Forum Groups
  All forums
    Help & Feedback
      Questions
      Work in progress
      Finished Art
      Non-Max related
    Community
      Offtopic
      News
    Hardware
    Photography



Maxunderground news unavailable

 
First page  Go to the previous page   [01]  [02]  Go to the next page  Last page
 
[Vray] - Simulating Fiber Optics?
show user profile  9krausec
If a person wanted to create a real world simulation of something along the lines of a fiber optic cable, would Vray be a good bet to do this?

Would the visual output be a good ballpark for real world results?

There is a project going on in Engineering and they need to create a solution to make light travel through some plastic shape. They've asked me if I could simulate their various designs to see which one would be most effective before tooling.


So GI, Caustics, Dispersion on and good to go, right? If the materials are accurate, the light intensity is accurate, then the resulting visual output should be accurate as well?




- Portfolio-




read 314 times
12/6/2016 5:19:38 PM (last edit: 12/6/2016 5:19:38 PM)
show user profile  herfst1
I'd use an sss material. I've only seen the end of a fiber optic but it's very gel-like.
read 304 times
12/6/2016 7:02:41 PM (last edit: 12/6/2016 7:02:54 PM)
show user profile  9krausec
Good call on the SSS. I don't know how well that will keep with keep with real world properties, but the way I figure is that if the material and the light source in the scene are the constant, and the only thing that changes is the geo, that it'll be fine to show what would work better.

This is all based around the assumption that Vray's light physics is solid (which it should be). Thanks.




- Portfolio-




read 300 times
12/6/2016 7:34:12 PM (last edit: 12/6/2016 7:34:12 PM)
show user profile  Nanne
I wouldn't bet on it. There is a lot of trickery going on under the hood of a renderer such as V-Ray. Especially in the SSS shader I would imagined, it is optimized for speed for rendering animation rather then physical accuracy.

I would run test first to compare it with real world images to see what you get.

Kristoffer Helander
www.kristofferhelander.se
  :: Works ::   |  :: Resumé ::   |  :: Contact ::   |  :: CG Blog ::   |  :: Links ::     
read 291 times
12/6/2016 8:57:16 PM (last edit: 12/6/2016 8:57:16 PM)
show user profile  ijzerman
Its going to be a bitch anyway but i think it can be done.
If you set up the ior correctly of the material vray should do the magic.
Glass fibre works by keeping the light inside the material by reflecting on the inside and not letting it out. If this is done by a coating or something you can do the same using a blend material.
Its not a scientific simulation tool in any way but it is physically accurate.
I made this a few years back just to see if vray could make a prism without trickery



It returns 3 solid colors which is correct since digital imaging works with 3 colors instead of analogues inifite colors.

--------------------------------------

Pushing buttons since "86
read 272 times
12/7/2016 8:45:09 AM (last edit: 12/7/2016 8:45:09 AM)
show user profile  Pil
Cant you just fake the output based on information of cable length and potential loss along the cable?

If there is a loss at all?

I would simple add a light to the end of the cable and calculate the output...
But disregard my post if it really need to be simulated :-)



read 256 times
12/7/2016 6:35:16 PM (last edit: 12/7/2016 6:35:16 PM)
show user profile  9krausec
Thanks guys. Cool video ijzerman!

It does need to be simulated. I'd 100% just fake this effect if all we needed were some pretty pictures. However, we are exploring the most effective shape to carry the light through (or something along those lines).

Solidworks can't simulate light, so they asked for my help with it.




- Portfolio-




read 245 times
12/7/2016 10:01:45 PM (last edit: 12/7/2016 10:01:45 PM)
show user profile  Pil
There is a, more or less, rule of the accepted angle in fiber optics for the beam to succesfully reflect on the inside of the tube.

Cant you just use this formular to stay within the limits?
Bend your shape how ever you like but put a limit/constraint on the curves!

This is really just math so I dont really see the need for a simulation.
Again if I dont make sense then just disregard my post.
I know how annoying it is when someone is trying to steer you into something you dont really want :-)




read 242 times
12/7/2016 11:57:28 PM (last edit: 12/7/2016 11:57:28 PM)
show user profile  ScotlandDave
My first thought would be that an unbiased engine would be more accurate with something like that, but in saying that, although most engines will take shortcuts or be quite selective in how they do path tracing etc i would think they do for the most part produce realistic if more approximate results. It might be worth if possible doing some single-frame tests to compare the results of an unbiased engine with the results from vray..


Website | Blog | Contact | Vimeo

read 238 times
12/8/2016 12:49:52 AM (last edit: 12/8/2016 12:49:52 AM)
show user profile  ijzerman
You can make Vray pretty much unbiased by setting adaptive amount to 0. under global DMC.
Does take ages to render tho. Its all in the correct setup of the material.
Have you checked vrscans?


--------------------------------------

Pushing buttons since "86
read 233 times
12/8/2016 9:05:55 AM (last edit: 12/8/2016 9:05:55 AM)
show user profile  joelr
Clayton, are you talking about a lightguide? im not sure because you wrote fiber optics...

if you guys are trying to light through some clear plastic its usually called lightguide.
i do a lot of those in my work.
im asking because i can send you a document that is about better design of lightguides.

also, in my experience, you will probably spend some time on this- it might be quicker and cheaper to just CNC the different plastic solutions and try them out (unless they are really big)

joel



read 219 times
12/8/2016 3:45:21 PM (last edit: 12/8/2016 3:45:21 PM)
show user profile  9krausec
Thank you all for the replies.

@joelr - A lightguide is exactly what I'm referring to. I wrote fiber optics as it was the closest technical name I knew for it (although I know fiber optics are in a different ballpark). My email is 9krausec AT gmail DOT com. It would be great if you could send me that document to review.

I'll bring up CNC'ing during our next meeting. We have the ability to do that at the office. Thank you.

@ijzerman - Good call on the vrscans. I know about it, but didn't consider it an option until you brought something up. I'll look into pricing today.

@ScotlandDave - I'm going to try and keep everything in vray, but this is good advice if vray doesn't do the trick in this instance.

@Pil - I'm a 3D Generalist Jim, not a scientist. I'm sure their are mathematics to assist in this, but I'll leave the math to the engineers on this one. Thank you for the guidance though. I'm not sure what shape they have in mind.. If it's a simple bent tube or if they want something a bit more complex. No cue as of now.


Thanks everyone. I'll post an update as to what I end up doing for this.




- Portfolio-




read 214 times
12/8/2016 4:36:20 PM (last edit: 12/8/2016 4:36:20 PM)
show user profile  joelr
sent you the email.



read 203 times
12/8/2016 8:49:48 PM (last edit: 12/8/2016 8:49:48 PM)
show user profile  jpedleham
I know you said vray but corona would probably do this quite well. a while back someone accurately built lenses with the physical glass and they built a working zoom lens with DOF in max. stuck it infront of a standard camera and it worked perfectly. i wouldnt be surprised if you could get a decent fibre optic effect out if it.
read 199 times
12/8/2016 10:08:39 PM (last edit: 12/8/2016 10:08:39 PM)
show user profile  jpedleham
I just had a quick go and yeah it seems to work remarkably well actually.

https://s23.postimg.org/gry69ll97/caustics_2.jpg
read 176 times
12/9/2016 1:36:27 PM (last edit: 12/9/2016 1:36:27 PM)
 
First page  Go to the previous page   [01]  [02]  Go to the next page  Last page
 
#Maxforums IRC
Open chat window


Support Maxforums.org