a. I did not find the sign convention for the strains. However, I did go the FLAC which provided the sign convention. I am assuming ADONIS uses the same convention for stresses, strains, ...etc., correct?
yes, ADONIS is very close to FLAC in terms of concept and almost identical in sign convention.
b. In your tutorials, you mention that the displacements need to be zeroed at the initial stage which is logical. The stresses of course are not. However, I was comparing ADONIS with PLAXIS ( which I also use ) and noted that for the initial stage, PLAXIS zeros all displacements and small strains. Do you have an opinion on this ? My thoughts is that the strains should not be zeroed and I could be mistaken .
First I need to clarify, You don't have to zero out the displacement. I do it because I want to track the differential displacement from previous stage. regarding your question, no there is no way to zero out the strain and honestly I don't know how PLAXIS handle that. you always can subtract the strain of current stage from previous ones to find the difference.
c. The stress results of running the initial stage with and without initializing the stresses are the same but strains are not. I am assuming that , without initializing, gravity is producing the correct stresses but not the strains. Any thoughts on this.
here is my thought, initial stress is automatically applied to the interface as well. Therefore, model reached to equilibrium quicker with less strain (less penetration in contact element). If you don't specify stress, interface is going to deform more in order to reach equilibrium as a result relaxes some percentages stress and strain in the vicinity of the interface.