Hello


inclinations on a compound parabolic concentrators
http://www.powerfromthesun.net/Book/chapter09/chapter09_files/image011.jpg
Where I attempt to render a light intensity map of point light source, after
placing a disc (act as a screen) under the reflector opening, I get some
convincing caustics;

however the my main aim is to test this concentrating collector model in
semi-real environment. ie.sun light with sun path modelling the sun movement in
the sky to be modelled from the elevation angle and the azimuth calculated
values, from sunrise to sunset, for the corresponding latitude and longitude of
the place where the collector will be  located.  And this is what
sun_postion.ini exactly does.

With the included POV script 'CPC_sun-2.pov' I set the collector and disc on the
'land' and I placed an orthographic camera quite far away from the receiver and
used high magnification to zoom in and render the image showing light
distribution.

it does not show any reflection on the disc (not collecting any light) and here
lies the problem.

Having tilted the reflected southward ,I get too much light distributed on the
collector and being all equally brightened!?


that there is a problem with the light source maybe?

See the links below for all files
http://ia600706.us.archive.org/30/items/reflector/sun_position.inc
http://ia600706.us.archive.org/30/items/reflector/CPC_sun-2.pov
http://ia600706.us.archive.org/30/items/reflector/CPC.inc
http://ia600706.us.archive.org/30/items/reflector/CPC_SUN-2.INI

image show the result i got when testing the collector with sun_position.inc

http://img89.imageshack.us/img89/6383/cpcsun253.png

testing the collector with point light source

http://img713.imageshack.us/img713/6086/pointlight.png

http://ia600702.us.archive.org/4/items/refion/CPC-1-1-1.pov






thanks
Zak

Post a reply to this message

> Hello
>

> inclinations on a compound parabolic concentrators
> http://www.powerfromthesun.net/Book/chapter09/chapter09_files/image011.jpg
> Where I attempt to render a light intensity map of point light source, after
> placing a disc (act as a screen) under the reflector opening, I get some
> convincing caustics;
>
> however the my main aim is to test this concentrating collector model in
> semi-real environment. ie.sun light with sun path modelling the sun movement in
> the sky to be modelled from the elevation angle and the azimuth calculated
> values, from sunrise to sunset, for the corresponding latitude and longitude of
> the place where the collector will be  located.  And this is what
> sun_postion.ini exactly does.
>
> With the included POV script 'CPC_sun-2.pov' I set the collector and disc on the
> 'land' and I placed an orthographic camera quite far away from the receiver and
> used high magnification to zoom in and render the image showing light
> distribution.
>
> it does not show any reflection on the disc (not collecting any light) and here
> lies the problem.
>
> Having tilted the reflected southward ,I get too much light distributed on the
> collector and being all equally brightened!?
>

> that there is a problem with the light source maybe?
>
>

> thanks
> Zak
>
>

With photons, you should use ether spacing OR count, not both.

Check the actual location of your sun. It can be on the wrong side of 
your reflectors. If that's the case, then all your photons are shoot 
away from the reflector.
Try moving your reflector so that it's as close to the origin as possible.
Make sure that your ground is at elevation zero.

I can't find Sun_Position.inc :( hard to test your scene.

Post a reply to this message

ZaK wrote:




this is difficult to compare based on your explanation, the scenes
are setup very differently (why is the object enclosed in a opaque
box in the second scene, where is disc?). Some peculiarities are the
distance of the sun (1.4 units) and the extreme color in the
first version (30000 * White?)

Also, using a sun position based on time is not suitable for
testing purposes. You should try to make a simple repeatable scene
based on the first scene with point area light, then use that light
source at a position specified by sun pos, if that still works
using the light_source from sun pos. don't change other parts
such as landscape unless it works. do visualize the position
of the sun as it might not be where you think it is.

Multifile samples are more conveniently provided as zip.

Post a reply to this message

>
> With photons, you should use ether spacing OR count, not both.
>
> Check the actual location of your sun. It can be on the wrong side of
> your reflectors. If that's the case, then all your photons are shoot
> away from the reflector.
> Try moving your reflector so that it's as close to the origin as possible.
> Make sure that your ground is at elevation zero.
>
> I can't find Sun_Position.inc :( hard to test your scene.

Hi Alain,

you can download all files from here

http://ia700600.us.archive.org/35/items/refcpc/cpc.zip

the problem seems to be that my scene is using real-world measures, and my light
source is too far away, i tried and divided the distance on the x,y,z
coordination by 10000, and now i got some good caustics.

// The light source representing the Sun.
#declare dist = r * ASTRO_UNIT;
light_source { <dist*Sun_X/10000, dist*Sun_Y/10000, dist*Sun_Z/10000> color
rgb<1,1,1>*1.5

parallel
         point_at <0,LOT_ALTITUDE+2,0> //the CPC location!

   photons {
     refraction on
     reflection on
   }

}


thanks for pointing out the count/photon issue

Post a reply to this message

Christian Froeschlin <chr### [at] chrfrde> wrote:
> ZaK wrote:
>

>
>
> this is difficult to compare based on your explanation, the scenes
> are setup very differently (why is the object enclosed in a opaque
> box in the second scene, where is disc?).

this is solely for visual purposes,also to eliminate the computation of the
outside of the reflector.

and i placed a flat box instead of the disc, to obeserve sun rays hitting nearby
the reflector, but the disc is there in the cpc-1-1-1.pov

>Some peculiarities are the
> distance of the sun (1.4 units) and the extreme color in the
> first version (30000 * White?)

i realised that real world measures won't work, as i was told " the scale of my
scene is somehow hitting a precision limit on POV-Ray implementation of photons.
Or perhaps it's just that it would need an insane count/spacing setting to work
at that distance" and (30000 * White?) i thought it might increase the
brightness of the light, so it would show more caustics.

> Also, using a sun position based on time is not suitable for
> testing purposes. You should try to make a simple repeatable scene
> based on the first scene with point area light, then use that light
> source at a position specified by sun pos, if that still works
> using the light_source from sun pos. don't change other parts
> such as landscape unless it works. do visualize the position
> of the sun as it might not be where you think it is.

interesting, i'll try and implement this approach thanks.

> Multifile samples are more conveniently provided as zip.

thanks, i uploaded the files as zip.

Post a reply to this message

>
>>
>> With photons, you should use ether spacing OR count, not both.
>>
>> Check the actual location of your sun. It can be on the wrong side of
>> your reflectors. If that's the case, then all your photons are shoot
>> away from the reflector.
>> Try moving your reflector so that it's as close to the origin as possible.
>> Make sure that your ground is at elevation zero.
>>
>> I can't find Sun_Position.inc :( hard to test your scene.
>
> Hi Alain,
>
> you can download all files from here
>
> http://ia700600.us.archive.org/35/items/refcpc/cpc.zip
>

Taking a good look at sun_position.inc and I see some problem:

In the INC, you unconditionaly set every variables. They should be 
conditionaly set to default values only when they are NOT set beforehand.

Use this construction:
#ifndef(VARIABLE) #declare VARIABLE; #end

Also, you need to set your variables BEFORE you include the INC file, 
NOT after.

As it is now, you can set the hour to noon, but have the sun placed as 
if it was midnight, then wonder why everything is so dark.




Alain

Post a reply to this message