Novel methods for the next generation of high-energy physics event generators

by Max Knobbe
Doctoral thesis
Date of Examination:2024-09-05
Date of issue:2025-03-14
Advisor:Prof. Dr. Steffen Schumann
Referee:Prof. Dr. Steffen Schumann
Referee:Prof. Dr. Frank Siegert
Persistent Address: http://dx.doi.org/10.53846/goediss-11149

 

 

Files in this item

Name:Dissertation_Max_Knobbe.pdf
Size:2.70Mb
Format:PDF

The following license files are associated with this item:

Abstract

English

Monte Carlo event generators are crucial for the success of the LHC physics program and provide accurate theory predictions for various applications. As experiments become more precise and undergo upgrades, event generators also need improvements to provide comparable theoretical predictions. This thesis introduces methods to overcome the limitations of the current generation of event generators and facilitates state-of-the-art simulations for current and future experiments. One major challenge is the computing time required to generate enough events, combined with a rapidly changing computing landscape largely inaccessible to the current generation of event generators. In this thesis, we introduce a novel production-ready leading-order parton-level event generation framework running on modern hardware and showing excellent performance for processes of the highest computational interest. Furthermore, we introduce novel techniques to eliminate numerical inaccuracies relevant for higher-order computations, allowing us to probe phase-space regions currently inaccessible. This is achieved via a set of exact, local modifications to the building blocks necessary to compute scattering amplitudes. Finally, we introduce a method for efficiently incorporating uncertainties introduced by non-perturbative modelling. All of the methods are implemented and tested in realistic scenarios, and novel frameworks are, if necessary, integrated into the current simulation toolchain.
Keywords: Monte Carlo; particle physics; high energy physics; Large Hadron Collider; LHC;
 
Statistik