Comparison of dynamic recrystallization and conventional controlled rolling schedules by laboratory simulation.

A seamlesstube mill rolling process was simulated with the aim of optimizing the ferrite grain refinement by decreasing the finish rolling temperature during the stretch reducing mill (SRM)stage.Tests were performed on one Nb-Vand two Ti-V micralloyed steels.For the Ti-V steels, maximum g~ain refinement is achieved whenthe SRM exit temperature is just abovethe Ar3 ; at a cooling rate of 3.5'C/s during transformation, ferrite ASTM grain size numbers of #1 3 and higher are produced.When the reductions are applied below the no-recrystallization temperature or Tnr, the strains are accumulated at first as a result of the absenceof static recrystallization during the short interpass times.At larger strains, dynamic recrystallization is initiated, due to the absenceof strain-induced precipitation du~ing the brief unloading inter- vals.Under these conditions, the rolling loads are lower thah when a conventional controlled rolling scheduie is employed.In the latter case, the interpass times are sufficiently long for the occurrence of appreciable strain-induced precipitation, which is responsible for the prevention of dynamic recrystallization, and the consequent accumulation of strain and pancaking of the austenite.For the Nb-V steels, whenthe SRM cooling curve passes through the nose of the CCP (continuous-cooling-precipitation) Ps curve, austenite pancaking takes place.By contrast, when the cooling curve passes above or be/owthe CCP curve, the dynamic recrystallization of austenite is initiated, Ieading to ferrite grain sizes which are generally finer than those produced by the pancaking process.