Does Nemesis Exist? Has it Ever?

[chew_on_this]

AND we've been looking long and hard for how long? Certainly not long enough by celestial standards.

 

[mikeemmert]

Hello mlorrey;<br /><br />I started looking at the Hills paper you linked to. I had trouble this afternoon accessing it and finally had to open a new window and copy/paste the url.<br /><br />It's a long paper with lots of complicated math. I still chewing on this. Sometimes I have to reverse and find factors discussed earlier. When I get through, I'll have to go over it again, having memorize what the subscripts on some of those "a"'s mean.<br /><br />So, here on my break, I looked back at the front of the paper. The first thing I noticed is that it was published in 1985. Has any evidence since then surfaced that can more properly evaluate the conclusions reached in the paper?<br /><br />Damn...can't get the "select text" tool to work! Anyway, he is trying to determine the probability of a "Nemesis" object passing through the inner solar system after being perturbed by "black dwarf" missing-mass material.<br /><br />At the time of publication, two facts pertinent to evaluations in this paper were not known: the 47 AU cutoff of the Kuiper belt, and the discovery of 160+ extrasolar planets. Hills only noted the lack of damage to our Solar system.<br /><br />But there is plenty of evidence of damage to other solar systems. This could easily have been produced by their Nemesis's (what's the plural of that?) invading the inner planetary system and creating all kinds of eccentric orbits, etc. This paper is in sore need of update in light of the new evidence.<br /><br />And then there's the 47 AU edge of the Edgeworth-Kuiper belt. It is only recently that the count of discovered KBO's has reached the 1000 mark that makes valid statistical analyses believable.<br /><br />Given that a certain mass object will create this edge at a certain distance, we could construct a graph of Nemesis size vs. perihelion passage. And it would be pretty precise. Nemesis's mass dictates how far away it is at closest approach given that 47 AU edge. It hasn't ripped up the Kuiper belt inside tha

 

[mlorrey]

I find the 1985 publication compelling, in that a number of his observational predictions are borne out by subsequent observations: the largest KBOs we know about tend to be highly inclined and highly eccentric, and there are none observed so far that are beyond the Pluto/Xena mass range: larger planets that may have existed there at one point have all been cleaned out.<br /><br />I am encouraged at reading your simulated Nemesiss captured some KBOs. I am envisioning some bodies with rocky cores and thick ice mantles becoming small oceanic/swampy worlds with thick atmospheres in orbit around a deuterium burning brown dwarf or small red dwarf of less than 100 jupiters mass. Denizens with pale grey skin and extremely large, dark irises that see in IR.... <img src="/images/icons/wink.gif" /> mwahahahaaa....<br /><br />

 

[mlorrey]

Another issue to consider: from what I've been reading, Neptune's gravitational influence could only have cleared out conflicting KBO's out to about 50 Au, yet the Kuiper Belt is thought to extend to ~20,000 Au, where the Oort Cloud starts. That is a rather huge swath of space to be cleared out by some unknown mechanism. Assuming that any Neptune sized object has a 50 AU reach, which has also likely been reinforced by the much closer in Jupter gravity field, a 19,950 AU wide zone seems to mandate either a large Brown Dwarf or multiple superjovians, at the very least. With a radius of influence of 9,975 Au, such an object would either need to be very massive, like a singificantly sized red dwarf, or else have a rather eccentric orbit such that its traverses through the region would result in high velocity non-circular force vectors that would make up for lower mass in causing perturbations.<br /><br />Given the phrase "absence of evidence is not evidence of absence" in dismissing outlandish pseudoscientific or mystical claims, Nemesis theory is actually based on evidence of absence: there SHOULD be a body there, just as observations of the perturbations of Saturn and Uranus led to the natural discovery of Neptune. Before Neptune was found, scientists recognised that it SHOULD be out there, so actually observing it was merely capping what was already mountainous evidence of its existence.<br /><br />Many of the critics of science will try to claim that things like subatomic physics, quantum physics, are not legitimate science because these things cannot be directly observed, as if direct observation is scientifically mandatory. It is not, but what is mandatory is 'evidence of absence", directly observed and calculated evidence that can only be if something exists which has not yet been observed. <br /><br />Theories as to what that something is can abound: witness the many and varied TOEs out there, the multiple string theories, and M theory which ties them all together.<b></b>

 

[mikeemmert]

