Hello brahm'lets. I've lately been doing some simulations of the feasibility of different physics measurements with three ALICE SDDs - Silicon Drift Detectors, a possible detector upgrade suggested by JJG. Since I'll be going to NBI for the next few weeks to do "other stuff" I present here a brief status report. I've not made this into an analysis note since nothing is final at this stage, but I think what I say below will exclude a few options so that if there is interest I can go into details on other topics... Please send me any and all questions and comments. Plots to back up my claims are readily availiable on request. (I promise not to make a habit out of this quasi-secrecy - I'm just a bit pressed for time ;-) What? ----- For half-gory details of the SDDs, see JJGs presentation at the last brahms coll. meeting: http://www.sdcc.bnl.gov/brahms/private/meetings/Jan2002/jacobsen/ALICE-FMD-SDD-PHOS.ppt Since the file is in such a weird format ;-), here is a brief summary: Subject to approval at the ALICE week in march, BRAHMS has been asked to test a prototype of an ALICE Silicon Drift Detector, intended for use in their Inner Tracker System (ITS). We would recieve three tracker planes (Si wafers) with the following vital statistics: Size (active): ~7.5 x 7.5 cm Pos. resolution in 2 dim.: ~30 micrometers Two-track resolution: ~600 micrometers These are small and light units that should be easily placible somewhere around the nominal vertex (see below) without adding very much to our background. How? ---- I've added the SDD planes to brag and played arond with different positions, particles etc. If anyone wants to have a look, check out my personal brag under ~bjornhs/brag/ on rcf. The extra detector is added under src/global and is called sdd (!). It is a simple virtual box volume (SDDV) which is turnable, and three active planes SDDA, SDDB and SDDC which are positionable within this box. Keywords (for those who do brag-stuff): set_tree sdd o set_parameter sdd angle 90 (angle of the planes, just like the spec. arms) set_parameter sdd first 10 (radial pos in cm of the first plane) set_parameter sdd space 2.5 (spacing in cm between consecutive planes) Also, I've added two extra particles: Nr. 42: D+, decaying to K-Pi+Pi+ Nr. 43: D0, decaying to K-Pi+ Results/status -------------- There are two main suggested physics topics for the SDDs - open charm and forward rapidity lambdas. I've looked a bit at both: Open charm ---------- Several theorists have made half-hearted attempts at predicting charm production at RHIC energies, mostly focusing on J/Psi production. They mostly agree that at 200GeV we can expect on the order of 20 ccbar-pairs per central collision. See e.g. hep-ph/0012292. Where relevant, I've also made the assumption that roughly a third of these end up as D0/anti-D0s (see hep-ph/9904441). These assumptions may, of course, be wrong by seveal orders... 1) Charged D-mesons. D+ decays to K-Pi+Pi+ with a b.r. of 9% (similar for D-). To identify this decay we would have to have a good PID for two of the daughters and see the third in the SDDs, and this I'm sorry to way we simply cannot do... Our acceptance for one kaon and one pion (or two pions) hitting the TOFW after going through the MRS is virtually 0 - I saw one such event for 100k D+'es at midrap. Hence I conlude that BRAHMS will not measure charged D's next year. (Not surprising maybe, but now it's been checked ;-) 2) D0s D0, decying to K-Pi+ with a b.r. of 3%, is what we can hope to measure. This decay has a typical opening angle of around 60deg, so the strategy would be to identify a kaon in the MRS and then place the SDDs at some optimal angle relative to this and treat all tracks as pion candidates. We can then find an invariant mass spectrum and hopefully select D0s. To find the expected number of reconstructed D0s in 1M central events, I do the following: a) Find out where the decay products actually go. I define a BIG active volume (a 2x2 meter SDD... Don't you wish...) to catch all relevant particles, look at the correlation of K and Pi angles and find the "best" positioning of MRS and SDD. For eta=0 D0s with 0.2<p<3 GeV/c this turns out to be MRS: 66 deg SDD: 123 deg b) Find my acceptance for D0s at this angle setting. I throw particles through all of phase space and see which ones are reconstructible (i.e. kaon hit in TOFW, pion hits in all SDD planes). This gives me roughly -0.33 < eta < 0.07 (not cent. on 0, because of magn. field?) |phi| < 8 deg c) I can now find the total number of D0s in this acceptance: Assume 1M central events w. |z| < 5 cm >From hep-ph/0012292 and hep-ph/9904441 this gives me 1M * 20 ccbar pairs * 0.3 = ~6M <D0+D0bar> 6M * 0.5 = 3M D0s in all of phase space. (My 1M events are for one field setting, so I assume I only