[jan]

Could you provide more information on your problem? Which example do you consider, how does your config.mk look like, etc?

My first guess would be that you’re only setting one cuboid in your simulation setup.

[jan]

Dear Evan,

due to the given output I assume that the lattice velocity () becomes too large. Try a smaller velocity (1/10 or even less just for testing). If the simulation is stable with the smaller velocity, then you probably need to change your discretization (Δt and Δx) in order to make the simulation work with the original velocity.

Also, do you still apply the gravitational force (it’s called external acceleration in the example)? This could cause an erroneous particle velocity.

Best,
Jan

[jan]

Dear Evan,

if you want to set the position of the particle, you can access the corresponding field directly. The following should work:

// get particle via id (starts from 0)
Particle<T, PARTICLETYPE> particle = particleSystem.get(particleID);
// set modified position
// you can use a similar method "getField" to get the content of the field
particle.template setField<GENERAL,POSITION>( modifiedPosition );

However, in my opinion it would be best to change the particle’s position before the coupling from the particle to the fluid via couple_particles_to_lattice. Therefore, you might have to separate the execution of the particleManager or even call the specified steps without it.

Best,
Jan

• This reply was modified 2 years ago by jan.

[jan]

Hello,

If you check, for example, the hlbmDynamics3D.hh (e.g., this), you can see an old model. However, this model most likely does not work out of the box. Additionally, it should only work for spheres and straight walls (it’s Hertz-Mindlin-based).

Cheers,
Jan

[jan]

Hello Julius,

Unfortunately, the functor you are looking for is indeed not implemented at the moment. Currently, I see two ways of handling this:

1. Porting SuperAverage3D and other necessary functors into 2D, as you already mentioned. For this, it should be sufficient to copy the 3d version and remove the last dimension, you are right.
2. Trying the SuperLocalAverage2D (see doxygen). This only limits the checked cells even further with a circle. However, a big radius will include your whole indicator and might work like a SuperAverage2D.

If you need any assistance, please let me now.

Cheers,
Jan

[jan]

Hello Julius,

Thank you very much for reporting this mistake in the code. I have included the fix in our development version. The next release will include this bug fix.

Regards,
Jan

[jan]

Hello,

Unfortunately, I cannot reproduce your problem. I downloaded the version 1.3-1 and unpacked it. Afterward, I used the ‘cd’ command to go to examples/laminar/bstep2d and then used ‘make’.
Now, bstep2d is compiling in precompiled mode. Could you maybe tell me more about the system and programs you are using? (OS, compiler + version) Additionally, did you change something before you tried to compile the example?

Usually, ‘there is nothing to be done for “all”’ indicates that there is already an executable and that this executable got created after the last source code changes. Is that maybe the case here?

Cheers,
Jan

[jan]

Hello Shuainan Zhai,

The problem is that the compiler cannot find SuperLatticeEpsilon2D, since it does not exist. Please try to use SuperLatticeGuoZhaoEpsilon2D instead. For more information about this functor see https://www.openlb.net/DoxyGen/html/db/da5/classolb_1_1SuperLatticeGuoZhaoEpsilon2D.html.

Cheers,
Jan

[jan]

Hello Pedro

have a look at chapter 2.1 “Lesson 1: Getting Started – Sketch of Application”, there you will find a description of how to start a program using the example bstep2d. You will immediately see the output on the terminal. Further information on input and output is given in chapter 8 of the User Guide and on visualization in chapter 9.

The User Guide and the provided examples are already a great way to get familiar with OpenLB. It’s also worth taking a look at the Spring School. There is a lab session where an introduction to OpenLB is given. Experienced tutors are also present to provide support and advice.

Let me know if you have any further quetsions.

Cheers,
Jan

[Author]

Posts