Solar Cooking Archive Discussion Forums

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

I have been looking for software to show how light bounces from parabola's when they are pointed a little away from the sun. Even 2 d would be helpful. You need to be able to move the "ray" to simulate the sun moving and it has to reflect off the surface and ideally be absorbed by the cooking pot shape.
Currently I am playing with a compound parabola where the focus is the same for both parabolas and it accepts light for 2 3 or 4 hours. (depending on how great the offset of the 2 parabolas is. I have to use a protractor to show myself how light bounces (which is pretty tedious.). Considering just how advanced even the free software (stuff like povray is miles ahead of what I need for instance) is now it should be extremely easy to find a simple program to do this. even one from 5 years ago would be fine.
Alas, i have not yet found one.
Any ideas?
Thanks
Brian White

Bill Bradley · Joined: 26 Nov 2006 Posts: 21 Location: Springfield MA USA

I have done some two dimensional ray tracing with Mathcad. It is a bit tedius, but works quite well. The idea is to write an equation for each ray and then use the graphing function to plot the ray. While I used Mathcad, this principle should work with any software that lets you write equations and plot the results. Mathcad has a neat animation function that allows you to iterate any variable such as the sun angle and produce an AVI video of the result. See some examples on my website
www.earthboundtech.com
_________________
Bill Bradley, EarthboundTech

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

Just a note that I used a hardware solution instead. I used a laser level, a page with drawings of the curves and a strip of mylar over the curves.
This worked really well and I have video at
Take a look at http://www.youtube.com/watch?v=uX9Z-nsUHiA

You can find info about compound parabolics at
http://www.fossilfreedom.com/increase-output.html
and at http://www.appropedia.org/Understanding_Solar_Concentrators
Brian White

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

I have been looking in forums and so far google sketchup, truespace and blender seem like possible options. Google sketchup have given the quickest answers (by far) in their forums. Blender have been the slowest with answers. I asked them first too.
A guy already has a parabolic dish and how to make the model on sketchup! And they have a thing where they can shine 3 different colours of light on different parts to show how the light gets collected. Check the sketchup forum for my solar cooker thread and find the link to how to do a parabola in the thread.
Once you can do a parabola, perhaps you can also do a panel cooker, or a compound parabolic dish! I am keen to see it. One part of the compound parabolic dish I didn't get to design was the bit directly behind the cooking pot. (I do not know how it should be shaped so I just left the 2 sides meet. This is probably not optimum. Anyway, it looks like sketchup but we shall see. blender and truespace are more mature programs with great power. Their community may spring into action.
I did get one reply on the blender forum from someone who wanted to do the same thing as I wanted so there is hope!
Brian White

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

In fairness truespace gave quick replies too. I was just not informed. One of the replies suggested which renderering programs I must use in conjunction to get proper reflection modeling (like real life).
I quote
"You would need to use caustics to calculate reflected light. This takes a lot of processing power so would be slower to render than not using caustics.

You could write scripts to handle the sun position, and move lights following that formula, and then render to see how the light is focused. I imagine writing the scripts to calculate where to place the lights dependent on geospatial location, time of year and time of day would be quite tricky, as there is a lot to that, but I can see no reason why it couldn't be implemented in tS via scripts.

You could use the SDK and so code in C# instead, which may allow use of existing libraries that solve the sun calculation issue.

Note that doing an animation with caustics could be very slow, eg if you want to see your solar cooker's performance over a period of time, you'd have to render multiple frames and not just one image, and with caustics being processor intensive and so taking a while to render each single frame, then that would all add up of course.

To use caustics, you'd have to use either Virtualight or V-Ray to render with.

