# [Getdp] Problem with induction heat geometry

Matti Pellikka matti.pellikka at tut.fi
Mon Feb 4 18:30:04 CET 2013

```On 4 February 2013 01:35,  <stefan.braun at smartcae.de> wrote:
> Hi Matti,
> thank you very much for your help. Now i come a big step further.
> I hope this is one of my last questions. Where i have to change the number
> of elements in radial direction.
>

Hi,

You get a finer mesh by decreasing the values of the lc variables in
the beginning of indheat.geo. However, if you increase the number of
elements tenfold in the radial direction, the problem size gets quite
large. To keep the problem size moderate, you might need to do some
drastic changes to indheat.geo to only have a fine mesh in the skin
depth region. Alternatively, you need to consider some other means to
approximate the eddy current losses in the workpiece and use that
instead as the source term for the thermal analysis.

> i try out your example. But if i look at the temperature out of tThe.pos the
> temperature is unchanced.
> Could you please give me a hint what i have to change.
> Thank you very much,

Weird, for me the temperature increases a few degrees during the
simulation when the coil current is at 1000 A. Try to increase the
coil current or directly alter the heat source term in indheat.pro
file.

--
Matti Pellikka, Researcher
Tampere University of technology, Electromagnetics

> Thank you very much,
> Best regards,
> Stefan Braun
>
> SmartCAE, Munich
>
>
>
>
> On Sunday, February 3, 2013 3:02:47 PM UTC+1, Matti Pellikka wrote:
>>
>> Hi, I've updated indheat.pro at
>> http://onelab.info/wiki/Magnetodynamics_with_cohomology_conditions
>> with coupled thermal analysis.
>> It first solves the time harmonic eddy currents in the workpiece and
>> then uses the heat generated by them as heat source for a transient
>> heat conduction problem. As boundary conditions, there are both
>> convection and radiation heat transfer terms present. To solve it, run
>> the command
>>
>> getdp indheat.pro -solve TheDyn -pos TheDyn
>>
>> --
>> Matti Pellikka, Researcher
>> Tampere University of Technology, Electromagnetics
>>
>> On 3 February 2013 13:56, Matti Pellikka <matti.p... at tut.fi> wrote:
>> > Hi Stefan,
>> >
>> > On 3 February 2013 01:41,  <stefan... at smartcae.de> wrote:
>> >> Hi Matti,
>> >> i have two additionally questions to the example.
>> >> If i change the frequence from 50 to 5000 Hz i see no real skin effect.
>> >> Could it be possible that the mesh is not fine enough?
>> >
>> > Yes, you should have a couple of finite element layers per skin depth.
>> > In this case, about ten times more elements in the radial direction
>> > ...
>> >
>> >> My second questions is, how is it possible to integrate the temperature
>> >> and
>> >> the nonlinear mu curve.
>> >> If found an example in the group
>> >>
>> >> http://www.geuz.org/pipermail/getdp/2000/000111.html
>> >> http://geuz.org/pipermail/getdp/2000/000125.html
>> >>
>> >> but the attachment of the working file in 000125.html is missing.
>> >>
>> >
>> > For nonlinear mu curve you can study the example at
>> > getdp, password: getdp). With nonlinearity however, the eddy current
>> > problem is no longer strictly time harmonic. Transient analysis should
>> >
>> > https://geuz.org/trac/getdp/wiki/ChainedResolutions is a simple
>> > example how to couple two problems, for example thermal and magnetic.
>> >
>> >>
>> >> Thank you very much for your help.
>> >> Best regards,
>> >> Stefan Braun
>> >>
>> >>
>> >>
>> >
>> > --
>> >
>> > Matti Pellikka, Researcher
>> > Tampere University of Technology, Electromagnetics
>> >
>> >>
>> >>
>> >>
>> >>
>> >>
>> >> On Tuesday, January 29, 2013 11:53:49 PM UTC+1, stefan... at smartcae.de
>> >> wrote:
>> >>>
>> >>> Hi Matti,
>> >>> thank you for your help. Now it works.
>> >>> Best regards,
>> >>> Stefan Braun
>> >>>
>> >>> SmartCAE, Munich
>> >>>
>> >>> On Tuesday, January 29, 2013 5:03:18 PM UTC+1, Matti Pellikka wrote:
>> >>>>
>> >>>> Hi,
>> >>>>
>> >>>> I was able to reproduce the issue with the latest nightly build on 64
>> >>>> bit Windows 7. I'm investigating the issue, but as a temporary fix,
>> >>>> running the mesher from Windows Command Prompt seems to work.
>> >>>>
>> >>>> Open the Windows Command Prompt and go to the directory where both
>> >>>> gmsh.exe and indheat.geo are. Run the command:
>> >>>>
>> >>>> gmsh indheat.geo -3
>> >>>>
>> >>>> This creates the mesh file indheat.msh, which you can examine
>> >>>> normally
>> >>>> using the graphical interface of Gmsh. To solve the induction heating
>> >>>> problem, have getdp.exe, indheat.pro and indheat.msh in the same
>> >>>> directory and in that directory run in the Command Prompt the command
>> >>>>
>> >>>> getdp indheat.pro -solve MagDynTOComplex -pos MagDynTO
>> >>>>
>> >>>> You can examine the resulting .pos files using the graphical
>> >>>> interface of
>> >>>> Gmsh.
