The multi-region critical problem can be transformed into an eigenvalue problem in the classical sense by using the method of Kuscer & Corngold and of Wing. This transformation is applied to derive a variational formulation for a reflected reactor.
An approximate critical value of the multiplying factor is determined by maximizing the Rayleigh quotient for radially and totally reflected cylindrical reactors. It is shown that this approximate critical value is an upper bound of the true critical value.
From the facts that the operator is self-adjoint and the eigenfunction is positive, an expression is derived for the upper and lower bounds of the true eigenvalue, by making use of the approximate distribution. The difference of the upper and lower bounds is an uncertainty of the presumption of the true critical value. It is found that we can compute the bounds to any required precision.
The narrow bounds are calculated for two radially and one totally reflected cylindrical reactors.

View full abstract

This paper concerns the application of the decoupling theory for a nonlinear system to the power control of the coupled core reactor. By using this theory, we can deal with the power change problem, the regulating problem for a system subject to a large perturbation, etc., which cannot be dealt with by the decoupling theory for a linear system.
But, since this theory assumes tacitly that all the states are measurable, the theory can not be applied directly to the nuclear reactor that has nonmeasurable state variables such as the delayed neutron precursor. In this paper, the above difficulty is overcome by introducing the observer theory for a nonlinear system.
The decoupling controller and observer resulted from these theories are applied to the coupled core reactor and the dynamic characteristic of the reactor is simulated by the digital computer. The simulation shows that the reactor power of each core is controlled satisfactorily in decoupling manner.

View full abstract

Experiments on reactor noise were conducted at KUR. Depending on the operating condition of the reactor, the cause of the noise are classified into the following four types.
(1) Zero-power noise source due to the branching process of fission neutrons and/or due to random bombardment of neutrons to the detector-under natural circulation of coolant and at essentially zero-power level.
(2) Coolant temperature fluctuation due to natural convection-under natural circulation and at relatively high power level.
(3) Flow induced vibration of shim control rods-under forced circulation of coolant and at low power level.
(4) Fluctuation of inlet coolant temperature-under forced circulation and near the maximum power level.
Vibration of a spare shim control rod and fluctuation of inlet coolant temperature were measured simultaneously with the neutronic noise. Then the noise sources of the types (3) and (4) were verified. The vibration of a control rod has a broad spectrum in low frequency region.besides the large peak at 14 Hz. The fluctuation of inlet coolant temperature is non-white noise and consists of large low frequency component. The theoretically predicted sink structures in the neutronic PSD relating to the transit time of inlet coolant temperature fluctuation through the core were not observed in the experimental results.

View full abstract

For the purpose of providing standard data for checking two-dimensional neutron penetration calculations, fast neutron spectra as well as thermal and epithermal neutron fluxes were measured over a two-dimensional (R, Z) space in water shield using an activation method. Threshold reaction rates were converted to fast neutron scalar flux spectra with the aid of the SAND-II code. These results agree within a factor of 2 with the calculations by a two-dimensional discrete ordinates code PALLAS-2D. Thermal and epithermal neutron fluxes obtained with the Westcott's method agree quite well with the calculated values by the PALLAS-2D code in which the diffusion equation was adopted for dealing with low energy neutrons to reduce the computing time. All experimental results are given in the absolute values.

View full abstract

Interstage flows are analyzed for ideal cascades composed of asymmetric separation elements. It is shown that, in such a cascade, the separative power is additive, that is, the summation of the separative powers of all stages equals the total separative power of the ideal cascade composed of asymmetric separators. This is proved by calculation establishing that the total sum of interstage flows is equal to that obtained by dividing the total separative power of the cascade by the separative power per unit flow of the elements. A similar additive property is also evidenced for separative powers relevant to the desired and undesired materials. As a concrete example, a 2-up 1-down cascade is discussed.

View full abstract

It is found from numerical analysis that a strong influence is exerted on the separative efficiency of centrifuges by the thermal convection arising from the temperature difference prevailing between the end plates of a centrifuge drum-known as "Ekman suction". The numerical method previously developed by Nakayama & Torii is used to compute the distributions of concentration resulting from considerations of both thermal convection and forced flow.
The effect of Ekman suction is represented by the parameter λ_E(≡(a²√Eρ_sVΔT)/(16lρDT₀)), different values of which were adopted in the calculations, as were other parameters such as λ(≡G_F/(4πlρD), G_F: feed rate) for the forced flow and θ representing the cut. The separative efficiency of an exemplified counter-flow type centrifuge is found to reach 58%, when λ_E ?? 0.7, λ ?? 1.0 and θ ?? 0.3, with a product port radius γ_P=0.5.
It is also shown that the thermal convection in the Stewartson layer near the side wall of a centrifuge drum affects the separative efficiency to relatively small but not negligible extent.

View full abstract

A portable chair type whole-body counter has been studied which is for internal contamination monitoring of the subjects working in nuclear facilities. A 3"φ × 3" NaI(Tl) crystal housed in a lead collimator is used. Findings from the study are as follows : the crystal size does not influence the screening level of internal contamination defined as a three standard deviation above the mean level of distributed counts obtained for the uncontaminated subjects. Reasonable shieldings of both detector and chair may be more than 3.8 cm of lead. A three-channel pulse height analyzer set in the γ energy region above 0.2 MeV will be suitable for the monitoring. The detection sensitivities for stomach, lung, kidney and small intestine to a ¹³⁷Cs source are not lower than in the case of the homogeneous distribution in the whole body revealed by phantom study.

View full abstract