·Table of Contents
Detection gross and Fine Leaks in Nuclear Fuel ElementsI.I.Loktev, V.V.Rozhkov, A.B.Aleksandrov, I.G.Chapaev
Novosibirsk Chemical Concentrates Plant (NZHK), Bogdan Khmelnitsky street 94,
Novosibirsk, 630110, Russian Federation, E-mail: email@example.com
Nuclear fuel tightness of light water reactor is a necessary requirement of their successful operation and safety. Therefore a series of strict non-destructive monitoring methods apply to constructional materials and joint welds of fuel elements on their production. Besides they use helium leak testing of all manufactured fuel rods. However this aspect of monitoring has the features, which need to be taken into account at its using.
The leak testing method has both low detection limit for fine leakage and upper limit for gross leakage. Each kind of leakage has its optimal testing conditions for detection. Sometimes preferably to carry out gross leaks testing by separate stage and may be done by another method, for example by manometric one.
Behavior of the leaky fuel rods of different types during manufacturing and operation is investigated completely. Therefore in some cases a fuel designers shell to carry out leak tests under elevated temperature. Hot leak detection has additional problems of suitability estimation of fuel elements. It is necessary to have temperature correlation for leakage flow rate.
The rigour requirements to fuel rod tightness and on the other hand the lack of a noticeable amount of rejected product provoke necessity of leak simulation and determination of their behavior under different conditions. There is special computer code developed by Moscow research institute VNIINM, which allow to do it in dependence of defect size. It is convenient to consider all questions of leaky fuel detection on base of conditional defect size which is equal to real one with same leak flow.
Some results on mentioned questions, obtained at Novosibirsk Chemical Concentrates Plant, which have allowed manufacturing of non-defective VVER fuel rods, are given in this report.
They usually utilize helium as a trial gas, existing under cladding in finished fuel elements or passed though defect during special pressurization. When leaky products are in vacuum chamber they loose helium during preliminary pumping out and staying there before registration of leakage. Whether be detected the residuary gas in fuel elements depends on size of defect in the cladding.
It can be shown, that for any leaky product with interior volume V, helium pressure P0, leakage flow Qt will be defined for any time t from the equation:
In this expression the coefficient of exponential decay VP0/Q0 is a period of time which is necessary for gas flow out from fuel element with preliminary flow rate of Q0. To estimate leak behavior in a time and choose the most convenient requirements of their detection, it is preferable to consider separately fine and gross leakage.
A fine leak has a small gas flow, which doesn't vary during some technological period of time Ttec, when fuel elements are in a manufacture stage. A gross leakage gives a large gas flow vanishing to zero during fuel production.
Next relations can describe these features:
|a gross leaks||(2)|
|a fine leaks|
Leaky element with VP0/Q0, approximately equal to Ttec, has leak with both kinds of mentioned features.
Expression (1) with preset t, for example t = Ttec, is a function of preliminary flow rate Q0, having dependence on defect size, more exactly, on gas passing capacity. From this point of view we can introduce into practice the size of cylindrical defect with the same capacity for gas flow, which have real defect. It gives us possibility to use theoretical means linking gas flows, which we register, and cladding defect, which we try to detect. This way are used by number of researchers for analyze of leaky fuel element behavior /1-6/.
When we control tightness of fuel element with permissible leak flow q after Ttec from sealing, suitability for production can be determinate by the next expression:
It is possible to show expression (3) graphically, not in scale, in figure 1 in axes "diameter of defect d" - "flow of leakage Q0". The curve 1 shows magnitude of a leakage to the monitoring moment.
In conformity with expression (1) flow tends to zero in two cases, when initial flow of leakage Q0 (as well as defect size) decreases to zero or increases to very large magnitude. It is shown in figure 1 that the method determines all defects in size range from dmin1 to dmax1.
The estimates define that these sizes range for VVER fuel rods is 0,1 - 10 microns. A darkened space under curve 1 (figure 1) is a working field for fine defect detection with initial helium in fuel elements as a trial gas.
Defects with size more then dmax2 begin confidently be determined by other NDT methods: ultrasonic, eddy-current, X-ray. However, more reliable overlapping of detection fields in this region can be reached by another leak test method not having upper threshold of defect detection, - manometric method.
The figure 1 is divided on five zones concerning types of leak behavior and method of detection them.
|Fig 1. The graph of flow and pressure change for leaky fuel rod depending on the defect size.|
In the figure 1 a curve 3 in axes "diameter of defect d" - "pressure P" displays not in scale a pressure change of capsulated helium under cladding of fuel rod depending on the defect size d during the technological time Ttec after sealing. NDT pressure detection method described for example in /7/, can define a pressure drop value after some Ttec, Sequence, it can find leaks in fuel element beginning from some Qmin, wich is estimated by expression:
where V and Pnom is volume and initial pressure of the fuel rod. Low leak detection threshold is defined by sensitivity of the device.
