• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Perpetual motion machine examination rules, please.

Nathan,
Sure, take a chill pill. I think the real problem is that you're using your monitor as an instrument and calculating the slope using pixels. The representation using pixels might not be accurate and although you can't do it if you would look at the difference between the top edge of the body and where it's selected you would see that it's a very small part of the height (0.3 meters). You can't scroll down the edge of that corner because it's magnified 5.942 e+10 times it's normal size and the view window you're looking at (and trying to measure in pixels) is 5.942 e-12 meters.

If I were to take the beam (I no longer have it set up) and set the magnification back to a level where I could see the entire beam height (0.3 meters) in the screen I'm sure I wouldn't see any rotation. I think the problem you're having is trying to use the montior as an instrument (using a pixel count) when a more accurate method would be to consider the entire edge of the beam compared to the very tiny portion that you can see selected. The ratio of that portion selected to the entire height would be a more accurate representation of the program's ability to rotate it that very slight rotation.

Gene
 
Ginarley said:
I have no problem with people pushing the boundaries of science to try and find holes. It is a win-win proposition because either established rules get further evidence or something fundamental is discovered. It is exactly on this basis that science has got where it has got today. Poke holes in logic all you want but don't poke holes in genuine effort.
Click to expand...
Which is exactly what is not happening here.

Long ago in this thread I compared what is going on here is equilvalent to somebody trying to get 2+2=5. Let's see 2.000000001 + 1.99999999 = 4.00000001. Yay, I'm making progress! No, my calculator is making a rounding error. Ad nauseum.

There's a couple of possibilities here. One, that there is some implication of our known physics that hasn't been discovered yet. However, straightforward analysis of the given formulas tell us this is not true - there is no free lunch. Likewise, if I analyse addition, the [SIZE=-1]commutative law, etc., I can show that 2+2 will always equal 4. There is no need, and no point to, exploring different ways of adding numbers to check if there is some case where I can get 2+2=5. I can prove that will never happen. Same with perpetual motion. If the laws are valid, then there is no perpetual motion.

So, maybe the laws are wrong. This has a small hint of hope, because we know laws have been amended before (Einstein). However, this simulation assumes the laws. So any output from the simulation will never discover this case.

So, what we see in this thread is nothing more than tilting at windmills. Well, yes, there is hopefully an educational element, in which the poster will finally realize why we disagree with him, the limits of simulation, etc. But there are so much easier ways to learn those lessons - by taking a course in math, in physics, and working out the consequences. But no, he insists on being willfully ignorant, deciding that he doesn't have to obey the laws of physics, and searches for something with absolutely no understanding.

That's the problem. Willful ignorance. There's plenty I don't know about many subjects; there is no shame in ignorance. It's not a perjorative term. But willful ignorance - refusing to learn a topic? Yes, we are going to continue to point out the folly of that approach, and not encourage it.

It's clear the poster has a burning desire, and I commend that. It's rare. I can't commend refusing to learn the basic tools needed to approach this problem. That's unfortunately all too common, and something the James Randi Educational Foundation, and the posters in the JREF forums, are not going to allow to stand without comment.
[/SIZE]
We are not putting him down, but trying to show him how to approach this topic with intellectual rigor. His response is sticking his index fingers in his ears and going "LA LA LA LA LA - I can't hear you".
 
rwguinn,

  • your 10^-22radians example is typical. Small differences in large numbers, large differences in small numbers--all of them screw up the math when you end up generating mass and stiffness matrices internal to the programs--these are things you never see, unless you know the mechanics and math that go into doing the modeling.

My contention that the ratio of the amount of the beam that is selected to the the entire length of the beam is a more accurate indication of whether the beam rotated and to what extent it is, not a pixel count. Now that beam is 5.510 e+10 times larger on the screen than it is in real life or 0.3 meters tall. I'd say the myth (as you put it) is that any monitor can be used as an instrument for determining slope based on pixel count.

I just measured that difference and it's 2.65cm. That difference is 5.510 e+10 times it's normal size (due to magnification) or 1.0273771100472280643770352842551e-9 meters. The actual question and it's answer isn't found by trying to use a pixel count in those particular extremes. Also assuming that there is a problem in wm2d ignores the other factors in the equation specifically the driver for the monitor. A good question is how does that driver deal with the output or file from wm2d that it's trying to write to the monitor. Different drivers handle things differently. If it were a plotter driver I think this entire pixel idea would be mote. In less extreme cases and with a good monitor (high res) and the appropriate driver you might be able to use a pixel count (if you could see them) and use that for calculations yet in this case I think it's an erroneous method.

