Intense parasitic nucleation has been observed at the surface of differential scanning calorimetry samples for various polymers,
whereas their crystallization traces exhibit complex shapes. Revisited overall kinetics theories and computer simulation,
taking into account small thickness of samples and transcrystallinity effects, allow to explain and reproduce experimental
‘double peaks’, currently observed with polyamide 6-6. The beginning of the transformation and the main peak are attributed
to surface and bulk nucleations, respectively. As a consequence, any DSC experiment should be followed by a microscopic observation
and more accurate models including thermal gradients and resistances should be developed for their interpretation.