415
-
#1713 by RFPlumber on 15 Jan, 2016 01:53
-
...
My solution was to have two nested frustrums, the inner one with RF energy being injected via waveguide (like Shell's build), and the outer one being separated from this by a thick layer of ice--see the attached diagram (I apologize in advance for my crude MS Paint drawings). This could be done by simply placing the inner frustrum into the larger shell, then pouring water into the gap and letting it sit in a freezer overnight; no expensive components required.
...
This would certainly help... however, one caveat is this design demands a perfectly hermetical frustum + perfectly hermetical routing of the waveguide through the outer shell... Ok, both are likely still feasible with a lot of silicon caulk.. A more serious concern is that there would still be air inside, and it could well be heated if not by the frustum itself, then by magnetron... And then all it takes is a single tiny gap somewhere around the magnetron to waveguide assembly in order tocreate an air jetobserve thrust.
... If I were to bother with an outer enclosure, and if there were no weight concerns, I would just go for a vacuum-tight chamber, and then would evacuate it to some reasonably low pressure before testing. Yes, I do have a vacuum pump...
-
#1714 by TheTraveller on 15 Jan, 2016 02:03
-
After much research, I've determined that this is the best way to construct a frustum...seamless at the small diameter:
This horizontal lathe was modified to include a spinning "form", or shape, that the copper sheet "folds" over. The left side of the lathe uses a compression disc that snugs the copper sheet onto the top of the form, which could be any shape. A frustum would be one of the more simpler forms to have constructed.
The brass-smith said he could make a bell/frustum but it would have a brazed seam that would have to be ground/polished away. More labor costs.
The cost is similar for both ways to do it. Looks like the spinning copper method is superior. Interior polishing would be needed.
Dave,
You also need to consider the frustum side wall to end plate interface. It also needs to appear to the Rf eddy currents as being seamless. Needs to look like 2 waveguide flanges mating together under high pressure to form a electrically seamless interface or bad stuff can happen and the Magic Smoke that makes all electronics work may leak out.
As the skin depth for copper at 2.45 GHz is approx 0.00000132947m, even a scratch on the interior surface may be deep enough to stop eddy currents forming across it and destroy a high Q. Roger's advice to me at the very start was to ensure there were NO scratches and / or dings / pits / bumps on any interior surface and that the interior surface was as mirror like as possible, with NO fingerprints.
-
#1715 by TheTraveller on 15 Jan, 2016 02:07
-
...
My solution was to have two nested frustrums, the inner one with RF energy being injected via waveguide (like Shell's build), and the outer one being separated from this by a thick layer of ice--see the attached diagram (I apologize in advance for my crude MS Paint drawings). This could be done by simply placing the inner frustrum into the larger shell, then pouring water into the gap and letting it sit in a freezer overnight; no expensive components required.
...
This would certainly help... however, one caveat is this design demands a perfectly hermetical frustum + perfectly hermetical routing of the waveguide through the outer shell... Ok, both are likely still feasible with a lot of silicon caulk.. A more serious concern is that there would still be air inside, and it could well be heated if not by the frustum itself, then by magnetron... And then all it takes is a single tiny gap somewhere around the magnetron to waveguide assembly in order tocreate an air jetobserve thrust.
... If I were to bother with an outer enclosure, and if there were no weight concerns, I would just go for a vacuum-tight chamber, and then would evacuate it to some reasonably low pressure before testing. Yes, I do have a vacuum pump...
One way to check for / monitor for air leaks (hot air jets) is to monitor frustum internal pressure, that is assuming your frustum is sealed. Which I will be doing. A thermal camera will also help to spot air leaks.
The EW alum frustum has a nice grove for a sealing O ring.
-
#1716 by zellerium on 15 Jan, 2016 04:34
-
...
The EW alum frustum has a nice grove for a sealing O ring.
I hadn't seen this picture from EW before, is it from a larger document?
-
#1717 by RFPlumber on 15 Jan, 2016 05:24
-
Here’s to successfully modified frustum which now happily resonates at 2,331 MHz. I removed 8 mm from the small end, resulting in a shorter central length and a bigger diameter of the small end, thus pushing the small end cut-off frequency further down and away from resonance. New cut-off freq is 2,256 MHz which is 75 MHz below test frequency.