Now that aforementioned actual difference as a ratio to 30,000 meters, which is the actual length of the beam is 3.4245903668240935479234509475171e-14, which is the tangent of the angle of rotation, or 1.9621457458017896925957865144105° e-10 or 3.4245903668240935479234509461783 e-14 radians. That isn't the rotation that I would expect based on what I entered into the property box but it's closer than Nathan's pixel approximation....

  • I see that image, it appears to have a slope of 2 pixels across half your screen. If your screen is about 1000 pixels wide, that means it's slope is 4e-3 radians.

I'd like to address that point you made above that I emboldened...
  • screw up the math when you end up generating mass and stiffness matrices
The entire point of this exercise was to answer the question 'to what degree does wm2d consider the position of a body that you construct?' Now if the internal stress was factored into a body 0.3 meters high and 30K meters long then the actual weight of the body (9000 kgrams) would have precluded looking at this particular example. That's why the factor of stress didn't even enter into the problem.

and now this point that you made...
  • And that knowledge is what differentiates the mythologist from the engineer and scientist.
You again over-look the idea that this entire exercise was to examine precisely 'to what extent does wm2d position bodies as you request them to be positioned.'

You've not done a very good job following the point but I don't think that makes you a mythologist (someone that studies myths) but I do think it points to your level of competence.

Now I wouldn't come to any firm conclusion on the ability of wm2d to accurately calculate values based on either the method I used or that Nathan used. I've found in practice when you get two different answers to the same question you still don't have an answer. I also think both methods ignore the driver of the monitor and how it parses the results that wm2d gives it.

Coming to hasty conclusions based on inconclusive results or an inability to follow a point aren't the habit of a professional whether they're a mythologist or an engineer yet they are more indicative of level of incompetency. You're scoring very high.

Gene
 
Last edited:
Roger,

Thanks for your thoughts. I have to say they were some of the finest example of reeking condescension I've ever read. I would like to address this manner you led into one particular rant with ....
  • It's clear the poster has a burning desire, and I commend that.
Please allow me to introduce my self,
my name is Gene.

If you were to have something to say to me specifically rather than a general rant to the entire thread I'd appreciate it if you'd actually use my name. I might be able to understand if that is beneath you. In any event if you care to make a specific point I'll certainly do my best to address it.

Again thank you for that fine example of reeking condescension.

Gene
 
rwguinn,

That was a rather long post so I'll distill my thoughts here. To suggest that I don't know what I'm looking at then to cite as proof an example where I'm investigating what a software is showing me is ....heck, I don't know how to put this kindly.

I hope you see the point.

Gene
 
For gravity to not be a conservative force, I thought we'd have to prove that the r^2 in the denominator of the gravitational force equation is not actually r^2, but r^2.000000000001 or something like that. Last I heard, that 2 has been measured EXTREMELY accurately to be 2, and not something else.

And if gravity propagates like other forces (mediated by particles), the r^2 is due to the fundamental nature of space.

So gravity not being a conservative force is a real long shot.
 
Almo,
rwguinn has made a call for logic so I thought I'd take a look at your conditional premise...
  • And if gravity propagates like other forces (mediated by particles), the r^2 is due to the fundamental nature of space.
I don't know if we actually know how gravity propagates and far as I know we're most certain that light (for example) either propagates by particles or waves. To draw conclusions based on uncertainty will give you an absolutely uncertain conclusion...
  • So gravity not being a conservative force is a real long shot.

I do think you make a fine point yet I don't agree how you arrived at it. I don't disagree that within the constraints of time/space as we know, the idea that gravity is conservative is a safe conclusion.

Well, I need to get back to some very strange and warped ideas that I have. Who knows, they might be the cause of me getting that million dollars from Mr. Randi. That would be a sweet day.

Gene

warped.gif
 
AgingYoung said:
I think the real problem is that you're using your monitor as an instrument and calculating the slope using pixels. The representation using pixels might not be accurate and although you can't do it if you would look at the difference between the top edge of the body and where it's selected you would see that it's a very small part of the height (0.3 meters). You can't scroll down the edge of that corner because it's magnified 5.942 e+10 times it's normal size and the view window you're looking at (and trying to measure in pixels) is 5.942 e-12 meters.
Click to expand...

You really are not paying attention are you? My technique is perfectly valid. I repeat 'rotation is magnification invariant'. If the beam was really only rotated the amount you claim, there should be no jaggies on the rendering. The fact that there are means one of three things (a) it is rotated more (b) your CAD package is worthless because it is adding artifacts that have no reason to be there -- I.e. a more accurate representation would be a horizontal line, and is perfectly renderable on your hardware (c) you are operating outside of its design constraints (probably because of the use of 32bit floating point numbers and the discrete reprentation they provide)

I have avoided telling you my qualifications for making that assertion, I will do so now. (a) I have experience designing graphics hardware, (b) I have experience with writing rendering engines at both the blitting level and the 3d transform level (c) I have experience with 3D modelling packages (d) I have experience designing floating point units (e) I have experience writing floating point emulation routines.

