My project this Summer was twofold: the first being working on the CMS inner tracker upgrade, and the second being data analysis on jets.

The goal of the CMS inner tracker upgrade is to increase the tracking capabilities with respect to granularity and detection coverage. The inner tracker is able to detect and track the resultant quarks from proton-proton collisions using silicon chips, which compose the tracker.

Fig 1. This is a diagram of CMS. To the far left, you can see the inner tracker. Each type of charged particle has a different path through the tracker, which allows them to be detected and identified.

These chips, composed of circuits, have been doped, a process which allows the particles passing by the chips to create detectable ionization currents. The team at Kansas State University has been working to test these chips before they are put into the inner tracker. These tests are conducted via a probe station, which physically touches down on the individual chips and conducts tests to see if the chips are functioning. Part of my project has been helping to install the probe station and prepare for the chip testing which will take place this Fall.

Fig 2. This is an example of the probe station that we are using to test the chips.

The other half of my project involves data analysis on jets. After two protons collide, they break up into quarks. Due to the strong force, quarks are attracted to each other and assemble into a detectable jet. We can fit functions to the data from these detected jets in order to recognize which particles created them.

The first function that I tried to use to fit the data is the Crystal Ball function. The Crystal Ball function is a Gaussian with a power-function tail.

Fig 3. The Crystal Ball function

Unfortunately, the Crystal Ball function was not a good fit.

Fig 4. This is an example of the Crystal Ball function fitted to the jets data. Clearly, the fit does not reach the peak.

Ultimately, I was able to fit the data using a function composed of three Gaussians. There are two versions of this function: one in which the parameters which indicate which of the three functions with which to fit the data are manually defined,

Fig 5. This is an example of the triple Gaussian with manual parameters fitted to the jets data.

and one in which the parameters are not

Fig 6. The manually defined parameters lead to a slightly more accurate fit, while the function defined parameters allow the function to be more universal.

While the function composed of three Gaussians worked well for the jets data in the Signal series, I needed to use a different function to fit the other set of data. For this set, I used a Gaussian with an exponential tail.

Fig 7. This is the function composed of a Gaussian and an exponential fitted to jets data

This fit works well on the set of data.

Acknowledgments

Firstly, I would like to thank the National Science Foundation for funding this research. I would also like to thank Dr. Flanders and Dr. Greenman, and Kim Coy. Without their work this REU would not be possible. I would like to thank Dr. Ivanov for being my research advisor. I would like to thank Braden Allmond, Sedrick Weinschenk, and James Natoli for being so welcoming and sharing their office. Finally, I would like to thank the rest of the REU cohort.

Final Presentation