These resonance cavities turn out to be remarkably predictable. Simulating new dimensions (in COMSOL) was showing a freq shift of +20 Mhz (2323 MHz-> 2343 Mhz). Actual as-measured freq moved by +19 MHz (2312 MHz->2331 Mhz). 80(?) years-old technologies rule.
New dimensions:
D_big: 264 mm (as before)
D_small: 162 mm (+4 mm)
L_center: 196 mm (-8 mm)
TE012 freq (MHz): Simulated: 2343, Actual: 2331.
Df: 0.69
Small end cut-off: 2,256 MHz.
Q factor (at -3 dB S11): 2300 (went down from the original 3100, not sure if this is due to coupling mismatch under new dimensions, or oxidation from minor torch work, or just aging since the last time it was measured).
If anyone knows why this frustum should not be producing thrust then speak now or forever hold your peace.
Ladies and Gentleman, you can now make your bets. (I have already simulated mode frequency shifts for difference thicknesses of HDPE disks at the small end, so this gives a good hint as to what my own prediction for this upcoming test is…)
-
#1718 by Tellmeagain on 15 Jan, 2016 05:30
-
Hi Mr.Li, we would appreciate your comments regarding the following:
Regarding your analysis of Lorentz forces, were your measurements carried out at (or did you analyze) power levels of 900 W as used by Zellerium ? (California Polytechnic State Univ., San Luis Obispo, Zeller, Kraft, Echols in http://emdrive.wiki/Experimental_Results )
Do you have any comments regarding whether the transverse force reported by Zellerium as 2 mN can be attributed to Lorentz force effects ?
For convenient reference, you can find all of Zellerium's papers, concerning his experiments, at the following link:
Hi Dr. Rodal,
Thank you for your calling. I took a look of the first paper of his three papers on his Linkedin page, and noticed from his Fig. 6 that his experiment suffered the same problem as that in the Tajmar experiment. Refer to my attached figure. From DC point of view, his circuit has two parallel "hot" wires at -5000V, which I labeled with one red line, that also serve as the filament power supply wires. He also has a ground (or return) route, which I labeled with a green line, at ground level or 0V. This great DC carrying loop will introduce Lorentz forces that will move his cylinder, in two aspects. One is that the current in the moving part of the circuit (labeled with blue dots) will interfere with the Earth's magnetic field. The other is that the current in the non-moving part of the circuit (labeled with yellow dots) will interfere with the Magnetron's magnets.
The reason why the cylinder only moves when in resonance is that, when out of resonance, the magnetron does not deliver much RF power, thus does not generate much RF power, thus does not consume much DC current. On the other hand, when the cylinder is in resonance, the DC current is much higher so Lorentz force is much higher.
He has a quick remedy to his experiment. Rfmwguy's approach is a good reference for him to make the change. 1. Remove that ground wire which connects the aluminium board to the microwave oven. 2. Add another ground wire, parallel to and twisted with, the two filament wires, that connects the microwave oven case to the magnetron case. Pay attention to the wire insulation voltage grade.
I predict after this change, his 2mN will be gone. Of course, only the experiment can tell.
-
#1719 by TheTraveller on 15 Jan, 2016 05:52
-
...
The EW alum frustum has a nice grove for a sealing O ring.
I hadn't seen this picture from EW before, is it from a larger document?
I made a Google drive archive of all Paul's NSF attachments.
You can find the image in the archive.
https://drive.google.com/drive/u/0/mobile/folders/0B7kgKijo-p0ifk9EakZfbW9aZGMwNWZMQ01xVnBON0tkM2w0Q1NLbmtjRFFwMXBuNVlVN0U/0B7kgKijo-p0ifi13QTRLNldVb2llY05DaU5XMXM1OHFrTHRYTlF3bWtKMVNKdTQyeGswTlE?usp=folder&sort=13&direction=a
). This could be done by simply placing the inner frustrum into the larger shell, then pouring water into the gap and letting it sit in a freezer overnight; no expensive components required.