SPH Inlets/Outlets Conditions
Altair EmployeeHello everybody,
Is anyone well informed about SPH Inlets/Outlets Conditions? I´m a beginner and have little knowledge, here is my situation: I´m making a simulation of a pipe. The fluid is air. The SPH that are defined at the Inlet should flow through the pipe and be eliminated at the Outlet.
First problem is that only a part of the SPH are eliminated at the Outlet.
Second problem is that my calculation time is very long even for simple simulations.
Some data on the simulation:
My units are: kg, mm, ms, kN
Formula used to calculate particle mass: c^3x density
I selected Simple Cubic for SPH creation, so the Smoothing Length was set automatically.
But on the help section of the homepage of Altair there are several formulas for calculating mp and h.
Is it correct the way I´ve done it?
SPH-Property: qa=1e-20 and qb=1e-20
For the air I chose material law LAW6 and I entered the following values:
Air
# RHO_I
1.22E-9 0
# Knu
.015
# C0 C1 C2 C3
0 0 0 0
# Pmin Psh
0 1.0132E-4
# C4 C5 E0
.4 .4 2.533E-4
I changed the orientation of the surface segments in a way that their normal vectors point towards the interior of the domain.
I defined DIST-inlet = 2xh
Density function defined: time: 0, 1e30 ; density: 1, 1
For Fscale I entered my actual air density.
I defined the velocity function: Time: 0, 500 ; Velosity: 50, 50
DIST-outlet = 2xh
Pressure function: Time: 0, 1e30; Pressure: 1, 1 ; For Fscale I entered 0.005.
I set for SPH the „SPH Particle Reserve“
Does SPHGLO-control card has an impact on my results and calculation time?
i've also attached the files.
Thank you.
Answers
-
Hi,
The higher the Qa, Qb values you have the more viscous behavior you see. As you go long on this, it will be time expensive.
If you are assuming air as an incomperssible fluid then you have to think about the C1. This will be density times speed of sound velocity square
Regarding simulation time, I would first play with contact. If that is not helping then I will go with changing the order in property.
0 -
The mass of the particles should relate to the density of the material and to the size of the cubic net. And it seems to be fine with you. As Prakash mentioned for incompressible fluid consider C1 value,.
0 -
First of all thanks for your tips, gentlemen. Mr. Pagadala could you please specify what you mean by 'changing the order of property'. Another general question: Is there a way to create SPH elements directly on a surface instead of a volume?
0 -
Moka,
There is a filed in PROP/SPH with name the order. Try to change and check the results.
0 -
Hello,
I tried to redo example 22- ditching using sph by Radioss ( I used hypermesh and hypercrash). I do not know which kind of surface I should select for the part that I show in the attached picture.
How can I create the appropriate Surface? If I create a usual surface, I will need to assign material and properties to this surface. Although, there is no information about these issue in the example.
I would really appreciate it if you help me.
Best ,
sara
<?xml version="1.0" encoding="UTF-8"?>
0 -
Hi Sara,
For this example the water mesh is surrounded on five faces by Outlet SPH boundary frontier absorbing conditions. A control surface is placed at a distance equal to 2 x ho.
At this approximate distance a structure like this is created using with only a single element and from the contact surfaces panel a csurf is created.
<?xml version="1.0" encoding="UTF-8"?>
This created csurf is recalled in /SPH/INOUT option.
0 -
Altair Forum User said:
Hi Sara,
For this example the water mesh is surrounded on five faces by Outlet SPH boundary frontier absorbing conditions. A control surface is placed at a distance equal to 2 x ho.
At this approximate distance a structure like this is created using with only a single element and from the contact surfaces panel a csurf is created.
<?xml version="1.0" encoding="UTF-8"?>
This created csurf is recalled in /SPH/INOUT option.
Dear George,
First, I maid a cubic solide and then I used SPH mesh to create the water. Then I deleted the solid. This is how I created the water
So I do not understand clearly what you said. you mean now, I have to create the red box at 2h distance, and then consider it as a one element as you showed on the picture? and then recall it?0 -
Hi Sara,
Yes, a separate box has to be created at 2h0 distance and has to mesh it with a single element. Please go through the video below where this is shown.
PS: This video is in .swf format, you can open it in GOM player.
0 -
Dear George,
I Finally i modeled example -22( ditching using SPH). I got Error ID 495. I checked material , property compatibility but it does not work. What should I do to fix this?
Best Regards,
Sara
0 -
Hi Sara,
The corrected model file has been sent to you via mail.
0 -
Altair Forum User said:
Hi Sara,
The corrected model file has been sent to you via mail.
Dear George,
Thank you so much. I got it.
would you please explain how you fix it? I have 3 parts but, in your model you have 2 parts. how you keep the csurf at SPH inlets /outles conditions without having it in part section.
Best,
Sara
0 -
Hi,
The elements we created were only for extracting the contact surface and which was used to apply this boundary condition. So there is no need to assign any mat/prop for the same.
Like a face extraction was only done.
0 -
Altair Forum User said:
Hi,
The elements we created were only for extracting the contact surface and which was used to apply this boundary condition. So there is no need to assign any mat/prop for the same.
Like a face extraction was only done.
Dear George,
I am sorry what is the face extraction? when I import your model in HyperCrash there are two parts, while; my model shows three parts. I attached two pictures which you can see.
I did not assign any material and property . But when I checked my model I got Errors related to this part
I appreciate your help.
Best,
Sara
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
0 -
Hi Sara,
I checked the same file in HyperCrash but it shows only 2 components and it runs fine without any errors in RADIOSS.
<?xml version="1.0" encoding="UTF-8"?>
The same file is attached for your reference.
0
