Aluminum oxide occupies a larger volume than the aluminum it contains would fill as metal, consequently, the assumption has been made that holes in metallic aluminum would close by a sufficient amount of oxidation. Therefore, we were asked to investigate the rate of plug formation under conditions to be expected in the pile. For the latter we were requested to approach the pile conditions as nearly as we could by employing the Chicago cyclotron. It seems to us that the problem divides itself into two separate questions: (1) under what conditions may holes be expected to close? (2) if holes do close how much corrosion of uranium may be expected before the closure becomes impervious to water vapor? In this report only the first question is considered. The experiments and theory coupled with the data collected by other workers on the project definitely define the limits within which pores in the aluminum cans may be expected to close by an oxidation process. Under the most favorable conditions only small holes may be sealed in this manner. In the large majority of the cases the holes not only fail to close but become larger.

ABSTRACT. When solutions of high specific alpha activity are to be assayed for product, it is often difficult to obtain reproducible results by diluting the solution sufficiently to remove a small volume from which a sample can be prepared and counted at 52% geometry. In order to assay such a solution, most reproducible results have been obtained with minimum effort in sample preparation by making use of low geometry chambers. Two types of low geometry chambers are described: 1) The air-screen type which reduces the geometry simply by placing screens of various transparencies above the sample; and 2) The vacuum low geometry chamber which accomplishes geometry reduction by increasing the distance between sample and collecting electrode thus decreasing the solid angle subtended by the sample and aperture through which the particles pass into the ionization chamber. This type chamber was developed concurrently by this group and the instrument group of Chemistry Section C-I at the Metallurgical Laboratory. Because the geometry of the air-screen type chamber is quite critical to sample spreading and be- cause 12 mm sample plates are required, it has found very little use. On the other hand, the vacuum chamber, because it is noncritical to sample spreading …

The temperature coefficient of eta has been measured with the thermal column of the Argonne pile using uranium foils of different isotopic ratios. The temperature change was effected by filtering neutrons through silver. The measured fractional change per degree centigrade is [formula].

Fermi, Anderson, and Nagle's experimental distribution of Indium resonance neutrons around a point source of fission neutrons in water has been fitted by analytic expressions which are source functions in the two-group, three-group, Fermi and Christy-Wheeler pile theory. The Christy-Wheeler function (exponential followed by a Gaussian) is the best fit; the two-group function (exponential) is slightly better than the Fermi Gaussian.

Various experimenters have shown (H.H. Gersman U.S. Patent 2,335,590, Nov. 30, 1943; Aluminum Co. of America, various technical papers; also CT-482) that when a billet is extruded by proper technique into a rod (or tube by a floating mandril) that flow of material is streamline and and the extruded article is essentially a space replica of the billet, with linearly distorted coordinates. Advantage is taken of this fact in the manufacture of alclad tubing in which a billet containing an inner core of one alloy with the outer part of another alloy cast around it is extruded together into an integral tube, e.g., to combine high corrosion resistance with high strength. The following experiments were carried out because of the desirability of obtaining a control rod which can be water cooled (or immersed in P9) and which contains boron. For some pile structures it may be desirable to have the major portion of the energy released by the neutron absorption of the control rod be spend in the rod itself by the nuclear reactions [formula] rather than in the surrounding media as is the case when absorption of neutrons is by cadmium according to the reaction [formula]. In the later …

It is usual to determine the density of neutrons by using foils of metals which become radioactive as a result of bombardment in the region under investigation. From the rate of disintegration of the newly formed radioactive element the number of neutrons absorbed can be calculated. By use of the capture cross section of the element of which the foil is composed the density of the neutrons can then be computed. By counting the alpha particles from a bismuth foil of known weight after exposure to neutrons the density of the neutrons which produced this activity can be calculated. The least accurately known value which enters into this evaluation is the capture cross-section of bismuth for pile neutrons. This value may be improved by future measurements. It is of importance only for absolute measurements of neutron density so that relative values can be measured with considerable accuracy on the basis of known data.

The resonance absorption of a lump is described in terms of three components. Calculations of resonance absorption are carried out using the model for the resonance spectrum of uranium recently derived; calculations are also made with variations of this model which involve local fluctuations in level strengths. For metal and oxide lumps the agreement with observation is satisfactory. For dilute mixtures, whose resonance activation was measured by Mitchell, computed values fall 20% to 30% above measured ones.

Technical report with short reports from the (1) Experimental Nuclear Physics Group; (2) Theoretical Physics Group ; (3) Lattice Design Group; (4) Pile Design Group; (5) Exponential Experimental Group; (6) Radiation Group; and (7) Shielding Group.

The transition region between two parts of a pile which have different compositions is investigated. In the case where the moderator is the same in both parts of the pile, it is found that the diffusion constant times thermal neutron density plus diffusion constant times fast neutron density satisfied the use pile equations everywhere, right to the boundary. More complicated formulae apply in a more general case.

It was desired to determine a method of evaporating boron to form a rather pure uniform tenacious coat of specified thickness. These coats are needed as monitoring films for neutron intensities, particularly in steel ionization cylinders. The most satisfactory method of evaporating boron employed a graphite filament. A mixture of amorphous boron and Carbenoid A was painted onto the filament which was then heated by electrical resistance method to 2300 degree C at which temperature the boron evaporated. Opaque films with purities up to 98% boron or better could be deposited by this method. Much heat was liberated by the filament, and it was found necessary to cool the steel cylinders during evaporation to prevent alloying of boron with the steel. Cathodic deposition also proved satisfactory for producing high purity films; this method has the advantage that little or no heat is produced during the process, but requires much time. Other less efficacious methods of depositing lighter films of born were developed; these are included in the following discussion.

