A possible R-phi geometry "a la VELO" for FTD

Description of R-sensors and phi-sensors

---

A idea is a set of 7 pairs of sensors in r and phi:

• The r-sensors will be placed in a concentric way with a constant separation of 38 microm (similar to minimum value of VELO design of LHCb).
• The phi-sensors are a little more complex. We will suppose silicon strip placed in radial way. As the pitch will be larger with r, we decided implement some “stages” in r (similar to VELO). In each stage, the pith will be change from 38 to the maximum value. The number of strips will be the maximum (filling the whole 360 degrees), and leaving always the minimum pitch of 38 microm.

* phi_sensors_1_3.jpg:

• phi_sensors_4_7.jpg:
  

• r_sensor_4.jpg:
  

Phi Parameters:

---

No attempt was made to decrease minimum pitch with respect to the LHCb 38 micron value.

• The máximum pitch will be 2.5*PITCHmin for 1,3,5,6,7 disks:
  • *PITCHmin*=38 micras
  • *PITCHmax*=2.5*38=95 micras
• The máximum pitch will be 2*PITCHmin for 2,4 disks:
  • *PITCHmin*=38 micras
  • *PITCHmax*=2.5*38=76 micras

Resolution in phi

---

For the 7 disks, the number of “stages” will be: 2,2,2,2,1,1,1, i.e, two stages in the 4 first disks and only one stage in the rest. In this way, the 4 first stations are much more uniform. In each of this stages, the phi resolution will be constant, (with ALPHA=1/sqrt(12)).

Some Remarks

---

• 1) we think the "VELO" geometry cannot really compete with any reasonable pixel option, for the first 3 disks. We clearly advocate, as a group, for a pixel option for these layers, and intend to pursue R&D along those lines. Pixels are superior (in material, resolution, pattern recognition), so performing a relative comparison "VELO vs. pixels" is probably not worth, at software level.

 

• 2) if however double-sided microstrip options (XY+stereo angles) are discussed as baseline for the first 3 disks, then a more thorough comparison "XY vs. VELO" deserves to be made on equal software foot. In the VELO, the ghost rejection is achieved by the "dog-leg" technique, easily implementable in the software, and viable for construction.

 

• 3) for the farmost 4 rings, a microstrip option is rather standard, and really using R-phi or normal-stereo should not give essentially different results.

  In summary: we do not advocate "VELO" for the first 3 disks, but in a microstrip-based scenario the software should desirably evaluate both options on equal foot, including backgrounds. We may have still time, if so decided.

-- Main.iglesias - 10 Feb 2009

to top