Re: [Vo]:Nature Editorial: If you want reproducible science, the software needs to be open source
I have another update concerning my on-going theoretical research into characteristics of celestial mechanic algorithms. Last Wednesday I mentioned the fact that another way to graph an elliptical orbit (an orbit that obeys Kepler's 1st, 2nd, and 3rd laws) would be to plot the satellite's distance from the central mass on the "Y" axis, while plotting equal time intervals on the "X" axis. What you end up with is a graph that looks like a bouncing ball. The bouncing part of the plot is where the satellite has made its closest approach to the attractive body, the perihelion point in the orbital ellipse. Most curiously, there appears to be a simple algorithm that simulates this x,y "bouncing ball" graph behavior, by incorporating it into a classic feed-back loop. The mathematical formula involves: Distance = 1/r^2 - 1/r^3. If you feed the current calculated vectors back into the formula you will generate the same bouncing ball plot. Keep in mind the squared (1/r^2) value is the attractive "force" whereas the cubed (1/r^3) value is the repulsive "force". If you employ this simple formula into a simple feedback loop you will end up plotting the exact same bouncing ball plot. The implication is that an orbiting satellite as it enters the perihelion phase of the orbit is effectively experiencing something akin to negative gravity, presumably due to centripetal forces that have temporarily overpowered the 1/r^2 attractive force. * * * As of today, Friday, I appear to have uncovered another suspicion of mine: What appears to be the generation of a perfect sine wave if you replace the "x" axis value (which previously contained a fixed time interval) with the accumulated vector value pertaining to the orbiting satellite. Said differently: As the orbiting satellite enters the perihelion phase of the orbit (closest approach to the body) the current vector will be significantly larger than when the satellite reaches its aphelion (farthest distance to the body). If you accumulate these individual slices of vector values and systematically plot them on the "X" axis proportionally, while simultaneously charting the satellite's distance on the "Y" axis, it seems to cause the charted line, which previously looked like a bouncing ball to transform into a perfect sine wave. Oh, by the way, in order to get this sine wave you need to return back to using just 1/r^2 for calculating the "Y" distance. My theoretical research continues. I have more suspicions. Maybe they will turn out to be right. ...or not. That's my Zen thought for the day! Have a good weekend. ;-) Regards Steven Vincent Johnson www.OrionWorks.com www.zazzle.com/orionworks
Re: [Vo]:Nature Editorial: If you want reproducible science, the software needs to be open source
Steven, I was puzzled because I took your bouncing ball metaphor literally. Thanks David and Robert. I guess the graph approaches the path described by bouncing ball as the ellipse becomes flatter. harry On Thu, Mar 1, 2012 at 12:01 PM, David Roberson wrote: > The orbital distance is changing faster when the object is closest to the > earth which would tend to look like a quick bounce. At the far spacing, the > change in orbital distance is slower depending upon the elliptical shape. > The mathematical equation defining the function of orbital distance versus > time should be available and in a closed form. I recall that equal orbital > areas are swept out in equal time, which is one of Kepler's laws as derived > by Newton. Wikipedia has a fairly good article on Kepler's laws. > http://en.wikipedia.org/wiki/Kepler's_laws_of_planetary_motion > > Dave > > > -Original Message- > From: Harry Veeder > To: vortex-l > Sent: Thu, Mar 1, 2012 11:25 am > Subject: Re: [Vo]:Nature Editorial: If you want reproducible science, the > software needs to be open source > > On Wed, Feb 29, 2012 at 12:50 PM, OrionWorks - Steven V Johnson > wrote: >> From Harry: >> >>>> From OrionWorks: >>>> What I can say is that the new system involves an alternative way of > graphing out a periodic orbit - where you plot an "elliptical" orbit on a > TIME-LINE chart. The orbital distance is the "Y" vertical value and the > horizontal "X" value is the time value. >>> >>> That graph should look something like a sine curveor not? >> >> You're on the right track. However the time-line looks more like a >> bouncing ball. > > I think I understand now. You are mapping a two dimensional distance > vector to the distance axis of your distance-time graph, so that a > perfectly circular orbit corresponds to a straight line. > This differs from a distance time graph in an introductory course in > physics where the distance axis represents the length of a one > dimensional vector so that a straight line in this graph corresponds > with a stationary body (and by implication zero velocity and zero > acceleration.) > > > > >> The "bouncing" part is where the satellite has reached the perihelion >> (closest distance) in the orbital period. > > I am puzzled by this. Why isn't there a "bouncing part" at the aphelion? > >> Ironically, at this moment >> in time I would