[M3sh_Master-Ti]

Dear all,

I am wondering what is the physical definition of the contraction coefficient. I am currently modifying my simulation model for an injector and found the experimental result is much higher than the simulation in terms of SMD. The current assumption is that the injector orifice quality is much worse than expected, and I wish to tune the simulation result by modifying the coefficient Ca, however, can this be a sufficient reason to tune this coefficient?

Thank you

[ksrivast]

Hello Zhe,

Please refer to the concept of "Vena contracta" for an understanding of contraction coefficient. If you're using the BLOB injection drop distribution model, the injected parcel/droplet effective diameters are calculated as nozzle_dia*sqrt(C_a). See image below.

We would recommend you set your nozzle_dia equal to the nominal nozzle diameter of your injector nozzle if you lack actual/precise measurements. The contraction coefficient affects not only the injected parcel diameters, but also (for non-LISA injectors) injected pressure/velocity for a user-specified injected mass and duration. The Discharge coefficient model can be enabled to tune the value of C_d to match target injection pressures/velocities. The expected injection pressures/velocities can be ascertained through the Spray Rate Calculator tool in STUDIO (see STUDIO manual) without having to run the simulation which also writes out the values in spray_rate_inj<id>.out.

If the SMD results don't match with correct injection pressures, we recommend you take a look at the break-up model settings and evaporation behavior (if enabled) to improve your predictions. An improvement in grid resolution along the spray plume (fixed embedding + AMR) should also be looked into.

While you can always use other Injection drop distribution models like Uniform (specified SMD) to tune the results to match, this is not recommended as you have shifted to curve fitting your results rather than predicting them.

Hope this helps.

Sincerely,

[ksrivast]

Image : (removed)

[ksrivast]

Image :

cvgbat50.jpg