I think there is a mistake in the implementation of the Mei lift force (and the paper of Kurose and Komori, Eq. 3.7), when I compare the function to the original data of McLaughlin 1991. Unfortunately, in the original paper of Mei 1992, a left parenthesis seems to be missing in the corresponding Eq 10, but it is there in Mei and Klausner 1994 (Eq 17), as well as in Loth and Dorgan 2009 (Eq 32).
The line in question is
https://github.com/CFDEMproject/CFDEMcoupling-PUBLIC/blob/d03549ba886fa6...
J_star = 0.3
* ( 1.0 + tanh( 2.5 * log10 ( epsilon + 0.191 ) ) )
* ( 0.667 + tanh( 6.0 * ( epsilon - 0.32 ) ) );
while it should be
J_star = 0.3
* ( 1.0 + tanh( 2.5 * ( log10 ( epsilon ) + 0.191 ) ) )
* ( 0.667 + tanh( 6.0 * ( epsilon - 0.32 ) ) );
Daniel Queteschiner | Thu, 12/07/2017 - 10:14
?
Can anyone confirm or disconfirm this suspicion?
TimMJN | Fri, 12/08/2017 - 09:35
Confusing
Dear Daniel,
There indeed is some confusion about the form of this equation. As you said, Loth and Dorgan (2009) presented the relevant part as:
log10(sqrt(omega*/Rep))+0.191
While it is implemented as:
log10(sqrt(omega*/Rep)+0.191)
However, to add some more confusion, the very same Loth (2008) presented it as:
log10(sqrt(omega*/Rep + 0.191))
(BTW the same goes for the other tanh term, which is presented as tanh(6*sqrt(omega*/Rep - 1.91)) instead of tanh(6*sqrt(omega*/Rep)-1.91))..)
So let's have a look where Loth got his information. In the 2008 paper he refers to a note by Mei, Klausner and Lawrence (1994) which is not even on the subject of lift forces. But let's assume he meant the paper of Mei and Klausner from 1994. Here, indeed, the term is given as
log10(sqrt(omega*/Rep))+0.191
In the 2009 paper Loth and Dorgan refer the the original paper of Mei (1992). Here a right parenthesis is missing, but indeed it appears that the form above is the correct one.
Based on all of this I agree with you. Both the implementation and paper of Loth from 2008 are wrong and the correction you proposed should be implemented.
I do have some other confusion about this lift model. It appears that the particle spin rate is not taken into account, only the vorticity of the fluid field. I wonder if this is implemented in some other version of the code, or if I should try my hand on implementing it myself.
Kind regards,
Tim
Eindhoven University of Technology