[Avilash]

Hi there,

I have been trying to do motoring validation in Converge using sector geometry for compression ratio of 15, but I am getting 5 bar higher peak pressure in simulation compared to experimental.

I have attached the image comparing simulation and experimental result.

Can you tell me any possible reason for this difference and how to rectify it?

Thank you very much.

[tmburton]

Avilash,

Are you confident that the experimental data was obtained at CR=15? Have you used the compression ratio calculator tool inside Studio to verify that the simulation achieves CR=15? I would also verify that trapped pressure in simulation matches the experimental data (it's a little hard to tell on the scale in your image), and that you aren't using adiabatic boundary conditions in your simulation.

Best regards,

Tristan

[Avilash]

Hello Tristan,

Thanks for your reply. I am confident that the experimental data was obtained at CR=15, and I have also verified compression ratio using tool for that (I have made the geometry using 'Make surface' tool for CR=15). I have initialized the trapped pressure using the pressure at IVC(closed cycle simulation) from experimental data. I am using isothermal boundary conditions(wall temperatures=300k).

I also want to mention that I have been trying different options, in one of the cases I reduced the compression ratio to 13.2 (just to see the effect) and it matched with the experimental data. I have attached a snap of the same.

[Blanco]

Originally Posted by Avilash

Hello Tristan,

Thanks for your reply. I am confident that the experimental data was obtained at CR=15, and I have also verified compression ratio using tool for that (I have made the geometry using 'Make surface' tool for CR=15). I have initialized the trapped pressure using the pressure at IVC(closed cycle simulation) from experimental data. I am using isothermal boundary conditions(wall temperatures=300k).

I also want to mention that I have been trying different options, in one of the cases I reduced the compression ratio to 13.2 (just to see the effect) and it matched with the experimental data. I have attached a snap of the same.

Hi,
how did you verify that the effective CR was 15? Have you measured it on the real engine?
Moreover, even if you're considering the experimental pressure@IVC this typically need some adjustment since a reference pressure is typically needed to convert measured pressure into an absolute pressure. How did you corrected the experimental pressure? Moreover, how did you choose the in-cylinder temperature?

[Avilash]

Originally Posted by Blanco

Hi,
how did you verify that the effective CR was 15? Have you measured it on the real engine?
Moreover, even if you're considering the experimental pressure@IVC this typically need some adjustment since a reference pressure is typically needed to convert measured pressure into an absolute pressure. How did you corrected the experimental pressure? Moreover, how did you choose the in-cylinder temperature?

Blanco,
I donot know the effective CR of the engine but geometric CR is 15 and it was verified by calculating volume of head and piston bowl. On other hand thanks for pointing out about reference pressure for measured pressure adjustment(I will be taking 1atm@BDC as reference pressure for NA engine). I have to correct it and see the change. By in-cylinder temperature I guess you mean wall temperatures, I have taken all wall temperatures as 300k because the motoring was done at cold-condition. Even changeing it to 330k had little effect on peak pressure.

[Blanco]

Originally Posted by Avilash

Blanco,
I donot know the effective CR of the engine but geometric CR is 15 and it was verified by calculating volume of head and piston bowl. On other hand thanks for pointing out about reference pressure for measured pressure adjustment(I will be taking 1atm@BDC as reference pressure for NA engine). I have to correct it and see the change. By in-cylinder temperature I guess you mean wall temperatures, I have taken all wall temperatures as 300k because the motoring was done at cold-condition. Even changeing it to 330k had little effect on peak pressure.

Ok, I asked about effective CR because typically real operating engine have CR which could be quite different than nominal CR (this is caused by gasket dimension for example, in addition to tolerances, etc.). In my experience the "real" CR could be lower/higher than geometric CR by something like 0.3-0.5 max.

As far as the temperatures are concerned, I was thinking to gas temperatures because the (p,T) you're using@IVC will set the in-cylinder mass in the compression stroke. How did you choose the gas temperature? Did you set the temperature to match an intake manifold airflow measurement?

The wall temperatures could affect the heat transfer during the compression, but as you mentioned since the engine was in motoring conditions I expect something near 300 K to be quite reasonable.

I would like to add that the correction to pressure signal should be as precise as possible, as little differences in p@IVC will be amplified by the CR during the compressoin stroke.

Hope this could help!

Best

[SamWijey]

Hello,

Have you considered flow in to the crevice? Does your surface file has crevice drawn? If not, you can use crevice.in to account for flow pushed in to crevice during compression. This will reduce your effective CR and will bring the pressure closer to that measured.

Thanks,

[Avilash]

Originally Posted by SamWijey

Hello,

Have you considered flow in to the crevice? Does your surface file has crevice drawn? If not, you can use crevice.in to account for flow pushed in to crevice during compression. This will reduce your effective CR and will bring the pressure closer to that measured.

Thanks,

Hello Sameera,

Yes, I have crevice drawn on my surface, but I am not using crevice model. So, if I want to use crevice model should my surface have crevice drawn in it, or the crevice model will take care of it? I have added the crevice in bowl profile itself(please see attachment).

[SamWijey]

Drawing a crevice in the geometry can account of additional volume but cannot account for any piston blow-by. If you want to use crevice.in, you should remove the crevice from the geometry. Some questions to consider:

1. You should use real gas equation of state
2. Do you have measured intake mass flow rates? If so, you should run the full engine case with valve motion to see if you can predict the same intake mass flow rate. We recommend the error to be below 2% of the measured mass flow.
3. If there's no piston blow-by in the experimental engine, if you have matched the intake mass flow rate accurately, using real gas eos, and using the correct intake flow gas composition and your motoring pressure is still bad: Then that points to issues with your geometry.

Thanks,

Sameera