Collection Efficiency in Fluent?

August 9, 2004, 18:28

Has anyone done any work to find collection efficiency for icing applications using Fluent? I appreciate any info.

Thanks, Jason

August 10, 2004, 09:16

solve the flowfield in fluent and then use a NASA code called LEWICE or LEWI3DGR

August 26, 2004, 23:30

I think you can use udf

aminrbspace · July 3, 2012, 17:23

Hi,
I have many document about computing collection efficiency with fluent, but now I deal with it and I couldn't take correct answer.

Quote:

Originally Posted by Jason
;114252

Has anyone done any work to find collection efficiency for icing applications using Fluent? I appreciate any info.

Thanks, Jason

Marta · November 23, 2012, 07:14

Hi all.
Hav eyou succeded in implementing a model/simulation using just Ansys for icing accretion, in the end?
If so, how were you able to do so?

Thanks a lot,

Marta

miss chemical · December 28, 2014, 08:12

Dear aminrbspace: can you provided us with your documents about computing collection efficiency with fluent . I trying to compute the collection efficiency in cyclone separator with fluent but I don't Know how to do that?
Thank you so much

Manathan · May 4, 2015, 11:34

It's impressive how there is no answer to these questions on internet. Couldn't find either a proper answer on how to compute the local collection efficiency through the "DPM accretion rate" of fluent. It seems that the notion of flux : kg/m2.s at injection is unclear, especially for 2D axisymmetric since the accretion rate is given in kg/m2.s while the face magnitude used is supposed to be in "m" since it is 2D.

`e` · May 4, 2015, 20:41

In 2-D cases, the third dimension is often set to 1 m which means if you have a mass flux of 23 kg/m^2.s then your mass flux is 23 kg/m.s.

What do you mean by "local collection efficiency through the 'DPM accretion rate' of fluent"?

The simplest definition of the collection efficiency is a trivial calculation of the ratio between the number of particles deposited and the number of particles injected:

$\eta = \dfrac{N_{deposited}}{N_{injected}}$

Each particle stream has a mass flow rate, say m1 [kg/s], comprised of parcels (unsteady tracking) each with a number of particles. If you inject this stream at a cell face with normal area, A1, then your mass flux of particles at injection is m1/A1 [kg/m^2.s]. Consider a deposition mass flow rate of m2 within an area A2 yielding a mass flux of m2/A2 [kg/m^2.s]. However, calculating the ratio of the deposited mass flux and injected mass flux is not equivalent to the above equation where the face areas of the injection (A1) and deposition site (A2) are unequal.

Determine the mass flow rate of deposited particles, m2, and then:

$\eta = \dfrac{m_2}{m_1}$

Practically, you would have a number of injected particles and deposition sites to accurately determine your collection efficiency.

Manathan · May 5, 2015, 04:18

yes I agree with you, this was clear for me but in my situation I cannot inject on each cell, I have to inject along a line using group type with much more injected points than cells.

So the only challenge is knowing what should be this injected area for a specific stream (since for each stream I should have a specific area). Since the macro DPM_EROSION gives me the kg/m2-s at each cell along my wall, its quite easy to divive by the constant flow rate (kg/m2-s) injected constantly along my line using a linear distribution of the injected mass flow rate.

At first for 2D axisymmetric I was using : area_injected = 4*pi*r*dr

with r being the position injected; dr half the distance between each point injected. But this means that I'm considering the area around 2*PI. Apparently, in DPM_EROSION, fluent is not using the definition of area around 2*PI but for 1readian only, so the (i think) correct area_injected per stream would be : 2*r*dr

Then If I fix lets say my; constant along the line, injected flow rate to 1kg/m2-s, I would know the mass flow rate distribution I need to put in FLUENT. (i'm using linear distribution since the area is evolving linearly with respect to r (dr is constant though).

