If I have a SQL database MYDATABASE and I import this into a new DDT project MYPROJECT I would expect it to read all the structures, indexes, PKs etc ( if I have check the right boxes on the import wizard )

Now, I make changes in DDT and I deploy to the server.  My changes are deployed.

Now the DBA makes some changes on the server.  I want to keep my metadata in DDT in sync.  If I import into the same project I get MYDATABASE_1 .  Fair enough.  I can now use that as the most current version of the metadata ( or use MYDATABASE to fix the server if the DBA has done something boneheaded. )

But I am not sure what is going on in the Strataframe database.  If I create a new DDT project - MYPROJECT2 - and import the MYDATABASE, am I going to confuse any of the data stored in Strataframe?  ie. are surrogate keys keeping all this straight?  Suppose I import MYDATABASE into MYPROJECT2 and then deploy a changed version of it to MYDATABASE2 on the same server? 

In other words, is there anything like this I have to be careful of when using DDT ?

TIA

No.  The SF database does not rely on the name alone.  It relies on the PKs of the profile (normalization ).  So you will not have any problems in the scenario that you described.  In fact, we have a number of profiles with the same DB names and slightly different structures.