Page 1 of 1

Numerical simulation of combustion

Posted: Tue Mar 14, 2006 5:25 am
by exp
Hi!

I programmed a tool to simulate combustion kinetics during firing a spud gun. During my stay in India last year I was so bored that I build a spud gun and had enough time to write this software too.
I did not validate the results because of a lack of time and equipment, maybe somebody can do that for me (I am now in Germany and I can go to jail for possession of a spud gun).

For the theory of the simulation: I am a mechanical engineering student from munich with profund knowledge of thermodynamics, fluiddynamics and combustion. the theory used here is veryfied and results are reasonable.

If you have any questions, post here or mail to exp3.14159@gmx.de

Have fun

Posted: Tue Mar 14, 2006 12:15 pm
by PCGUY
Hey, I updated your file to a .zip (many people don't have WinRAR for RAR's and it's just easier to use a zip).

I also added a small HTML file for help when people get an error. Many people don't have .net framework either so I included a link to download it from Microsoft in the HTML file.

Posted: Tue Mar 14, 2006 1:09 pm
by exp
@PCGUY:


thanx!

Posted: Tue Mar 14, 2006 2:38 pm
by saladtossser
u miss spelled butane with buthane, but i like it

Posted: Tue Mar 14, 2006 4:42 pm
by drac
Oh salad, do you have to nitpick? I for one think it's incredible, this guy's got some mad skills.

Posted: Tue Mar 14, 2006 6:18 pm
by boilingleadbath
I ran the program with the specifications of latke's L3 launcher; .3m/s front velocity.

The results did not come close to matching latke's experimental values, with energies being about 1/4 of those reported.
I had some luck changing the front velocity to about 3 m/s and using pressure gain of 2.5... matched the entire curve fairly well, acctualy.

Posted: Tue Mar 14, 2006 8:22 pm
by drac
BLB, perhaps that is because the program's combustion modeler is based on Butane, not propane?

Perhaps an Emperial version is in order? It would make it easier for those of us not good with metric (Me).

Posted: Tue Mar 14, 2006 9:06 pm
by boilingleadbath
The energy/volume for a stociometric mix is nearly the same... butane is going to be giving you something like 1% more fuel/fuel-air mix volume, but is going to burn with slightly less energy/mol of O2.

Posted: Wed Mar 15, 2006 1:42 am
by Rambo
Why the program changes the energy of the projectile when useing 1kg and 2kg for not a big gun? (V*V*M)/2=const (for hevy spuds)
Why don't we use GGDT with 40psi, helium gas and burst disc for a valve?

Posted: Wed Mar 15, 2006 6:35 am
by exp
@ boilingleadbath: today I am not able to open Latke's page to check his results. A pressure gain less than 4 is not realistic because the mixture will simply not ignite. Does Latke use any kind of turbulator in his combustion chamber to explain a front velocity of more than .37?
Maybe a CFD (Computed Fluid Dynamic) simulation should be done on that matter*.
Of course the real process will be different because of 3D flame distribution and flux towards the barrel entrance. Maybe I should do a version 2 with a simple 3D model...

@ drac: There is really no significant difference between propane and butane.

@ Rambo: The projectile energy is the integral of the pressure over the barrel length. A heavier projectile causes it to stay in the barrel longer, therefore the burnout is higher and the pressure too -> more projectile energy.

@ all friends of empirial units: I for myself am very happy with SI units especially for more complicated mathematics. If somebody wants to edit it to inches/feet/pounds/etc. I will send you the source code (it's C#).


*but it is a huge amount of work and computation power...

Posted: Thu Mar 16, 2006 7:56 pm
by boilingleadbath
Yeah, he has a fan in the chamber, which probably causes some degree of turbulance. Of course, I forget wheather he has the ignition behind or infront of the fan, so that may not be the cause.
(turns out that it's movable, so no word on where it was during the tests)

A chamber fan adds about 40%-60% more muzzle energy acording to latke tests.