Available Bachelor and Master Theses
Available Bachelor and Master Theses
The Science for Nuclear Diplomacy group offers Bachelor and Master Thesis projects in the field of nuclear verification and disarmament for physics students. Students will conduct physics simulations and statistical data analyses related to current research projects in the Science for Nuclear Diplomacy Group.
The analyses use Python and C++ for which prior knowledge is not required and can be acquired during the thesis project. The supervision takes place in Darmstadt or Frankfurt. Information on possible topics can be found below. If interested in a topic, contact the respective supervisor listed below for additional information.
Nuclear Archeology
M.Sc.
In nuclear archaeology, we are interested in original irradiation parameters of a nuclear reactor. It is possible to calculate these parameters with knowledge of the reactor design and forensic isotopic composition measurements. However, this analysis entails solving an inverse problem which poses a challenging task. One approach for this problem is the calculation of a posterior distribution of the parameters with the Machine Learning technique "Markov Chain Monte Carlo". This method has proven useful in the past but several challenges occur which might be tackled by applying a special type of neural networks, conditional invertible neural networks, to the problem. The thesis will focus on the implementation of both techniques and a comprehensive comparison of them in order to assess advantages of using either of them in an nuclear archaeology context.
What you will learn:
- Solving inverse physics problems
- Applying advanced machine Learning and Deep Learning techniques
- Working on a High Power Compute cluster
Fabian Unruh
Doctoral Researcher
B.Sc. & M.Sc.
At the heart of every nuclear weapon lies a core of so-called fissile material – usually plutonium produced in a nuclear reactor. Having a good grasp on the operations of said reactors is thus highly relevant from a disarmament perspective. Nuclear Archaeology is a field trying to establish this grasp retroactively, by measuring and analyzing traces left behind by the operation in the reactor itself – e.g. through irradiation of reactor tubes. Investigating this problem does not only include simulating the creation of said traces by using a computer model, but also the extraction of the relevant information and inverse reconstruction of reactor operations. One approach that can be used for this reconstruction is the statistical method of Bayesian inference. After being introduced into the topic and physics and familiarizing yourself with the simulations, you will investigate different aspects, depending on thesis level, of how information available for the reconstruction influences the achieved result.
What you will learn:
- Physics behind reactor operation
- Reactor modelling & simulation
- Solving inverse problems
- Bayesian inference & MCMC-methods
- Using the cluster
Lukas Rademacher
Doctoral researcher
B.Sc.
Nuclear reactors extract energy by inducing fission inside the reactor fuel. Neutrons released in these events travel through the reactor, interact with the surrounding materials, and some of them cause new fission events, driving the chain reaction at the heart of the reactor. A physical description of reactor operations thus consists of an interplay between neutron movement and material changes caused by interactions – an interplay so complex that exact analytical calculations are impossible. Instead, computer models and simulations must be used to derive numerical descriptions of these processes. This thesis will revolve around practical considerations when implementing such simulations. After learning about the underlying physics, both in their real-life and numerically approximated form, and setting up your own test simulation, you will investigate the influence of choices inherent to the simulation on its results. Answering questions such as: ‘How many neutrons should be simulated?’, ‘How fine-grained should the time scale be?’, and more.
What you will learn:
- Physics behind reactor operation
- Reactor modelling & simulation
- Using the cluster
Lukas Rademacher
Doctoral researcher
Cosmic Ray Spectroscopy
M.Sc.
Cosmic rays are an ubiquitous signal in all surface-level detectors, as highly energetic muons with multiple GeVs of energy constantly reach the surface at a rate of ca. 1 Hz per square meter. These cosmic muons are highly penetrating, meaning they are even present in most buildings or lightly shielded areas. As these muons are produced in the upper atmosphere, their characteristics are dependent on meteorological conditions, such as air pressure - potentially leaving an unique "fingerprint" in what is usually discarded or suppressed as background or noise.
In the course of this project, the flux, energy, and angular dependence of the incident cosmic rays with respect to barometric conditions will be investigated using a computer simulation. This study will then evaluate if and how this "fingerprint" can be used to make verification measurements tamper-resistant by correlating them to weather conditions.
What you will learn:
- Cosmic ray physics
- Muography techniques
- Monte Carlo simulations (GEANT4)
- C++/Python
Dr. Yan-Jie Schnellbach
Researcher
M.Sc.
Cosmic ray muons are natural but highly penetrating background radiation. By measuring their occlusion through materials, it is possible to image the interior of structures using a technique called muography. Traditionally, muography either measures changes in the muon rate (one-sided measurements) or scattering of muons inside a structure (two-sided measurements). However, the energy information of muons is usually not used, as the stopping path of muons is usually several metres.
Future concepts for monitoring nuclear power plants or facilities, however, foresee the use of antineutrino detectors, which tend to be several cubic metres in volume. This project will use computer simulations to understand the muon energy spectrum after passing through different materials in nuclear sites, such as concrete, steel or nuclear fuel. By combining this energy measurement with regular muography, it will be determined whether this technique can be used to distinguish different material efficiently and accurately.
What you will learn:
- Cosmic ray physics
- Muography techniques
- Monte Carlo simulations (GEANT4)
- C++/Python
Dr. Yan-Jie Schnellbach
Researcher
Nuclear Material and Warhead Authentication
M.Sc.
In nuclear disarmament verification it is crucial to be able to authenticate nuclear warheads prior to the dismantling stage and thus ascertain that the object to be dismantled is indeed a nuclear warhead and not a deceiving replica. An approach that can be applied for such purpose is using their passive radioactive emissions, in the form of gamma rays or neutrons, as identifying signatures. This requires that the signature obtained from a specific combination of measuring instrument and measured object-type must be unique. Otherwise, the possibility of cheating would arise, in which an inspected actor could replicate the signature with different materials or geometries. This project thus investigates the geometrical effects of varying warhead-detector configurations on the obtained neutron signatures and their uniqueness. Using simulation tools, the neutron multiplicity counting technique is implemented for this purpose.
What you will learn:
- Neutron multiplicity counting technique
- Conducting Monte Carlo simulations (including Geant4)
- Using the cluster
Dr. Luis Pazos Clemens
Researcher
M.Sc.
Nuclear Resonance Fluorescence (NRF) is a process in which a nucleus is excited after absorbing a gamma ray and subsequently de-excites through the emission of one or more resonant gamma rays. The energy of the emitted gamma ray(s) depends on the distribution of the nuclear energy levels and can therefore be used as a signature to identify specific elements and even isotopes. Samples can thus be probed by using a photon beam to excite the nuclei present in them and by detecting the gamma rays thereby scattered or transmitted. Since the energies of the gamma rays are high enough to travel through shielding materials, this technique can be used to non-destructively study the interior composition of objects. In nuclear disarmament and non-proliferation verification this presents the potential to inspect objects such as canisters, warheads or facilities. This project focuses on the development and application of a Monte Carlo simulation framework to study the use of NRF in relevant verification scenarios.
What you will learn:
- Nuclear Resonance Fluorescence technique
- Conducting Monte Carlo simulations (including Geant4),
- Using the cluster
Dr. Luis Pazos Clemens
Researcher