So, checking fuel element pressure is monitoring of important quality parameters of the fuel - value of pressure and presence of gross defects.
Obviously, Qmin>>q, and this manometric method of leak detection can not replace totally helium leak testing, but it reveals through defects from some dman.(darkened space under curve). A manometric method has not upper threshold of defect detection. There is rather large region of overlapping with helium leak testing. As it shown on the figure a combined method of leak checking demands only pressure indicating, not measuring of it. This circumstance brings down requirements to sensitivity of a pressure-checking device.
Thus, during organization of quality production process on part of leak-checking as it is circumscribed above, necessary to have in mind:
It is possible to estimate a consequences of replacement actual leak sizes by equivalent ones. A knowledge of defect size give us possibility to solve next tasks:
The actual defect has a developed shape and can be considered rather slot then cylindrical channel. The degree of an aberration from the exact shape can be characterized by a ratio of the maximal and minimum sizes of a channel cross-section, or in another words, by ratio of a defect disclosing a to its width b. As a rule, a manufacturing defects have slot-type feature, for example crack in cladding, fig. 2, or welding with poor penetration, having the size relation close to 10. Depending on rate of splitting or kind of joining parts this relation can be close to 100.
|Fig.2. Crack in a cladding.|
Hence, it is possible to expect that, at least, one of the sizes of defect cross-section can be on the order or two more then equivalent diameter
Following to known expressions for gas flowing and fluids can be output relations between these parameters for:
|fluid flowing and viscosity gas flowing :||(5)|
|molecular gas flowing :||(6)|
For example, a crack on the figure 2 with disclosing 300 microns and width 3 mm has equivalent diameter 720 microns.
Magnitude of the disclosing defines an opportunity of defect detection by defectoscopy: X-ray, ultrasonic method, and can be appreciated too.
The performance of a gas permeability of a defect is given by its equivalent size by definition. Using gas lows it is possible to calculate an equivalent diameter of the channel in view of leak flow change in a time and parameters of Gas State. However it is necessary to take into account, that the change mode of gas flowing from viscosity to molecular is probable with diminution of a channel width.
The water permeability of the channel is connected with its capillary action, which inversely depends on smallest size of channel cross-section only .
So, the estimation allows consider that the performance of defects in product on base of equivalent diameter will be more conservative, concerning both material tight and leakage value estimation.
Behavior the leaky fuel rods of different types at manufacturing and operation are not clear and investigated completely. Therefore a fuel designers demand in some cases to carry out leak test under elevated temperature. Hot leak detection has additional problems of suitability estimation of fuel elements. It is necessary to have temperature correlation for flow rate of fine leakage.
As is shown by investigation /8/ a recalculations of molecular flow values, given, for example, at temperature T0, must be made by formula (7) for any temperature T.
If tested workpieces are heated only in a defect zone, for example in a weld joint, the recalculations should be carried out by formula (8).
It is necessary to have in mind sharp magnification of desorption gases flow from metal items at the temperature higher then 150°C in vacuum.
The rigour requirements to fuel rod tightness and lack of a noticeable amount of rejections provoke necessity of leak simulation and determination of their behaviour under different conditions. The Moscow All-Russia Research Institute of Inorganic Material (VNIINM) had developed mathematical code of prediction any scenario of behavior leaky items depending on an initial conditions and defect size /9/. Physically feasible simulate only gross leakage.
All described above methods of leak checking are applied on Novosibirsk Chemical Concentrates Plant (NZHK):
Long-term manufacture and exploitation experience of the NZHK VVER fuel proves on the one hand non-defect production and, on the other hand, reliability of applied leak-testing techniques. It had been detected three kinds of through defects of fuel rods: cladding cracks, poor welding, porous end caps. A corresponding measures which was timely realized, helped to do non-defect production of nuclear fuel rods and exclude producing of leaking fuel even in single cases.
Understanding of leaky fuel element behavior and particularities of leak-checking techniques, is important condition for reliability of production quality control.
Using of a helium leak-testing demands an estimation of leakage detection diapason. In case of multi-stage checking by different methods for exclusion of defect omission it is need overlapping or connection of detection diapason each methods at least. It can be made on base of defect size definition with using simulating code.
|© AIPnD , created by NDT.net|||Home| |Top||