News from the GRETINA TRACKING group New in tracking: We can track Geant4 (G4) simulated data Using both the (1) original AGATA code simulations (with GRETINA geometry) and the (2) US adapted version of the Geant4 code. We can show that they give the same results We can show the GT and AGATA tracking is about the same

Lets start with some experimental 60Co data from MSU Not as close packed as we would like; but.. Not background subtracted

Typical tracked spectrum, FOM<0.8 Sum of central contact energies Tracked spectrum

FOM spectrum, a measure of how well the interaction angles and interaction energies follow the Compton scattering formula Single hits, FOM==0 Single interaction over range Over flow ‘mostly bad guys’ ‘ mostly good guys ’ Typical FOM cut

We have to make the usual P/T vs Photo peak efficiency compromise Clearly we would like to do better... Not a calibrated source, so this axis is arbitrary Fom<0.3 Fom<1.0 No Fom cut Using 10 keV low cut

We will use Geant4 to simulate 60Co to find an answer Is it the data or the tracking that has a problem? So, the question is:

THE NEW THING: Tracking Geant 4 data Use G4toMode2.c to ‘translate’ G4 data to Mode2 format (as if it came from decomposition task, timestamps and all) Tracking code stays the same!! Also ‘packing’ the hits within ~5mm + smearing positions and energies + separate into crystal hits (as decomp task)

Geant4 simulation, NSCL geometry, 60Co This data was supplied by Amel Korichi and is from the AGATA Geant4 code

Typical G4 tracked spectrum, FOM<0.8 Looks cleaner

Tracked Geant4 data from AGATA G4 code It seems the GT tracking code does significantly better with G4 simulated data And just as well as the AGATA tracking code AGATAtracking

Part way conclusions: Using simulated data, we have shown that the GT tracking code is working fine and it works as well as the AGATA tracking code We think the problem with tracking real data is that the input from the decomposition task needs to be improved....some evidence?:

Radius spectrum G4 Simulations Real data frontback Decomp task prefers to put interaction points at the segment boundaries Special worldmap

Narrow radius SMAP, 20.9<r<21.1 Decomp seems to favor center of crystal

Conclusions Now have G4 data to track, really tests the tracking code in a controlled fashion! Managed to compare to AGATA tracking!! We can optimize tracking parameters Still have problems with real data because the decomposed data favors segment boundaries Have a new G4 simulation group: tl, Lew Riley (Ursinus), Con Beausang/Keegan Sherman (Richmond), Amel Korichi (IN2P3/ANL), Augusto Macchiavelli(LBNL), Vikram Prasher(UML) You can download the GT TRACKING code from: “ Any help is welcome! New people in the US tracking team: Edana Merchan (UML) [will add pair interactions], Ragner Stroberg (MSU).

Extra extra extra....

Close pack geometry at ANL, what can we expect? Geometry Suggested By Shaofei Zhu

Geant4 simulations by Amel Korichi and the AGATA team; GT tracking, No Single hits

G4 simulations:..Full GT, ideal GT

Important test of tracking codes ability to find the first interaction point for proper Doppler correction Use: 850 keV line emitted at 40% of v/c Use: current MSU geometry Try G4 simulation data with high v/c

worldmap NSCL geometry, but rotated in azimuth angle, but that should not matter much (we are working on the problem) Notice the Relativistic aberration Simulation Code from Lew Riley

G4 spectrum and tracked spectrum BLUE: spectrum from G4, with position and energy spearing RED: tracked spectrum, FOM<0.8, excluding single hits In real life we don’t do so well... Geant4Tracked Photo eff9.6%7.7% P/T FWHM1.18%1.26%

todo O check the packing of G4 data O G4: check/improve smearing of positions O G4: check/improve energy resolution O try to get single hits into play O add pair creation to tracking code O see if we can quantify the decomposition tasks preference for interfaces O different target chambers, beam pipe in G4 simulations

Polar angle spectrum Simulations Real data

Excluding single hits: to do or not to do? If interest is only at high energy, cut them Is it fair to not include single hits when we quote P/T? YES, we have a new tracking detector array and using new information from it to cleanup spectra is fair and encouraged! We may be able to get single hits in play in the tracking by giving them a finite FOM We have a energy dependent max range for cutoffs of single hits, that does help!

Radius spectrum for individual crystals

Depth, r1<r<r2, selected world map TBD, think perhaps I can catch a ‘strange’ worldmap distribution by selecting a radius cut where things go wrong, ie. at the boundaries between segments Did not work so far

IDEAL GRETA == 30 modules

Real GRETA == 21 modules

The P/T vs Photo efficiency for different FOM cuts, 1 MeV gamma ray No FOM cut FOM<0.1 FOM<0.2 FOM<0.8 THE USUAL TRADEOFF G4 simulated data tracked, including single hits

Last December 2012 run geometry

Augusto stuff From AOM