In case of failure scenario, the SSD is not active within the HFT.

To simulate this effect, we ran the bfc chain over Hijing files , and by removing the SSD from the tracking chain ; so instead of SSD+IST+PIXEL1+PIXEL2, there is only IST+PIXEL1+PIXEL2.

I used option "NoSsdIt" (="-SsdIt") in the following chain :    

-ssdfast,upgr15,gstar,trs,Simu,pixFastSim,StiRnd,PixelIT,IstIT,StiPulls,NoSvtIt,NoSsdIt,Idst,IAna,l0,bbcsim,emcY2,EEfs,GeantOut,IdTruth,MakeEvent,miniMcMk,McAna,

CMuDst,tpcI,fcf,ITTF,genvtx,event,analysis,EventQA,tags,Tree,Evout,StarMagField,FieldOn,fzin,McEvOut,clearmem

a.  with single pions (100 pions/event)

The number of tracks reconstructed in TPC should be independent of any requirement of inner detectors in or not

    1. ssd out (Figure 1)

      2. ssd in( Figure 2)

b.  with hijing events (auau@200GeV)

Reconstruction is done with pile-up hits ; use Global Tracks

The 2 next plots show the pT distribution for reco. tracks with :

• pT >0.1
• |eta|<0.1
• TPC Fit > 10

The blach curve is for all reco tracks, eg no cuts on idTruth nor number of Silicon hits ; this should be tracks reconstructed in TPC and independent of the presence of Si detectors.

The read markers are for tracks with TPC Fit > 15 && pixeI_1_idTruth == tpc_idTruth && pixeI_2_idTruth == tpc_idTruth : this should define the numerator for the single track efficiency.

1. no ssd (Figure 3)

 2. full chain (Figure 4)

From these 2 plots. I take the ratio and obtain the following plots :

1. no ssd (Figure 5)

2. full chain (Figure 6)

3. Double ratio

Finally, i take the efficiency for no ssd (Fig. 5 )over efficiency for full system (Fig. 6)