batanony,
although Professor Crisi has given an - as always - outstanding explanation of the "visible" or "measurable" magnetic behaviour of an iron or Iron Carbon material (Iron-Carbon-Phase-Diagram), respectively, and additional you by yourself have given us the background information on what the reason were for asking your question I am now hardly considering if it wouldn't be needless to treat this interesting topic here further on. But on the other hand I am person (unfortunately) who is always interested to have a devotional look behind the nature of things and thus - as you can imagine - likewise behind the nature of magnetism in solid matter.
I am certain that it isn't necessary to talk about the basics of magnetism, since there are some excellent links, been attached, and of course the comments been made by the others herein.
However, I would hold my promise to write some sentences and this, what I would like to write subsequently, is - different to the experts insight of e.g. Professor Crisi - just the understanding of things of an impassioned layman using his own words. I request your understanding.
I want to start - as usual - with a more detailed view on an iron-atom, which means that the particle is presumed to be electrical neutral, and as it can be found e.g. in an iron vapour atmosphere.
When we are going to use the imagination that the main electrical positive charged matter (protons) and mass (protons + neutrons) of an atom is concentrated in its nucleus. By the way, I am going to use here "BOHRs atomic model" and not the quantum mechanical atomic model, since this is far above my head (due to DIRACs and HEISENBERGs mathematics). We assume furthermore that the electrical negative charge carriers, namely the electrons, are spinning around the atoms nucleus and creating thus, depending to the number of protons, the atoms neutrality. When we now assume that every movement of an electric charge does generate a magnetic moment and we know that the electrons "move" around the core we can also presume that there has to exist a such a magnetic moment as well. This, so far as I know, is really the case. In German language it's called "Bahnmoment". In English I would call it, "Magnetic Moment"(?).
The second important moment existing to define the behaviour of an electron, is the so called "Spin-Moment".
What is important in general, is that all existing moments referring to an atom are in charge for the magnetic behaviour of physical matter.
When we now assume an outer magnetic field acting on an electric circuit one can observe that this field induces a current which acts contrary to the already existing current (LENZ' law). This behaviour counts also for the electron paths, since moving electrons are - as well known - being seen as to be the "electric current". This again means, that the electrons magnetic moments are reduced likewise and this effect is called by using the term "diamagnetism". Diamagnetic shares are shown in any atom and as far as no other effect can be found one speaks of a "diamagnetic material". Its susceptibility, which is the ratio of magnetization (J) and field strength (H), is commonly not depending to the height of temperature. When we assume now the sum of the existing electron moments in an atom is equal to Zero, we have a pure diamagnetic material, which is the case when we have a symmetrical electron structure, where the distinctions of the electron moments in the atom are compensated to Zero. In case of an unsymmetrical electron structure of the atom there is also a distinction in the compensation of the mentioned moments. The sum is then different from Zero since the single moments are, uncompensated by each other.
O.k. by using a magnetic field, acting from "outside", one can "adjust" now the single free magnetic moments of an atom. In other words, one can increase the flux line density within the material, which is the "Paramagnetism" phenomenon. Paramagnetic appearances do only occur in substances where the atomic moments or their interactions can be normally neglected, as, for instance, in gases. Due to that the elements (periodic table of elements) are as well-known defined by the atomic nucleus and depending on this electron structure. Elements having an unsymmetrical electron structure i.e. all elements except the inert gases (Helium, Neon, Argon, Krypton, Xenon, Radon) do now have a paramagnetic behaviour due to their sum of "moments par atom" (different from "Zero"). By the way, elements or ions having the "inert gas character" by having fully saturated "electron shells" (compare BOHR's atomic model) have as well a diamagnetic character as inert gases too.
Coming back now to the "ferromagnetic" behaviour of iron.
"Ferromagnetism" is a specific range of "Paramagnetism", where the height of Permeability - which is the ratio of the induced flux line density within the material (Induction "B") to the acting outer field force ("H") - relatively high (1... 10^6). To have a better understanding for the subsequent items we should have to have a closer look on another important fact, and therefore it is necessary to use another technical term, which I would call the "atomic moment" (Atom-Moment in German language). The magnetic saturation of matter is depending to the height of the total atomic moments. This again is being defined by an atomic mass unit, called: "BOHR Magneton". For atoms having more than half filled up sub shells, the height of the BOHR Magneton is equal to the number of absent electrons. In case of iron (3d-shell) thus the number would be "4", since four electrons would be necessary to completely filling up the shell, although the 4s-shell is already filled up with two valence electrons.
