Collider Phenomenology of Composite Higgs Models

Composite Higgs models provide a well-motivated solution to the naturalness problem plaguing the Standard Model of particle physics. We study the collider phenomenology of a class of composite Higgs models with an underlying fermionic description. The Higgs appears as a pseudo Nambu-Goldstone boson (pNGB) originating from the spontaneous breaking of the global symmetry when the strongly interacting composite sector condenses. There will be further pNGBs, however, both colour singlets and coloured states. We collect and derive bounds on the masses of the coloured pNGBs for all popular cosets. For the purely electroweakly charged states we determine simplified model bounds on Drell-Yan pair production with decays into vector bosons or third generation quarks, covering all possible processes within this model class. As a complete model we study the SU(5)/SO(5) coset due to its rich particle content including a doubly charged scalar $S^{++}$. We first apply the simplified model bounds and then perform simulations of the full model to determine bounds on several mass scenarios. We show that this model features the process $pp \to S^{++} S^{--} \to W^+ t\bar b W^- \bar tb$. We design a search proposal for this process employing deep learning techniques to train neural networks to separate signal from background events. The best performing network is a combination of a convolutional neural network using jet images and a multilayer perceptron on kinematic variables. We assess the discovery reach and expected exclusion limit for this process at the high-luminosity LHC. The models also contain colour triplet fermionic resonances, required to give mass to the top quark through the mechanism of partial compositeness. We systematically classify the production and decay channels and derive bounds on pair production for a few sample processes. Finally we turn to the spin-1 resonances, focusing on the coloured states. Among them there is a ubiquitous vector octet $\mathcal V_8$ that mixes with the gluon, allowing for single production. For all models, we calculate the relevant Lagrangian, derive bounds on $\mathcal V_8$ production, and explore which signatures can be expected from vector pair production at future colliders.