Update on BES light nuclei w.j. llope I Analysis Mtg, 7/16/2013, Purdue physics directions: Tree, analysis, PID, spectra, and ratios codes all in place… Still improving corrections…. Since results shown at QM2012 Added 3 He, anti- 3 He, and α Five track cuts sets for systematics: none, loose, tight, cross-terms Detailed event QA (a la bulkcorr) Feeddown – added more events, centrality dependence TOF match eff – improving systematic treatment Extend/Simplify implementation of corrections in analysis code Recent work Finalized absorption corrections, now includes nuclei too Added some new embedding sets to the table Significant improvement in TOF Match Eff approach

ppbarddbarttbarhhbar SL10h_emb SL10h_emb SL10k_emb SL10k_emb SL10k_emb 182  SL10k_emb SL10k_emb SL10k_emb 185  SL1hk_emb SL10h_emb SL10k_emb SL10k_emb SL10k_emb 182  SL10k_emb SL10k_emb SL10k_emb 185  SL11d_emb SL11d_emb SL11d_emb SL11d_emb 253  SL11d_emb SL11d_emb SL11d_emb SL11d_emb 253  SL10k_emb 186 (jetcorr) SL10k_emb 187 (jetcorr) SL10k_emb 183  SL10k_emb SL10k_emb 182  SL10k_emb SL10k_emb SL10k_emb 185  SL10k_emb 162 (jetcorr) SL10k_emb 166 (jetcorr) SL10k_emb SL10k_emb SL10k_emb 163  SL10k_emb SL10k_emb SL10k_emb 165  200 (r10 ) SL10k_emb 162 (jetcorr) SL10k_emb 166 (jetcorr) SL10k_emb SL10k_emb SL10k_emb 163  SL10k_emb SL10k_emb SL10k_emb 165  200 (r11 ) RM-matchedRM-matched &/or Abs-scaled scale by RM, Abs, and r11/r10 …scale up by 4-9% depending on cuts Reconstruction Eff from Embedding

He3 scaling not great even for |y|<0.1, and there’s a dip at pt~1.75 Have been discussing with Gene recently... pt< <pt< <pt< <pt<3.0 40<RM<80 160<RM< <RM< <RM<500 RefMult Eff pt He3 “dip

He3 for |y|<0.1 & 0.1<|y|<0.3 – now fitting explicitly RM<40 520<RM<560

Absorption... Now with all brand new embedding for p and pbar at all root-s… eff(p)/eff(pbar) PTPT Absorption is independent of root-s (as it must be) factor is o.k., but 0.45 is better

Absorption For the antinucleus half of my analyses, all I’d have to do is the usual xsec-scaling approach but use the factor of 0.45 instead of But…. 1. What am I supposed to do about the nucleus absorption? 2. is this pbar to Abar xsec-scaling STAR uses for antinucleus absorption consistent with more recently measured cross-sections? These questions can be answered with G EANT 4. G EANT 4 has extensively validated cross-sections for light (anti)nuclei based on experimental data… Google ‘geant4 validation’ etc…See also e.g. Do a G EANT 3 simulation with a STAR mock-up Do a G EANT 4 simulation with the same STAR mock-up. And then compare. As I am only interested in the absorption itself, the STAR mock-up used in both only needs to be good, not perfect ( STARSIM )... This can be tested (in two ways)! Embedding still handles all the real-life reconstruction aspects (resn, merging, cuts etc)

STAR Geometry from STARSIM, Y2010/11 “final” STARSIM X 0 values used to define the G3 & G4 geometries… First test is to use the “material” functions in G3 and G4 to show that the expected X 0 values are reproduced.. ✔ pipe (Be) C-fiber support rods IFC (Al/Cu) P10 OFC (Al/Cu) G4 code with visualization

Throw 10 particles (p, pbar, d, dbar, t, tbar, h, hbar, a, abar) flat P T distribution |η|<0.5, or, as in my analysis, |y|<0.1, 0.1<|y|<0.3, 0.3<|y|<0.5 Wasted some time comparing different G4 physics lists. All but one were nonsensical. Then I learned FTFP_BERT is recommended. And this is the one that made sense! Let G3 and G4 swim these tracks, and deposit hits in a “TPC” and a “perfect TOF” If the particle exits the TPC (or makes hits the “TOF” layer) this track is “not absorbed” Form the ratios of “not absorbed”/“generated” vs. P T for each particle thrown… This defines an “not absorbed efficiency” -- This is precisely what I need. Make ratios of these efficiencies to compare to the Struck approach… G4 events were generated on the DAVINCI cluster at RICE. …several 10’s of millions of events were generated… G3 events were generated on my laptop (10M events in ~5 hours)

Second key test… pbar absorption w.r.t. protons from G3 and G4 G3 G4 Struck-style p/pbar parameterization using factor 0.45… Struck abs(dbar) G3 and G4 mock-up geometries reproduce the expected p/pbar absorption from embedding… Important. PTPT Struck abs(tbar,hbar) Struck abs(abar)

Eff(A)/Eff(Abar) G3 reproduces the expected values of Eff(p)/Eff(pbar), as expected G3 gives Eff(A)/Eff(Abar) = 1 independent of PT, which is nonsense, as expected G4 shows that Eff(Abar) < Eff(A), as expected Eff(p)/Eff(pbar) d/dbar t/tbar h/hbar a/abar PTPT G3 G4

Now forming the ratios that I need: Eff(p)/Eff(X)… Green lines are a pol8 fit and define the absorption corrections I will use…

Improved TOF Match Efficiencies (this is an animation – full screen to view)

…Summary Expanded table of embedding data with some new requests Absorption corrections for antinuclei and nuclei now finalized using Geant4 Significant improvements in TOF Match Efficiency calculations …Time scale… To be honest, this analysis isn’t my highest priority, but I am keeping codes running… Two of the 5 cuts sets have been processed so far Will have some results to QA in a couple of weeks …“new” physics directions… - use of light nucleus ratios significantly reduce (  B,T) uncertainties in stat. model fits… -d/p ratios possibly sensitive to spinodal decomposition near 1 st order phase transition