 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Natural Science Forum / Physics / Optics / December 2004dispersion in prism
I am writing a forward ray tracer to model dispersion, now in a prism, later in diamonds. For the moment I use a triangle with refraction angle of 60 degrees (I work in 2D now) and I work in Java I want to know if my setup is correct, because I get a strange effect. When I send 1 ray of light to the prism, I split the ray in R, G and B (I can do more, but just testing now) and each trace them regarding their speed of ligth and with some interpolation I get a nice spectrum at the screen I placed where the rays exit the prism. But when I send for instance 5 rays, I get 5 times the spectrum on the screen. I get the impression that the difference in deviation of each color is too small to mix up the colors when reaching the screen, but my formulas to calculate angles and the refractive indexes are what they have to be... Thankx! Don't really quite understand what you mean by 5 times the spectrum. Do you mean that the spectrum ends up being 5 times longer? If so, you are not clearing a variable when you do the calculation as any particular color should end up at the same point no matter how many rays you trace. Trace each color by itself and verify that the tracing is still the same when others are included. -- Why isn't there an Ozone Hole at the NORTH Pole? > Don't really quite understand what you mean by 5 times the spectrum. > Do you mean that the spectrum ends up being 5 times longer? No I mean it repeats itself 5 times. I put a little example online, so you can see it better. The way one ray follows is drawn. You can also see the very small difference between the way red, green and blue follow. The example can be found at http://users.pandora.be/fabioccopolis/vb.jpg. Grtz Let me start out by saying that the prism doesn't work the way that you are showing in that picture. By entering the prism at the 90 deg. direction, you don't get any dispersion (angle is 0) and then you do an internal reflection which doesn't do any dispersion and finally, you are exiting at the same 90 degrees which again produces 0 dispersion. Finally, I don't know what the 4 yellow lines on the right side of the prism are as 3 of them seem to just appear from nowhere. FWIW, normal procedures on drawings are to draw the path from left to right. -- Why isn't there an Ozone Hole at the NORTH Pole? > Let me start out by saying that the prism doesn't work the way that > you are showing in that picture. By entering the prism at the 90 deg. [quoted text clipped - 5 lines] > nowhere. FWIW, normal procedures on drawings are to draw the path from > left to right. Hm, I think here is something wrong. The rays on the right you see are not perpendicular to the prism, but something around 15 degrees to the normal at the side of the prism. The lines in the prism are the bent dispersed rays, but that's just one the problems, the difference between the different colors is extremely small. The 4 yellow lines are 4 light rays, but I display here only one that is traced for clarity. Finally, I know that from left to right is more usual, but I don't want to rewrite it for this, the result should be the same. I made this "mistake" in the beginning. Well, I didn't pull out a ruler to check if the lines were exactly perpendicular to the face of the prism but just observed that they were pretty much perpendicular to the faces. Put the ray so that it hits one of the faces of the prism and then goes to the other face without hitting the bottom of the prism and you will be getting a much better dispersion of the light. The prism should take a ray set (all colors to one spot on the prism) and disperse it into the different angles of the dispersion of the colors. There shouldn't be a repeat of the colors in the spectrum unless you take and make a second set of rays and put them in at a different location. This will indeed make a second spectrum as this set is a different set than the first set. If you do this in reality, you set the resolution of the spectrum by the width of the input light beam. I'd also spend the few minutes needed to change the direction of the light rays so that the program agrees with convention. It will make it a lot easier for you to understand others when you do the same basic processes. -- Why isn't there an Ozone Hole at the NORTH Pole? > Well, I didn't pull out a ruler to check if the lines were exactly > perpendicular to the face of the prism but just observed that they [quoted text clipped - 13 lines] > it a lot easier for you to understand others when you do the same > basic processes. Ha, thanks, I'll try that. Someone also told me that I have to calculate the interference of the light rays to have a good result. > "Put the ray so that it hits one of the faces of the prism and then goes to the other face without hitting the bottom of the prism ..." I believe you are describing the following setup, which is indeed how dispersing prisms are typically used: http://images.amazon.com/images/P/B000002U82.01.LZZZZZZZ.jpg -- Mark "Fabiocco" wrote in > I am writing a forward ray tracer to model dispersion, now in a prism, > later in diamonds. For the moment I use a triangle with refraction angle [quoted text clipped - 11 lines] > > Thankx! If I understand your rather sketchy sketch, your input consists first of 1 ray which yields 1 spectrum.Fine. Then you proceed with input rays which are parallel and spatially separated. If this true you will obtain as many distinct (possibly overlapping) spatially separated spectra. Nothing abnormal, your software is fine... as to your understanding of elementary geometrical optics I am not sure. I suggest some study of a book such as Hecht's (look up refraction, prisms, spectroscopic optics). My apologies if I am wrong or mistaken. Any special reason to use a prism under conditions which are very far from maximum dispersion? Why the reflection? > If I understand your rather sketchy sketch, your input consists first > of 1 ray which yields 1 spectrum.Fine. > > Then you proceed with input rays which are parallel and spatially > separated. If this true you will obtain as many distinct (possibly > overlapping) spatially separated spectra. You understand it very well. I put a screenshot online at http://users.pandora.be/fabioccopolis/vb.jpg > Nothing abnormal, your software is fine... as to your understanding of > elementary geometrical optics I am not sure. This could be. I do computer sciences, no physics :) > I suggest some study of a book such as Hecht's (look up refraction, > prisms, spectroscopic optics). I went to the library of the university for some books to see if my setup was right, but there they don't tell much about the ray itself. The only thing I see there is that they use a very very small ray (by sending light through a slit), but they don't tell something how to simutate this with a computer (not that strange, it's a book about physics). > Any special reason to use a prism under conditions which are very far > from maximum dispersion? Aha, is a prism with an angle of 60 degrees far from maximum dispersion? Then I read something wrong. What are the best conditions? > Why the reflection? I think you looked the picture I put online. Total internal reflection occurs when the angle between the incident ray and the normal is bigger than a value depending on the material the prism is made of, but I think you know this, so is there something wrong? Grtz |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2009 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.