EM Drive Developments - related to space flight applications - Thread 7

[Rodal]

Where do a couple of posters get the impression that sidewall tolerances are as important as endplate tolerances? If someone has an emdrive theory, please share it.
...

OK, we are in thread number 7 of the EM drive, let's recapitulate the EM Drive theories discussed in previous threads:

*Shawyer (the initiator of the EM Drive and frequently quoted by TheTraveller)
*McCulloch (who you have invited to post again in these posts)
*Notsosureofit
*DeAquino


All these theories have the claimed anomalous force/inputPower proportional to the Q quality of resonance.

All of them have this simple relationship, and as it is trivially shown, it is due to the Q that it is claimed that the EM Drive can exceed the force/PowerInput of a photon rocket.  We have gone over these myriads of times in this thread.

The quality of resonance relationship to the geometry (including the side walls) is known from countless textbooks.   Look at Collin's and Jackson's textbooks for example.  Furthermore, this has been shown in these EM Drive threads: as COMSOL Finite Element Analysis and FEKO Boundary Element Method analyses have shown the dependence on all geometrical parameters.  (Even the COMSOL analyses run by NASA have shown these dependence).

Let's start by asking, since you are posting about your DIY efforts, about your Q quality of resonance measurements.  Are you concerned about reaching a high Q in your tests? Did you measure Q for your tests? If so, what was the Q quality of resonance for your tests?

And looking at your future tests you are posting about, what % of the theoretical Q quality of resonance are you aiming for your in your DIY tests? 

As previously discussed, the discussion about tolerances has to do with its influence on the Q, quality of resonance.

Let's remmber that this is prompted by TheTraveller's post that Shawyer says that a tolerance of 13 micrometers is needed (for present EM Drive geometry and testing frequency) to get 75% of the theoretical Q quality of resonance and that's how this discussion of tolerance started. 

[Rodal]

Currently this site hosts half a dozen builders of EMDrives. Some are on their first or second or third  build (some are keeping quiet and just building), but the common thread throughout it all is the quality of the build has increased dramatically in the last year.  This is due to no small part by the mass contributions of so many here pushing the envelope of theory, creativity, and DYI quality. I never could have made it this far if it wasn't for the contributions and guidance and the heartfelt giving of others.

We need to thank those who offer their time, their expertise and encouragement to quantify whether this drive will be a curiosity in the annals of time or reinventing fire for humanity.

Thank You.


Shell

Shell, what do you think of rfmwguy's post below ?

Dr rodal, you seem to be bothered by diy methodology. I suggest no diyer has a correct methodology and neither do you. This topic has advanced well beyond being a roger shawyer comment thread to diy and theory beyond one person. My advice is to build one yourself and show others where errors are being made. Either that or publish a diy guide you would recommend to new diyers.

Rfmwguy posted this upon my questioning the 1 mm wall thickness he is using while simultaneously polishing the ends like a mirror, and after I saw TheTraveller quoting Shawyer's advice that the tolerance should be 13 micrometers.  I thought that the NSF threads were here for the user community to discuss NSF posts like these between NSF readers, rather than  "look at what I'm doing, but don't comment on what I'm doing, if you don't like what I'm doing, go and conduct your own DIY or go and write your own DIY guide."

What do you think of such comments on your DIY construction?

Is it worthwhile for readers not involved in DIY EM Drive to spend time on comments on DIY construction?

[Monomorphic]

Here is a comparison of distorting the dimensions of end-plates vs side-wall vs both. I would not have believed the results had I not run them myself: Resonance got stronger with side-wall distortion before it got weaker.

Not surprising, distorting the end-plates had an immediate effect on the strength of resonance. The effect was obvious with distortions as small as 1.5 - 2.0mm having up to 17% reduction in frustum E-field power. This effect seems to follow a linear pattern as distortion increases.