see one species. This is to be conservative...) For now I assume a flat distribution in momentum, which is of course not right. However, it will not make a huge impact on this analysis - it becomes more important when determining realistic "best" angles. I assume a gaussian rapidity distribution of D0s (sigma of 1.7), and find that for -0.33 < |eta| < 0.07 we see ~10 % of the particles. (+-3%, dep. on wether you believe the distr. has a plateau or not... My estimate is for a quite small plateau, a pure gaussian would give a higher percentage.) Combined with my |phi| < 8 deg constraint I see 3M * 0.1 * 16/360 = 13k D0s within my acceptance for 1M cental events. Of these 3% decay into our detectible channel, so we will see 13k * 0.03 = 390 detectible D0s in 1M central events. Not very uplifting... d) Finally, I of course don't see all the detectible D0s within this region. Throwing 500k D0s into this phase-space only 127 (!) of these are reconstructible (again this means kaon with hit on TOFW and pion with hits in all SDD planes) giving me an expected number of SEEN D0s: 390 * 127/500k = 0.1 So you see the problem... While I see a kaon in TOFW for every 100 decays and a pion in all SDD planes for every 15th decay, the combined acceptance is just too small to give reasonable numbers. I'm a bit sceptical of my geant simulation since it seems to contain some events without particles(!) and I'm sure that the angles and positions of the detectors can be even better tuned, but even if this number were to go up by many orders of magnitude we will have problems seeing it since the background from non-charmed processes will be enourmous. So - unless someone can tell me that I'm very wrong on one or more counts (and please do if I am!!!) I'm afraid we simply don't have the acceptance to do open charm measurements with MRS + SDDs... I know there are a lot of weak links in my chain, but I don't see how doing the analysis thorougly at all points could raise the numbers enough. (A shame... It would have been a nice contribution from brahms, but I guess I'm not really surprised at the answer.) Lambdas at forward rapidities ----------------------------- I've not yet done this analysis to the level of detail shown above (due to some problems with brag...) but I can still give a general overview: The hope is that by placing the SDDs some short distance in front of D1 we can * track a proton through FFS (or the full FS) * find an optimal angle relative to the FFS acceptance where the pion would go if the proton came from a lambda decay * assume each sdd track is a pion * find hypothetical secondary vertex, momenta, invariant mass etc. I have made some tests with the SDD planes at radial distances from the nominal vertex of 60, 70 and 80 cm. A full HIJING simulation shows that the first plane will be hit by an average of 30 particles per event and the last plane by 15-20, so we should expect at most ~50 tracks per event. This occupancy is pretty low, so we can assume the nominal position resolution of 30 microns (i.e. few two-track hits). This in turn gives us a very good track position and direction, so the uncertainty in the decay position will come from swimming the proton back through D1. This is where I got some BRAG problems so I don't yet have acceptance tests and the likes, but a quick test shows me that for a given SDD position it is possible to see a pion track for every third proton in H1 coming from a lambda. Since the predictions of Lam/p production at 200GeV range from 0.1 to 1 this does not seem all that bad. Of course we will have lots of background from other particles, but with the angle resolution of the SDDs we should be able to exclude primary paricles to a certain degree (perhaps by doing a vertex determination from the SDDs? ;-) There's still some work to be done here checking possible yields, resolutions etc., but at least this measurement is not yet excluded. As of yet it actually looks a bit promising to a young and (over?)confident experimentalist ;-) Of course this could also be extended to the MRS if we want midrap. lambdas instead, but I kind of expect that if we are to add this detector we want to measure something as unique as possible? Conclusion ---------- Open charm: No. Sorry, JJG.. Lambdas: Maybe. Looks good, but the yields are not yet estimated. If this is still interesting enough to go for the SDDs I can probably have some more details ready by the ITS meeting. Now for a day off before I go to NBI. Ping :-) -- Bjorn H. Samset Phone: 22856465/92051998 PhD student, heavy ion physics Adr: Schouterrassen 6 Inst. of Physics, University of Oslo 0573 Oslo \|/ ----------------------------> -*- <----------------------------- /|\
This archive was generated by hypermail 2b30 : Sat Feb 23 2002 - 11:09:43 EST