CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Modelling Valveless Pulsejet (http://www.cfd-online.com/Forums/fluent/75155-modelling-valveless-pulsejet.html)

husni April 18, 2010 05:27

Modelling Valveless Pulsejet
 
hi all. i'm husni. i need ur help.
i'm simulating a valveless pulsejet wit Fluent. first i try to run with density based, steady, reynold stress viscous, and species transport with propane-air. after 800 iterations, it's convergent. but i'm not sure with my result.
i use 2 pressure inlet BC for fuel and air and 1 pressure outlet. is that quite right?

and for transient simulation, i don't do that yet because i don't know what things to do in transient simulation. i look in fluent tutorial guide about transient compressible flow on nozzle and i find that we have to use UDF for one BC (for this tutorial it's for pressure outlet).
i don't know how to make an UDF, and for my case on valveless pulsejet, where should i use an UDF?
and should i use ignition/spark on my combustion chamber?
and last, how to calculate Thrust that valveless pulsejet produce?

thank you very much for the answer. i really appreciate it.

AMusteykis April 21, 2010 05:49

Hello, husni!
First of all, are you sumulating pulse detonation or just deflagration?
You should definetly go to transient solution because this case has transient physics. You can learn how to create UDF for your case but you can also do the following: solve flow of components in combustion chamber, stop solution, change inlet bc's into wall, because you anyway won't have inlet flow during combustion, ignit your mixture using spark or zone of high temperature and solve combustion process. The main problem is to know when you need to switch your inlet flow again, it should be correlated with pressure value near inlet. So, you'd better create UDF for this case. It will do it all for you.
How to create UDF you can look here: http://my.fit.edu/itresources/manual...f/main_pre.htm
And moreover, what model are you using for combustion? For such case I suppose premixed combustion model fits best because you anyway need to solve the propagation of flame front.

husni April 21, 2010 12:18

thanks a lot Musteykis.
i'm new in CFD and Fluent.
i try to running steady to know the mixture of propane-air so i can put the ignitor on the right mixture fraction. i read on tutorial that we need to calculate steady condition for initial solution for unsteady case. is that necessary or should i run transient directly without steady condition before?
i'm not quite sure i simulate pulse detonation or deflagration. i just want to see the numerical result about my pulsejet design. i want to know that if it will give same Thrust from analytic calculation (about 10 Lbs).
my lecturer said this case is non-premix model. i modeled it with two inlet for air and for fuel than those will mix in combustion chamber and ignited.
is that correct/suitable for my case?

until now i can't get the convergence. i think the problem is on boundary condition. should i model far-field condition on my case? because my model just axis-symmetric pulsejet. for combustion model i can only use species transport because i choose density based solver. when i try pressure based, divergence problem occurred.
with species transport, how to model combustion process? for now i just want to model the flow and mixture of propane-air and then will continue with combustion process.

thank a lot for your answer. i really appreciate it.

AMusteykis April 21, 2010 14:12

Hello again!
You can actually model either steady or unsteady state for filling process. But I would recommend you to use unsteady simulation because in this way your case will be logically divided into stages: filling, combustion, exhaust and refilling with exact times needed for every stage. It will help you further to evaluate overall performance of your designed pulsejet.
Well, about combustion it is not so clear as I thought. Yes, you can't use premixed combustion because you have inlets and that's why you won't have ideal distribution.
If you use spieces transport you can model combustion enabling Volumetric reactions with one reaction for propane-air. They can be based on finite-rate Arrenius equation or on turbulence dissipation. For more about it: http://my.fit.edu/itresources/manual...ug/node601.htm

Hope it helps.

husni April 22, 2010 00:42

thanks for the answer Musteykis.
perhaps i will run it in unsteady condition directly.
do you ever running pulsejet simulation before? if you don't mind can you give some tutorial guide with pulsejet CFD simulation?
i still don't understand how to model pressure or velocity function in UDF. i want to know how the function is and how to get the number. can you give me some guidance and example of calculation maybe?
please send me on my email : husni_firmansyah@yahoo.co.id
thank you very much, you help me much.


All times are GMT -4. The time now is 01:01.