Multiphysics? That's a new word on me.. Unless you think Asimov.
jn
Pretty common these days with some of the big FME programs, we use COMSOLVE where heat flow and EM effects are co-simulated.
Interesting.
I don't normally think of E and M as different physics, so never considered "multiphysics"
So far, we've been able to consider heatflow, heat capacity, modulus and the such independently of excitation forces, but certainly understand how complex the systems get.
As we move into high ramp rate items, I suspect that will gain importance. Medical beam delivery stuff is one such app, but about a decade away. Hope I'm not dead before I need it..
jn
The big snag with the electrodynamical "elephant in the room" is that it permits the geometry to change. If you clamp the voice coil so all you have is some coupled inductors then all becomes linear again. Sorry to have to shoot the elephant.
Linear EM with linear media in a fixed geometry cannot create second-order. Jn keeps asserting, as though it were true, that skin effect is proportional to rate of change of current. If you look at how the equations actually work out in it turns out that the solution is separable: at every point the fields/currents etc. are proportional to the product of a geometry factor (which is a function of frequency but not of magnitude) and a frequency/phase factor and an overall magnitude factor. This is not a low field approximation (provided that the media remain linear). Twice the magnitude simply means that everything gets doubled - the geometry factor remains unchanged because skin effect does not depend on the magnitude of the fields. We know this because in order to calculate skin effect you just need to know the material properties and the frequency. If jn was right then skin depth would depend on the incoming field strength or the current in the conductor. Now if jn is claiming that all the EM textbooks get this wrong and their solutions to Maxwell's equations are wrong then he ought to write his own EM texbook and put all the physicists straight.
As I have said before, I can see how someone might get the erroneous impression that skin effect depends on rate of change, as rate of change is the product of frequency and magnitude. If rate of change is doubled by doubling the frequency then you get a different skin depth, but if rate of change is doubled by doubling the magnitude then you get the same skin depth. In either case rate of change is doubled. For a sine wave excitation it is simply false to claim that the skin depth penetration varies over the cycle, however much this might appear to be the case from a naive extrapolation from pulse observations. Actually, you should go the other way: to understand pulses you need to do a Fourier analysis and treat each frequency component separately.
I am losing count of how many times I and a few others have tried to teach jn some basic EM physics but he still sticks to his own ideas.
Linear EM with linear media in a fixed geometry cannot create second-order. Jn keeps asserting, as though it were true, that skin effect is proportional to rate of change of current. If you look at how the equations actually work out in it turns out that the solution is separable: at every point the fields/currents etc. are proportional to the product of a geometry factor (which is a function of frequency but not of magnitude) and a frequency/phase factor and an overall magnitude factor. This is not a low field approximation (provided that the media remain linear). Twice the magnitude simply means that everything gets doubled - the geometry factor remains unchanged because skin effect does not depend on the magnitude of the fields. We know this because in order to calculate skin effect you just need to know the material properties and the frequency. If jn was right then skin depth would depend on the incoming field strength or the current in the conductor. Now if jn is claiming that all the EM textbooks get this wrong and their solutions to Maxwell's equations are wrong then he ought to write his own EM texbook and put all the physicists straight.
As I have said before, I can see how someone might get the erroneous impression that skin effect depends on rate of change, as rate of change is the product of frequency and magnitude. If rate of change is doubled by doubling the frequency then you get a different skin depth, but if rate of change is doubled by doubling the magnitude then you get the same skin depth. In either case rate of change is doubled. For a sine wave excitation it is simply false to claim that the skin depth penetration varies over the cycle, however much this might appear to be the case from a naive extrapolation from pulse observations. Actually, you should go the other way: to understand pulses you need to do a Fourier analysis and treat each frequency component separately.
I am losing count of how many times I and a few others have tried to teach jn some basic EM physics but he still sticks to his own ideas.
You still ignore basic understandings. The vc force is a product, regardless of movement or compliance.
The math is the same in both examples, even if you cannot understand.
See Lenz's law.
Proximity effect is a consequence of the rate of change the magnetic field caused by the rate of change of the current driving the system.
Sigh...armchair physicists armed with google...
