|
![[Up]](arrow.marble.up.gif){kind=icon}
|
![[Next]](arrow.marble.right.gif){kind=icon}
Previous page
| Up to the index
| Next page
| |
|
|
|
| Previous page
| Up to the index
| Next page
| |
What if the entire MVD is offset transversely from the beam axis? The official word is that the central magnet, when it is rolled out, and then back in again, could be off by as much as 1/8", or 3mm from its former location. We are mounted onto this magnet. Could the MVD operate if it were 3mm off-axis?
On a previous
slide, , it was shown that the vertices could be found out to many mm,
albeit with bad efficiency. I added a feature to the software so that we can
use the mean beam position, found in a 'calibration pass', as input to the
algorithms, and recover full efficiency.
Below are shown results of this method applied to a set of events that were
generated 2mm off-axis in both x and y (r~1/8").
|
For high MVD hit multiplicity, the 'pseudo-tracking' agorithm is used
(left column),
and for low multiplicity, the 'correlation method' (right column).
Horizontally, the input x,y,z of the event vertex is plotted, and vertically the reconstructed x,y,z is plotted. If the method were perfect, all points would lie on the dashed diagonals. You can see that in this case, the correlation method is not as good as the pseudo tracking method.
|
|
these are the projections of the above 6 plots along the direction of
the diagonals, giving the resolutions in x,y and z.
|
|
These are the same plots as above, with the MVD aligned on the z-axis.
You can see that the correlation method does better in this case.
|
|
these are the projections of the above 6 plots along the direction of
the diagonals, giving the resolutions in x,y and z.
The tails that were seen in the correlation method for the 1/8" offset case above are no longer there.
|