Howdy, mlorrey;<br /><br />Well, here's a case where I'm glad I didn't pop off at the mouth. When I read Hills' paper and he said Nemesis had to be 10 or more Jupiter masses, I said, "Nah, no way. Jupiter's got to be big enough".<br /><br />But in fact, when I plugged it into GravitySimulator, it turns out that a 300 Earth mass object doesn't look to be big enough to tear up the Kuiper belt and leave an edge. Just ain't got the cojones. I changed that simulation a little. I don't feel like I need a 5000 AU semimajor axis because I'm trying to see if a certain mass object will tear up the Kuiper belt and it takes too long. I used 500 AU instead. I used e=0.8 and I need to use a little less when I try again.<br /><br />A little info about "Black Dwarves", using information I remember from print sources about white dwarves. A star between 0.8 and 5.0 solar masses will create a white dwarf of 0.65 solar masses. I would guess that the reason they all turn out the same size is that that size is where fusion really gets going and blows the outer shell of the star off. If the original star was the size of the Sun or smaller, it would still be active and would have easily been found, in fact it would be one of the brightest stars in the sky. To form a black dwarf, the star would have had to have started at the upper end of that size range. There's no sign of a planetary nebula around the Sun, so a black dwarf is indeed unlikely.<br /><br />There have been all kinds of star surveys that I believe would have detected a red dwarf by now. They're not exactly invisible. The smaller ones are flare stars so the American Association of Variable Star Observers would probably have spotted it by now.<br /><br />We're left with a brown dwarf as the most likely candidate. I'll edit this message with a result from the Gravity Simulater run after I get through frying some chicken up.

 

[mlorrey]

Thanks for the feedback. I will admit that it is likely a brown dwarf, but while all observable red dwarfs have been catalogued, according to Muller 5,000 have never had their distances measured by parallax, so we have no idea how far away they are. This is something he's been working on the last few years, and as I've not heard of any results on this, I expect he's still at it. Given the perceived higher risk of earth-crossers, more of his time has been dedicated to that program, but he seems optimistic that Nemesis is in fact a Red Dwarf. Given his experience, I'm not inclined to gainsay it, though others think it is in fact a Brown Dwarf. <br /><br />Given the difficulties in finding a brown dwarf 1.0 ly out, I think it would be helpful to narrow down potential candidates we already know about but don't know the distance of before haring off on a brown dwarf chase. By the time we know for sure whether or not its a red dwarf, we'll have technology developed to make finding a brown dwarf easier.<br /><br />What speed is your computer? I've got a pretty fast machine here. Is GS freeware/shareware?<br /><br />Here is another question: is it possible to just scale the solar system down to speed things up? Perhaps change the inverse square law or something to compact things down.... just thinking aloud... <br /><br />Oh, here's some more reading, from Muller's book on Nemesis: http://www.muller.lbl.gov/pages/nemch1.htm

 

[mikeemmert]

Howdy, just got back from the Jupiter mass simulation. After some time evolving, seven orbit are decidedly eccentric, and most of them show signs of being disturbed. 242000 years have gone by. So a jupiter sized object can disrupt a Kuiper belt, but it takes a while. My orbital period in this sim is about 12,000 years, so that's 20 orbits.<br /><br />Since there <i>was</i> some distortion with the Jupiter mass, I think we're safe in assuming that the 16 J mass object represents a likely figure. So we are looking for a brown dwarf.<blockquote><font class="small">In reply to:</font><hr /><p> What speed is your computer? I've got a pretty fast machine here. Is GS freeware/shareware? <p><hr /></p></p></blockquote>It says on the front plate that "processor speed is 2.93 Ghz" (? Where does that fit on the scale? I have no idea.). GravitySimulator is real easy to find. In fact, I've run into it idly googleing.<blockquote><font class="small">In reply to:</font><hr /><p> Here is another question: is it possible to just scale the solar system down to speed things up? Perhaps change the inverse square law or something to compact things down.... <p><hr /></p></p></blockquote>You can't change the inverse square law! That's physices. GS has things you can do. There is the time step. That's the time between the calculated positions. If you're modeling Earth's moon, you only want about a 16 second time step, otherwise everything will move so far that the calculations become inaccurate. For simulations involving the solar system including Neptune, about as fast as you can run it is 65536 seconds. I ran this simulation with a time step of 6.0681 days and probably could have run it faster. When I was doing flybys of Triton/Xena past Neptune, I had the approach to Neptune at 64 seconds; when the binary got within a million kilometers I ramped that down to 16 seconds, then four. I wanted to see what was going on, so at closest approach I had it to a quarter of a second and occasionally less. <</safety_wrapper>

