What’s your dt? I assume that you’ll see something squirrely happen if you boost your dt high enough.
Umm, just a side note, but that’s 50,000…not 5,000,000 and I don’t see variable ‘i’ being used anywhere else. Are you iterating the entire function 100 times or 5,000,000 times or is that taken into account by the delta time variable?
I changed the iterations down to something reasonable immediately after it ran though so that I didn’t actually run it and lock up my system for another hour. Just a precaution.
i isn’t anywhere, I’m just running that loop however many times I set i to. That’s the only reason for the loop. I can run it only 1000 times and get a half circle because it didn’t have enough iterations to make a full one.
I believe I was using .001 as my dt. I have a feeling it has something to do with dt because I just chose that arbitrarily and didn’t calculate what it should be. But still, after 5 million iterations one would think it wouldn’t matter. I’ll calculate what it should be and let you guys know what happens.
So the circular orbit is with dt = 2pi/100 and the cool looking ring is with dt = 2pi/10. Obviously higher dts introduce more error, but I wanted to find instabilities in the orbit to see it precess which is what I would have liked to see…
From the Wikipedia article on apsidal precession:
There are a variety of factors which can lead to periastron precession, such as general relativity, stellar quadrupole moments, mutual star–planet tidal deformations, and perturbations from other planets.
I’m assuming that your equations don’t attempt to emulate/simulate any of those factors. It all looks pretty symmetrical to me.
I should have precession not because of my equations but because of the limitations of how my algorithm is calculated since I have a finite amount of precision with floating point and because I must choose whether to calculate the force vector or the velocity vector first.(force first will cause it to spiral inward, velocity first spirals out word. If you make a force diagram you can see how after many iterations this would happen). I guess precession was the wrong word. I want to see the instabilities of the algorithm.
I already know a method to bring the error down to third order but I just want to see it first.
I’m looking at this as a software engineer, not a scientist, so I can’t speak to anything in your equations.
Single precision values have between 6 and 9 significant decimal digits. That’s far below pixel resolution on a 1920x1200 monitor. So you may well have significant drift in the orbit, but it’s not going to be visible at the pixel level.
Try larger dt values - and I mean by an order of magnitude at a time. Also, display only 1 of every N orbits. Start with 1 in 10, then 1 in 100, etc. You’ve got a lot of orders of magnitude to fiddle with.
Well, I was, and am able to see some drift…but it’s only the drift in the actual precession of the orbit(You’ll see what I mean). Precession is absolutely what I meant once I could differentiate between N orbits.
I was playing with it today with another Prof and we got some pretty cool things(which to my own frustration I cannot reproduce) showing the precession, but I find with any dt that is on the order of 1/1000 I have 0 instability. Above that and we get some.
It was so cool and I can’t reproduce it T.T
Edit: Red orbits -> Green -> yellow -> blue
1 Like
The instability is caused by over/underestimating of the centripetal force due to inaccuracies in the position of the planet. If dt is very small, the change in position will be very small, and the inaccuracies in force will also be exceptionally small. I’m guessing that you’re approaching the limit where you’re approximating simultaneous updating of your variables.
“Seal spotted surfing humpback whale in Australia” I think hes facing the wrong way!
2 Likes
1 Like
I’m so happy Japan exists.
1 Like
Does anybody on here have experience with Barbed Fittings?
I would like to know how permeable they are. Barbed fittings hit the spot between low pressure compression fittings and (very) high pressure power tooled crimp fittings. I know those high grade ones are really tight (and too expensive for my application), but how do barbed fittings compare to those?
I’m talking about leakage per year at a certain pressure (2 MPa). I couldn’t find any data on that on the web, I suppose because it really depends on how exactly the barbed fitting is shaped, the type of hose used, the support piece that is used to secure it and how that is applied.
Why are you considering leakage at 2MPa? Surely any leakage at that pressure would ruin the seal by forcing the tubing wall outwards, causing the fitting to slip and rocket out of the tube.
I didn’t expect that seals were considered near perfect and any leakage neglectable. If that is the case, that would explain lack of mention in the documentation.
So, I “just” need to assure that the seal is properly established and that it withstands the pressure? Doing a first check for leaks after assembly to prevent later (but inevitable) explosive decompression, or would that happen right the first time anyway? What’s the point of leak detection then?
In considering this a bit more, I think you still have to worry about leaks. Gene could probably give you a bunch of insight into this stuff.
A perfect fitting and hose would either maintain a connection or just fail catastrophically. Imperfections in fitting and hose would allow for leakage without ejecting the fitting. Imagine a hose or fitting that has aged and which has been incrementally eroded or corroded by the fluid. Tiny cracks and such would allow fluid to move past the seals without attacking the overall integrity of the seals. Well, those same cracks could be in the fitting or hose from the time of manufacture. Also, it may be in the nature of the fitting or hose to compress or expand asymmetrically under pressure in minute amounts, again exploiting imperfections in the connection. Beyond that, there’s the mechanical strains of any changes in pressure, possibly moving the hose on the fitting or temporarily breaching a rib.
This is why I like software; bits are far less temperamental than atoms.
3 Likes
So I was talking to a friend about I:BS, telling him to check it out and stuff.
“They built an engine? Yeah right, it’s probably just a version of Unity. I’ll hack it using stuff off steam, break it, and they’ll be all like ‘Your banned, please fix our stuff for $30 an hour’”. 
I’ve been telling a buddy about Infinity for the last 5 years and he almost always rolls his eyes and says “you like a game that isn’t even a game”. Haters gonna hate.
1 Like
Can’t really fault this assumption thesedays with how many “projects” are just flipping unity store assets. Steam’s bar of quality hasn’t existed for a few years since they opened the floodgates. Just send him the 2010 video and he should get it.
1 Like
My coffee maker is broken…this is very disheartening to me…
How can I even get to class tomorrow? 
1 Like