Any package that draws the line you showed and claims it is rotated 10^-19 radians wrt the X axis of the view frustrum is either buggy or being operated outside of its design constraints.
 
Nathan,

I'm fascinated with your qualifications. My brother was a member of a siggraph (acm), I'm not sure if he still is. You might have seen some of my work if you watched the premier of the rca selectivision commercial released during the superbowl some time ago. We did that work with the movie byu. Ancient history but now to the present.

Would you disagree that the tangent of the angle of inclination is better determined by the difference between the tops of the two beams divided by it's length than by a representation of pixels? I do agree that the line should be straight and I really have no idea why it isn't. It is possible that the degree of magnification distorts the picture and is a limitation of the software.

I've further looked at this with wm2d and I'm inclined to believe the level of magnification is indeed a factor in the way that the beam is displayed. I looked at beams 100K meters yet only magnified them to the extent that the height would cover the screen. Also when I entered a slight rotation wm2d attempted to rotate it and when it finished (a) said it couldn't because of constraints but (b) did show a pic of what it would look like rotated and there were no jaggy lines. That leads me to think that there is distortion when wm2d magnifies images to the extent I did.

Considering your qualifications it amazes me you don't have a windows platform. In light of your conclusion (c) being a possibility (and I'm inclined to agree with that yet the parameter that is being pushed beyond the limit is the magnification function) it's hard to conceive that you'd even consider (b) as being a possibility.

This is a rather lengthy discussion on a matter that I don't now think to be a factor. Your original point was the accumulative effect that approximations have vs. what actually happen thereby giving the effect of pm. I'm mostly using wm2d to consider parts and not the whole idea. I'm also using it to consider specific ideas thru a limited range of motion. For what I'm using wm2d for it's very useful. I don't have an E&S graphics viewing station with a driver (rendering engine) residing on a vax 11780 running a simulation on a cray but then I'm not running an intergalactic flight simulator here.

You do have a qualified opinion and the line should be straight. It might be that looking at bodies magnified 5.510 e+10 giving you a view at the atomic level is beyond wm2d's ability to accurately render.

Gene
 
Last edited:
  • You do have a qualified opinion and the line should be straight. It might be that looking at bodies magnified 5.510 e+10 giving you a view at the atomic level is beyond wm2d's ability to accurately render.

I've been thinking about this and now I disagree with myself. :) I'm only guessing but I'd say there is a file that has the actual model with it's shape, size and position in the cartesian plane. I'm almost certain that the magnification of the model has no effect on the model. That magnification is merely matrix multiplication (where the problem might be when taken to extremes of the atomic level) that generates another file that is output to the screen. That magnified file is not used for simulation and in no way changes your model.

Now before I would unequivocally state there was a problem with the magnification feature of wm2d when taken to the atomic level I'd need to see a plot of that corner on e-sized paper wrt the x-axis. It's possible that wm2d generates one file to print to the screen yet it goes thru a few more steps when printing a blueprint. In any event it's not an issue with my use of the code.

Gene
 
AgingYoung said:
Roger,

Thanks for your thoughts. I have to say they were some of the finest example of reeking condescension I've ever read. I would like to address this manner you led into one particular rant with ....
  • It's clear the poster has a burning desire, and I commend that.
Please allow me to introduce my self,
my name is Gene.

If you were to have something to say to me specifically rather than a general rant to the entire thread I'd appreciate it if you'd actually use my name. I might be able to understand if that is beneath you. In any event if you care to make a specific point I'll certainly do my best to address it.

Again thank you for that fine example of reeking condescension.

Gene
Click to expand...
I suspect you are being disingenuous in your thanks, Gene. But no matter.

No condescension was intented. I never once called you a name, explicitly or implicitly. You may have noticed I was addressing another poster, which is why I didn't address you directly, and spoke of you in the third person. I'm honestly sorry if you took offense over that, but fully support my right to talk to another poster. However, I assert everything in my post was factual. Perhaps it wasn't. Please point out the errors so we can all learn.

Fact is, you have argued with qualified physicists, mechanical engineers, graphic designers, all from a basis of ignorance. You have dismissed the work of fairly good scientists (Newton, et. al.) with no basis.
 
AgingYoung said:
rwguinn,

That was a rather long post so I'll distill my thoughts here. To suggest that I don't know what I'm looking at then to cite as proof an example where I'm investigating what a software is showing me is ....heck, I don't know how to put this kindly.

