Nuclear Physics Forum

The LBNL Nuclear Physics Forum (NP Forum) is a weekly (more or less) seminar series covering topics in low-energy nuclear science, including nuclear structure, nuclear chemistry and heavy element research, applied physics and much more. The NP Forum series involves the nuclear structure group, heavy element research group and research groups from the UC Berkeley campus. NP Forums are typically scheduled for 11am on Thursdays in the Building 88 2nd floor conference room area, with refreshments (coffee + cookies) available from 10:45am.

2020-2022 Pandemic NP Forum Series

Upcoming Forums...

Friday, April 15, 2022, 9:00 AM

Zoom ID: 961 4839 3619 Password: 714891

Dr. Yonggi Park, Texas A&M University - Corpus Christi

"Recurrent Neural Networks for Dynamical Systems: Applications to Ordinary Differential Equations, Collective Motion, and Hydrological Modeling"

Classical methods of solving spatiotemporal dynamical systems include statistical approaches such as autoregressive integrated moving average, which assume linear and stationary relationships between systems' previous outputs. Development and implementation of linear methods are relatively simple, but they often do not capture non-linear relationships in the data. Thus, artificial neural networks (ANNs) are receiving attention from researchers in analyzing and forecasting dynamical systems. Recurrent neural networks (RNN), derived from feed-forward ANNs, use internal memory to process variable-length sequences of inputs.

This allows RNNs to applicable for finding solutions for a vast variety of problems in spatiotemporal dynamical systems. Thus, in this paper, we utilize RNNs to treat some specific issues associated with dynamical systems.

Specifically, we analyze the performance of RNNs applied to three tasks: reconstruction of correct Lorenz solutions for a system with a formulation error, reconstruction of corrupted collective motion trajectories, and forecasting of streamflow time series possessing spikes, representing three fields, namely, ordinary differential equations, collective motion, and hydrological modeling, respectively. We train and test RNNs uniquely in each task to demonstrate the broad applicability of RNNs in reconstruction and forecasting the dynamics of dynamical systems.

Past Forums...

Wednesday, February 20, 2019, 11:00 AM

Prof. Michael Block, GSI Helmholtzzentrum fur Schwerionenforschung

"Precision Measurements in the Region of the Heaviest Elements"

Thursday, February 28, 2019, 11:00 AM

Dr. Boris Pritychenko, NNDC, Brookhaven National Laboratory

"New Approach for Determination of Stellar Nucleosynthesis Abundances"

Thursday, March 7, 2019, 11:00 AM

Prof. Kyle Leach, Colorado School of Mines

"The Current Status of Vud and the Top-Row CKM Unitarity Test: How Did We Get Here?"

Thursday, March 21, 2019, 11:00 AM

Dr. Darren Bleuel, Lawrence Livermore National Laboratory

"Neutrons at the 88-Inch Cyclotron"

Thursday, April 18, 2019, 11:00 AM

Prof. Ritu Kanungo, St. Mary's University

"Reaction spectroscopy unveils new features in rare isotopes"

Thursday, May 2, 2019, 11:00 AM

Dr. Alex Zylstra, Lawrence Livermore National Laboratory

"Studying Nuclear Astrophysics with Laser-Generated High-Energy-Density Plasmas"

Friday, May 17, 2019, 11:00 AM

Prof. Umesh Garg, University of Notre Dame

"Nuclear Incompressibility: How Collective Excitation Modes of a Nucleus Characterize Astrophysical Processes"

Thursday, May 23, 2019, 10:00 AM

Dr. Jack Henderson, Lawrence Livermore National Laboratory

"Nuclear Collectivity and Deformation"

Thursday, November 14, 2019, 11:00 AM

Prof. Alfredo Poves

"Shape Coexistence and Islands of Inversion"

Tuesday, December 10, 2019, 11:00 AM

Dr. Bethany Goldblum, University of California Berkeley

"Nuclear Science for National Security Applications"

Thursday, December 12, 2019, 10:30 AM

Jonas Karthein, CERN & Universitat Heidelberg, Germany

"High-precision mass measurements for nuclear structure and astrophysics studies"

Monday, December 16, 2019, 11:00 AM

Dr. Fnu Krishichayan, TUNL and Duke University

"Nuclear data activities at TUNL relevant to NNSA interests and isotope production"

Thursday, August 27, 2020

11:00 AM

ZOOM ID: 932 0238 3806; Password: NPForum27

Dr. Michael Skulski

Department of Physics, University of Notre Dame

"Development of 129I AMS at the NSL for Measurements of the Great Lakes Region"

