4. Relation between Various Types of Remanent Magnetization

[37]  The dependence of magnetic states of single-domain grains on various factors (temperature, pressure, chemical composition, time, magnetostatic interaction of particles, and others) inferred above can be used to describe and compare various types of the remanent magnetization in terms of a unified model.

2005ES000184-fig02
Figure 2
[38]  Thus, comparison of thermoremanent ( Irt ) and anhysteretic ( Iri ) magnetizations in ensembles of single-domain particles coinciding (narrow spectrum) and differing (wide spectrum) in coercivity [Afremov and Kharitonsky, 1986a, 1986b] shows that Rt = Irt/Iri < 1 for a system of weakly interacting grains and Rt > 2 for strongly interacting grains (Figure 2). The dependence of the ratio between chemical and anhysteretic magnetization Rc = Irc/Iri on the factor F defined by Sholpo [1977]
2005ES000184-fig03
Figure 3
is plotted in Figure 3. The curves Rc = Rc(F) are similar in behavior to Rt = Rt(B), which is quite natural because Fapprox B/H max at Fll1.

[39]  Comparing Rt and Rc, it is easy to show that, with increasing Hc max, the parameter Rc reaches an ultimate value at greater values of B than the parameter Rt does. This is due to the fact that the chemical and thermoremanent magnetizations are mainly due to, respectively, low- and high- coercivity grains. Note also that, in the experimentally observed range of interaction fields ( Bapprox 3-30 Oe) [Ivanov et al., 1981; Ivanov and Sholpo, 1982], the maximum value of Rc is Rcapprox 1, whereas Rt depends on coercivity.

[40]  Knowing the response of each particle to an external effect (variations in Hc and Is and the corresponding displacements of representative points in a diagram), we can easily calculate all of the aforementioned types of remanent magnetization and establish the relations between them.

[41]  The function g(|H|) can also be applied for the estimation of detrital magnetization in a system of large particles whose orientation is modified by a magnetic interaction field rather than temperature. This estimate is directly related to the so-called cluster model of depositional magnetization developed in [Shashkanov et al., 1989, 2003

].

[42]  Using the distribution density given by (4), one can show [Belokon and Nefedev, 2001] that, before the decrease in the tilt angle of a certain portion P of elongated or flattened particles, the detrital magnetization is given by the formula

eq027.gif(26)

where H is the external field. After the particles become less tilted, we have

eq028.gif(27)

eq029.gif(28)

Here (p2- vartheta0) is the inclination of Ir0 and vartheta0-vartheta0 is its error.

[43]  In conclusion, we can note the following.

[44]  (1) Knowledge of values of Rt and Rc is insufficient for the identification of thermoremanent and chemical magnetizations in an ensemble of single-domain particles. This identification requires additional information on the coercivity and the intensity of magnetostatic grain interaction that can be obtained from laboratory experiments.

[45]  (2) Spontaneous magnetization can change with time due to diffusion processes [Afremov and Belokon, 1972]. This can lead to stabilization of the vector Is and a rise in the magnetic moment of the system (diffusion-induced viscous magnetization).

[46]  (3) The formation of chemical magnetization can be interpreted in terms of a mechanism [Belokon et al., 1995] that is an analogue of the thermoremanent magnetization formation mechanism and differs from the crystallization mechanism proposed by Haig [1962].


RJES

Powered by TeXWeb (Win32, v.2.0).