EDIT: Original image had a duplicate in end-plate distortion. Image was corrected

[Rodal]

Here is a comparison of distorting the dimensions of end-plates vs side-wall vs both. I would not have believed the results had I not run them myself: Resonance got stronger with side-wall distortion before it got weaker.

Not surprising, distorting the end-plates had an immediate effect on the strength of resonance. The effect was obvious with distortions as small as 1.5mm having up to 17% reduction in frustum E-field power. This effect seems to follow a linear pattern as distortion increases.

Thanks so much for running these comparisons.

More light and less heat 

This is enlightening.  We learn a lot from these comparisons.

Could you please further discuss what is the meaning of these numbers 0.1, 0.2, 0.4, 0.6 and how do they relate to distortion in mm ?  Did you run a random distortion? What is the expression used by FEKO to model a geometrical distortion and what was the maximum and average (mm) distortion for each of those numbers?

Thanks again

[Rodal]

Here is a comparison of distorting the dimensions of end-plates vs side-wall vs both. I would not have believed the results had I not run them myself: Resonance got stronger with side-wall distortion before it got weaker.

Not surprising, distorting the end-plates had an immediate effect on the strength of resonance. The effect was obvious with distortions as small as 1.5 - 2.0mm having up to 17% reduction in frustum E-field power. This effect seems to follow a linear pattern as distortion increases.

Why is it that the side wall distortion looks much more regular, and less random (more smooth) in the run for only side-wall distortion and it looks more random and higher amplitude for the run having both side wall and end plate distortion?

Is the amount of actual side wall distortion different for both runs?

Can you instead fix the amount of distortion so that the Max and Average distortion, and geometrical frequency (not time frequency) of distortion is the same in all runs?

[Monomorphic]

Could you please further discuss what is the meaning of these numbers 0.1, 0.2, 0.4, 0.6 and how do they relate to distortion in mm ?  Did you run a random distortion? What is the expression used by FEKO to model a geometrical distortion and what was the maximum and average (mm) distortion for each of those numbers?

I build the geometry and add the distortions in another program. 0.1 distortion is equal to 1.75mm and 0.6 distortion is right at 1cm. The distortion is random in all 3 dimensions applied to each vertex. 

[Monomorphic]

Why is it that the side wall distortion looks much more regular, and less random in the run for only side-wall distortion and it looks more random and higher amplitude for the run having both side wall and end plate distortion?

Is the amount of actual side wall distortion different for both runs?

Can you instead fix the amount of distortion so that the Max and Average distortion, and geometrical frequency (not time frequency) of distortion is the same in all runs?

The end-plates have more triangles (and vertices) in the geometry. So the distortion has more noise in it there.  I could add more geometry to the side wall.

Just noticed I have two identical images in there. Must have copied the wrong one. Will update the image with correct version.

 

[Rodal]

Why is it that the side wall distortion looks much more regular, and less random in the run for only side-wall distortion and it looks more random and higher amplitude for the run having both side wall and end plate distortion?

Is the amount of actual side wall distortion different for both runs?

Can you instead fix the amount of distortion so that the Max and Average distortion, and geometrical frequency (not time frequency) of distortion is the same in all runs?

The end-plates have more triangles (and vertices) in the geometry. So the distortion has more noise in it there.  I could add more geometry to the side wall.

Just noticed I have two identical images in there. Must have copied the wrong one. Will update the image with correct version.

I understand that the comparisons shown, only show the effect of geometrical tolerance on the electromagnetic fields.

Can you also run these geometrical comparisons to show the effect on theoretical Q quality of resonance?

[Flyby]

Could you please further discuss what is the meaning of these numbers 0.1, 0.2, 0.4, 0.6 and how do they relate to distortion in mm ?  Did you run a random distortion? What is the expression used by FEKO to model a geometrical distortion and what was the maximum and average (mm) distortion for each of those numbers?

I build the geometry and add the distortions in another program. 0.1 distortion is equal to 1.75mm and 0.6 distortion is right at 1cm. The distortion is random in all 3 dimensions applied to each vertex.