To come to "real" ferromagnetic behaviour of iron one has to leave the field of existence of its vapour or gaseous condition, respectively, and must move to the "crystalline" condition. In this condition one can no more treat the single atom as "independently" from others (as in gaseous condition) but to be "disturbed" by its "neighbours". The nearer one atom is approached to each other, the higher their mutual interference.
What does this mean?
The physicists use the model of "overlapping" energy levels. Whereas in a gaseous and "single atom condition" one could speak of stringent separated energy levels, existing in stringent distances from the atomic nucleus, in solid state of matter these separated energy levels - in particular in the outer shells - can no more be observed but they are "broadened" by overlapping one to each other. It is a risk to try to describe it as I do it here since in reality it is much more complicated and only to be explained by using the quantum mechanical mathematical formalism developed by famous people having got the Nobel prize. Who am I? However, let's resume. When we look on iron we can see, that the 3d-shell is partially saturated with 6 electrons although fill saturation is achieved by 10 electrons. And we have the 4s-shell which is fully saturated by 2 electrons. Hereby the iron "achieves" the best possible energy level for itself. Under different temperature heights one can achieve different "overlapping conditions" which is crucial due to its importance to your question if there is a temperature-magnetism dependence in iron, which has of course already been replied by Prof. Crisi.
This here deals only with the questions:
"Where does ferromagnetic behaviour actually come from?" and "Why is there even a difference in the magnetic behaviour of iron to be observed when exceeding a specific, namely "CURIEs temperature?"
Up to now one could state.
It has to do with the electrons and the electron structure of elements (e.g. iron) whose "shells", according to BOHRs model, are incompletely filled up with electrons to not receiving the inert gas character.
Let's stay at the electron levels which are overlapping when the iron vapour is being solidified. When the iron atoms are "ordered" by changing its state of aggregation (gaseous to solid) the 3d-shell and the 4s-shell are overlapping by "mixing" their electrons. The probability density of in which "overlapped shell" the regarding 8 electrons are actually to be found, can be calculated by using the fraction of electrons under using the BOHR Magneton again. But this I do not want to treat further on herein. However, what's important, is that by changing the state of aggregation of iron from gaseous to solid, the number of BOHR Magnetons is reduced (4 to 2.2). By the way batanony, you are almost right when you say that the path-breaking surveys in this subject were conducted approx. 100 years ago, so e.g. in the year 1911 by P. WEISS and G. FOEX (Ann. Phys. Paris Volume 1, 1911, pp.64).
Beside different other details it is important to know that the sum of atomic moments within iron or elements of the "iron-group" of the table of elements, is being reduced down to only the "spin-moment" which is, by that, the most pivotal factor for the magnetic behaviour of the element. This is founded on the fact that the crystal forces (in solidified iron) can compensate the other electron path moments.
Now we come slowly but surely to the "subjects core".
WEISS has in [Le Magnétisme. Paris 1926] stated that the ferromagnetic condition of an element is no property of a single atom but the more a property of an atomic, I would call it, "cluster" which means that a specific region of atoms provides the ferromagnetic behaviour of the material. This again is defined by a kind of "parallel orientating" the "spin-moments" of the involved electrons (overlapped shells). But this parallel orientation of the spin-moments again does nothing less mean than "ferromagnetism" (!).
This should normally mean that iron should show magnetic behaviour without any further necessary outer field acting on the electron spins to orientate them parallel to each other, what is again the reason for the ferromagnetism of it which can be easily shown by approaching a magnet (-ic field) to it. But why isn't iron then permanently magnetic? Well, honestly one can find in the range of the atomic "clusters" actually a permanently magnetic behaviour. In German language this behaviour is called "Spontane Magnetisierung". Unfortunately I don't know how to translate this term into English language (I must think on the term "Inverse Bremsstrahlung" in the field of Power Beam Welding Processes where a translation isn't found yet as well). Perhaps I would translate it by using the term "Spontaneous Magnetism"(?). However, this means, a very small concentration or minimum number of atoms (even the cluster-region - according to H.KÖNIG: "Naturwissenschaften Volume 33/1933" this minimum number is 64 atoms) is a kind of a permanent magnet. The only reason for iron in a macroscopic scale is not permanently magnetic is the fact of that the "cluster" regions, as mentioned above, are sure parallel orientated in their spins but due to the immensely high number of "clusters" we have a statistical distribution of orientations. This again means nothing less than that the magnetic forces of each "cluster" do compensate themselves from one region one to each other. Normally one should go deeper into the very interesting physical backgrounds here, since there are some elementary coherences between the atom's electron structures in relation to its nucleus, but I request your understanding when I am avoiding this for not overloading the response.