The radionuclide 129I, with a half-life of 15.7 × 106 years, is produced naturally on Earth in the atmosphere through cosmic-ray-induced reactions on xenon isotopes, as well as through spontaneous fission of terrestrial uranium. These contributions to the 129I content on Earth pale in comparison, however, to the amount that has been released by spent nuclear fuel reprocessing centers, having contributed approximately 60 times the natural content. 129I can move very efficiently through the environment, and because of its “point-like” releases from reprocessing centers, 129I has the potential to be a powerful environmental tracer. For this capability to be realized, the 129I distribution throughout the environment needs to be established – previous measurements have primarily focused on nuclear facilities and nuclear disaster sites, leaving a large area of the globe unmeasured. This inspired the Collon group of the Nuclear Science Laboratory (NSL) of the University of Notre Dame to contribute data on 129I levels, specifically in the Great Lakes region. Water samples were collected from Lake Michigan and rivers throughout Michigan and analyzed for their 129I content through the technique of Accelerator Mass Spectrometry (AMS) for separation from the primary interference, 127I, using the time-of-flight technique. The AMS program at the NSL, the development of 129I AMS at the NSL, results of the water survey, and future work on 129I AMS and other AMS projects will be discussed.

Wednesday, May 5, 2021

9:00 AM

ZOOM ID: 948 3122 3509; Password: Nuclei

Mr. Axel Frotscher

Institut fur Kernphysik, TU Darmstadt

"The (p,3p) two-proton removal from neutron-rich nuclei and the development of the STRASSE tracker"

The knockout of nucleons from nuclei is a powerful tool to investigate nuclear structure. It was observed in several occurrences that different final states in a nucleus are populated when produced from one nucleon knockout (p,2p) or from two nucleon knockout (p,3p). The understanding of the latter could provide a new tool for nuclear spectroscopy.

Two experimental campaigns conducted at the RIBF in RIKEN, Japan, were investigated. The scattered proton angular distribution from several neutron-rich medium-mass nuclei were analysed. The radioactive nuclei were impinging onto a 100-mm long liquid hydrogen target. The protons issued from the reaction were measured with the MINOS time-projection chamber surrounding the target, giving access for the first time to angular correlations of the three protons in the final state. The obtained proton distributions were benchmarked against kinematical models assuming three different reaction mechanisms.

To improve the vertex-resolution and enable missing-mass spectroscopy, STRASSE is being developed. The silicon tracker array is configured in a hexagonal shape in two layers around a liquid H2 target. In the future STRASSE with its excellent vertex resolution can be combined with other γ-detector arrays (such as AGATA or GRETA) to perform high resolution in-beam γ-ray spectroscopy using low intensity fast beams.

Wednesday, May 19, 2021

9:00 AM

ZOOM ID: 938 6185 8697; Password: Nucleus

Mr. Anton Samark-Roth

Department of Physics, Lund University

"Spectroscopy along Decay Chains of Element 114, Flerovium"

In the wake of the discovery of superheavy elements, nuclear spectroscopy experiments aim at providing anchor points at the uppermost end of the nuclear chart for nuclear structure theory, which otherwise had to solely rely on extrapolations. In two runs in 2019 and 2020, such a nuclear spectroscopy experiment was conducted to study α-decay chains stemming from isotopes of flerovium (element Z = 114).

The experiment conducted at the GSI Helmholtzzentrum für Schwerionenforschung, Darmstadt, Germany, employed an upgraded TASISpec decay station placed behind the gas-filled separator TASCA. The fusion-evaporation reactions 48Ca+242Pu and 48Ca+244Pu provided a total of 32 flerovium-candidate decay chains in effectively 18 days of beam time. Two and eleven decay chains were firmly assigned to even-even 286Fl and 288Fl isotopes, respectively. The – admittedly unexpected – observations include (i) an excited 0+ state at 0.62(4) MeV excitation energy in 282Cn, and (ii) a Qα=9.46(1) MeV decay branch (1 out of 51) from 284Cn into 280Ds [1]. Both observations indicate that there is hardly any shell gap at proton number Z = 114 - at least not at neutron numbers N ≈ 172-174. The remaining decay chains stemming from 289Fl indicate the presence of α-decay fine structure as has been theoretically predicted for odd-A Fl-decay chains. This is the focus of an ongoing analysis [2].

[1] A. Såmark-Roth et al., Phys. Rev. Lett., 126, 032503 (2021).

[2] D.M. Cox et al., to be submitted to Phys. Rev. Lett.

