[laurentD]

Hi foamers,

i am trying to define a boundary condition, between my fluid and a solid, on the temperature. This BC needs a spatial dependent value of the convection coefficient h and a time-dependent value of the solid temperature.

To define the spatial dependence of the h coefficient i have used the groovyBC utility and i have succeed to make my job work with a constant value of Tp.

But for Tp, can i use timeVaryingMappedFixedValue, inside the groovyBC?
Because i don't see another way to do the time-dependence.

Thank you evrybody.
Laurent

[gschaider]

Originally Posted by laurentD

Hi foamers,

i am trying to define a boundary condition, between my fluid and a solid, on the temperature. This BC needs a spatial dependent value of the convection coefficient h and a time-dependent value of the solid temperature.

To define the spatial dependence of the h coefficient i have used the groovyBC utility and i have succeed to make my job work with a constant value of Tp.

But for Tp, can i use timeVaryingMappedFixedValue, inside the groovyBC?
Because i don't see another way to do the time-dependence.

Thank you evrybody.
Laurent

No. You can't use one boundary condition inside of another. But you can always use the time() function to get the current time. If your time dependence is from a data file you can use a lookup-table in groovyBC. For an example see Examples/groovyBC/wobbler

[laurentD]

Thank you for this answer.

Can i imagine use the timeVaryingMappedFixedValue.C in the groovyBC block ?
I know that it implies to make available the various dependencies of it but the first step is : is it possible ?

Maybe by an include, pointing on the timeVaryingMappedFixedValue.C, previously copied in the right directory...
I feel that swak4foam is much more powerful that it seems.

Thank you very much for your help, i couldn't imagine a better person to discuss about this topic.

Laurent

[laurentD]

To be as clear as possible, i would like to plug T values directly on points for each time, like with timeVaryingMappedFixedValue, but the process of plugging on points is a bit too complex to be written in the groovyBC block, isn't it ?

It is why i would like to use the existing C++ file to this step.

[gschaider]

Originally Posted by laurentD

To be as clear as possible, i would like to plug T values directly on points for each time, like with timeVaryingMappedFixedValue, but the process of plugging on points is a bit too complex to be written in the groovyBC block, isn't it ?

It is why i would like to use the existing C++ file to this step.

I don't quite know what you mean with "plugging on points"? You mean: setting the T-value depending on the time and the position? That is possible.

[laurentD]

That is possible

Yes i think so...

My current idea is to re-use the data files used by 'timeVaryingMappedFixedValue' (with a file with temperatures for all points of a patch for each time-step, associated with another file which contains the global coordinates of these points) in a groovyBC process using timelines and lookuptables.

Maybe it will need a light adaptation to make them compatible with groovyBC but curently i think i can do something good.

I have succeed to use groovyBC with lookuptables this summer but i have never used the 'timelines' utility. So i will begin my work by an understanding of the way of use of it.

I will warn you about my adventures.

Thanks a lot.
Laurent

[laurentD]

Does a complete documentation of the groovyBC exist ?

[laurentD]

To achieve my objective, i have to define a value for the temperature for each time and for each point.
How can i do it ?

Because the 'lookuptable' utility doesn't want file with ((x,y,z) T) for example.

So can i use the numbering of the points ?
Or another method that anybody knows ?

Thanks.
Laurent

[gschaider]

Originally Posted by laurentD

I have succeed to use groovyBC with lookuptables this summer but i have never used the 'timelines' utility. So i will begin my work by an understanding of the way of use of it.

"timeline()" is basically a shorthand for "lookup1D(time())"

Originally Posted by laurentD

Does a complete documentation of the groovyBC exist ?

There is a complete documentation of the expressions (which is the most important part) in the Documentation-folder (the incomplete reference guide). All other parts (description of function objects etc) will get written when I have the time

Originally Posted by laurentD

To achieve my objective, i have to define a value for the temperature for each time and for each point.
How can i do it ?

Because the 'lookuptable' utility doesn't want file with ((x,y,z) T) for example.

So can i use the numbering of the points ?
Or another method that anybody knows ?

The maximum I can offer ist the 2D-lookuptables. Look for lookuptables2D in the reference guide (you seem to need 4 dimensions)

Using point numbering is not a good idea as that would break with a change of the mesh (or even a simple renumbering. Or decomposition). You can get the face number with "id()" but I don't think this is practical

[laurentD]

Originally Posted by gschaider

"timeline()" is basically a shorthand for "lookup1D(time())"

Ok i understand.

Originally Posted by gschaider

There is a complete documentation of the expressions (which is the most important part) in the Documentation-folder (the incomplete reference guide). All other parts (description of function objects etc) will get written when I have the time

Hope you will have the time because your tool is very powerful but need a documentation to spread in the OF community. It is not a critic, i know time is expensive.

Originally Posted by gschaider

The maximum I can offer ist the 2D-lookuptables. Look for lookuptables2D in the reference guide (you seem to need 4 dimensions)

Yes i have seen that. Maybe a part of the soluce.

Originally Posted by gschaider

Using point numbering is not a good idea as that would break with a change of the mesh (or even a simple renumbering. Or decomposition). You can get the face number with "id()" but I don't think this is practical

I am thinking about a special numbering based on the coordinates x, y and z and in a scalar-form. I will inform you if i have results.

For the moment, two questions :
- Can i ask to a valueExpression to use an integer part of a variable ? (int() doesn't work)
- I am trying to use pts() but i have an error (unexpected TOKEN points). What do i have to do to use it ? Because my input datas used to define the BC are defined on points...

Have a good day,
Laurent

[laurentD]

Concerning the integer part, the fonctionality i have to use is floor(). But it needs to include math.h.
Can i do that ? I mean inside groovyBC.

[laurentD]

Concerning the integer part, the fonctionality i have to use is floor(). But it needs to include math.h.
Can i do that ? I mean inside groovyBC.

[laurentD]

Concretely, i have to take into account the time but in seconds, so the integer part of the current time. How can i do that ?
Any ideas are welcome.
L.

[gschaider]

Originally Posted by laurentD

Concretely, i have to take into account the time but in seconds, so the integer part of the current time. How can i do that ?
Any ideas are welcome.
L.

There currently is neither ceil nor floor implemented in swak4Foam as 'I rarely see the need for it. You can work around it using the modulo-operator % (note that a%b maps to [-b/2,b/2]). Getting the interger part of time() should work with something like "time()-((time()-0.5)%1+0.5)" (but check that before really using it)

[laurentD]

Thank you very much, i will take a look on it.

By my side i was thinking about a kind of tabular with (1 1) (2 2).
By this way if i define the second value depending of the first by B(A), i can expect an integer as a result.

But i need a constant result (and not a ramp) between two bounds and i don't really trust on the outOfBounds utility.
For example when i use "outOfBounds toto;"
it seems to work...

L.