Rather I would like to try to come to an end and therefore I have to come now slowly to the influence of the temperature on the (ferro-)magnetic behaviour of iron, since as you have stated in your initial post (quote):
"...the demagnetisation process start to increase with the appearance of Austenite at 723C....What about before that? I mean from 0C- 723C? what is the relationship between the magnetic permeability of iron and temp.? decrease gradually? no change till this temp? or what?..."
The "temperature" - which should mean nothing less than heat induced movement of particles - has significant influences on the magnetic behaviour of the material. This is founded on the fact, that the heat induced movement of atoms or ions works contrary to the inner "ambition" to parallel orientate the electrons spin-moments. This means, the higher the temperature the material (iron) is subjected to, the higher has to be the outer magnetic field to achieve an increase in field line density, or in other words, the higher the temperature the lower the sum of equally parallel orientated atomic regions. On the other hand this means, the lower the temperature, the higher is the probability of achieving parallel spin orientation and thus achieving a "ferromagnetic" behaviour. At the lowest point of temperature (0 Kelvin) thus there must be found a maximum magnetism-value specific to the material to be observed.
Let us state up to here.
The temperature - or let's say, the heat induced particle movement within the solid matter - strives to work against the electron spin-orientation which is again the necessity for the materials ferromagnetic behaviour.
In regard to iron this is - of course - even the case. When we are talking about "CURIE-temperature" we are actually talking about the point of temperature where the ratio between both forces - the parallel spin-orientating forces on the one hand and stochastical heat induced atomic movement forces (working against the first named) on the other hand - is quite balanced. The material loses its "ferromagnetic" behaviour and attains pure "paramagnetic" behaviour instead of this. CURIE has surveyed this already in 1908 - you see almost 100 years ago(!) - and I would like to attach a diagram which shows the dependence between the temperature's height and the height of magnetisation (or demagnetisation as you have called it) in case of iron, please see also the attached magnetisation_dependence.pdf.
Concluding now and thus - of course - confirming Prof. Crisis excellent explanations one could state:
"Yes, there is a dependence between the temperature the material - in this case iron - is subjected to and its magnetisation or in opposite demagnetisation behaviour. This dependence occurs not promptly but gradually up to a specific point of temperature (CURIE-temperature) where the ratio between the "magnetisation inducing" forces and their "counterparts", the heat induced particle movement causing forces is in balance and the material is loosing its ferromagnetic share and becomes paramagnetic."
By the way, this peculiar behaviour is also described by the "CURIE-WEISS-law".
Well, now I am at the end and I hope that I have found the right words for describing something what is hard to describe for a layman like me. Some good spirits may perhaps mean that this what I have stated is too simple explained or that I have forgotten too many things to mention due to the subject of matter is much more complex. And if so, I would agree with those ones, they are surely right by saying so. But however, it is truly fascinating from my personal point of view and I thank you for asking such a great question. It gave me the opportunity to busy myself a bit more in depth with this interesting matter than I have already done before up to now, and finally I can always apologize what I have "forgotten" to explain by saying:
"I am lastly no physicist but only a welder and I am proud of being one!" :-)
By the way and then I will shut up. I had the joy and honor to once speak to an outstanding expert in Underwater Welding (Barry Richards from Great Britain) and have asked him if there are no problems with arc- or welding-process stability when SMA-Underwater Welding. He answered "No!" On the contrary! Due to the increased surrounding pressures under water you have a significantly improved process stability compared with welding under atmospheric conditions." And just a few weeks ago I have listened to a wonderful presentation been held by Prof. Bill Lucas from England's The Welding Institute (TWI). He has reported about the state of the art in SMA Underwater Welding and has given some very impressive and excellent instances for the highly advanced results by using a meanwhile excellent variety of welding consumables (Stick Electrodes). These consumables yield both outstanding inner as outer welding results which are truly unbelievable good. The last development is an underwater welding stick electrode been developed between the welding experts from TWI and a British company (unfortunately I do not know the name of the manufacturer). This electrode, so Prof. Bill Lucas, is the actually secret of the meanwhile possible to perform SMA Underwater Welding operations.
So far my humble try of an explanation.
Best regards,
Stephan