 

[mlorrey]

Well, my system is 2.8 Ghz, so your system is pretty good. The amount of RAM and the speed/memory of your video card will also impact sim run speeds, also whether your chip has a math coprocessor or not.<br /><br />What I would conclude from the simulation you've done is that for the particular orbit and mass, there can be no object jupiter sized or larger at the current time without seeing perturbations. What you are doing I believe helps to discount theories of Planet X: there shouldn't be anything taking up residence in the Kuiper Belt that is that big or bigger. Depending on what future perturbations that are observed in various KBOs may indicate whether this is valid or not.<br /><br />What I'd really like to see is a program that takes known objects and their orbits and makes predictions about objects in other orbits which wouldn't perturb known objects significantly. This would be akin to a program that solves Pac-Man for you mathematically: figuring out a consistent path that gets you safely through a simspace without encountering the steeper parts of gravity fields of other objects. If you (i.e. hypothetical object P) or your gravity field touches something else or its gravity field, you die. I know there are methods to solve Pac-Man without simming all possible paths....

 

[mlorrey]

I was just browsing the sedna sim for gravitysimulator. It is a compelling theory that could be more evidence for Nemesis.<br /><br />http://www.orbitsimulator.com/gravity/articles/sedna.html<br /><br />of course, the percent of the Brown Dwarf's orbiting objects that get lost to the sun is dependent upon the mass of the dwarf.

 

[mikeemmert]

Remember, the sedna sim was for a parabolic or hyperbolic orbit - a single flyby.<br /><br />I'm beginning to think that a single flyby would not explain the 47 AU cutoff of the Kuiper belt. While the flyby is going on, there will be all kinds of objects on the other side of the Sun, and they will continue their orbits without being affected. This calls for more simulations.<br /><br />Meanwhile, I assume they are going to look at other dust belts. Some of them do not appear to have a cutoff.<br /><br />I just did a sim with Nemesiss = 1000 Earth masses and it's messing up the Kooiper belt. I gave that one 10 degrees inclination. I'll be trying 45 degrees. I messed up and didn't put enough objects closer in to the Sun. Another modification for the next sim will be coloring the objects differently for different distances so I can see what I'm looking at better.

 

[mlorrey]

Neptune is supposed to have influence out to 50 AU, so the Kuiper Belt may be thinner, or more well groomed, in that region. Beyond it, out to the 20k AU inner boundary of the Oort Cloud, should be a region that was once cleared out by Nemesis, but now is only occasionally messed with when it does its 26 million year pass throughs.<br /><br />Muller says the major axis should be 2.6 ly, which according to him is less than the limit at which passing stars would interfere with it more than once or twice in solar system evolution.<br /><br />Coming as close as 40-50 AU with a herd of Oort bodies picked up on the way in is likely the mechanism for how the comet death showers happen: Nemesis picks up a passel of them in the Oort, each time it heads in, and when it reaches inside the 50 AU limit, the Sun and Neptune strip a large fraction of them away from Nemesis orbit and draw them into the inner system. What we see now for KBOs are former Oort bodies that are largely former Nemesis orbiters that managed to get into resonant or other stable orbits in transneptunian space.

 

[mlorrey]

Yes, I have, Steve, AND I've read Mullers comments as to their mistakes. Have YOU ever read anything by Muller, Raup, Sepkoski, Hut, or Hill? I don't think so. There is a growing body of mutually affirming evidence.<br /><br />Of course, such a body is never going to be observed if head-in-the-sand fools keep yapping about why it is a waste of time to look for it.... you are being entirely unscientific.

