Research On Measurement Technology Of Tritium Retention Depth Distribution In Fusion Reactor Facing Plasma Material

Precise measurement of the content and distribution details both on surface and in bulk of tritium retained in plasma-facing materials(PFMs)is of great significance for fuel balance and occupational safety in magnetic fusion reactors.Tungsten is one of the most important candidate PFMs and is highly possible to be employed in ITER and CFETR,since its extraordinary thermodynamics performance and minimal fuel retention.However,the direct measurement for tritium content and distribution in tungsten are limited,especially for the research on tritium analysis in real PFMs at present.Real tungsten PFMs are quite different from the tungsten materials in laboratory.On the one hand,tungsten tiles will produce a large amount of radioactive neutron activated products(NAPs)by neutron irradiation from DT discharge,which could emit both ? and ? rays to affect the subsequent tritium measurement process.Thus,the existence of NAPs would be a new challenge of current tritium analysis approaches.On the other hand,the tritium 2-D concentration(the concentration of tritium at the plane which is parallel to the tile surface)in PFMs is usually non-uniform,and it might influence the measured results of tritium.Besides,the reconstruction results of tritium from BIXS method(tritium ?-ray induced X-ray spectrometry)should be verified by reliable direct measurements since the ill-posed problem in its reconstruction process.Therefore,the main purpose of this thesis is to evaluate the influence of NAPs and non-uniform tritium distribution on tritium measurement and establish a reliable tritium depth profile measurement system.For the consideration of feasibility and research cost,we employed chemical etching method(CE)and BIXS for obtaining tritium depth profile and tritium imaging plate technique(TIP)for 2-D tritium distribution.In the analysis of the influence of NAPs on tritium measurements,an elaborate model of ITER was established firstly to obtain the neutron flux at different position,including plasma-facing components(divertor and the first wall),CuCr alloy layer,steel backboard,steel substrate etc.And the amount of NAPs at a specific location in the divertor and first wall was calculated using FISPACT code.Then the quantificational influence of NAPs in chemical etching,BIXS,TIP and tritium calorimeter(TC)was evaluated by Monte Carlo(MC)and numerical methods.With the assumption that the reactor power was 1 GW and the operating time was 1 year,118 kinds of radioactive NAPs at the cooling time of 0 s were produced in divertor according to our computation,and their total activity were 5.2E14 Bq/kg at the depth of 25?m.Then we defined a factor ? to describe the effect of NAPs on tritium measurement,and it was mainly dominated by three quantities,including cooling time,sample thickness and tritium depth profile.Results shows that after cooling time from 3 to 60 months,W-185,W-181,Ta-179 and Ta-182 mainly dominates the influence of NAPs.The effect in chemical etching method was minimal,?CE=8.9%when the cooling time was 3 months,and ?CE<1%after 16 months.While NAPs strongly influence the measurement of tritium in other three methods.The estimation of BIXS method shows that the proportions of counting rates in ROI(1-16 keV)of NAPs and tritium equals to 2.4E2 with 3 months' cooling time and decreasing tritium distribution,and the sample thickness was 0.1 mm.Then this value reduced below 10%after 50 months.In TIP technique,?TIP equals to 130.140,140 and 140 with decreasing depth profile and the sample thickness of 0.1,1,3 and 5 mm,respectively.In TC analysis,when the cooling time was 3 months and the sample thickness was 0.1,1,3and 5mm,?TC equals 12.140.430 and 730 and this influence will reduce to 10%after 26.40,49 and 54 months.Therefore,NAPs should be taken into consideration when using BIXS.TIP or TC technique to analyze tritium retained in PFMs shortly after shutdown.In the evaluation of tritium 2-D non-uniform distribution,the influence on CE.TIP and BIXS techniques were calculated.Since the result of CE is the total amount of tritium in an eroded layer,the plane distribution in this layer has no influence on CE.In TIP technique,as the range of tritium(3 rays are extreme short,the result is barely the tritium surface distribution.Hence,the 2-D distribution of tritium has no influence on TIP either.And in the analysis of BIXS method,as the effective region of the LEGe detector and experimental devices are nearly centrosymmetric,X-rays emitted from the same radius will create same signals in the detector theoretically.Thus,only radial and axisymmetric distributions of tritium were taken into consideration in our calculation,including Gaussian,linear decreasing,linear increasing,exponential decreasing,exponential increasing,and uniform distribution.And an elaborate model based on our actual experimental BIXS device was established to evaluate the influence of this effect by MC method.The materials of specimens were set to be beryllium,carbon and tungsten.The results indicate that the intensities of response spectra derived from different tritium distribution vary significantly both in high-Z and low-Z materials,and the tritium depth profiles would deviate from the real amount of about-30%to 60%in beryllium and carbon with our assumptions.It is indicated that if this effect is not considered in the calculation of response spectra,the deviation reconstruction results will deviate from the real amount of about-30?60%.And in tungsten,this ratio is-40?100%.In addition this diversity changes as the depth increases in high-Z materials,which might alter the reconstruction results of tritium depth profiles.While tritium with different distributions only has a little influence on the shapes of response spectra,except for the case that most tritium distributes at the margin or central area of the sample.Therefore,to improve the accuracy of BIXS method,the auxiliary works should be introduced to reduce this influence.Furthermore,the specific experimental procedure and parameters of CE were analyzed in the experiments in detail since chemical etching method has not been employed for tritium analysis in tungsten.And the experimental parameters and influence of internal bremsstrahlung X-rays(IB)of BIXS were evaluated by MC method.It is found that aqua regia(conc.25.8%HCl,18.4%HNO3)is appropriate for eroding tungsten.And based on the selection of etchant,the anticorrosion material and experimental parameters(including erosion time,erosion temperature,amount of etchant,etc.)were tested and discussed in the preliminary experiments.It is indicated that the etching thickness was 1.2?m when the corrosion time was 4 hours and the temperature was 30?.And the liquid scintillate counter could provide stable and reliable results while the dilution multiple was larger than 50.Besides,in the analysis of BIXS method,we established a MC model based on the real experimental system,selected the appropriate parameters and estimated the influence of IB in carbon and tungsten materials.Results show that argon is a kind of suitable work gas with the pressure of 1 atm and the distance from the sample surface to Be window is 6 mm.In addition,the influence of IB on the intensity of response spectra was 12.6%in carbon and 0.95%in tungsten.And the influence of IB in the shape of response spectra in both materials was slight that could be neglected.Finally,the tritium depth profile and surface distributions after each etching of 3#tungsten sample were measured based on the three approaches mentioned above.The result shows that the tritium depth profile was approximate to exponential decreasing.distribution and the 2-D distribution at different depth was diverse.In addition,the results of CE,BIXS and TIP methods were in good agreement.In this thesis,we evaluated the influence of NAPs and 2-D non-uniform distribution on tritium measurement in real PFMs in detail and provided several advices to improve the accuracy of tritium analysis in PFMs.And it could be a part of reference points for selecting the appropriate tritium measurement method in CFETR.Besides,it was the first time to obtain the depth profile of tungsten sample by CE method,and its results could be used as basic data to verify other nondestructive approaches.It is of great significance to improve the accuracy of tritium reconstruction depth profiles in nondestructive methods.