conjecture that it would not be incorrect to stipulate >> that the orbiting satellite is behaving as if it's being influenced by >> a NEGATIVE gravitational field. That's where the 1/r^3 (cubed) part of >> the algorithm comes into play. It influences the direction the >> satellite is taking by pushing it away. Traditionally speaking, we are >> used to interpreting that aspect of the orbit as the influence of >> centripetal action. It's all a matter of interpretation! The cubed >> (negative forces) influence only comes into play in close proximity to >> the planet for which the satellite is orbiting around. At farther >> distances, the normal 1/r^2 (attractive forces) take over. >> >> It's really kind of a nifty perspective, if not a little wacky! ;-) >> >> Regards >> Steven Vincent Johnson >> www.OrionWorks.com >> www.zazzle.com/orionworks >> >
Re: [Vo]:Nature Editorial: If you want reproducible science, the software needs to be open source
The orbital distance is changing faster when the object is closest to the earth which would tend to look like a quick bounce. At the far spacing, the change in orbital distance is slower depending upon the elliptical shape. The mathematical equation defining the function of orbital distance versus time should be available and in a closed form. I recall that equal orbital areas are swept out in equal time, which is one of Kepler's laws as derived by Newton. Wikipedia has a fairly good article on Kepler's laws. http://en.wikipedia.org/wiki/Kepler's_laws_of_planetary_motion Dave -Original Message- From: Harry Veeder To: vortex-l Sent: Thu, Mar 1, 2012 11:25 am Subject: Re: [Vo]:Nature Editorial: If you want reproducible science, the software needs to be open source On Wed, Feb 29, 2012 at 12:50 PM, OrionWorks - Steven V Johnson svj.orionwo...@gmail.com> wrote: From Harry: >> From OrionWorks: >> What I can say is that the new system involves an alternative way of raphing out a periodic orbit - where you plot an "elliptical" orbit on a IME-LINE chart. The orbital distance is the "Y" vertical value and the orizontal "X" value is the time value. > > That graph should look something like a sine curveor not? You're on the right track. However the time-line looks more like a bouncing ball. I think I understand now. You are mapping a two dimensional distance ector to the distance axis of your distance-time graph, so that a erfectly circular orbit corresponds to a straight line. his differs from a distance time graph in an introductory course in hysics where the distance axis represents the length of a one imensional vector so that a straight line in this graph corresponds ith a stationary body (and by implication zero velocity and zero cceleration.) The "bouncing" part is where the satellite has reached the perihelion (closest distance) in the orbital period. I am puzzled by this. Why isn't there a "bouncing part" at the aphelion? > Ironically, at this moment in time I would conjecture that it would not be incorrect to stipulate that the orbiting satellite is behaving as if it's being influenced by a NEGATIVE gravitational field. That's where the 1/r^3 (cubed) part of the algorithm comes into play. It influences the direction the satellite is taking by pushing it away. Traditionally speaking, we are used to interpreting that aspect of the orbit as the influence of centripetal action. It's all a matter of interpretation! The cubed (negative forces) influence only comes into play in close proximity to the planet for which the satellite is orbiting around. At farther distances, the normal 1/r^2 (attractive forces) take over. It's really kind of a nifty perspective, if not a little wacky! ;-) Regards Steven Vincent Johnson www.OrionWorks.com www.zazzle.com/orionworks
RE: [Vo]:Nature Editorial: If you want reproducible science, the software needs to be open source
Speed increases as the satellite orbits closer to its parent, and slows as the orbit is extended. As the x-axis is a linear representation of time, the changes in speed during orbit serve to "compress" the wave troughs and expand the wave peaks. Thus the wave resembles more of a bouncing ball than a simple sine. > Date: Thu, 1 Mar 2012 11:25:02 -0500 > Subject: Re: [Vo]:Nature Editorial: If you want reproducible science, the > software needs to be open source > From: hveeder...@gmail.com > To: vortex-l@eskimo.com > > On Wed, Feb 29, 2012 at 12:50 PM, OrionWorks - Steven V Johnson > wrote: > > From Harry: > > > >>> From OrionWorks: > >>> What I can say is that the new system involves an alternative way of > >>> graphing out a periodic orbit - where you plot an "elliptical" orbit on a > >>> TIME-LINE chart. The orbital distance is the "Y" vertical value and the > >>> horizontal "X" value is the time value. > >> > >> That graph should look something like a sine curveor not? > > > > You're on the right track. However the time-line looks more like a > > bouncing ball. > > I think I understand now. You are mapping a two dimensional distance > vector to the distance axis of your distance-time graph, so that a > perfectly circular orbit corresponds to a straight line. > This differs from a distance time graph in an introductory course in > physics