 

[mikeemmert]

<blockquote><font class="small">In reply to:</font><hr /><p>And did you EVER bother reading the astronomical reports which state that Nemesis was completely unsubstantiated and unlikely to be real? That's the point & it's pretty well supported by the lack of ANY observations which support the existence of a large body of that type.<p><hr /></p></p></blockquote>Well, Steve, that was then (circa 1993) and this is now (early 2006) and as has been pointed out in this post and in the article that started it all (see the first post of the thread), time has changed what we are looking at.<br /><br />We now have a statistically meaningful sample of KBO's. We also have Muller's paper on the lunar spherules, which maybe isn't as precise as we would like, but is indicative nonetheless. Both have been mentioned before on this thread.<br /><br />Now we have the Stardust samples, which haven't been analyzed yet, but will be.<br /><br />I wish you hadn't shouted on that "ANY" word. I feel there is enough evidence to warrant further investigation. What I was trying to do, and what mlorrey was trying to do, was to figure out what the new evidence is and how much credence to put into it.<br /><br />So we have a theory here, one which doesn't have a lot of support, franky, which has more support today than yesterday. It needs to be falsified (wikipedia). That's what we're trying to do here, including with the simulations. We're trying to come up with some set of observations that either #1) prove the hypothesis is correct; #2) Demonstrate that more evidence is needed, or #3) prove the hyposthesis is incorrect.<br /><br />From the wikipedia article: <font color="yellow">"Falsificationists claim that any theory that is not falsifiable is unscientific. Psychoanalytic theory, for example, is held up by the proponents of Popper's philosophy</font>

 

[mikeemmert]

Moving right along, we have what we need to look for from the one million comet particles, most from Wild 2. One candidate would be deuterium which, according to wikipedia, <font color="yellow">"... is a stable isotope of hydrogen with a natural abundance of one atom in 6500 of hydrogen. "<font color="white"><br /><br />But that's on Earth! It's known that the natural abundance on Jupiter is different; it's higher than here. So I looked around and had to resort to Google to find the FUSE mission. The unfortunate answer is that we don't know:<font color="yellow"> "...astronomers believe that the total amount of deuterium in the Universe is decreasing as matter gets cycled through stars, but they do not know how fast it is decreasing or how much deuterium has already been destroyed. <br /><br />This is where FUSE enters the quest to understand our cosmic origins. Astronomers will use FUSE to search for deuterium in the interstellar medium near the Sun, in gas clouds in the far reaches of the Milky Way, and in distant intergalactic clouds between galaxies."<font color="white"><br /><br />(mlorrey, you had a post on another thread on Bussard ramjets. I imagine the hot wind from the stars of the Hercules cluster is devoid of deuterium. Check this link.)<br /><br />Do flares from brown dwarves contain deuterium? I don't know. Lithium? Unknown. Can the flares from brown dwarves, if those flares even exist at all, be expected to paint comets and dust particles? That's not in my encyclopedic knowlege.<br /><br />FUSE wasn't specifically designed to answer Nemesis questions, but the answers it gives will help answer Stardust questions.</font></font></font></font>

 

[mlorrey]

Okay, I've installed gravitysimulator. Its interesting. I've run the oort-kuiper belt sim, adding a Nemesis that goes by Mullers rough description: major diameter 2.8 ly, eccentricity .5-.65, inclinations from 10-30 degrees. I've run it several times at various masses: 10,30,50,100, and 200 Jupiter Masses.<br /><br />I don't know how well the 150 objects this sim has represent the distribution of objects in the Oort Cloud and Kuiper Belt, but here is my results:<br /><br />a) all Kuiper Belt Objects are thrown out of the solar system before Nemesis makes a first pass, so apparently their orbits are not stable over the long term, they are likely recent visitors from the Oort, since the last Nemesis pass 5 Mya.<br /><br />b) Nemesis' perhilion cruises through the outer region of the Oort bodies the author of this particular sim included, and over 10-20 orbits, seems to clean out the outer Oort of a lot of bodies. Some are given gravity boosts out of the solar system on interstellar journeys, while others are slowed down and wind up dropping into the inner solar system for several passes until they are ejected on fast interstellar trajectories or very wide orbits around the binary Sol-Nemesis system. At large Nemesis masses, some objects are captured in Nemesis orbit for 1-4 orbits of Sol before being lost again to Sol. This would tend to indicate that some bodies currently in orbit around the sun may have spent considerable amounts of time (26-104 million years) in orbit around Nemesis and possibly within its heliopause.<br /><br />c) Another interesting phenomenon: some objects perturbed by nemesis wind up in extremely extreme eccentric orbits around the sun, essentially travelling back and forth between the two foci of the Nemesis orbital ellipse on what appear from this scale to be nearly linear paths (i.e. 99.99+% eccentricity) until they are ejected on very fast interstellar trips. Given that each Oort object in this sim likely is a stand-in for more than a million real

