[nik.rapons]

Hi,
I'm working with Converge in order to simulate a PFI SI engine (with SAGE) fueled with a mixture of hydrogen and air with an equivalence ratio of 0.25, with both of turbulence models (LES and RANS).
I'm having problems with the LES (viscous one equation) simulation (either in single species diffusion than mixture - averaged diffusion), because I cannot ignite the mixture, while RANS case shows combustion.
The cases are the same, except for the turbulence models.
I'm simulating first both the intake (from -360 °CA) and compression phase (to -25 °CA), for LES and RANS cases, with a base grid of 4mm and some fixed embeddings: lev 3 around valves and piston skirt, lev 2 around tumble fixture, head, liner, piston and injector zone.
Then, with the two map files from -25 °CA, I run simulations for combustion (to +30 °CA): the base grid is 1mm, while the fixed embeddings are lev 1 for piston skirt; lev 2 for head, liner, piston, injector zone; three sphere embedding around spark of lev 2, lev 3 and lev 4.
I'm using AMR too.
I'm igniting the mixture with two energy source of 40mj in a sphere with 0.0015m radius (while converge example case use 20 mJ in a sphere with a radius of 0.0005m) and with some radicals (O and H, even in a sphere radius of 0.0015m), in order to simulate the ionization due to the laser ignition.
I'll be glad if someone of you could give me some advice.
Thanks!

[arajendran]

Originally Posted by nik.rapons

Hi,
I'm working with Converge in order to simulate a PFI SI engine (with SAGE) fueled with a mixture of hydrogen and air with an equivalence ratio of 0.25, with both of turbulence models (LES and RANS).
I'm having problems with the LES (viscous one equation) simulation (either in single species diffusion than mixture - averaged diffusion), because I cannot ignite the mixture, while RANS case shows combustion.
The cases are the same, except for the turbulence models.
I'm simulating first both the intake (from -360 °CA) and compression phase (to -25 °CA), for LES and RANS cases, with a base grid of 4mm and some fixed embeddings: lev 3 around valves and piston skirt, lev 2 around tumble fixture, head, liner, piston and injector zone.
Then, with the two map files from -25 °CA, I run simulations for combustion (to +30 °CA): the base grid is 1mm, while the fixed embeddings are lev 1 for piston skirt; lev 2 for head, liner, piston, injector zone; three sphere embedding around spark of lev 2, lev 3 and lev 4.
I'm using AMR too.
I'm igniting the mixture with two energy source of 40mj in a sphere with 0.0015m radius (while converge example case use 20 mJ in a sphere with a radius of 0.0005m) and with some radicals (O and H, even in a sphere radius of 0.0015m), in order to simulate the ionization due to the laser ignition.
I'll be glad if someone of you could give me some advice.
Thanks!

Hi Nik,

Generally, we recommend going at least two levels finer for your LES case compared to your RANS simulation.
If you would like us to take a deeper look at this, please email CONVERGE support at support@convergecfd.com