I attach here three plots showing the Local and Cartesian positions for the ETMY HEPI for the last 72 hours.  I conclude that, at least in this instance, the high frequency excitations (100-1000Hz) which ran this morning, destresses the platform and exhibits strains of significant magnitude.

The first plot is the four vertical Inductive Position Sensors showing the zeroing of the IPS after Alignment approval Tuesday afternoon. When Mitchell pulls ACB elements Wednesday Morning, there is vertical shift of 100s of nm; as we expected, not too much to worry about.  Toward the end of the plot, the start of TFs beginning at 500 to 1000Hz and then continuing with 100to 500Hz as well shows small steps.

The second plot however shows the four horizontal IPS and much larger steps in the IPS readouts at the start of the TFs.  The third plot are the trends of the Cartesian channels.  The Wednesday AM Matrix changes I made are very insignificant but not so the horizontal shifts from the TFs.  The 0.3mm translations of X & Y are ignorable but the Rz of -15urads may not be. With Jason's alignment of ETMY indicating a 20urad CCW error, and if we've all got our signs correct, the -15urads should actually put us closer to zero error.

Notice that the H2 IPS had the largest shift.  I attempt to push it back and was successful with a 1000ct local offset but it went back as soon as I removed the offset.  In the last plot attached you can see all the position senors responding to the stopping of the excitations this morning and then to  the offset and all pretty much returning to their previous values when the offset is removed.

So I would suggest that a destressing exercize of the HEPI may be a valuable step in the alignment process and of course an explicit zeroing (or at least a logging) of the IPS value are in order right after alignment, something we haven't always done.  This also stresses the importance of getting the posirtion loops turned on.