Figure 8 shows the proposed location of the Roman pots that
will comprise the Forward Proton Detector, where
A refers to the outgoing anti-proton side, P the outgoing proton side,
Q represents the low beta quadrupole
magnets, D the dipole magnets, and S the electrostatic separators.
The dipole spectrometer consists of two Roman pot detectors
(
and 
)
located after the bending dipoles about 57 meters downstream
of the interaction point on the outgoing
arm. The other Roman
pots in the figure are components of the quadrupole spectrometers
discussed in the next section.
The dipole spectrometer pots are located inside the Tevatron
ring in the horizontal plane to detect scattered anti-protons
that have lost a few percent of the original beam momentum.
These are the equivalent positions of the CDF pots
(E-876) [35]
which were added at the end of Run I.
There are no known obstacles to implementing this portion of
the FPD as the optics
are roughly the same at CDF and DØ, and there is space available
at the equivalent location near DØ. It is not possible to instrument
the outgoing proton side with a dipole spectrometer without major
modifications to the accelerator (not being considered).
Figure 8: Placement of Roman pot detectors near
the DØ interaction region. The horizontal scale shows
the distance from the interaction point in meters.
Each of the independent momentum spectrometers
consists of two Roman pots (represented by black rectangles)
in combination with the machine magnets as described
in the text.
Some of the physics topics mentioned in Sec. 2.2
are accessible to the dipole spectrometer, which for CDF had
almost full acceptance for anti-protons with |t|<3 GeV
and 
[12].
These include studies
of diffractive jet production, diffractive W boson production, and pomeron
structure.
CDF has preliminary results using their new (anti-)proton detector
and sees events consistent with diffractive jet production, although there are
background uncertainties due to their limited 
acceptance [12].
They also have a few diffractive W boson
candidate events with a track in their
detector.
A single dipole spectrometer with acceptance characteristics
similar to that of the Run I CDF spectrometer
has two principal limitations:
hard double pomeron exchange cannot be studied using p and \
tags since only the arm is instrumented,
and the acceptance is restricted to a relatively
large region
where the backgrounds from other processes are large
and hard to understand.
To remove these limitations, the FPD discussed in this document is optimized to improve the acceptance and also includes quadrupole spectrometers.