You should try designing, building, and testing hardware for the physics community. At best, you would learn something...at the least, you'd meet physicists who do this for a living, they could teach you what you refuse to learn from me.
jn
SY:
Well, yes, but someone might misunderstand and think that the issue was electrical vs. mechanical when the real issue is fixed vs. variable geometry. Before you can shoot an elephant you first have to tie it down!
If skin effect really does depend on rate of change then it would disappear twice in each cycle - we would momentarily have DC current distribution and so DC resistance. Don't you think the RF guys would notice if all their wires, inductors and transmission lines started generating significant amounts of second harmonic? It would have to be a major feature of all RF textbooks. Strangely, I have never seen it - I have several RF textbooks at various levels from technician/hobbyist up to graduate.
Well, yes, but someone might misunderstand and think that the issue was electrical vs. mechanical when the real issue is fixed vs. variable geometry. Before you can shoot an elephant you first have to tie it down!
Lenz's law says that the induced current opposes the incident field. It is the EM version of 'no perpetual motion'. Lenz's law arises from Maxwell's equations and is primarily about the sign of a response.jneutron said:
No, real physicists armed with real EM textbooks.
I would be very surprised to meet a real physicist who misunderstands skin effect in the way you do, but I suppose anything is possible. On the other hand, people who build equipment for scientists or engineers can sometimes misunderstand the science they are using or investigating. A rule of thumb may be helpful when building stuff without actually being true.
If skin effect really does depend on rate of change then it would disappear twice in each cycle - we would momentarily have DC current distribution and so DC resistance. Don't you think the RF guys would notice if all their wires, inductors and transmission lines started generating significant amounts of second harmonic? It would have to be a major feature of all RF textbooks. Strangely, I have never seen it - I have several RF textbooks at various levels from technician/hobbyist up to graduate.
Last edited:
That's it? Well, yes? edit: ah, you were responding to sy...sorry bout that..
A fixed, clamped coil within a time varying magnetic field produces forces on the coil wire which is proportional to the product of the current and magnetic field. Vary both, the force is STILL the product. The force is at double the frequency when both are excited with the same current. edit: IT HAS NOTHING TO DO WITH MECHANICAL MOVEMENT
SMPS coils do the same thing. As the frequency increases, the location of the current centroid moves such that the resistance goes up. The current centroid moves as a consequence of the absolute value of the slew rate of the magnetic field.
You might, I have not. You're strawman argument means nothing, you ignore that the resultant entity (vc force) is a consequence of a product of two time varying entities regardless of movement, in the case of proximity effect current crowding, you ignore everything.
jn
Last edited:
Skin effect -
Looking up the diagrams/maps of field density for various configurations (or calculating the fields ) will show you where the L is lowest... in electronics, this is called 'skin effect'.
It has been studied extensively.... but the one's who finally did the 'unified' theory is this one --
View attachment Skin1.pdf
View attachment skin2.pdf
View attachment skin3.pdf
Look it up.... [the Internet is a wonderful thing if its used]
THx-RNMarsh
Looking up the diagrams/maps of field density for various configurations (or calculating the fields
) will show you where the L is lowest... in electronics, this is called 'skin effect'.It has been studied extensively.... but the one's who finally did the 'unified' theory is this one --
View attachment Skin1.pdf
View attachment skin2.pdf
View attachment skin3.pdf
Look it up.... [the Internet is a wonderful thing if its used]
THx-RNMarsh
Last edited:
Looking up the diagrams of field density for various configurations (or calculating them
It has been studies extensively.... but the one's who finally did the 'unified' theory is this one --
View attachment 429580
View attachment 429582
View attachment 429583
Look it up....
THx-RNMarsh
Thanks RN,
I note they state ""in which the elements are frequency independent have been used, but tend to product very large ladder circuits."" Does this mean they are using frequency dependent elements?
I'll have to look up Yen and his buddy "et al", as well as this paper in it's or original format. I'm thinking that I've actually seen this paper before, but the date (today's) on the upper left confuses me..
jn
If you want to think in terms of forces then you have to determine and add up all the forces: the magnetic force pushing outwards, the Coulomb force pushing back inwards, the applied voltage causing the original current etc. Even then you just have a net force - you still have to determine motion and displacement. I look forward to seeing you work through all the calculations.
