It is an ongoing process where one has to continuously go back and check and double-check.
As the great mathematician George Polya wrote in his classical book "How to Solve it":
1) First, you have to understand the problem.
What are you asked to find or show?
Can you restate the problem in your own words?
Can you think of a picture or a diagram that might help you understand the problem?
Is there enough information to enable you to find a solution?
Do you understand all the words used in stating the problem?
Do you need to ask a question to get the answer?
2) After understanding, then make a plan.
Guess and check
Make an orderly list
Eliminate possibilities
Use symmetry
Consider special cases
Use direct reasoning
Solve an equation
Look for a pattern
Draw a picture
Solve a simpler problemUse a model
Work backwardUse a formula
Be creative
Use your head/noggin
3) Carry out the plan.
4)
Look back on your work. Check and double check everything !
5) How could it be better?
At aero's suggestion, I had meep generate at image at the end of the resonance calculations. This is interesting because the large-end diameter is 2.4 wavelengths and that is about how many cycles of color I see.
At aero's suggestion, I had meep generate at image at the end of the resonance calculations. This is interesting because the large-end diameter is 2.4 wavelengths and that is about how many cycles of color I see.
Some questions
1) What view does the image represent? Why does the image have rectangular boundaries? (the cross sections of Shell's frustrum of a cone are circumferential, while the planar views should be trapezium-shaped instead of rectangular).
2) What field is being displayed ? The E (or D) field? The H (or B) field?
3) Is the image representing a contour plot of the field? If so what are the numerical magnitudes of the contours?
1. The image is square because all PNG images are rectangular. I see a circular edge there, though there seem to be leaks in the upper left and right corners. It is very soft-focus, probably because the meep resolution is only 100 for this run, to speed it up.
2. It is the H field.
3. It is a contour plot. The basic 'h5topng' utility does not add a scale legend. There is another way to make the image that can do that.
Here it is using a different way to generate the image from the same raw data.
Here it is using a different way to generate the image from the same raw data.
Thank you, I understand now that the image represents the H field, but I don't understand the following
1) the contour plot represents the scalar magnitude of what vector component of H?
2) the coordinate directions of the image.
To interpret the image, it would be helpful if you could define something like this (replace as needed)
magnitude of H vector component in the Z direction (Hz)
Frustrum of a cone geometry:
Longitudinal direction (height of the frustrum): coordinate Z
_________________
Coordinates of the two-dimensional planar image of H:
Vertical Direction: Y
Horizontal Direction: X
Out-of-plane Direction: Z ( image taken corresponds to Z = ?, or "Image corresponds to Z at the small end", or "Image corresponds to Z at the large end")
Here it is using a different way to generate the image from the same raw data.
The image shows relatively low magnitude values of the H field (10^(-7)). Perhaps this field in this direction (whatever direction it is supposed to be in) has a low value and it is not relevant. Have you looked at all the fields in all directions: Ex, Ey, Ez, Hz, Hy, Hz and determined that this is a relevant field?
The fact that the plot shows lack of symmetry is either indicative of:
1) Not one of the relevant field directions
or
2) If this is the result of a time-marching solution, not enough time has gone by for the fields to be well-formed into a standing mode well developed pattern
or
3) Unsymmetric field distribution is due to unsymmetric excitation (assuming a symmetric geometry of the frustum)
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
Welcome to the forum
Your suggestion would also address the possible problem of porosity leading to lower conductivity (assuming that the silver sheets are fabricated such as resulting in a non-porous sheet)
On the other hand, there maybe a possible problem of waviness/non-flatness of the sheet due to being so thin, as the surface needs to be smooth for reflection purposes.
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
Welcome to the forum
Your suggestion would also address the possible problem of porosity leading to lower conductivity (assuming that the silver sheets are fabricated such as resulting in a non-porous sheet)
On the other hand, there maybe a possible problem of waviness/non-flatness of the sheet due to being so thin, as the surface needs to be smooth for reflection purposes.
Dr. Rodal,
If the sliver plating is greater than skin deapth, shouldn't that avoid the problems in the earlier linked article on the affect of plating? It seems the only issue at that point should be the purity of the plating...
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
Welcome to the forum
Your suggestion would also address the possible problem of porosity leading to lower conductivity (assuming that the silver sheets are fabricated such as resulting in a non-porous sheet)
On the other hand, there maybe a possible problem of waviness/non-flatness of the sheet due to being so thin, as the surface needs to be smooth for reflection purposes.
