## current pulse heated sample

General discussions about how to formulate a script for FlexPDE.

### current pulse heated sample

Hello to all,
I am new to this forum but as I am stuck with my problem I hope someone here can help me out.
short description of the problem:
There is a metallic sample clamped on the top and on the bottom inside a vacuum chamber (so no convection).
By using a current pulse of 1kA for 0.5sec it gets heated up to short under melting temperature.
After this 0.5sec the current gets cut off immediately and the sample can cool down freely by radiation and conduction to the clamps.
The sample shape is a tube with a black body like hole on one side (if necessary I can provide some schemes).
I want to measure the temperature inside this hole AND on the other side of the tube alongside the sample.
So the measurement shall be location-dependent AND time-dependent.

Short description of the simulation I want to do:
For the first tests the sample shape can be just a cylinder of defined height.
The top and the bottom plate shall have ambient temperature (300K) to generate heat flux towards the endings.
Also there shall be the voltage for current be applied (top and bottom plate).
As I previously said convection can be neglected.
So the equation formula includes the following:
Heating up the sample using the current pulse for 0.5sec.
The heat is flowing through the sample towards the clamps depending on the location.
The only loss parameter is the radiation corresponding to temp^4.
The heat is stored inside the sample with the heat capacity depending on time.

Please can someone help me figuring out why the FlexPDE script is not converging.
all the best
Max
Attachments
Schematic description of the experimental setup
pulse_heating_new_setup.jpg (248.31 KB) Viewed 2103 times
Probe_Tubus_Zeichnung.pdf
detailed description of the sample
zylinderprobe_e_transient_forum.pde
FlexPDE 6.50 script describing the problem
gessner.max

Posts: 6
Joined: Wed Jul 05, 2017 5:48 am

### Re: current pulse heated sample

Ok, so as the simulated sample is not heating up while applying a voltage on top (v_max/2) and on the bottom (-v_max/2) I assume something is wrong with my equations.
The help files just provide "contact" resistance heating, but I need just resistance heating of a random shaped metal by applying a voltage drop to it.
The boundary conditions I assume should be:

BOUNDARIES
SURFACE "cyl_bottom" value(temp)=temp0 value(voltage)=voltage_down
SURFACE "cyl_top" value(temp)=temp0 value(voltage)=voltage_up

Am I right or does it have to be different?
current flow is initiated by:

EQUATIONS
voltage: div( -sigma * grad(voltage)) = 0

Heating is done via:
natural(temp) = sigma * (magnitude( -grad(voltage))^2)

gessner.max

Posts: 6
Joined: Wed Jul 05, 2017 5:48 am

### Re: current pulse heated sample

Professor Backstrom has a similar problem at "Samples|Backstrom_Books|Fields_of_Physics|D_Heat|heats2.pde".
moderator

Posts: 776
Joined: Tue Jan 11, 2011 1:45 pm

### Re: current pulse heated sample

Thank you for the quick response.
I looked at that particular problem already.
As it turned out it just covers heating up a sample via current in steady state conditions.

My problem on the other hand is highly dynamic.
Adding a time dependence to heats2.pde just solves the steady state problem in every time step.

Any other suggestions where my mistake in the script is?
gessner.max

Posts: 6
Joined: Wed Jul 05, 2017 5:48 am

### Re: current pulse heated sample

Just for the record, I solved the problem.
If anyone is interested in the solution feel free to ask me directly.

Only thing that bothers me now is the initial condition of the sample which should be at 300K but rises to 400K at start and slowly cools down to 300K after 0.1 sec...
gessner.max

Posts: 6
Joined: Wed Jul 05, 2017 5:48 am