hy mbm i hope i got it right
the hydraulic cilnder has a basic travel, from point A (left) to point B(right); A<B; this travel is fully executed when there is no chuck
if you put a chuck on it, and you wish to use it for OD clamping ( and that is pretty common ), than the travel will change :
... variant 1 : from point A1 to point B1, where A<A1<B1=B ( clamping position is somewhere between A1 and B1 )
...... it is recomended that A<A1, so the hydraulic cilinder to have a travel longer than the jaw travel : thus it i possible to clamp regardless of master jaws position
...... it is recomended that B1=B, thus, when the chuck is opened, the pull bar is in a relaxed position, and the master jaws should have a bit of clearance : you can check this by hand : if all jaws are sliding a bit when the chuck is opened, than it means that the chuck is not under pressure
... variant 2 : from point A1 to point B1, where A<A1<B1<B
...... it means that the pull bar never reaches its full travel, thus the hydraulic cilinder is always in tension, especially when you left it open there is a tendency to throw the chuck away from the spindle : this is equivalent to ID clamping with low pressure as a result, it means that a chuck left open over the night, will always be in tension some guys do not consider this to be ok ( for example those guys that believe that the mill spindle should be left empty over night ); however, also chips play a major part in this, because as they acumulate inside the chuck, stuff may happen try to move the jaws by hand when the chuck is opened : if at least one jaw is stucked, than this phenomen is present
so far, i am trying to say that what you are seing when clamping ID may have been manifested for a long time when leaving the chuck empty after clamping OD
when you clamp ID, this phenomen becomes stronger, because it manifests at the clamping position, where the pressure is high, because the travel is long : as a result, the tendency to throw the chuck away still exists, but how the chuck is secured with nice screws, than the hydraulic cilinder may slide back, because there is a rigidity issue : so the movement will continue until the pull bar reaches its relaxed position, or something breaks
in my opininion, something is missing ( and if it is breaked, i hope it is easy to fix ), and that's why the pull bar reaches its relaxed position ( this may be an explanation why "there is not much pressure at all to hold the part" ) : but you may still clamp OD without worries
hydra cilinder + pull bar + chuck : must be rigid at both compresion ( clamping OD ) and extension ( clamping ID ); you may not have rigidity at extension ; also, you can not see this if you take down the chuck
hmm : take down the chuck and push the pull bar with the turret : is the hydraulic cilinder going back ? or better, leave the chuck on the spindle : is the hydra cilinder going back when clamping ID ?