Dr. Rodal,
If the sliver plating is greater than skin deapth, shouldn't that avoid the problems in the earlier linked article on the affect of plating? It seems the only issue at that point should be the purity of the plating...
It depends on the porosity morphology and whether the features range from sub-micrometer to hundreds of micrometers. Difficult to say anything without knowledge of the actual structure of the porosity.
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
That would be one solution for achieving the highest Q possible, but the Fine Silver sheet would have to be mounted on some kind of backing. Alone it would be almost as flimsy as household Aluminum foil.
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
I wore gloves and a mask and in a well ventilated area just to be safe.
Here is their MSDS if you would like to read.
Shell
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
Welcome to the forum
Your suggestion would also address the possible problem of porosity leading to lower conductivity (assuming that the silver sheets are fabricated such as resulting in a non-porous sheet)
On the other hand, there maybe a possible problem of waviness/non-flatness of the sheet due to being so thin, as the surface needs to be smooth for reflection purposes.
Dr. Rodal,
If the sliver plating is greater than skin deapth, shouldn't that avoid the problems in the earlier linked article on the affect of plating? It seems the only issue at that point should be the purity of the plating...
I asked the Tech Support on the quality of deposited silver and he assured me it is 99.9% silver. How much silver you lay down is dependent of time plating. With the little wand that comes in the kit it's very slow but you can increase the time by using a SS 3/8" rod and making a bath for your part ~3=4" deep and drop the rod about 2" into the soultion and about 2" up from the bottom.
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
Welcome to the forum
Your suggestion would also address the possible problem of porosity leading to lower conductivity (assuming that the silver sheets are fabricated such as resulting in a non-porous sheet)
On the other hand, there maybe a possible problem of waviness/non-flatness of the sheet due to being so thin, as the surface needs to be smooth for reflection purposes.
Dr. Rodal,
If the sliver plating is greater than skin deapth, shouldn't that avoid the problems in the earlier linked article on the affect of plating? It seems the only issue at that point should be the purity of the plating...
I asked the Tech Support on the quality of deposited silver and he assured me it is 99.9% silver. How much silver you lay down is dependent of time plating. With the little wand that comes in the kit it's very slow but you can increase the time by using a SS 3/8" rod and making a bath for your part ~3=4" deep and drop the rod about 2" into the soultion and about 2" up from the bottom.
I remember you mentioned this earlier and now I will do the forbidden and assume...., that when electroplating the silver is laid down as atoms or close to it. Using the stainless steel electrode there would be no possibility of eroding clumps of silver from the electrode. The question would be whether the stainless leaches any iron or ?? into solution. But stainless steel does seem to be one of the suggested materials.
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
That would be one solution for achieving the highest Q possible, but the Fine Silver sheet would have to be mounted on some kind of backing. Alone it would be almost as flimsy as household Aluminum foil.
There are many plating companies out there who will gladly take your frustum and electroplate it with a pure silver layer and even with a gold flash to protect. Make sure you research the ones that specialize in waveguides for the semi industry that maybe local for you.
I looked into it at the start of this project but decided that O2 free copper was good enough until I found I could inexpensively do my own silver electroplating.
Shell
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
Welcome to the forum
Your suggestion would also address the possible problem of porosity leading to lower conductivity (assuming that the silver sheets are fabricated such as resulting in a non-porous sheet)
On the other hand, there maybe a possible problem of waviness/non-flatness of the sheet due to being so thin, as the surface needs to be smooth for reflection purposes.
Dr. Rodal,
If the sliver plating is greater than skin deapth, shouldn't that avoid the problems in the earlier linked article on the affect of plating? It seems the only issue at that point should be the purity of the plating...
I asked the Tech Support on the quality of deposited silver and he assured me it is 99.9% silver. How much silver you lay down is dependent of time plating. With the little wand that comes in the kit it's very slow but you can increase the time by using a SS 3/8" rod and making a bath for your part ~3=4" deep and drop the rod about 2" into the soultion and about 2" up from the bottom.
I remember you mentioned this earlier and now I will do the forbidden and assume...., that when electroplating the silver is laid down as atoms or close to it. Using the stainless steel electrode there would be no possibility of eroding clumps of silver from the electrode. The question would be whether the stainless leaches any iron or ?? into solution. But stainless steel does seem to be one of the suggested materials.