Must say , I'm rather surprised about the "robustness" of the resonance patterns. I was expecting it to be much more sensitive...
which leads me to draw conclusions on the micron-meter requirements... I don't think it is required to fall within those extreme precisions, unless you're going for the super, super efficiency...
But those are not the aim of most DIY builds. As it is now, we still need to establish whether or not it delivers a force...

[Rodal]

Why is it that the side wall distortion looks much more regular, and less random in the run for only side-wall distortion and it looks more random and higher amplitude for the run having both side wall and end plate distortion?

Is the amount of actual side wall distortion different for both runs?

Can you instead fix the amount of distortion so that the Max and Average distortion, and geometrical frequency (not time frequency) of distortion is the same in all runs?

The end-plates have more triangles (and vertices) in the geometry. So the distortion has more noise in it there.  I could add more geometry to the side wall.

Just noticed I have two identical images in there. Must have copied the wrong one. Will update the image with correct version.

So, as I understand it, the mesh has more nodes on the end-plates than on the side wall?

Could you run these comparisons with a mesh having the same (mm) distance between mesh nodes in the vertical longitudinal direction as the distance between nodes in the horizontal direction along the diameters?

[Monomorphic]

Can you also run these geometrical comparisons to show the effect on theoretical Q quality of resonance?

Sure.  I will need to figure out something with the antenna, as a distorted frustum wall does not make a very good connector! So I skipped that problem and didn't request S-parameters.  I can probably do it but will need a little time.

[Rodal]

Can you also run these geometrical comparisons to show the effect on theoretical Q quality of resonance?

Sure.  I will need to figure out something with the antenna, as a distorted frustum wall does not make a very good connector! So I skipped that problem and didn't request S-parameters.  I can probably do it but will need a little time.

This is really excellent work and very illuminating !

(Also great comment from Flyby about the robustness of the fields to distortion)

[Monomorphic]

Could you run these comparisons with a mesh having the same (mm) distance between mesh nodes in the vertical longitudinal direction as the distance between nodes in the horizontal direction along the diameters?

This is what the mesh looks like. Looks like all I need to do is double up on the horizontal edge loops to do this. I will run one or two and see if it makes a big difference.

[Rodal]

Could you run these comparisons with a mesh having the same (mm) distance between mesh nodes in the vertical longitudinal direction as the distance between nodes in the horizontal direction along the diameters?

This is what the mesh looks like. Looks like all I need to do is double up on the horizontal edge loops to do this. I will run one or two and see if it makes a big difference.

Does that mean doubling up the number of nodes in the vertical direction? (it looks like the distance between the nodes in the vertical longitudinal direction is smaller than the distance between nodes in the radial horizontal direction)

[Monomorphic]

Does that mean doubling up the number of nodes in the vertical direction? (it looks like the distance between the nodes in the vertical longitudinal direction is smaller than the distance between nodes in the radial horizontal direction)

This is the difference. It is impossible to get exact since a frustum expands towards one end.

[SeeShells]

Currently this site hosts half a dozen builders of EMDrives. Some are on their first or second or third  build (some are keeping quiet and just building), but the common thread throughout it all is the quality of the build has increased dramatically in the last year.  This is due to no small part by the mass contributions of so many here pushing the envelope of theory, creativity, and DYI quality. I never could have made it this far if it wasn't for the contributions and guidance and the heartfelt giving of others.

We need to thank those who offer their time, their expertise and encouragement to quantify whether this drive will be a curiosity in the annals of time or reinventing fire for humanity.

Thank You.


Shell

Shell, what do you think of rfmwguy's post below ?

Dr rodal, you seem to be bothered by diy methodology. I suggest no diyer has a correct methodology and neither do you. This topic has advanced well beyond being a roger shawyer comment thread to diy and theory beyond one person. My advice is to build one yourself and show others where errors are being made. Either that or publish a diy guide you would recommend to new diyers.