Measurements made in a uranium cylinder result in a mean value of 1.55 cm. for the diffusion length L for distances of 1 to 4 cm. from the base of the cylinder. Calculations give a value which agrees with the experimental result and show further that L increases from 1.40 to 1.63 cm. as the neutrons diffuse a distance of 5 cm. into the uranium.

Abstract. Recently (CP-1989) it has been shown that the neutron density, and thus the heat production, at the end of a slug with an A1 cap may be appreciably higher than that at the middle of the slag. A re-examination of earlier temperature calculations is made in light of this effect.

Abstract. The thermal conductivity of uranium as a function of temperature is measured by comparing its conductivity with that of a sample of brass. An absolute measurement is then made of the conductivity of the brass sample. The thermal conductivities as a function of temperature of two samples of extruded uranium rod are given.

Abstract. Temperatures are measured at two points in an internally heated jacketed slug which is cooled by water flowing in an off-center annulus around the slug. The variations of these temperatures with water velocity, water temperature, power, and the amount the annulus is off center are measured. The temperature near the slug surface which approaches and comes in contact with the water jacket does not increase as much as anticipated. The temperature near the slug surface on the side which is moved away from the water jacket increases nearly as much as on the side approaching the slug. An explanation for this unexpected increase of temperature as on side moving away from the jacket is given.

Abstract. Steady-state temperatures were measured in an electrically heated brass cylinder which was water cooled in a four-ribbed Al tube. Thermal transfer coefficients h were calculated from the measured temperatures. Curves of h values are presented for the regions of high and low ribs. the h curve is found to show a much sharper break in the region of high rib than in a low rib region. There is some evidence that h decreases linearly with annular thickness in the range considered. The low rib is found to exert a much greater effect than the high rib in decreasing the h value at points distant from the rib.

Abstract. Given the maximum intrinsic curvature of a rod, the maximum force required to confine it between parallel rigid walls is computed.

Abstract. The probability of exceeding assigned values of displacements or forces in simple rod models assembled from slugs picked at random is discussed, and some estimate for the statistical gain resulting from use of shorter slugs is obtained. The probability that a rod with self-warping will touch the top of the tube appears great enough to justify a recommendation of the use of top ribs.

Maximum curvatures and forces set up in taking a rod in a tube are considered.

A good determination of the resonance absorption of uranium when admixed with various substances is necessary so that calculations can be made on the efficiency of homogeneous piles. The original experiments along these lines were undertaken by Creutz, Jupnik, and Snyder (C-116) and consisted of experiments on the resonance capture in pure U, UO2, U3O8, and one experiment on capture in a mixture of U3O8 and graphite. These experiments were done at the cyclotron in Princeton and consisted of determining the ratio of the twenty-three minute activity of uranium to an iodine monitor placed in some position with respect to the sample. By using spheres of different radii both the volume and the surface absorptions were measured. It was considered very desirable to extend these measurements to mixtures containing large amounts of graphite to uranium and also to investigate other substances containing hydrogen and deuterium. In addition it was decided to use both iodine and gallium monitors in the experiments. While the experiments were in progress, an absorption curve for uranium metal was made by Marshall using iodine and gallium detectors, which showed that there was a certain amount of overlapping of resonance levels of uranium and iodine.

Prior to operating a chain reacting, graphite moderated, water cooled uranium pile at Hanford, the corrosive action of water on uranium was studied. It was necessary to enclose the uranium cylinders in Al jackets. Preliminary data were available which indicated Al jacket corrosion might be influenced by the water temperature. Therefore, in order to specify safe operating conditions, it was desirable to know the temperature at any position on the Al jacket of a slug in the tube. Because of the experimental difficulties encountered in the measurement of the correct surface temperature, the problem has been attacked primarily from a mathematical standpoint. Mathematical formulae have been developed for most of the proposed Hanford designs. The purpose of this report was to evaluate and summarize the theoretical and experimental information for calculation of Al jackets surface temperatures for the design now installed at Hanford. Also, the summarized results were to be put into a form suitable for use in routine calculations. As a result of this survey, the Al surface temperature for the Hanford tube and slug design may be calculated with 15 degrees C by employing routine methods and certain simple factors and equations contained in this report. The factors …

Warping of a rod by a variation in heat transfer coefficient around the periphery is considered. Estimates for some simple cases are given, and a more precise numerical calculation is carried out in ne instance.

Calculations were made to determine the temperature at various points of a thick disc and cap, on the assumptions that the heat production rate is uniform throughout the slug and that the slug is in perfect thermal contact with the cap. Results obtained by different methods of computation are compared with each other and with results obtained by Drew on the Paschkis electrical network computer at Columbia University.

This technical report discusses a method proposed to measure neutron energies by accurately determining the momenta of recoil protons in a magnetic spectrometer specially designed for this purpose. The main consideration in the design of the spectrometer is the shielding of the detector from the high radiation background associated with intense neutron beams.

The present report is concerned with the electronic relay which, upon the reaching of a pre-determined value of ionization current in the associated chamber, actuates the safety rod release mechanism. The are here proposed two independent suggestions: (1) An improved trip circuit of such sensitivity and stability as to allow it to be actuated by a signal directly from the chamber without the necessity of interposing intermediate amplifier stages. The value of chamber current at which tripping occurs is stable to +- 0.4%. (2) A triple-wound relay to prevent shut-downs of the pile which would otherwise be caused by development of defects in the safety circuits.