where the distance axis represents the length of a one > dimensional vector so that a straight line in this graph corresponds > with a stationary body (and by implication zero velocity and zero > acceleration.) > > > > > > The "bouncing" part is where the satellite has reached the perihelion > > (closest distance) in the orbital period. > > I am puzzled by this. Why isn't there a "bouncing part" at the aphelion? > > > Ironically, at this moment > > in time I would conjecture that it would not be incorrect to stipulate > > that the orbiting satellite is behaving as if it's being influenced by > > a NEGATIVE gravitational field. That's where the 1/r^3 (cubed) part of > > the algorithm comes into play. It influences the direction the > > satellite is taking by pushing it away. Traditionally speaking, we are > > used to interpreting that aspect of the orbit as the influence of > > centripetal action. It's all a matter of interpretation! The cubed > > (negative forces) influence only comes into play in close proximity to > > the planet for which the satellite is orbiting around. At farther > > distances, the normal 1/r^2 (attractive forces) take over. > > > > It's really kind of a nifty perspective, if not a little wacky! ;-) > > > > Regards > > Steven Vincent Johnson > > www.OrionWorks.com > > www.zazzle.com/orionworks > > >
Re: [Vo]:Nature Editorial: If you want reproducible science, the software needs to be open source
On Wed, Feb 29, 2012 at 12:50 PM, OrionWorks - Steven V Johnson wrote: > From Harry: > >>> From OrionWorks: >>> What I can say is that the new system involves an alternative way of >>> graphing out a periodic orbit - where you plot an "elliptical" orbit on a >>> TIME-LINE chart. The orbital distance is the "Y" vertical value and the >>> horizontal "X" value is the time value. >> >> That graph should look something like a sine curveor not? > > You're on the right track. However the time-line looks more like a > bouncing ball. I think I understand now. You are mapping a two dimensional distance vector to the distance axis of your distance-time graph, so that a perfectly circular orbit corresponds to a straight line. This differs from a distance time graph in an introductory course in physics where the distance axis represents the length of a one dimensional vector so that a straight line in this graph corresponds with a stationary body (and by implication zero velocity and zero acceleration.) > The "bouncing" part is where the satellite has reached the perihelion > (closest distance) in the orbital period. I am puzzled by this. Why isn't there a "bouncing part" at the aphelion? > Ironically, at this moment > in time I would conjecture that it would not be incorrect to stipulate > that the orbiting satellite is behaving as if it's being influenced by > a NEGATIVE gravitational field. That's where the 1/r^3 (cubed) part of > the algorithm comes into play. It influences the direction the > satellite is taking by pushing it away. Traditionally speaking, we are > used to interpreting that aspect of the orbit as the influence of > centripetal action. It's all a matter of interpretation! The cubed > (negative forces) influence only comes into play in close proximity to > the planet for which the satellite is orbiting around. At farther > distances, the normal 1/r^2 (attractive forces) take over. > > It's really kind of a nifty perspective, if not a little wacky! ;-) > > Regards > Steven Vincent Johnson > www.OrionWorks.com > www.zazzle.com/orionworks >
Re: [Vo]:Nature Editorial: If you want reproducible science, the software needs to be open source
>From Harry: >> From OrionWorks: >> What I can say is that the new system involves an alternative way of >> graphing out a periodic orbit - where you plot an "elliptical" orbit on a >> TIME-LINE chart. The orbital distance is the "Y" vertical value and the >> horizontal "X" value is the time value. > > That graph should look something like a sine curveor not? You're on the right track. However the time-line looks more like a bouncing ball. The "bouncing" part is where the satellite has reached the perihelion (closest distance) in the orbital period. Ironically, at this moment in time I would conjecture that it would not be incorrect to stipulate that the orbiting satellite is behaving as if it's being influenced by a NEGATIVE gravitational field. That's where the 1/r^3 (cubed) part of the algorithm comes into play. It influences the direction the satellite is taking by pushing it away. Traditionally speaking, we are used to interpreting that aspect of the orbit as the influence of centripetal action. It's all a matter of interpretation! The cubed (negative forces) influence only comes into play in close proximity to the planet for which the satellite is orbiting around. At farther distances, the normal 1/r^2 (attractive forces) take over. It's really kind of a nifty perspective, if not a little wacky! ;-) Regards Steven Vincent Johnson www.OrionWorks.com www.zazzle.com/orionworks
Re: [Vo]:Nature Editorial: If you want reproducible science, the software needs to be open source