HTH!
Tom"

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

Blender institute replied to me and said that although blender does not have that functionality, it could be added or scripted in. (a function to have the suns rays shine on a 3d model and show where they end up as the sun travels across the sky).
so I am still hopeful.
Brian

mcaK · Joined: 24 Dec 2008 Posts: 1

You might want to investigate the Wolfram Mathematica software. I don't have hands on experience with this app myself, but it appears to tailored for the type of analysis you're talking about. Functions can be defined & used as input for other functions, etc. The variables can even be set to sliders so out put can be changed by regular Joe users. Of course the results can be presented in tabular form, statistically or graphically. Particularly for a 2D solution, it looks quite doable. (This is of course completely ignoring the learning curve for the software.) Wink

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

I got some useful replies in this thread, and didn't seem them till recently.
http://blenderartists.org/forum/showthread.php?t=139485&highlight=solar+cooker
I am useless with software but perhaps others might be able to use the information in their work
Thanks
Brian

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

Art of illusion have been extremely helpful! I only discovered it a couple of days ago and already I have had replys from the originator and several people in its forums.
http://sourceforge.net/forum/forum.php?thread_id=2915545&forum_id=47782
Someone even did a very quick tutorial on the spot!
Art of illusion has 4 panels for elevation, plan, sideview and perspective. A little like technical drawing. Also some aditional panels for adjusting material, colour etc.
If you want to make a curved dish, just plot out the curve and use the lathe tool to turn the curve into a dish! I still have to figure out reflectivity, etc but they are helping me in the forums. Perhaps art of ilusion is the one. It has a simple interphase, it runs as a java application (not an applet) so it runs on linux and on windows and it is a free download.

webmaster · Posted: Fri Jan 30, 2009 9:42 pm Post subject: Great information!

Could you post this information on the Solar Cooking Archive Wiki?

http://solarcooking.wikia.com

Thanks,

Tom

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

Thank you, Tom, I will.
I have been learning the software and I have set up little tests with a parabolic, and several of my different "compound parabolic" shapes and a hemisphere. Surprisingly, EVERYTHING is about the same when they are pointing directly at the sun. (No advantage to parabolic because the cooking vessel is much larger than a dot!)
The winner so far is the Hemisphere. It goes off focus slower than anything else.

Bill Bradley · Joined: 26 Nov 2006 Posts: 21 Location: Springfield MA USA

I too am interested in compound parabolics. Several ray traces that I have seen show up to 3 reflections before hiting the pot. I am interested in what can be done without requiring multiple reflections.

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

http://www.instructables.com/id/SHDDXHZFR0ZHCZX/ has my first results and some useful files to get you started with art of illusion.
You might like to ask your question in their forum. They are very helpful.
I have barely scratched the surface of what their software can do!
I also made a little model dish which I test with a laser level. It shows where the "up to 3 reflections" happen and gives clues about how to fix the problem.. When the compound parabolic dish is directly pointed at the sun, almost everything hits the pot with one bounce.

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

I have an "instructable" devoted to using the art of illusion software now.
http://www.instructables.com/id/Use_software_to_design_a_better_solar_cooker/
I have a downloadable zipped folder. The folder contains an art of illusion scene file, (this scene file has shapes for 5 different solar cooking dishes ready made). You can then make and compare your own new reflector shape. The folder has a html file which can automatically tabulate 10 stills to show the differences between the reflective abilities of different dishes over time.
I used seamonkey composer (like the olde netscape composer but updated) to make the tables). You can use it to edit the names of dishes,etc in your own tables.
It was a lot of work, and it is not finishd.
I also have 2 explanitory videos on youtube.
http://www.youtube.com/user/gaiatechnician and they are on page 2 of my videos. They are not very well made but help explain things quite well, I think.
Brian

gaiatechnician · Joined: 06 Jan 2008 Posts: 32 Location: Victoria BC

I have several new videos to help anyone using art of illusionsoftware for solar design. They have got favorable comment even from the art of illusion forums so if anyone wants to dive in and invent a new reflector type, install art of illusion on your computer, download the scene file (8 or 10 kb! small) and start designing and comparing with other solar cooker designs already included in the scene file.
Best way to get the tutorials is to got to http://www.youtube.com/user/gaiatechnician and click on my recent videos, then choose the ones you think most suitable.
Brian