The CERN Experience Part 2: Fahmi Ibrahim

Basically, my assignment at CERN was divided into two sections:

1. Determination of time differences between two hits in the same fiber.

   I installed gAn (gravity analysis) in offline and used it to analyse the data from the fast annihilation cryogenic tracking (FACT) detector. A FACT detector is used to detect the annihilation of antihydrogen in the trapping production. Firstly, I need to understand the digital response of FACT by studying the time over threshold (TOT) distribution which exhibits a two-peaks distribution. I must understand the origin of those two peaks by first looking at the correlation between the time difference of one peak to the other, and the TOT of the second peak. When antihydrogen is trapped inside the detector, at certain times it will collide with the detector’s wall and annihilate to produce two pions. Each pion will pass through the scintillating fiber and it creates an event called hit. The light intensity proportional from radiation to the deposited energy makes it possible to differentiate pions produced by annihilation from other types of particles going through the detector. The light intensity produced inside a scintillating fiber that was hit by either a pion or a number of photons (gamma burst) is estimated.

   One of the first tasks given by my supervisor is to develop the code for the measurement of time differences between two hits in the same fiber (ToT vesus time differences between two hits in the same fiber). This task is quite challenging, where most of the code used for the pointer and reference are in C++ which was not taught in my university. At the same time, I need to refer to the source of gAn code that has already been developed by other developers which I will then use the functions contained within to measure the time differences between two hits in the same fiber after the antihydrogen annihilation process.

2. Determination of a non-biased estimator for the efficiency of FACT fibers.

   From prior studies proposed by Francesco Guatieri, only three- and four-layer tracks were considered, assuming that I have a reasonable sample of all detected tracks and an efficiency value for the fiber. There are several possibilities for N-tracks passing through a certain fiber that it should be seen:

   1. full tracks or tracks missing one fiber that is not the studied one;
   2. tracks missing only the studied fiber;
   3. tracking having our fiber but missing at least two of the others; and
   4. tracks missing the fiber and one or more of the other three.

   However, cases c) and d) are not detectable in the real world case. Hence, the this method produces a biased estimator, which is the inverse of efficiency. The efficiency given by a rational function with an asymptote around 0.66, however this technique will fail for fiber efficiencies below 2/3. Regardless of the presence of the studied fiber, I have chosen to observe only events in which all three fibers are present in order to overcome this.

   Currently, the idea by Philip Hackstock is to reconstruct tracks and to count the hits and misses for each fiber. This way has successfully produced an unbiased estimator with a low uncertainty. Two cases 4/4 and 3/4 are possible fibers firing were considered. Generally, this approach is not limited to FACT; it can be implemented to any pixel-like tracking detector in order to measure its efficiency on the level of its smallest unit. The current method proposed by Philip Hackstock is the best method.

   For my second task, I need to come up with a new method to prove the non-biased estimator for fiber efficiency based on my assumption and mathematical derivations. I need to brainstorm and discuss with my supervisor to ensure all the equations are valid and can be implement in the code of analysis. However, this task is under progress to prove it.

Almost every week, I spent my free time by travelling and visiting places of interest in Switzerland and France. Before leaving to travel for the weekend, I need to get permission from my supervisor and get some recommendations from him about the place that I should visit. We must ask permission from our supervisor because they might need help for about six hours on a shift in the control room to make online analysis during the experiment. So far, I don’t have any shifts on the weekends. Weekend is the most suitable time for traveling because I have enough time to travel within 18 hours and I can manage my time to travel any places that I want to visit. On weekdays, I spend most of my free time after work to prepare my dinner and sometimes I hang out with other summer students in Geneva.

The CSSP programme provides lectures and introductory courses about antimatter at low energy in order to ensure that they get some fundamental knowledge. This is important for me as I am working on AEgIS which requires using FACT detector to study and detect antihydrogen produced in the antiproton trap.

I am grateful that CERN is positively impacting my current studies, especially in learning how to analyse and interpret data, which gives me a better understanding of what kind of process and interactions are happening inside the detector. Through analysis, several questions will arise, which will tell us more about what’s happening inside the detector. In addition, analysis also provides information about the efficiencies of fibers inside the detector.

I must gain more knowledge and learn more on how to analyse data of particle physics using C++ and Python which are quite useful tools. Analysis and interpretation skills that I learnt from CSSP will give me a huge advantage for my future studies on the Master’s level.

Now, I have found my passion in a field that I am interested in: I would like to pursue my interest in nuclear physics for my Master’s Degree next year.