Author Options:

Pulse detonation(numerical simulation) Answered

Hello everyone iam a Msc student now, i have done many projects over pulsejet engines which you could see in my instructables list. Now i have decided to do a project over pulse detonation engines, before beginning with the real one i thought that making a fluent simulation will be better. I have several tries to create the detonation front and atlast i succeeded in creating one. Take a look at the attached pictures which shows the development of a detonation front inside a long tube of 100cm. Further works will be updated soon.

Want to see the sequence of deto wave formation please follow the link...


if anyone interested in seeing the pressure contour animation please click the following link.

update: simple video of detonation wave propagation....

Is this a fully turbulent 3D simulation, or just 2D (r-z)? Your pictures are all slices through the tube, but I can't tell if they're supposed to be rotationally symmetric (i.e., 2D) or just representative slices.

Also, what chemistry are you simulating in Ansys? Are you doing multiple hydrocarbons mixed with ambient air, or something simple like just propane-oxygen?

what do u think about this?? the pic of the pressure waves.. Looks very fine..


Looks excellent; congratulations again. With the ANSYS notations included I see that you're doing a 2D (r-z) simulation. Obviously, a full 3D model with turbulence will take "infinitely" more CPU :-), but once you get all of your configuration parameters nailed down, it might be worth getting some time on your local batch cluster just to see how much difference (especially in the trailing environment) the turbulence makes.

You are correct for just 2d simulation and 1 step reaction mech my processor fells like dying. Next time if i plan to do 3d i will try to use supercomputer which my univ has got. Just for 2000 iterations my system took a time of about 5.5hrs.

Update: Here is the pic of the point at which the detonation wave forms and accelerates. The cj velocity was found to be 1760 approx 5.17mach.. First image is the pressure and the second the temp distribution.


That doesn't look like the same scale as your earlier images. How thick is the shock front you show in the first image (say, from the leading edge back to the lightest blue region with the two "notches")?

The problem with the previous shockwave was, it wasnt strong enough to travel along the tube but the recent pic which i have uploaded now is a strong detonation front which travels at CJ velocity till the end of the tube(verified).

Cool. Congratulations on your success!

Cool! Did you write the Ansys code yourself, or do you use one of the standard combustion libraries? If so, which one? Also, can you provide some details on the parameters of your simulation? This would make a very interesting I'ble.

No coding used the available resources.... :D