If you have not found your PCM in the
database
you could use the following template to define a custom PCM and
generate source code for it.
Select language:
and
or
Usage notes
Phase transitions and temperature range
Non-isothermal phase transition behavior is described by
the lower and upper bound of the phase transition range,
and by the peak shape.
Different Ansatz functions are available to model the peak:
Smooth Step function (symmetric peak)
Sigmoid function (symmetric peak)
Gumbel Minimum Distribution function (asymmetric peak, left-skewed)
Log-Normal Distribution function (symmetric or asymmetric peak, left or right-skewed),
with user-defined 'peak shift parameter', a value of zero corresponds to
the Normal Distribution function with a symmetric peak
All peak functions (except the smooth step function) show an asymptotic behavior,
which means that the peak curves are never exactly zero,
instead they approach zero when the temperature tends to infinity
and minus infinity.
The integral of the peak function is the liquid mass phase fraction ξ(T) which indicates the
phase change progress.
Because of the asymptotic behavior of the peak function, the phase fraction function
never gives exactly
ξ=0 g/g (material is completely solid) or ξ=1 g/g (material is completely liquid).
As a solution, we define the start and end
of the phase transition temperature range by:
T_min where ξ = 0.001 g/g, and T_max where ξ = 0.999 g/g.
Symmetric vs. asymmetric peaks
Measured peaks are often asymmetric.
Asymmetric peaks can be modelled either by the Gumbel Minimum Distribution function, or by the
Log-Normal Distribution function.
The (user-defined) peak shift parameter (peak shift: -80 % ... 80 %) works only for the Log-Normal Distribution!
Try it out here:
Note: Some of the above user-defined models are also included in
the Modelica/Dymola library
GitHub-slPCMlib,
see the Package
slPCMlib.Media_generic.