This is dependent on the thickness of the steel sample. So it would be a plane strain condition, i.e., thick sample, on the steel sample in which strain associated along z-direction is ZERO or it would be a plane stress condition, i.e., thin sample, on the steel sample in which stress along z-direction is ZERO.
Even though I didn't know the history of the steel sample, yet I would suggest that the delamination phenomenon on steel fracture was due to probably a residual stress. Has the steel sample preheat treated prior to the fracture or post heat treated after the fracture that might let to the probably residual stress.
Delamination phenomenon usually is attributed to band structure of steel. Your guess (residual stress) may be other cause of delamination and this should be investigated.
But my question is unsloved yet:
Why z-stress is developed in thick specimen?
As you know, this state of stress isn't dependent on kind of material.
Try to do X-Ray structure analysis in order to determine the structural parameters (constants of the unit lattice) of the upper layer (microstructure phase) of the steel surface. This would give you an idea whether the steel surface has attained any residual stresses especial when you compare the obtained structure parameters with the ones in reference literatures.