Wednesday, June 9, 2021

8:00 AM

ZOOM ID: 974 0263 1730; Password: Nucleus

Ms. Carlotta Porzio

University of Milan

"Configuration mixing investigation in germanium isotopes through measurement of E0 transition strengths"

Experimental and theoretical studies of the germanium isotopes point increasingly toward the emergence of triaxiality, configuration mixing, and shape coexistence. Studies of the E0 strengths, which can provide a direct measure of the amount of configuration mixing, are lacking. Thus, determining E0 transition strengths is essential for an understanding of the evolution of structures in the Ge isotopes.

Beta-decay experiments populating excited states in the 72,74,76,78Ge isotopes were performed at the TRIUMF-ISAC radioactive beam facility. The GRIFFIN spectrometer combined with the PACES silicon array enabled us to perform both gamma-ray and electron spectroscopic investigations, to measure E0 strengths between states of J>0. Preliminary results from this study will be discussed.

Thursday, July 8, 2021

8:00 AM

ZOOM ID: 970 9765 0085; Password: Nucleus

Dr. Özge Aktas

KTH Royal Institute of Technology, Stockholm, Sweden

Gamma-Ray Spectroscopy of Neutron-Rich 111Mo, 85,87Ge and Self-Conjugate 88Ru Far From Stability

Excited states in the neutron-rich nucleus 111Mo were studied following nucleon knockout reactions. 7 γ-ray transitions were identified for the first time using the DALI2 and MINOS detector systems at the BigRIPS and ZeroDegree electromagnetic fragment separator at the RIBF, RIKEN, Japan. Total Routhian surface (TRS) and Particle-Plus-Rotor calculations have been performed to investigate the predicted shape coexistence and its effect on the structure of nuclei in this region of the nuclear chart. Following the results of the calculations, theoretical level schemes are proposed for positive and negative parity states and compared with the experimental findings.

Gamma-ray transitions have been identified for the first time in the extremely neutron-rich (N = Z + 25) nucleus 87Ge following nucleon knockout reactions studied at the RIBF, RIKEN, Japan. Previously unknown γ-ray transitions between excited states in 85Ge were also observed and placed in a tentative level scheme. The results are compared with large-scale shell model (LSSM) calculations and potential energy surface calculations based on the total Routhian surface formalism.

The neutron-deficient self-conjugate (N=Z) nucleus 88Ru was populated via the heavy ion fusion evaporation reaction 54Fe(36Ar, 2n)88Ru in an experiment performed at the GANIL accelerator laboratory in France. Using the AGATA γ-ray spectrometer together with ancillary detectors, prompt γ−γ−2n coincidence and charge particle anticoincidence analysis was performed for the low-lying energy spectrum of 88Ru. The results confirm the previously assigned γ-ray cascade and extend it to the 14+ level. The level scheme is consistent with a deformed rotational system. However, the rotational frequency of the alignment of the valence nucleons has a significantly higher value than what is predicted by theoretical calculations performed without isoscalar neutron-proton pairing. By including isoscalar pairing, an agreement is obtained with the experimentally observed delayed rotational alignment.

Friday, July 9, 2021

8:00 AM

ZOOM ID: 967 4554 8394; Password: Nucleus

Dr. Xiaoyu Liu

KTH Royal Institute of Technology, Stockholm, Sweden

Experimental Studies of the Neutron Deficient Atomic Nuclei 88Ru and 87Tc, and the Diagonalization of the Pairing Hamiltonian

The low-lying excited states in the N=Z nucleus 88Ru and N=Z+1 nucleus 87Tc were studied via fusion-evaporation reactions induced by 115 MeV 36Ar ions bombarding 6 mg/cm2 thick metallic 54Fe target foils at the Grand Accélérateur National d’Ions Lourds (GANIL), Caen, France. The prompt γγ-neutron and charged-particle coincidences from the de-excitation of the reactions were measured by the AGATA γ-ray spectrometer coupled to the auxiliary NEDA, Neutron Wall, and DIAMANT detector arrays. The results for 88Ru confirmed and extended the previous level scheme to a tentative (14+) state. The constructed level structure exhibits a moderately deformed rotational behavior but shows a band crossing at a significantly higher rotational frequency compared with neighboring nuclei with N>Z. The observation of a “delayed” rotation alignment in the deformed N=Z nucleus is consistent with theoretical predictions related to the existence of strong isoscalar neutron-proton pair condensate. The yrast band in 87Tc from the (9/2+) state to the (33/2+) state was established based on six mutually coincident γ-ray transitions. The constructed yrast band exhibits a sharp back bending at ω ≈ 0.50 MeV. In the odd-A isotonic chains around N = 44, approaching the N = Z line, the observed decrease in alignment frequency and increase in alignment sharpness were proposed as an effect of the enhanced isoscalar neutron-proton interactions. In addition to the yrast band in 87Tc, six new mutually coincident γ-ray transitions were identified by comparing the γ-ray intensities in the spectra gated under different reaction channel selection conditions. The constructed level scheme was compared with the shell model and TRS calculations. The results indicate that these low-lying states exhibit spherical behavior different from the previously identified oblate yrast band, and the band might be built on a (7/2+) ground state. In the second part, two OpenMP parallel Fortran programs, PairDiag and PairDiagSph, for the diagonalization of the pairing Hamiltonian were developed. The principles of the method will be introduced.

