[HuifengGong]

Hi,

I am confused about the drop size distribution. I could define the drop size distribution, but what kind of sizes of drops are injected during simulation? For example, I use the Rosin-Rammler distribution, and give the SMD and the RR parameter to determine the distribution, then, during calculation, what the exact minimum/maximum drop size of the spray is? and how many different kinds of drop sizes (or size bins) will be used between the minimum and maximum sizes?

Thank you!

[jlemoine]

Huifeng,

The Rosin-Rammler distribution has two inputs smd_dist and q_rr. Once q_rr and smd_dist is defined in spray.in the injected drop radius is determined from:
r = gamma_function*(1-(1/q_rr))*(0.5*smd_dist)*zeta
with zeta = r / r_bar.

Please refer to CONVERGE manual, Chapter 12.1.3 for more information. Thanks,

[HuifengGong]

Originally Posted by jlemoine

Huifeng,

The Rosin-Rammler distribution has two inputs smd_dist and q_rr. Once q_rr and smd_dist is defined in spray.in the injected drop radius is determined from:
r = gamma_function*(1-(1/q_rr))*(0.5*smd_dist)*zeta
with zeta = r / r_bar.

Please refer to CONVERGE manual, Chapter 12.1.3 for more information. Thanks,

Thank you very much, but I just wonder the actual size of drops that are injected during calculation. I know the distribution function, which is a continuous one, but I think when doing the calculation, the values of drop size should be discrete and the number of values should be limited. For the function of Rosin-Rammler distribution, the values of zeta could be selected from 0 to zeta_max, continuously, but I think the calculation won't use the values continously, so, what is the minimum and maximum zeta? and how many individual values will be used between minimum and maximum zeta? how these values are determined?

[jlemoine]

Huifeng,

This information is not available in the manual, let me see if I can retrieve this information for you.

[jlemoine]

Here's the information that you are looking for:
bins = 100
zeta_min = 0
zeta_max = ln(1000)^(1/q_rr)
I hope this helps,

[HuifengGong]

Originally Posted by jlemoine

Here's the information that you are looking for:
bins = 100
zeta_min = 0
zeta_max = ln(1000)^(1/q_rr)
I hope this helps,

Thank you very much! This really helps. It seems that there are always 100 different kinds of sizes of drops injected, and it always covers (1-1/1000) of the drop size range. Can I change the bins and the zeta_max by myself? I think in some cases, 100 bins are too many, and zeta_max could be smaller to cover less. This will be beneficial for computing cost.

Further questions:

a. The function is about the drop mass over radius?

b. The 100 size bins are distributed evenly between the min and max, or in other forms like log / ln?

c. During calculation, will the drops be injected at each timestep? What is the mass of drops that are injected for each drop size? for the nth size bin, the injected mass is the sum of drops between radius r(n-1) to r(n)?

d. How many parcels will be injected for each drop size at each timestep? the number of parcels are the same for all sizes?

Many questions... and many thanks to you!

[jlemoine]

Can I change the bins and the zeta_max by myself? No it is harcdoded and it cannot be changed through UDF

Further questions:

a. The function is about the drop mass over radius? The function is the cumulative probability function over non-dimensionalized radius by R_max. The drop mass is not part of the function.

b. The 100 size bins are distributed evenly between the min and max, or in other forms like log / ln? The 100 bins are distributed evenly between zeta_min and zeta_max.

c. During calculation, will the drops be injected at each timestep? What is the mass of drops that are injected for each drop size? for the nth size bin, the injected mass is the sum of drops between radius r(n-1) to r(n)? Parcels represent a group of identical drops (i.e same radius, velocity, temperature...).The parcels or drops will be injected every timestep. Every parcel initially has the the same mass, the drop radius is determined by the cumulative probability function, then the number of drops in a parcel is computed based on the radius and the parcel mass. The injected mass is treated as the mass between r(1) to r(n), and each bin will then gives a fraction of that injected mass.

d. How many parcels will be injected for each drop size at each timestep? the number of parcels are the same for all sizes?
The total number of parcels injected at each time is determined by how much mass is being injected. We use random numbers to decide the number of parcels with a fixed radius to satisfy statistical properties, and this number needs to comply with the user-specified total number of parcels.