Improving activation cross section data with TALYS

Needs for accurate (n,x) activation cross sections for fusion technology have been considered with reference to the current status of the TENDL library. The current work is focused on improving activation cross section data for nuclear reactions relevant mainly for fusion and astrophysical needs. The fits have been performed with the TALYS-1.8 code by means of nuclear model parameter variation, mostly for the optical model and level densities, followed by comparison to recent experimental data taken from EXFOR and other evaluated nuclear databases. The updated cross section data are going to be adopted into the new version of TENDL. The improvements have been performed both for differential as well as integral data sets.


Introduction
Nuclear data forms the basis of many nuclear technology studies.The area of their applications is expanding and penetrating deeper into various fields of research.One of these is in fusion research where accurate activation cross sections enable efficient development of fusion design concepts.Comprehensive computer modelling of experiments and installations can be achieved using reliable information on cross sections from nuclear data libraries.
The implementation of many research programs requires well-qualified nuclear databases.For instance the Monte Carlo codes must have available a wide range of reliable nuclear reaction cross sections and decay properties for all the materials of interest in the nuclear devices.However despite a large amount of existing nuclear data the problem of inconsistency between various data libraries as well as a lack of those still remains and that requires more efforts and new approaches to solve.Thus the requirements to the quality of those data are quite high.The current work was aimed to verify current requests, following up by improving that set of (n,x) cross sections mainly for the fusion relevant materials [1].Analysis of available data indicated rather big deviations for more than 100 reaction channels [2] and effort has been applied to perform a better evaluation.

TALYS improvements
All calculations of nuclear reaction excitation functions of interest, in the range of (0-30) MeV, have been performed with the TALYS-1.8 code [3].The fitting procedure of activation cross sections included a variation of the following nuclear model parameters: • Diffuseness and radius in optical potential • Model of level densities • Densities of exciton model constituents a e-mail: dzysiuk@nrg.eu • Pre-equilibrium gamma emission • Branching ratios • Probability of α/d/t-particle formation.
Keeping a reasonable balance between competitive channels was the crucial issue.The adjustment was targeted at achieving the best agreement between the latest experimental data, taken from the EXFOR database, and presently calculated TALYS curves.Special attention was paid to fitting the cross sections of isomer production reactions.In this particular case, there was a need to analyze the level structure of the final nucleus thoroughly and to vary the probability of transition between isomer and ground states.Where the information on that transition probability was not given in experimental data tables the values could be estimated.In the fitting procedure it was attempted to fit to the most recent experimental data, to control the shape of excitation curve, and to adjust parameters within acceptable physical limits.In several cases there was a need to study the publications related to measurements in order to understand why some points are outliers when compared general trends.However, in many cases publications did not contain enough details which made process of analysis difficult.One of the most reliable ways of nuclear data validation is a comparison of evaluation to integral data.In the current work, C/E values are used for this purpose, where E -is the real neutron spectrum folded with experimental integral cross section data, C -is the same neutron spectrum folded with the TALYS curve.This ratio can be used as an indicator of the fit quality, and allows for the possibility to analyze each individual channel.In Table 1 there is information on several reactions and C/E values defined using the latest TALYS calculations showing the performed improvements of TENDL-2015.There is a comparison of the latest TALYS fit and data taken from EAF [4] as well as TENDL-2012, 2015 data libraries [5].In the last column TENDL-2017* is related to a new release of TENDL which is expected at the end of 2017.It is a re-evaluation of TENDL-2015 plus the performed improvements of the cross section fittings.These reactions have been selected   because for them there are only one or two C/E available in the report of J. Kopecky [6], when there are more values then it is difficult to get an agreement with them all simultaneously and statements about certain improvements are not justified.For example for the 51 V(n,γ )V 52 reaction it was possible to make validation at two different energy ranges due to availability of corresponding integral experimental data.Below there is an example of improved fit performed for the 90 Zr(n,p)Y 90 and 90 Zr(n,p)Y 90m reactions (Figs. 2, 3).Zirconium is an important construction material therefore such data are of a high priority.
As could be seen in Fig 3, the TALYS curve (black solid line) is in agreement with the latest experimental data taken from EXFOR [7].On the right side starting from 15 MeV and up TALYS follows the experimental trend and is essentially better than TENDL-2015.
In addition the shape of this excitation function became more physical.
In the case of the capture channel, it was possible to mainly vary the high energy part of the excitation curve.A special attention was paid to the 28 Si(n,γ )Si 29 reaction since it was mentioned in several literature sources  stressing out this certain problem at energies higher than 1 MeV [8].Now the current evaluation agrees with other evaluated data libraries as well as with the few experimental data at 14 MeV (Fig. 4).Also the problem with iron has been risen in several publications [9].With respect to those requests in the current study the cross section for the 56 Fe(n,xα) reaction was considered and significantly improved (Fig. 5).

ND2016
It should be emphasized that for (n,α), (n,xα), (n,d), (n,t) reactions it is relevant to take into account the mechanism of formation of particle in output channel.In TALYS this option is implemented by means of a model.Applying the "cstrip" parameter [3] is very efficient to get a better cross section fit for mentioned reactions.
Many new evaluations have been performed for various reactions on chromium isotopes (Figs. 6, 7).Unfortunately the experimental data are very scarce and new measurements are required.
In Fig. 8 there is a fit done for the 100 Mo(n,p)Nb 100 reaction.After studying the papers dealing with those measurements the current evaluation was performed based on the lower experimental point since the other one can be overestimated due to incorrect half-life used for cross section deducing.
In case of the (n,2n) reaction channel it is quite important to choose the level density model.In Fig. 9 there is the evaluation done for the 183 W(n,2n)W 182 reaction.The current fit is in agreement with the only experimental set.

Figure 1 .
Figure 1.Connections between cross sections and practical applications.

Table 1 .
C/E for some reactions.