Thursday, January 20, 2022, 12:00 PM

Zoom ID: 982 0333 4147 Password: 642577

Dr. Mustafa Abdallah, Purdue University, Department of Electrical and Computer Engineering

"Artificial intelligence and machine learning frameworks for securing cyber-physical systems"

Modern cyber-physical systems (CPS) have been facing increasingly sophisticated technical problems due to the growing complexity of these systems and the associated attack surface. Therefore, managers of these systems have to both judiciously allocate security resources efficiently and detect failures (anomalies) in order to reduce their cyber risk and operational cost. In this work, we propose novel artificial intelligence (AI) and machine learning (ML) models for achieving such goals. In particular, we evaluate our frameworks on interdependent CPS with real-world attack scenarios, and with real data collected from deployed sensors in smart agriculture and smart manufacturing systems. Taken together, we show how in these two application domains our frameworks can be leveraged for paving the way for enhancing the future applications of these systems.

Monday, January 24, 2022, 9:00 AM

Zoom ID: 956 7779 4705 Password: 050412

Dr. Behzad Ghazanfari, University of South Carolina

"Machine Learning-Based Decision Making in Autonomous Systems"

Novel perspectives are presented in three separate parts in which lead to better performance and solving some shortcomings.

Abstraction in Reinforcement Learning (RL): We proposed SARM-HSTRL based on sequential association rule mining to extract the hierarchical optimum policy task structure of multiple policies like multi-task RL or Multi-agent RL autonomously in MDPs and FMDPs.

Error Fission: In current machine learning techniques, the error is considered as one scalar value in the literature, we introduce error fission as a novel perspective in which error breaks up into several values. We do error fission as a representation learning perspective of error by breaking up error in form of several features --- highly abstract features of representation of the error.

Quasi-periodic Time Series Classification: The proposed method translates quasi-periodic parts by some identical functions and then uses some functionality on the quasi-periodic parts to magnify the difference and remove similarities.

Monday, April 11, 2022, 9:00 AM

Zoom ID: 916 6707 8390 Password: 183579

Dr. Victor Watson, Commissariat à l’energie Atomique / Université Bretagne Sud

"Statistical tools for extracting information"

Technological advancement generalized access to large amounts of data. The increase in resolutions, the multiplicity of sensors, and also the capacity to store huge volumes of data give the possibility to obtain more accurate information about the phenomena studied in physics. However, this information is not always easy to extract. Indeed, if the information one seeks does not represent the main component of the data, retrieving it is not straightforward.

In this presentation, I will address two different cases where getting the in- formation from the data is complex. First, I will present a case involving basic Machine Learning techniques, statistical regressions, used to retrieve information about stellar physical parameters from high resolution spectroscopy. For this specific case, Sliced Inverse Regression (SIR) has been adapted to be used with ill-conditioned covariance matrices and non-linear link between the data and the explanatory variable.

Second, I will describe an approach to detect pollutants from a small set of sensors distributed over a monitored area. To detect weak signals, Generalised likelihood Ratio Test (GLRT) techniques such as Cumulative Sum (CUSUM), are well studied in univariate cases but the extension to multivariate data is still a matter of research. The event can also be transiting over the monitored area such that when the last sensors start monitoring the signal, other sen- sors which monitored it earlier have stopped being exposed. We propose the Temporary-Event-CUSUM to tackle the synchronicity problem caused by the transiting event, without losing the recursive computation, and keeping a low computational cost.

These two examples show how data sciences have become necessary to physics. Statistics allows us to make powerful tools using Machine Learning for estimation and detection.

NP Forum Series Archives: 2015 2016 2017 2018

For further details, or more information on the NP Forum series, please contact current organizers: Jackie Gates or Heather Crawford.

Internal NP forum documents