Research On Attribute Analysis Of Metal Uranium Components Based On ²⁵²Cf Source-driven For Nuclear Weapon Verification

Nuclear weapons are destructive and lethal on a massive scale.The existence and development of nuclear weapons poses a serious threat to global security and the survival of humankind.Therefore,a number of treaties or agreements had been signed to ban nuclear tests,reduce nuclear weapons and combat nuclear proliferation.Verification is the key problem whether the implementation of the relevant treaties on nuclear arms control is effective.Highly enriched uranium components are the main raw materials in the manufacture of nuclear weapons.Therefore,nuclear arms control verification technology has been trying to identify and judge the properties of uranium components quickly,accurately and effectively.It is not only about being proactive in the nuclear arms control verification and even in the international political struggles,but also about the security interests and survival of a country.In view of the sensitivity and high radioactivity of using highly enriched uranium components,it is particularly difficult to carry out practical and experimental studies.With the funding support of the National Natural Science Foundation Projects?No.61175005 and No.11605017?,the basic scientific research operating expenses of the Central University of Chongqing University?project number:CDJZR11120006?and Laboratory Research Project of China Institute of Engineering Physics?Project No.676?and other scientific research projects,our study seeks to the essential attributes of the metal uranium component,which constitutes the"heart"of nuclear weapons or atomic bombs.We carried out the simulation research using the Monte Carlo?MC?method to the essential attributes of the metal uranium components by neutron time-correlation signals,based on the ²⁵²Cf neutron source-driven method.The research attempts to achieve the identification or judgment of metal uranium components and"inverse analysis"the metal uranium component properties?such as enrichment,quality,and geometry,and so on?for nuclear arms control verification.The main research contents of this paper are as follows:?1?Studied the detection principle and technology of verification system based on²⁵²Cf neutron source-driven nuclear material.In essence,²⁵²Cf neutron source-driven method is a collection of pulse neutron transport processes and signal processing method.Built the mathematical physical model of the neutron transport process and simulation system rely on the PC platform and the Monte Carlo?MC?method.The neutron pulse time-correlated signals of the intrinsic properties of the metal uranium components measured,which include information on the enrichment,quality,geometry of metal uranium components,and so on.?2?Carried out the study on the enrichment identification of the metal uranium components.Whether or not it is a highly enriched uranium component,which is an important part ofnuclear arms control verification.Based on the ²⁵²Cf source-driven uranium component detection system,the neutron time-correlation functions of the same quality and different enrichment of uranium components are obtained,and the recognition mechanism and method of the enrichment by time correlation functions were studied.Furthermore,the validity of the correlation analysis for the identification of uranium component enrichment is verified.In view of the particularity and difficulty of using uranium component,Monte Carlo method was chosen to study the analysis and identification of uranium component enrichment properties.?3?The quality properties of highly enriched uranium components were studied.The neutron energy distribution is in a considerable range during the processes of highly enriched uranium components induced by ²⁵²Cf neutron source.So the single-energy model is obviously applicable no longer.For this reason,by using the energy partition theory,the neutron energy region is divided into multiple energy regions,and the neutrons in each energy region are processed and analyzed according to the single-energy model.For uranium components of different quality,energy partition theory is used to analyze the neutron time-correlated counts in different energy regions.Choosing the energy range with most neutron time-correlation counts or selecting the appropriate delay time,which all can analyse and identify the quality properties of highly enriched uranium components.Without changing the structure of the system and affecting the accuracy,this is a method to extract the attribute labels for reducing measuring time.?4?The study on the geometry recognition of highly enriched uranium components had been carried out.Highly enriched uranium components with only the difference of the geometric shape,it is very difficult to distinguish directly using neutron related signals.For this reason,based on the neutron signal analysis method of the PCA-LSSVM,the analysis and recognition of geometry can be effectively realized.At the same time,by removing the influence of scattering neutrons?removing the"accidental"related interference in the neutron count?,and using the fission neutron time-correlated function,the purpose of geometry resolution is achieved.However,because the number of fission neutrons is relatively small,in order to achieve more sensitive and accurate recognition of the geometry of uranium components,the time correlation matching integral count was adopted,which can optimize the ability of system attribute recognition.?5?The reflection layer effect of highly enriched uranium components was studied.The highly enriched uranium component has a reflective layer,which can prevent more neutrons from leaking from the system.Uranium component with a reflective layer brings difficulties to the extraction of radiation signals.A series of feature labels are obtained by introducing the neutron time correlation counting and energy partition theory into the identification system.Through the comparative analysis of the reflection layer effects of different materials and different thicknesses,the detection and recognition of the uranium components with the reflecting layer are effectively realized.Based on the mathematical physical models of neutron transport process of Monte Carlo method,given the sensitivity and strong radioactivity of using metal uranium components,this paper constructs a simulation measurement system of metal uranium components attribute parameters based on ²⁵²Cf source-driven noise analysis for nuclear arms control verification.Neutron pulse time-correlated signals of the metal uranium components are obtained,which will contain information such as the enrichment,quality,geometry,and so on.Based on this,we have deeply and systematically studied the enrichment,quality,geometric shape properties,and reflection layer effects of metal uranium components through the neutron correlated signals.This has opened upnew methods or approaches for the study of the properties of highly enriched uranium components,which are particularly difficult to be used and tested.