>> >>>>
>> >>>> Best regards,
>> >>>>
>> >>>> --
>> >>>> Matti Pellikka, Researcher
>> >>>> Tampere University of Technology, Electromagnetics
>> >>>>
>> >>>>
>> >>>> On 29 January 2013 13:12,  <stefan... at smartcae.de> wrote:
>> >>>> > Hello Matti,
>> >>>> > thank you very much for your help.
>> >>>> > I tried the new build 2.6.2 gmesh and the bulid of getdp 2.2.2 and
>> >>>> > i
>> >>>> > come
>> >>>> > little bit further.
>> >>>> > I have no error messages and the mesher starts.
>> >>>> > When the mesher reaches reducing cell complex the mesher crashes.
>> >>>> > Do you have any idea how to solve this problem,
>> >>>> > Thank you very much,
>> >>>> >
>> >>>> > Best regards,
>> >>>> > Stefan Braun
>> >>>> >
>> >>>> > SmartCAE, Munich
>> >>>> >
>> >>>> > On Monday, January 28, 2013 10:02:52 AM UTC+1, Matti Pellikka
>> >>>> > wrote:
>> >>>> >>
>> >>>> >> Hi Stefan,
>> >>>> >>
>> >>>> >> The cohomology computation commands on those lines didn't make it
>> >>>> >> to
>> >>>> >> the current release version 2.6.1 of Gmsh. However, those commands
>> >>>> >> should be available in tomorrow's nightly builds of the
>> >>>> >> development
>> >>>> >> version.
>> >>>> >>
>> >>>> >> Alternatively, you can compile the current development version by
>> >>>> >> yourself, check that the build option ENABLE_KBIPACK is ON to
>> >>>> >> enable
>> >>>> >> the cohomology computation.
>> >>>> >>
>> >>>> >> I suspect that you also need the latest development version of
>> >>>> >> GetDP
>> >>>> >> to compute the example.
>> >>>> >>
>> >>>> >> --
>> >>>> >> Matti Pellikka, Researcher
>> >>>> >> Tampere University of Technology, Electromagnetics
>> >>>> >>
>> >>>> >> On 27 January 2013 00:57,  <stefan... at smartcae.de> wrote:
>> >>>> >> > Hello Group Members,
>> >>>> >> > i want to try out a example of onelab.info about induction
>> >>>> >> > heating.
>> >>>> >> >
>> >>>> >> > If i import the *.geo file in gmsh i get the following error
>> >>>> >> > messages.
>> >>>> >> >
>> >>>> >> >
>> >>>> >> >
>> >>>> >> > http://onelab.info/files/getdp/Magnetodynamics/GMSH_GETDP/indheat.geo
>> >>>> >> >
>> >>>> >> >
>> >>>> >> > Info    : Reading 'C:\...\indheat.geo'...
>> >>>> >> > Info    : Starting subloop 1 in Line Loop 543 (are you sure
>> >>>> >> > this?)
>> >>>> >> > Error   : 'C:\...\indheat.geo', line 176 : syntax error (()
>> >>>> >> > Error   : 'C:\...\indheat.geo', line 177 : syntax error (()
>> >>>> >> > Error   : 'C:\...\indheat.geo', line 186 : syntax error (()
>> >>>> >> > Error   : 'C:\...\ndheat.geo', line 187 : syntax error (()
>> >>>> >> >
>> >>>> >> >
>> >>>> >> >
>> >>>> >> > Thank you very much help on this problem.
>> >>>> >> > Best regards,
>> >>>> >> > Stefan Braun
>> >>>> >> >
>> >>>> >> > SmartCAE, Munich
>> >>>> >> >
>> >>>> >> > _______________________________________________
>> >>>> >> > getdp mailing list
>> >>>> >> > ge... at geuz.org
>> >>>> >> > http://www.geuz.org/mailman/listinfo/getdp
>> >>>> >> >
>> >>>> >>
>> >>>> >> _______________________________________________
>> >>>> >> getdp mailing list
>> >>>> >> ge... at geuz.org
>> >>>> >> http://www.geuz.org/mailman/listinfo/getdp
>> >>>>
>> >>>> _______________________________________________
>> >>>> getdp mailing list
>> >>>> ge... at geuz.org
>> >>>> http://www.geuz.org/mailman/listinfo/getdp
>> >>
>> >>
>> >> _______________________________________________
>> >> getdp mailing list
>> >> ge... at geuz.org
>> >> http://www.geuz.org/mailman/listinfo/getdp
>> >>
>>
>> _______________________________________________
>> getdp mailing list
>> ge... at geuz.org
>> http://www.geuz.org/mailman/listinfo/getdp

```