[LaurentDelaon]

Hi Adrian

I will read and try to do it.

thank’s for your prompt reply.

regards.
Laurent.

[LaurentDelaon]

ok I understand.

I try to do it.

thank’s for your prompt reply.

regards
Laurent.

[LaurentDelaon]

HI Adrian,

I precise : not working with olb1.4 …

What I can do ?

regard’s.
Laurent.

[LaurentDelaon]

I precise that is it compile with olb1.2 and not with olb1.4…
each cases under ubuntu 18.04.

Laurent.

[LaurentDelaon]

Hi Everybody,

Today with the update of Ubuntu 18.04lts
example/laminar/cylinder2d” with MPI will works fine!

DO entirely compile olb1.3 and cylinder2d.
So I have a doubt about the code correction…
(because it’s the same olb1.3 release that last time for me…)

Laurent.

[LaurentDelaon]

Hi Mathias,

you don’t understand what I want to say…
how I can send you a picture?
can I use your email ?

Laurent.

[LaurentDelaon]

HI

how did we dorealise the fix correction ?

regard’s.
Laurent.

[LaurentDelaon]

Hi Mathias,
I have doing like this.
It is not an instability!

My simulation didn’t realise the time changement of state : when we open the valve we obtain a step changement of pressure. This is not realistic value. When opening the valve pressure seem change immediatly…
We measure it on a real system : the time changement is about 0.15s this is not immediatly change.

I use BGK to calculate but maybe I do use MRT to avoid this phenomen ?

regard’s
Laurent.

[LaurentDelaon]

hi Mark

thank for your reply
I test this works fine
And I have also found informations in doc1.2 unfortunatly I read the 1.1…
regard’s
Laurent.

[LaurentDelaon]

Hi Robin,
I have progress in my case studies based on cylinder2d.
I’m arrive to measure pressures;
But I want to know how measure Static Pressure ? How obtain it in 2d ?
Is the pressure given by GetPressure is the same that the total pressure ?

Regards.
Laurentotal

[LaurentDelaon]

Hi Marcus,

is it possible to test pre-release 1.3 with GPU for a K40 ?
I only use cylinder2d base example.

Laurent.

[LaurentDelaon]

Thank’s Marcus.

[LaurentDelaon]

HI Markus,
ok I waiting for it!
I have imagined that mpi can benefit of th K40…Seem not true.
A question : with mpi I benefit better way to use my processors.
I have 31cores but this is work better with -np 15(wich give the simulation quicker).
How explain that ?
machine use a XEON E5-2620v4@2.1Ghz with 31 cores

[LaurentDelaon]

thank’s for your reply.
I have really progress. We have pb to the realistic of the fluid (air) simulation comportement.
I have starting using the cylinder2d modify example. We have transform cylinder in thin vertical cuboid with theta of 45°.

Code:

AnalyticalConst2D<T,T> rho(converter.getLatticeDensityFromPhysPressure(200*frac[0]);

Work fine with inlet pressure 1 and outlet pressure 0.98 but not the delta_P is inlet 200pa outlet 0pa…
I have empty y range error adjusting ….
Trying augmenting N 20->100 nothing better.
No simulation result.

My first objective is to compare simulation with the real system (pressure inlet and also the velocity outlet), this is the origin of my question…

on the real system we have pin:100500 pout100400 vout:1.8m/s.
cylinder L=280 diameter 25.4mm valve diameter 25.0 thickness 4mm in the middle , based on cylinder2d but not same dimensions!

Laurent..

[LaurentDelaon]

hi thank’s for your response.
I have look at documentation but I have think that I can miss a specific move function.
I not expert in C++. I am just using the code to try to simulate air flow.
My objectives are not to rewrite your code but to use it as it…

Regards.

[Author]

Posts