 

[mikeemmert]

Howdy, SpaceKiwi. Sorry it took so long to get back to your question:<blockquote><font class="small">In reply to:</font><hr /><p>If I can also ask a newbie-type question ... presumably the most 'likely' place to find a companion would be in the same plane as the Solar System? But, I guess astronomers have had a look everywhere, including perpendicular locations for example?<p><hr /></p></p></blockquote>Well, I'm not <i>really</i> sorry it took so long. Better to get good answers. If you use GravitySimulator, you have to resist the temptation to use too fast a time step. Time step is the time between the calculations of the objects on the board. It starts with the positions of the various objects. Then it calculates where they are, say, 48.5452 days later. This takes about 1/25 of a second I believe (I don't know how many frames/second a computer does. That's what an old CRT TV did, 25 frames/second).<br /><br />If your average object has a vast distance, like what we're simulating here, you can use a long time step. However, if you are simulating, let's say, the Earth-Moon system in orbit around the sun, a 48.5452 day time step represents over 1 1/2 orbits of the Earth around the Moon!! This will obviously introduce mathematical errors. <br /><br />When I first got GravitySimulator, it took a while to learn about time step. I destroyed the Solar System a few times because of impatience with the time step, even though I was warned about it. It takes a while to figure out how to fly GravitySimulator.<br /><br />I have digressed. Sorry, I'm thinking another poster is now going through the difficult process of learning how to fly this infernal machine. It <i>can</i> be worse than Defender or Robotron or Joust.<br /><br />Anyway -<br /><br />I have been running a simulation where a 13 Jupiter mass object is orbiting the sun with a perihelion of about 140 AU with an inclination of 45 degrees to the plane of the ecliptic. Aphelion on this object is />~1000 AU. That's not Nem

 

[mikeemmert]

I've been watching that stupid ad with the guy smashing his monitor with his keyboard for about 10 minutes. What is he doing? Defender? Robotron? Gravity Simulator?<br /><br />...Flight Simulator... <font color="yellow"><i><b>????<br /><br /><font color="red"><i><b>OH, NO!!!!!</b></i></font></b></i></font>

 

[mlorrey]

Ah, perhaps the KBO's in my sim were tossed out by math errors from too large a time step? Seems reasonable. There ought to be some scale by which we can get reasonably accurate orbital simming for orbits of x distance from a star, we can use y time step.<br /><br />Given that Jupiter is the primary filter on objects tossed into the inner system, and we know Jupiter has an orbital period of 11.82 years, what is the largest fraction of this orbital period can we use as a time step before introducing significant math errors?<br /><br />I am planning on building from my original sim a system that has the Jovians, KBOs, Oort Cloud, and Nemesis. I want to increase the number of objects in the Oort and Kuiper zoones significantly, having one object orbiting for every 100,000 actual objects theorized to be orbiting. Given Muller's math saying a 1 million comet death shower should produce 1-3 comet strikes on Earth following a Nemesis perhilion, if we can produce ~ ten objects heading inward for each Nemesis pass, then we have a viable orbit for Nemesis.<br /><br />Another question: is there a function by which I can easily change my perspective view of the solar system in mid-sim? I'd like to be able to move to various views from 0-90 degrees from the ecliptic to get a better view of all orbits.

 

[mlorrey]

Objects that get tossed inward tend to be in orbits retrograde to Nemesis' orbit or which pass in front of Nemesis in the same direction. Do we know if all Oort objects should be on orbits consistent with the inner solar system, or if there should be a relatively even distribution of prograde and retrograde objects?

 

[tony873004]

You probably don't want a time step of more than 1024 if you want to retain some accuracy from a close Jupiter encounter. But it will probably be painfully slow to run the simulation at this rate.<br /><br />A simple and crude method for determining if your time step is too fast is to zoom in on Jupiter or Nemesis and watch a close encounter. Is the object tracing a solid trail, or a trail of dots. A trail of dots is bad, escecially if they're spaced far from each other. That is because whatever Jupiter or Nemesis' gravity potential is at one point is used as the gravity potential for all the space between that dot and the next dot. This is where the math error comes from that causes the unrealisically-high speed ejections. But if it's tracing a solid line, which is still a series of dots except that they overlap each other, then each dot is close enough to the preceeding dot that the gravitational potential between them is not significant. <br /><br />You could just delete any objects that get flung out of the solar system at high speeds. It is the math error that causes such high ejection speeds. But if you ran it at a reasonable time step, the object probably would have had one of two fates upon making such a close pass to Jupiter or Nemesis. It would either have been ejected from the solar system, but at a lower speed, or it would have collided with Jupiter or Nemesis. In either case, it's gone. Just delete them.<br /><br />To change the perspective, use the scroll bar on the right of the screen.<br /><br />

