How to set critical points on a dilatometric curve-determination of CCT of a steel

Hello everybody,
I have just started a PhD lately; the project is about characterisation of microstructure of steels with bainitic and acicular ferritic structure; I need to create CCT diagrams from dilatometric curves of the aforementioned materials. Now, setting those critical points indicating the start /beginning of the transformation seems somewhat fuzzy to me; could anyone give me some hints/references/articles about the way it's done?
Thanks in advance,
Metalo

snipped-for-privacy@gmx.net wrote:
  Quoted Text Here  
Thermally activated processes start immediately after passing the equilibrium temperature.
AFAIR "beginning" means "1% transformed".
Michael Dahms

Thank your for the replies.
Now, more specifically, having the dilatometric curve and its derived one, I normally put the tangents on the curve to get the critical points. However, from time to time the derived curve is amazingly rough and oscillating, thus it's pretty hard to place a line on the curve. Do you have any trick to overcome this?
Thanks, Metalo

snipped-for-privacy@gmx.net wrote:
  Quoted Text Here  
You need a physical model in order to obtain reasonable physical data from a curve. Just doing math isn't worth much.
Michael Dahms

  Quoted Text Here  
What model do you use to treat experimental results?
Is the paper
http://www.msm.cam.ac.uk/phase-trans/abstracts/eutectoid.html
is similar to your work?
Evgenii

At the moment I use no model, just "by sight" I lay lines on the curve.
But thanks for your hints, I will have a look at the article you sent me.
Metalo

snipped-for-privacy@gmx.net wrote:
  Quoted Text Here  
You shouldn't do that.
Michael Dahms

snipped-for-privacy@gmx.net wrote:
  Quoted Text Here  
Look at the article by George Eldis titled "A critical review of data sources for isothermal transformation and continuous cooling diagrams" in the symposium book "Hardenability Concepts with Applications to Steel", (edited by D. V. Doane and J. S. Kirkaldy TMS-AIME 1978, pages 126-157
Pittsburgh Pete

  Quoted Text Here  
Hi Metalo,
A good starting point could be an equilibrium phase diagram that can be computed from Gibbs energies. Have a look at Thermocalc (www.thermocalc.se).
Best wishes,
Evgenii Rudnyi

Hi! The tangent method should work pretty well. I've used this method before and its fairly accurate. If you have a quench dilatometer, quench at different points of the curve to try to capture the phase forming by metallography. Although this may be hard for acicular and bainitic structures and work intensive. The other thing to look at is the gain setting on the LVDT, I don't quite know why your curve is choppy. It should be smooth linear thermal contraction or expansion followed by the phase changed. Otherwise there may be controler issues.

Oh yeah, I know of two research groups to look up.
1) Prof. Van der Zwaag has published some neat papers on dilatometry. 2) Prof. Boyd has done some excellent work on the characterization of bainitic and acicular microstructures using dilatometry.