August 9, 2004, 18:28	Collection Efficiency in Fluent?	#1
Jason Guest Posts: n/a	Has anyone done any work to find collection efficiency for icing applications using Fluent? I appreciate any info. Thanks, Jason

August 10, 2004, 09:16	Re: Collection Efficiency in Fluent?	#2
delic Guest Posts: n/a	solve the flowfield in fluent and then use a NASA code called LEWICE or LEWI3DGR

August 26, 2004, 23:30	Re: Collection Efficiency in Fluent?	#3
Doris Guest Posts: n/a	I think you can use udf

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Two questions on Fluent UDF	Steven	Fluent UDF and Scheme Programming	7	March 23, 2018 04:22
How to open Icem mesh in Ansys Fluent?	emmkell	FLUENT	27	February 6, 2018 04:34
fluent parallel problem in win7 x64 system	dunga82	FLUENT	8	April 19, 2012 21:23
Fluent 6.3 32bit vs Fluent 12.0 64bit	ibex7	FLUENT	7	April 18, 2011 03:44
Collection efficiency	Jen	FLUENT	0	November 16, 2005 11:16

November 23, 2012, 07:14		#5
Marta Member Marta Lazzarin Join Date: Jun 2009 Location: Italy Posts: 71 Rep Power: 17	Hi all. Hav eyou succeded in implementing a model/simulation using just Ansys for icing accretion, in the end? If so, how were you able to do so? Thanks a lot, Marta

December 28, 2014, 08:12		#6
miss chemical New Member nuha Join Date: Jun 2014 Posts: 9 Rep Power: 12	Dear aminrbspace: can you provided us with your documents about computing collection efficiency with fluent . I trying to compute the collection efficiency in cyclone separator with fluent but I don't Know how to do that? Thank you so much

May 4, 2015, 11:34		#7
Manathan New Member Join Date: Apr 2015 Posts: 28 Rep Power: 11	It's impressive how there is no answer to these questions on internet. Couldn't find either a proper answer on how to compute the local collection efficiency through the "DPM accretion rate" of fluent. It seems that the notion of flux : kg/m2.s at injection is unclear, especially for 2D axisymmetric since the accretion rate is given in kg/m2.s while the face magnitude used is supposed to be in "m" since it is 2D.

May 4, 2015, 20:41		#8
`e` Senior Member Join Date: Mar 2015 Posts: 892 Rep Power: 18	In 2-D cases, the third dimension is often set to 1 m which means if you have a mass flux of 23 kg/m^2.s then your mass flux is 23 kg/m.s. What do you mean by "local collection efficiency through the 'DPM accretion rate' of fluent"? The simplest definition of the collection efficiency is a trivial calculation of the ratio between the number of particles deposited and the number of particles injected: $\eta = \dfrac{N_{deposited}}{N_{injected}}$ Each particle stream has a mass flow rate, say m1 [kg/s], comprised of parcels (unsteady tracking) each with a number of particles. If you inject this stream at a cell face with normal area, A1, then your mass flux of particles at injection is m1/A1 [kg/m^2.s]. Consider a deposition mass flow rate of m2 within an area A2 yielding a mass flux of m2/A2 [kg/m^2.s]. However, calculating the ratio of the deposited mass flux and injected mass flux is not equivalent to the above equation where the face areas of the injection (A1) and deposition site (A2) are unequal. Determine the mass flow rate of deposited particles, m2, and then: $\eta = \dfrac{m_2}{m_1}$ Practically, you would have a number of injected particles and deposition sites to accurately determine your collection efficiency.

May 5, 2015, 04:18		#9
Manathan New Member Join Date: Apr 2015 Posts: 28 Rep Power: 11	yes I agree with you, this was clear for me but in my situation I cannot inject on each cell, I have to inject along a line using group type with much more injected points than cells. So the only challenge is knowing what should be this injected area for a specific stream (since for each stream I should have a specific area). Since the macro DPM_EROSION gives me the kg/m2-s at each cell along my wall, its quite easy to divive by the constant flow rate (kg/m2-s) injected constantly along my line using a linear distribution of the injected mass flow rate. At first for 2D axisymmetric I was using : area_injected = 4pirdr with r being the position injected; dr half the distance between each point injected. But this means that I'm considering the area around 2PI. Apparently, in DPM_EROSION, fluent is not using the definition of area around 2PI but for 1readian only, so the (i think) correct area_injected per stream would be : 2r*dr Then If I fix lets say my; constant along the line, injected flow rate to 1kg/m2-s, I would know the mass flow rate distribution I need to put in FLUENT. (i'm using linear distribution since the area is evolving linearly with respect to r (dr is constant though).