[steve42]

Hello,

I am doing Simulations of a ICE Engine with Hydrogen-PFI.
I am having a problem with using the G-Equation Combustion Model in Converge 3.1.9. To be more exact, I am having a problem with the Source model.

My Settings are:
Source: Passive, G_EQN_TRANSPORT
Limit Passive: 0.00035
Value: 1e+18 J/(m^3 s)

(Taken from a Converge Example file.)

Ignition timing is set from 709° to 715°, so a duration of 6°CA for Ignition should be more than enough, as I used a Duration of 1°CA before. (CA=Crank Angle).

These Settings did work before I switched Turbulence models from RNG-k-eps to k-w-SST.

Maybe anyone has an Idea as I am running out of ideas to fix it. I do not even know where else to look in the models.
Best regards,
Bjoern

[angela.wu]

Hi Bjoern,

Please have a look at chapter 16 - Ignition approach for G-equation. What is the grid size where your source is located? With G_EQN_TRANSPORT, you should set a max passive amount that is typically 2x your local grid size. Another way to source is to add energy source instead of G that is large enough to exceed combustion temperature offset.

Best,

Angela

[steve42]

Hello Angela,

thank you for your kind reply. I reconfigured everything now, following the manual again. I will start the Case soon.
As you said, max. Value should be double the local grid size. I am using 0.004m Grid, but during the Sparktiming I am using embedding with level 6 and 5. So should I set it to 0.008m, as it is double the usual grid size, or should I set it to double the value of my embedding grid size (respectively 0.000625 or 0.000125)?

I already tried the energy ignition option, but my supervisor is favoring the direct sourcing of G. But if I can not fix this I will try to convince him otherwise

Thank you and best regards,
Bjoern

[angela.wu]

Hi Bjoern,

I would set it to the grid size with the level 6 embedding, and use double that value to limit G_EQN_TRANSPORT. So if it's a 4 mm base grid, 2^6 embedding around the source is 0.0625 mm, and you should try a G_EQN_TRANSPORT of 1.25e-4. However, looking at the value you used (3.25e-4), it is already larger than the recommended value. Have you tried larger passive values?

I have noticed that when switching turbulence model, it affects the G-equation combustion modeling. Is there a reason why you switched from RNG K-E to k-w sst?

Best,

Angela

[steve42]

Hello Angela,

thank you for your kind reply. I now switched to an energy induced ignition, as it already works. I tried 0.00025, but stil got no ignition. I also tried 0.005 once, and it did not work.

I switched turbulence models as the thought process behind this is the following: H2 burns drastically more near to the walls, so to catch the wall-fluid interactions better, I switched to k-w-SST. Also other people in my institute are already using this turbulence model, so now we can better exchange data and changes between each other.

Best regards,
Bjoern