Perhaps it would be helpful to know if you have some EBSD or X-ray pole figures to this image. At least it would be clearer if you deal with some effect from the surface treatment or texture. Moreover, some more detailed data on temperatures for rolling and annealing will be important too . 

This might be shear band formation (will depend on hot rolling temperature), as the austenitic steel have very low stacking fault energy. Low stacking fault energy makes cross slip difficult, so there might be shear band formation in the rolling direction. 

Dear Oksana,

If the banding you mean are the vertical groove-like features in your image, than it is most probably a result of chemical heterogeneity kept from casting up to rolling of the steel. You may confirm it by EDS or better by WDS linescan across the bands, tracking especially Cr, Ni, Si and N.

Dear Ondrej Man,

Thus, the bands form during casting? Why does the chemical heterogeneity occur in shape of bands?

Dear Oksana,

Yes, the heterogeneity originates during the solidification process. Some elements "dislike" the others and move off them thus forming volumes of different chemical composition. These volumes are not originally in the shape of bands, but are formed into bands during subsequent forming, which is done most widely by rolling. The particular shape and orientation of the chemically distinct volumes is mostly driven by the forming technology.

Dear Oksana Klok,

1. I think your question regards rather a possible cause and perhaps an explication than a reason.

2. 316L comprises two variations: X2CrNiMo17-12-3 and X2CrNiMo18-14-3.

3. The alloying elements are in solid substitutional solution; given the details of the heat treatment you gave, the presence of an intermediate phase can be ignored, as is confirmed by your micrograph.

4. The micrograph represents the alloy after hot rolling and in a solution-annealed condition. 

5. Hot rolling implies instantaneous recrystallization. It is noted that in this austenitic species of stainless steel with very low carbon content the alloying elements are always in solid solution; hence, the indication 'in annealed condition' is sufficient. 

6. Given the combination of hot deformation, i.e. above the recrystallization temperature, and the annealing treatment, my first guess would be that the distribution of alloying atoms is rather uniform. Then it seems less probable to relate the more or less banded microstructure to heterogeneities originating from the solidification process. However, it might be possible that alloying atoms show a certain preference for positions along the hot rolling directions.

7. An explanation might be that the hot rolling deformation process is heterogeneous yielding different deformations compared to the critical deformation for recrystallization. Then, on solution annealing differences might occur in grain growth. This combination might also explain that the bands in your micrograph show a distinct direction. Could this direction be linked to the direction of the hot rolling? 

Dear P. Van Mourik, the cross-section which is shown on the picture was taken perpendicularly to the rolling direction. This fact confused me a lot.

Dear Oksana Klok, 

Being retired it is not possible for me to check the following suggestions. Nevertheless I dare to make the next remarks to stimulate the discussion and hopefully they help you in further thinking. Your micrograph represents the microstructure perpendicularly to the rolling direction, so when I understand you correctly, it shows the said cross section of the rolled plate. During hot rolling the plate becomes much longer while in the cross section the deformation remains very small as compared to the deformation in the length of the plate. This might imply rather small deformations in the cross section just above the critical deformation for recrystallization with rather small grain orientation differences (your micrograph is not contradictory to this statement). Then, such cross sections may show more or less columnar microstructures as shown by your micrograph. The basic point in these remarks is the difference in deformation along the longitudinal direction and in the cross section of the plate.

Dear Oksana,

Since there have appeared a "competition" of two opinions, one telling that the bands are caused mainly by chemical heterogeneity, and the second one putting more stress on strain developed during the rolling process, I would suggest - if you have the opportunity - to make an EBSD scan over several bands and grains in your microstructure. Some 200 x 200 microns with 1 micron pitch would be enough to see, whether there is some decrease in EBSD pattern quality inside the bands or not. Even more straightforward would be, if you could grab also the EDS counts during EBSD to be able to overlay chemical maps and "image quality" ones to see directly if there is a correlation either in chemical composition gradient or an increase of stored strain with the bands. This would give you rather qualitative information, nonetheless decisive enough to tell which opinion was true.

Dear Oksana Klok and Ondrej Man,

The maxime of the famous Dutch Nobel prize laureate Kamerlingh Onnes was: by measuring towards knowing (in Dutch: door meten tot weten). So, new experimental data are always welcome!

This occurs when the steel is subjected to cold deformation