Physicists usually just solve field equations, and trust that the correct solution to the correct equation already includes all the forces. We find this works. For skin effect it tells us that the geometry of the effect depends on frequency alone. This is what the textbooks say. This is what RF experience tells RF engineers. If you believe, as you appear to do, that the EM textbooks are wrong then you should say so. Others can then form their own conclusion on the likelihood of this being the case.
As has been said before, if you can model skin effect with linear elements like inductors and resistors then this is powerful evidence that the effect is linear or at least only weakly non-linear. If skin effect depended on current magnitude then it would be strongly nonlinear. Rate of change depends linearly on both frequency and magnitude, thus one would expect (if jn is right) that skin depth would depend on both frequency and current in the same way - a square root response. Why has this not been seen in any experiments? Why do coax cables not become much less lossy for small signals, as skin effect disappears due to the small currents? It would be wonderful if this were true, but it is not.
Physicists usually just solve field equations, and trust that the correct solution to the correct equation already includes all the forces. We find this works. For skin effect it tells us that the geometry of the effect depends on frequency alone. This is what the textbooks say. This is what RF experience tells RF engineers. If you believe, as you appear to do, that the EM textbooks are wrong then you should say so. Others can then form their own conclusion on the likelihood of this being the case.
As has been said before, if you can model skin effect with linear elements like inductors and resistors then this is powerful evidence that the effect is linear or at least only weakly non-linear. If skin effect depended on current magnitude then it would be strongly nonlinear. Rate of change depends linearly on both frequency and magnitude, thus one would expect (if jn is right) that skin depth would depend on both frequency and current in the same way - a square root response. Why has this not been seen in any experiments? Why do coax cables not become much less lossy for small signals, as skin effect disappears due to the small currents? It would be wonderful if this were true, but it is not.
Last edited:
This is unquestionably true for isolated conductors and for symmetry situations (e.g., coaxial round conductors). Is it true for non-isolated, non-symmetric situations?
That it is..
I remember this one from a while ago..
This technique is an approximation technique, not an exact. They describe iterations to get close. Note that for a twinline, proximity effects, while included, still give 20 to 22% errors in resistance. edit:note the errors shown on page 10, 21, and 22. And that's for single conductors, no multiple turn enhancement..
Note also, they do not consider coils, and point out that this metholology is only used for straight t-lines.
http://weewave.mer.utexas.edu/MED_f...ct_Ldr/MTT_96_poster/MMT_96_skn_crct_pstr.pdf
jn
Last edited:
Your "Hey look, a squirrel" schtick doesn't work on me.
Calculate the forces on a locked voice coil driven by a sine in a magnetic field that is caused by that sine. Sine squared.
If you want to also include coulomb forces where voltage is in the tens of volts, but magfield is in the tesla range, be my guest.
But diversion, sorry, doesn't work
jn
I don't think so.SY said:
We are being asked to believe that an effect which is alleged to be proportional to signal rate of change is not seen at all in RF situations (where the wide dynamic range and desirability of linearity is important) yet is present (albeit at low levels) in audio? Something is not seen at 20MHz or 200MHz or 2GHz (where inductors and wires and waveguides are linear) yet it appears at 20kHz? This would be plausible for a low frequency effect, but skin effect and proximity effect are high frequency effects.
You misapply what you've read. case in point:
So then, an iterated model using linear elements, which gives 20% errors (according to the authors) for the simplest case, is your "proof"?
Sorry, my standards of proof are a tad higher.
jn
Last edited:
Here's your problem in a nutshell.. see what you wrote in red? I didn't say that, you have messed up attribution...
And, if you recall, the proximity/skin effect screen capture I provided first (the impetus of your conflag), is a cull from a discussion of a VERY asymmetrical system, that of coils in a time varying magnetic field of a generator.
I recommend you stop with misattribution, mischaracterization, and diversions.. Standard internet ploys, pretty transparent.
jn
Last edited:
- Status
- Not open for further replies.