I need cooperation on this theoretical work (it would be interesting to include some experimental results and additional suggestions before submission). It follows a preliminary abstract:
The Barkhausen Noise (BN) is attributed to a series of jumps in the magnetization of a ferromagnetic material when it is exposed to a varying magnetic field. When a coil is employed to capture the flux variation, these changes induce a voltage that may also be processed and transformed into acoustic noise. This noise manifestation has a vast practical application, especially for noninvasive material characterization techniques and their importance for understanding the relation to the magnetic domain scale. The aim of this paper is to propose an extend or modified Preisach model for magnetic materials. It is capable of representing rate-dependent hysteresis with macroscopic hysteresis loops. It also produces the time domain signal of the BN. The model is devised in a discrete N-dimensional state space form, where each hysteron is not confined to only two possible amplitudes and is also associated with a second-order dynamical system. The global model order and statistical distribution of the corresponding parameters define the temporal and spectral features of the noise signal. These parameters may also be related to microstructural properties or stress. This proposed hysteretic-stochastic dynamical model may simulate the response to arbitrary magnetic field signals over an entire hysteresis cycle and provides a platform for developing theoretical and practical studies of rate dependency and BN.
That sounds interesting. Not sure I'd be able to contribute substantially, though. Would "dynamically" include the field sweep rate?
Some tima ago I have embarqued on simulating that for single domain particles with a given model for the anisotropy, by integrating corresponding rate equations employing Runge-Kutta integration.
- clearly, this is not dealing with Preisach models
- I'm not an expert in numerics (actually I am more an experimentalist with some interest in numerical simulation), so I was unable to deal with variable time steps without significantly disturbing the solution.
- Getting to the time scales I was interested in turned out to be prohibitive in terms of computing time given my resources (essentially scilab on a PC)
Besides the above: ho do you (intend to) deal with interactions between the domains represented by individual Preisach objects? Can this adequately be done with a purely statistical approach?
'We' have resorted to explicite modelization in one piece of work, see the links below. These simulations (aiming at interpreting our experiments) were done by Mihai Cerchez at the time. Again, I'd think that the methodology was quite different to what you're doing but there might be interesting links? Anyway, BN was not in our focus, since we measured M(H) cycles and intended to cover these by the calculations.
http://journals.aps.org/prb/abstract/10.1103/PhysRevB.70.214412
Article Effective exchange interaction in a quasi-two-dimensional se...
Dear Kai,
Thanks for sending the link and your valuable comments.
Yes, "dynamically" include the field sweep rate, which changes the BN and the hysteresis shape. Since the model comprehend stiff systems, it requires stiff solvers and it is working at acceptable computational cost. Classical Runge-Kutta integration may not apply.
I do not intend to deal with interactions between the domains. They are represented by individual modified Preisach operators.
The spectral representation of the noise I am currently obtaining with this model is pretty close to those presented in the literature. So I believe the proposed approach can adequately be done with a purely statistical (dynamic) approach. In fact, the way it is devised, the model may be interpreted as a result of the Karhunen–Loève theorem and Wiener process, under the framework of Itô calculus.
Hope to have made your questions clear.
Regards
In my theoretical studies of electromagnetic phenomena in ferromagnetic, including for the purpose of non-destructive testing, I kept returning to the question of formation of the magnetization curve. I even managed to write a few simple, primitive theories that helped to find the correlation dependence of coercitive force, saturation magnetization....with mechanical properties of some steels. These questions are of interest for me.
But I do not understand what you mean when you write "I need cooperation ....".
If you have any ideas - I would be interested to discuss them.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Material Characterization