When I did my first antenna 1/4 WL antenna (simple copper solid 12g wire) it took a long time to lay down what I figured was a good layer. I went online with Caswell Tech support and gave me the directions on how to do it a little faster. They made sure I was using a 316 SS rod (need to double check my notes but I believe it was 316 SS), that way you do not contaminate your surface with other metals.
Non cyanide silver plating solutions are fairly toxic so I was thinking that the parts could be fabricated out of fine silver sheet. I looked up todays prices on the Rio Grande site and found that a .010" thick 6" X 24" sheet costs about $130.
Welcome to the forum
Your suggestion would also address the possible problem of porosity leading to lower conductivity (assuming that the silver sheets are fabricated such as resulting in a non-porous sheet)
On the other hand, there maybe a possible problem of waviness/non-flatness of the sheet due to being so thin, as the surface needs to be smooth for reflection purposes.
Dr. Rodal,
If the sliver plating is greater than skin deapth, shouldn't that avoid the problems in the earlier linked article on the affect of plating? It seems the only issue at that point should be the purity of the plating...
I asked the Tech Support on the quality of deposited silver and he assured me it is 99.9% silver. How much silver you lay down is dependent of time plating. With the little wand that comes in the kit it's very slow but you can increase the time by using a SS 3/8" rod and making a bath for your part ~3=4" deep and drop the rod about 2" into the soultion and about 2" up from the bottom.
I remember you mentioned this earlier and now I will do the forbidden and assume...., that when electroplating the silver is laid down as atoms or close to it. Using the stainless steel electrode there would be no possibility of eroding clumps of silver from the electrode. The question would be whether the stainless leaches any iron or ?? into solution. But stainless steel does seem to be one of the suggested materials.
Chapter on porosity (as well as micro-cracks and their effects) from a book, on the Internet. (I'm not responsible for the horrible highlighting
which does not highlight anything in particular)
http://www.tau.ac.il/~chemlaba/Files/Electrodeposition/13208_08.pdf
The negative Q values are a known artifact of the computational model used inside meep under certain conditions. It has nothing to do with physics.
Correct. It has nothing to do with physics.
It has nothing to do with reality.
Discussing this is similar to previous discussions incorrectly attaching physical significance to:
previous Meep models which included a Finite Difference mesh outside the EM Drive, which showed numerical magnitudes of the electromagnetic fields to be 24 orders of magnitude smaller than the fields inside the EM Drive. People (unfamiliar with numerical methods like Finite Difference solutions) were unaware that such numerical methods entail the solution of simultaneous equations and therefore that the magnitude of the fields outside the EM Drive instead of displaying a perfect zero will routinely display very small numbers associated with the numerical precision and ill-conditioning of inversion of matrices. Some posters started to discuss in these threads whether these extremely small numerical fields outside the EM Drive were revealing "negative energy". Nothing to do with reality. Everything to do with the numerical solution of simultaneous equations.
Your correct Dr. Rodal, but it needed to be discussed and resolved even if it is a numerical hiccup.
I'm glad it's behind us and the fractals shown in the boundaries of some of the first meep models are as well. We have come a long way to making sure meep can model our models. When we get solid data from the drive and solid data from meep we can begin to compare and run cross checks between the two. I hope the combining of both data sets will give us some tools to dig a little deeper, answer some questions and open up better ones.
This is my hope.
Shell
Yeap, the "fractal" contour images associated with very small numerical magnitude of fields is another example.
Whenever electromagnetic fields are displayed, they should preferably be displayed with a numerical table identifying the numerical value of the contours (to prevent people from looking at nonsensical small values that are numerical artifacts and thinking that they are significant)
Another example are Meep Q's of 10 million (due to https://en.wikipedia.org/wiki/Garbage_in,_garbage_out ).
In numerical experiments, just as in physical experiments, everything needs to be double-checked over and over.
I would be willing to ignore artifacts of a model if I had the foggiest notion of how the thing worked. Since I can't grasp why it seems to work, I'm not willing to ignore anything. Kind of the reason I put the TurboEncabulator reference in the post.
In my somewhat small mind ( compared to the group here) I can't help noticing aa parallel to the speaker in a 1929 Philco highboy, a device for turning electromagnetic energy into air movement. Not very efficient, but once something like that works, the room for improvement is massive.
I'm banking on Shell to make some thrust that is beyond the realm of noise so I can come in and make it sing.