Rfmwguy posted this upon my questioning the 1 mm wall thickness he is using while simultaneously polishing the ends like a mirror, and after I saw TheTraveller quoting Shawyer's advice that the tolerance should be 13 micrometers.  I thought that the NSF threads were here for the user community to discuss NSF posts like these between NSF readers, rather than  "look at what I'm doing, but don't comment on what I'm doing, if you don't like what I'm doing, go and conduct your own DIY or go and write your own DIY guide."

What do you think of such comments on your DIY construction?

Is it worthwhile for readers not involved in DIY EM Drive to spend time on comments on DIY construction?

Dr. Rodal,

Undoubtedly there are some great theories out there proposed here by very sharp minds, also residing in this forum are great techs, engineers and PHd's and just dang smart people.

You really think I'd be here if the highlights of the day was chatting nonsense on the current picture of what I had for lunch?  The question is, would you be?

Sure I have friends with Facebook accounts and they post lovely pictures of a cat or a bird outside of their window and that's truly ok with me, but dude you know the same disagreeing and bickering and he said she said are present in both sites. It's like IQ or sharpness has nothing to do with normal human behavior.

Dave you were a little over the top criticizing Dr. Rodal like that, he has provided a wealth of information and we wouldn't be here in the same way if it wasn't for him.

Dr. Rodal, you can get your panties in a wad sometimes when someone steps on your toes. Dave is doing a great job trying to sift through all the information to do a great build.

You both have contributed way over the top and if I was closer I'd make you stand in the corner for 10.

We have way bigger things to discover here. The stars or not. So let's lighten up and listen and be respectful.

Now my cute puppy says please smile because I'm cute.

[rfmwguy]

Paul Kocyla in Aachen, Germany is beginning to test the 24 GHz emdrive on a rotary flotation pad. Unquantified force measurements, appears to be calibration tests:

[SeeShells]

Could you please further discuss what is the meaning of these numbers 0.1, 0.2, 0.4, 0.6 and how do they relate to distortion in mm ?  Did you run a random distortion? What is the expression used by FEKO to model a geometrical distortion and what was the maximum and average (mm) distortion for each of those numbers?

I build the geometry and add the distortions in another program. 0.1 distortion is equal to 1.75mm and 0.6 distortion is right at 1cm. The distortion is random in all 3 dimensions applied to each vertex.

Must say , I'm rather surprised about the "robustness" of the resonance patterns. I was expecting it to be much more sensitive...
which leads me to draw conclusions on the micron-meter requirements... I don't think it is required to fall within those extreme precisions, unless you're going for the super, super efficiency...
But those are not the aim of most DIY builds. As it is now, we still need to establish whether or not it delivers a force...

Thank you for this wonderful comment.
 

[Rodal]

Paul Kocyla in Aachen, Germany is beginning to test the 24 GHz emdrive on a rotary flotation pad. Unquantified force measurements, appears to be calibration tests:

Would appreciate somebody familiar with this testing program to clarify:

1) are the battery and the mini-EM-drive integrated together on the testing platform for the Kocyla test?  DeltaMass and I had agreed that by far the best proposed test was TheTraveller's proposal to have a battery and the EM Drive on a rotary platform together (rather than having the power be fed from a stationary source to a moving EM Drive which has a big testing flaw: the center of energy-mass is outside the moving EM Drive, therefore measuring an acceleration in such a test is flawed since in space the source of power would need to be in the same spaceship as the EM Drive)

2) What is the present testing platform arrangement?

[Monomorphic]

It's definitely the big end that can cause all the problems! Also, the higher frequency the distortion (how rough it is) the better. This is because 2.45Ghz cannot "see" features smaller than 1/2 the wavelength (~6cm). This means that frustums that are skewed or warped in large ways, perform worse than those with a rough surface of equal displacement. In my opinion gross geometric tolerances of 0.5mm (or less) should be a goal for DIYers.