On Wed, Feb 29, 2012 at 8:54 AM, OrionWorks - Steven Vincent Johnson wrote: > What I can say is that the new system involves an alternative way of graphing > out a periodic orbit - where you plot an "elliptical" orbit on a TIME-LINE > chart. The orbital distance is the "Y" vertical value and the horizontal "X" > value is the time value. That graph should look something like a sine curveor not? Harry Both the traditional AND the new alternative algorithms seem to work using this alternative X/Y chart. I overplayed both the traditional and alterative versions on top of each other and they fit like a glove. The implication is what appears to be an alternative (and possibly a more realistic or practical perspective) that suggests something akin to an interplay positive AND negative gravitational/centripetal forces that influence a typical elliptical orbit depending on where the satellite is located in its orbital period. Sorry, that last sentence was a mouthful, wasn't it. ;-) > > PS: I came up with the idea after reading up on one of Miles Mathis essays on > the physics of orbital periods. > > http://www.amazon.com/unified-Field-other-problems/dp/1452005141 > The book is self-published. I'm sure nobody wanted to be associated with what > he wanted to talk about. > > http://milesmathis.com/ > > Some consider Mathis to be a "crank", but he DID give me some ideas! > > Regards, > Steven Vincent Johnson > www.OrionWorks.com > www.zazzle.com/orionworks >
RE: [Vo]:Nature Editorial: If you want reproducible science, the software needs to be open source
>From Michele: > Have you played with celestia? > > http://www.shatters.net/celestia/ No, I haven't. I'll take a closer look at "the tour" when I get some time. > Back to the topic. The original article from Nature: > http://www.nature.com/nature/journal/v482/n7386/full/nature10836.html Thanks for the original Nature article as well. Following up on my previous commentary, yesterday for the first time I performed a computer simulation that showed me what appears to be an alternative way to graph (or simulate) a typical elliptical orbit. Traditional algorithms employ a basic feed-back loop based on 1/R**2 of the distance. However, the alternative feed-back loop algorithm I started experimenting with is based on combining both 1/r**2 AND 1/R**3. Obviously, this new algorithm might sound counter intuitive at first glance. I'll try to explain how I arrived at such an audacious algorithm when I get a little more time in a couple of days. What I can say is that the new system involves an alternative way of graphing out a periodic orbit - where you plot an "elliptical" orbit on a TIME-LINE chart. The orbital distance is the "Y" vertical value and the horizontal "X" value is the time value. Both the traditional AND the new alternative algorithms seem to work using this alternative X/Y chart. I overplayed both the traditional and alterative versions on top of each other and they fit like a glove. The implication is what appears to be an alternative (and possibly a more realistic or practical perspective) that suggests something akin to an interplay positive AND negative gravitational/centripetal forces that influence a typical elliptical orbit depending on where the satellite is located in its orbital period. Sorry, that last sentence was a mouthful, wasn't it. ;-) PS: I came up with the idea after reading up on one of Miles Mathis essays on the physics of orbital periods. http://www.amazon.com/unified-Field-other-problems/dp/1452005141 The book is self-published. I'm sure nobody wanted to be associated with what he wanted to talk about. http://milesmathis.com/ Some consider Mathis to be a "crank", but he DID give me some ideas! Regards, Steven Vincent Johnson www.OrionWorks.com www.zazzle.com/orionworks
Re: [Vo]:Nature Editorial: If you want reproducible science, the software needs to be open source
Steven, Have you played with celestia? http://www.shatters.net/celestia/ Back to the topic. The original article from Nature: http://www.nature.com/nature/journal/v482/n7386/full/nature10836.html mic Il giorno 28 febbraio 2012 15:53, OrionWorks - Steven V Johnson ha scritto: > > Some here might sympathize with the following editorial: > > http://arstechnica.com/science/news/2012/02/science-code-should-be-open-source-according-to-editorial.ars > > In my own case I continue to tinker away with s/w I have developed > over the years that simulate the characteristics of celestial > mechanics. I enjoy delving into the realms of theoretical research. > It's cheap! Cut's down on expensive lab equipment! I realize it is > somewhat maniacal for me to say this but on the surface it appears to > me as if I might have uncovered some additional "laws" that could be > added to Kepler's famous laws... or not. (I realize I could just be > fooling myself! ;-) )Time will tell. I'm still in the process of > trying to construct and run more accurate modeling simulations to > either verify or falsify my suspicions. If the results turn out to be > encouraging I'll need to publish the data along with the s/w that > reproduced the results. In the end one's work must be reproducible. > Otherwise, it all just here say. > > Regards > Steven Vincent Johnson > www.OrionWorks.com > www.zazzle.com/orionworks >