Background Data in Nuclear Interactions
Names: Sandhya Ravikumar, Lawrence Free State High School, Lawrence, Kansas
Research Teacher Mentor: James Deane, Ottawa High School, Ottawa, KS
Research Mentor: Prof. Philip Baringer, University of Kansas, Lawrence, KS
Purpose: This research is being conducted in an effort to reduce background data from the data collected from the CMS experiment at the LHC. Whenever experiments are conducted at the LHC, there is always a certain amount of redundant data that detracts from the useful data collected. I am hoping to eventually create a general fit function to eliminate the unnecessary data that comes with nuclear interaction experiments in any plot. Fit functions are essentially exponential functions created to match experimental data and adjust hardware measurements. The fit function serves the purpose of reducing accidental data collected by the sensitive instruments of the CMS pixel shield. It also adjusts the positioning of the machinery itself, as the machines shift from heating and cooling.
Methods: I began by familiarizing myself with the ROOT data analysis software. Anna Kropivnitskaya, a contact at CERN, provided me with Monte Carlo plots and plots from the CMS experiments I was also given a rough fit function program to refine to the graph. After fitting one graph, I moved on to the next, continuing with several graphs. I ran into several technical issues, mainly with software bugs and compatibility issues. ROOT operates best using a Linux or Mac OS, and having a Windows computer, I had to create a bootable USB drive using Ubuntu, a Linux distributor. Ubuntu had issues starting up with my computer, after which I had further issues getting it to work with ROOT. However, all issues were rectified relatively quickly. I had to use an older version of ROOT and use a different method of turning the USB drive into a bootable drive.
Results: I have created fit functions for various graphs, effectively eliminating background data and leaving a cleaner and more functional data set. By changing the fit program’s parameters, values, and operations, I have created several plots that are far more usable than they were in their original state. I achieved the results I hoped for, but I would like to continue working to create a more generalized fit function that would reduce background in any plot. I believe with more time and practice, the “universal” function should come rather easily.
Meaning to Larger Project: This research will help get to the truly useful and necessary data collected by nuclear interaction experiments. Once preliminary data is collected at the LHC, the unneeded and superfluous data can be eliminated by using various fit functions. The remaining data can be utilised easily and properly, without the obstruction of background. If a general fit function is created, the data and fit programs will be even more useful and applicable. The function also can be used to adjust the positioning of the equipment of the CMS tracker itself. As the machine heats and cools with use, it expands and contracts, causing components to shift. A version of the function can be used to measure and readjust the physical positioning of the trackers pixel shield and support beams.
Future Research: I have had success in creating individual fit functions for various plots by adjusting the parameters and values of the function. The logical next step is to create a universal fit function that would eliminate background from any CMS plot. By continuing to create fits for individual graphs, it would be relatively simple to establish a pattern and create a function that would only need to be adjusted slightly for differing graphs.
Eilish Gibson: Undergraduate Student, University of Kansas, Lawrence, KS
Emily Smith: Undergraduate Student, University of Kansas, Lawrence, KS
Anna Kropivnitskaya: Research Associate, University of Kansas, CERN, Geneva, Switzerland