 

[mlorrey]

Hmmm, understood. Optimally, if you instead used a dots position to adjust all other objects positions from 50% before the present increment to 50% after, this backward correcting should negate most all math error. This probably takes more processing time than just using a smaller increment, though.<br /><br />What do you mean by 1024? Minutes? Hours?<br /><br />I've got 8 various PCs in my basement. I've been thinking of setting them up into a Beowulf cluster. This would be a good application to run. Is there a free version of GravitySimulator I can run on a Beowulf?

 

[SpaceKiwi]

Hi Mike, thanks for your detailed and informative response. I'm really enjoying the discussion in this thread too, it's a tantalising initial question. <div class="Discussion_UserSignature"> <p><em><font size="2" color="#ff0000">Who is this superhero?  Henry, the mild-mannered janitor ... could be!</font></em></p><p><em><font size="2">-------------------------------------------------------------------------------------------</font></em></p><p><font size="5">Bring Back The Black!</font></p> </div>

 

[mlorrey]

BTW: Why can't the sim just do an integration? You seem to be implying that the sim is using something less than calculus...

 

[tony873004]

<blockquote><font class="small">In reply to:</font><hr /><p>Optimally, if you instead used a dots position to adjust all other objects positions from 50% before the present increment to 50% after, this backward correcting should negate most all math error. This probably takes more processing time than just using a smaller increment, though. <br /><p><hr /></p></p></blockquote><br />I'm not quite sure I understand. It sounds like you are describing halving the time step. Also, there's a method called Runge-Kutta 4 which takes half time steps and averages them for a more accurate result. But it takes longer for the computer.<br /><br />By 1024, I mean seconds.<br /><br />I imagine that networking 8 computers together would introduce a lag time that would negate any benefit. I could be wrong.<br /><br /><blockquote><font class="small">In reply to:</font><hr /><p>Why can't the sim just do an integration? You seem to be implying that the sim is using something less than calculus... <br /><p><hr /></p></p></blockquote><br />It is doing a numerical integration. There is no solution to the n-body problem when n /> 2. There are some estimations that work well for orbits whose pertabutions are well-known. But for simulations like the ones described in this thread, they would fail, leaving numerical solutions as your only choice.<br />

 

[mikeemmert]

<blockquote><font class="small">In reply to:</font><hr /><p>There is no solution to the n-body problem when n > 2. <p><hr /></p></p></blockquote>Hate to nitpick, but there are two stable and three unstable solutions if n=3, of course the Lagrange solutions (which is how I got into this whole project, anyway). Lagrange won a huge prize from the French Academy of Sciences for that.<br /><br />But if n />3, forget it, <i>including</i> the case where both L4 and L5 are filled. If mlorrey can find such an integral, he wins a huge prize. I hope he does. I think he'd have better luck hooking his computers together. I know little about that, but I imagine you would break the problem into parts.<br /><br />If I had 8 computers, I would have them working on different simulations. Several times during the Xena project I wish I had one computer doing small object Lagrange clusters on don't plot while I was doing binary flybys on another, which requires a great deal of supervision, and post flyby on a third. Now I'd want another one to model Nemesis.<br /><br />On my next post, I will try to address a more general audience. I would hope somebody will volunteer something like, "If x is observed (or x has been observed) then we can draw a conclusion from that" and that includes thowing the whole model out.<br /><br />Even stevehw made an astute comment, that these simulations are no good after 100,000 years. I think he knows something, but he was probably thinking of inner solar system simulations, and that sounds only a little bit pessimistic. Question: isn't it more a matter of how many total calculations are made? If so, then one can figure out when to shut the simulation down and start another one based on what you have learned from the last one. Wouldn't that entail time step (which depends on distance, see earlier posts) and how many objects are in your model?<br /><br />One should not hesitate to do math in conjunction with a program like this. I figured out the coordinates of t