| 
 | 
 | 
 | [Contents] | [Index] | 
Collaborator: Th. Jurke, O. Klein, J. Sprekels
Cooperation with: P. Krejcí (Academy of Sciences of the Czech Republic, Prague), S. Zheng (Fudan University, Shanghai, China)
Supported by: DFG: ``Hysterese-Operatoren in Phasenfeld-Gleichungen'' (Hysteresis operators in phase-field equations)
Description:
To be able to deal with phase transitions, one has to take into account hysteretic phenomena that are modeled by hysteresis operators. Moreover, methods derived for dealing with hysteresis operators also allow to derive results for equations formulated without a hysteresis operator.
In a number of papers (see, for instance,  
[1, 2, 
4]
and the references given therein), 
integrodifferential (nonlocal) models for isothermal phase transitions with 
either conserved or non-conserved order parameters have been studied, leading to
a number of results concerning existence, uniqueness, and asymptotic behavior
of solutions. In the recent papers [3,
10] the more difficult non-isothermal case has been treated, 
modeling the phase transition 
by considering the time  evolution 
of an order parameter
 and of the absolute temperature
 and of the absolute temperature  .
In these papers, 
one uses a free energy density 
containing a logarithmic part that forces the order parameter to attain values
within the physically meaningful range [0, 1].
.
In these papers, 
one uses a free energy density 
containing a logarithmic part that forces the order parameter to attain values
within the physically meaningful range [0, 1]. 
Within the covered research period,
the results of [10] have 
been complemented 
by investigating the case when the logarithmic part is
replaced by the indicator function I[0, 1] of the interval [0, 1], 
see [8].
Considering the phase transition 
within a
container 
 
  
  that forms an open and bounded domain, and
denoting  with T > 0 
some final time,
the following  system   has been considered
in
 that forms an open and bounded domain, and
denoting  with T > 0 
some final time,
the following  system   has been considered
in  
 x (0, T):
 x (0, T):

 [0,
[0, ),
appropriate functions
),
appropriate functions  , F1, and F2, and
positive constants CV and
, F1, and F2, and
positive constants CV and  .
In [8], this system has been investigated   by 
introducing the generalized freezing index
.
In [8], this system has been investigated   by 
introducing the generalized freezing index 
| w(x, t) = w0(x)  -  ![$\displaystyle \left[\vphantom{\frac{1}{\mu(\theta)} 
\left(\theta F_1'(\chi)+
F_2'(\chi)+ Q[\chi]\right)}\right.$](img265.gif)   ![$\displaystyle \left(\vphantom{\theta F_1'(\chi)+
F_2'(\chi)+ Q[\chi]}\right.$](img267.gif)  F1'(  ) + F2'(  ) + Q[  ] ![$\displaystyle \left.\vphantom{\theta F_1'(\chi)+
F_2'(\chi)+ Q[\chi]}\right)$](img268.gif) ![$\displaystyle \left.\vphantom{\frac{1}{\mu(\theta)} 
\left(\theta F_1'(\chi)+
F_2'(\chi)+ Q[\chi]\right)}\right]$](img269.gif) (x,  ) d  , | (4) | 
 (x, t) =
(x, t) = ![$ \frak{s}_{{[0,1]}}^{}$](img271.gif) [
[ (x), w(x, . )](t)
with
(x), w(x, . )](t)
with 
![$ \frak{s}_{{[0,1]}}^{}$](img271.gif) : [0, 1] x C[0, T]
 : [0, 1] x C[0, T] C[0, T] 
being the stop operator for the interval [0, 1].
This has been used to eliminate
C[0, T] 
being the stop operator for the interval [0, 1].
This has been used to eliminate  from
(1)-(3), leading to a system
for 
(w,
 from
(1)-(3), leading to a system
for 
(w, ) involving hysteresis operators, which is
of the same form as the system considered in 
[7, 9]
except for the nonlocal term.
The lines of argumentation used 
in [7, 9]
have
been adapted to deal also with the nonlocal term and, 
in [10], this has been used to prove
results concerning existence, uniqueness, and asymptotic behavior
for 
t
) involving hysteresis operators, which is
of the same form as the system considered in 
[7, 9]
except for the nonlocal term.
The lines of argumentation used 
in [7, 9]
have
been adapted to deal also with the nonlocal term and, 
in [10], this has been used to prove
results concerning existence, uniqueness, and asymptotic behavior
for 
t +
 +  , 
resembling those established in [10]
for the smooth case.
The results are even more complete than those of [10]
since
a certain crucial assumption is not needed
in [8].
, 
resembling those established in [10]
for the smooth case.
The results are even more complete than those of [10]
since
a certain crucial assumption is not needed
in [8].
It has been shown in [5, 6] that one can derive uniform estimates for the solutions to some partial differential equations involving hysteresis operators, if these operators are ``outward pointing hysteresis operators''. For scalar Prandtl-Ishlinskii operators and generalizations of these operators, appropriate conditions that allow to check if these operators are pointing outward have been formulated in [5, 6]. Within the covered research period, it has been tried to formulate also appropriate conditions for Preisach operators, but the derived conditions are not yet satisfactory, and further investigations are required.
References:
| 
 | 
 | 
 | [Contents] | [Index] |