I think you are talking about low and upper yield point in low carbon steels. As you said it is prominent in low carbon steel and for high carbon steel in my opinion excess of carbon makes the material to deform in brittle manner and so you do not really observe that phenomenon.

It may be due to high pearlite or carbide content in high carbon steels.

The concept that "yield elongation occurs due to pinning of dislocations by interstitial atoms" is a special case and limited to certain alloys only. Also it depends on the history of the alloy. The general cause of the yield elongation or Portevin Le-Chatlier (PLC) effect has been attributed to the heterogeneous deformation of the alloy.

The presence of PLC effect depends on the distribution of dislocation in the material being tested. Alloys which show this effect are known to yield non unifromly and those which do not show this are known to yield uniformly. Distribution of dislocation is generally not uniform and hence some portion of material yield rapidly as compared to other portions leading to formation of Luders' band. With the application of strain, the luders' band propagate in to whole material. Since to propagate, the band has to cross the grain boundaries of the poly-crystalline material the serrations in stress can be observed. This leads to the appearance of PLC effect. If the initial dislocation density and strain are uniformly distributed across the specimen being tested this phenomenon may not take place and material may yield uniformly.

Dear Mr. Varshney. Thanks for your answer. But i guess heterogeneous deformation behaviour is the outcome or the manifestation of PLC phenomena, rather than being the fundamental reason for PLC (for e.g dislocation pinning by solutes)

Comparing the ratio of activation energy of interstitial movement to activation energy for dislocation movement among the various alloys might give a clue. Or you may need a better measure of mobility of solutes and dislocations, perhaps the velocity of dislocation glide compared to the spatial velocity of solute diffusion, which would depend on temperature and concentration gradient.

Please see notes (attached) prepared by me on this topic of dislocation -solute interaction (PLC effect/dynamic strain ageing) and this may be useful to you. 

Best wishes.

Danish,

Please read section 9.7.3 - 'Effect of grain size' in "The Plastic Deformation of Metals" by R.W.K. Honeycombe, pub. Ed. Arnold, 1975. Prof. Honeycombe puts forward a reasonable explanation for the grain size effect based on consideration of the interplay between dislocation locking stresses*  and implications of the Hall Petch equation in this situation. (or you can subsittute the modern take of new dislocation generation, think it still works).

For a fine grained  structure in the ultra low carbon steel (plus locked disloactions) you can also think  of those fine grains giving you a high initial yield.  BTW Give the interstitials a bit more thermal energy, (slightly elevated temperature tensile testing), and they can re-lock moving dislocations giving a serrated yield (PLC effect).

re why it is associated with ultra low carbon steels, my picture is that for the continued flow at the lower yield point you need a soft sort of compliant matrix, for high carbon steels the matrix is quite different even if fully annealed, so you go straight into general work hardening. 

In general the effect is attributed to solute atoms like for example Mg in Al-alloys and C or N in steels. Distinct atoms in the crystal lattice lead to handicaped dislocation glide. Due to this pinning effect of these atoms/"atom clouds" on dislocations, a higher stress is necessary for dislocation movement (upper peaks in the sigma-epsylon-curve). When the critical stress is reached, dislocation movement can take place and at the same time the critical stress is lowered. This results in a temporarely higher strain rate. This effect repeats several times during deformation which results in the observed "jerky flow".

I hope this answer is helpful.

Best wishes