defining pressure as inlet boundary

Numerical methods and mathematical models of Elmer
Post Reply
inspektor2211
Posts: 3
Joined: 25 May 2011, 16:16
Antispam: Yes

defining pressure as inlet boundary

Post by inspektor2211 »

helo
im new to Elmer and to CFD in general.
i would like to ask the forums assistance in a particular case im investigating.
i wish to analyse the flow of water through a small orifice, with the possibility of cavitation an particularlly the occurance of
what in english (i think) is called hydraulic flip or separation of the fluid from its confining walls, as it passes the orifice.
though experimental measurements i did obtain the flow and its velocity, so i have managed to define input velocities etc, no problem.
but if i define pressure in the, for example, inlet boundary, the solver apparently does not take this pressure into account.

in other posts i have read that using Gradp Discretization=true and div Discretization=true is the way to go about solving this problem is this correct?
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: defining pressure as inlet boundary

Post by raback »

Hi

You could also define the normal component of the stress tensor directly by "External Pressure" or "Pressure i", i=1,2,3. When viscous forces are small this quite accurately sets the pressure. Any bc where the normal velocity is not defined implicitely sets the external pressure thus to zero.

-Peter
inspektor2211
Posts: 3
Joined: 25 May 2011, 16:16
Antispam: Yes

Re: defining pressure as inlet boundary

Post by inspektor2211 »

thank you for your response

i tried your solution and it works partially. what i mean by this is that the solver takes pressure into account, but the results seem a bit odd.
Believing that elmer uses Pascal as a pressure unit and defining an inlet boundary condition of 3.0E5, elmer's results along the entire structure in analysis, and specially near the inlet is a negative pressure of around 1.0E5 Pa.

first why a negative pressure unit? i have tried using an negative input pressure but the result seems the same.

and second why is the pressure near the inlet so low, this does not happen when i define the inlet velocity? should i define an initial condition that also includes pressure, not just velocity?
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: defining pressure as inlet boundary

Post by raback »

Hi

Perhaps you could share the case. Note that with the "Pressure i" and "External Pressure" the standard discretization schemes should be ok.

-Peter
inspektor2211
Posts: 3
Joined: 25 May 2011, 16:16
Antispam: Yes

Re: defining pressure as inlet boundary

Post by inspektor2211 »

hello again

so i will share with the forum my case files.
in summary i wish to model the flow through a 7.5mm orifice with a wall thickness of about the same length.
me and my colleges conducted experiments to determine the discharge coefficient for different pressure states inside the reservoir,
and to do this we obtained flow and velocity values.
these velocities, i believe, can be used so set or verify the results of the numeric simulation, but we also believe that using the
inlet boundary pressure would be a must for a correct numerical simulation, specially because the pressure states inside the reservoir
are much more accurate than our measurements of flow and velocity.

the mesh is 2D because the orifice is circular and to reduce computation time
the mesh was created in salome 5.1.5 in a .UNV format
Attachments
test press8.rar
(1.85 MiB) Downloaded 438 times
MatteoR
Posts: 3
Joined: 09 Jan 2018, 12:42
Antispam: Yes

Re: defining pressure as inlet boundary

Post by MatteoR »

Hello,
I understand this post is dated 2011 and we are now in 2018..but
-Has anyone ever succeded in making a simple flow simulation (e.g. in a pipe) that has PRESSURE as inlet bc? (I mean the pressure inside the initial part of the pipe)
-What other bc are then needed as minimum to make the simulation work?
-Must the option "gradp" be turned on? Any other option to turn on?

Really hope someone can help, 'cause I can't find any clear info on this basilar topic.

Thank you,

Matteo
Post Reply