I hope you see the point.

Gene
Click to expand...

your point was that you are trying to find where the program says X=0.
You went way past it.
"Now that beam is 5.510 e+10 times larger on the screen than it is in real life or 0.3 meters tall. I'd say the myth (as you put it) is that any monitor can be used as an instrument for determining slope based on pixel count"
was an answer to me? Who cannot follow the thread?
I said nothing about pixel counts--but I do remember an older engineer who kept trying to get dimensions off his CAD screen with a rule...
The screen is an approximation of the mathematical approximations the the program uses.
And go look up in an analytical geometry book the term "Ill-conditioned Matrix". If you have terms of 10^6 alongside 10^-22 in simultaneous equations, you are really sc*^%$d.
 
The jaggies are likely intentionally drawn by the CAD software to simply indicate that the line is *not perfectly* parallel (or perpendicular) as it has been rotated off zero. Let's move on. I'd like to get beyond the simulation so this thing can get built! That's where the rubber meets the road.
 
roger,
  • No condescension was intented. I never once called you a name, explicitly or implicitly.
I said condescension, not ad hominem.

Gene
 
rwguinn,
Thanks for reminding me what my point was....
  • your point was that you are trying to find where the program says X=0.
Thanks for reminding me. From now on your in charge of what in the heck's going on.

Gene
 
Grounded said:
The jaggies are likely intentionally drawn by the CAD software to simply indicate that the line is *not perfectly* parallel (or perpendicular) as it has been rotated off zero.
Click to expand...

That is a possibility, Grounded.

I spent the night working on this element I'm designing and it's looking pretty good. I wouldn't mind defending it in an open forum and explaining what I'm thinking but there is that small matter of the million dollars Mr. Randi said he'd give me if my ideas had any merit (ie prove that gravity isn't a conservative force). Well, back to work.

Gene
 
Ginarley said:
I have no problem with people pushing the boundaries of science to try and find holes. It is a win-win proposition because either established rules get further evidence or something fundamental is discovered.
Click to expand...

Would be true if he was truly pushing the boundaries of science. But he is not. He's trying to make a computer simulation of a person lifting himself up by his own hair and when it succeeds, due to the limitations of the digital approximation, claims that he has discovered something real. He hasn't, he is pushing no boundaries, he is wandering around in the Middle Ages.
 
Hi, professional physicist de-lurking here. There's something I don't understand, and pardon me if this has already come up. Newton's Laws inherently include energy conservation. Every force (mechanical, chemical, electromagnetic ...) can be calculated as F = -dE/dx: the force in any direction is equal to the rate of change of stored potential energy with respect to displacements in that direction. If you have a system with a lever sticking out of it, and the system's energy goes up by 50 joules when you displace the lever through 2 meters, then the lever will resist that displacement with a force of 25 N. If there isn't a conserved energy which changes, there won't be a force. For example, the gravitational force F = -mg can be derived from the gravitational potential energy, E = mgz +C (where z is the height, g is the acceleration due to gravity (taken to be positive)), so Fz = -dE/dz = -mg.

Ditto for kinetic energy: if you exert a net force on an object, and maintain that force F while the object changes position by dx, you change the object's kinetic energy by dE = F dx. Notice that, to actually apply that force, you had to decrease the potential energy in some other system---and that the KE increase exactly equals the PE decrease. If there wasn't a change in PE, then there wasn't a force. If there wasn't a force, then there wasn't an increase in KE.

This is true even for frictional forces: the force exerted by a brake pad is equal to the increase in thermal energy of the pad and rotor. The only difference is the irreversibility; you can increase thermal energy as much as you like, but you can't freely turn it back into kinetic or potential energy.

So, AgingYoung, you've got a simulation package with Newton's Laws built in, laws which do nothing but conserve energy, since they never exert any forces at all unless they have a potential energy store to sap. If you think that some aspect of your simulated system is picking up energy---well, I presume you must think that there's a subsystem, linkage, or doohickey for which F is not equal to -dE/dx.

Is this true? If so, what makes you think F=-dE/dx would break down in your system? What makes you think Working Model includes this breakdown in their underlying physics code, rather than actually calculating the conserved potential energy for each of your components, and calculating energy-conserving forces from there, thereby guaranteeing that the full system will conserve energy?

It seems to me that your task---twisting a Newton's Law simulation and hoping a non-Newton's Law result pops out---is rather like playing with a calculator and trying to find two odd numbers whose product is even. Sure, you may find some round-off error or approximation that gives the appearance of success, but why bother?
 
You must log in or register to reply here.

Back
Top Bottom