Described in this paper are the test results of mechanical properties of several kinds of concrete mixes at very low temperatures and an investigation into the mechanism of change in their properties, for the purpose of obtaining design data for concrete structures exposed to very low temperatures, such as liquefied natural gas storage tanks and refrigerator warehouses. It was learned from the preliminary tests that the strength of concrete under the temperatures of the range of -1O'C to -7O'C was affected by moisture contents; the larger was the moisture content, the higher the rate of the strength increase was, and that the increase of concrete strength corresponded to that of ice at very low temperatures. Under lower temperatures of -1O'C to -196'C, it was verified that compressive strenth, modulus of elasticity and tensile and bond strengths of concrete increased with the decrease of temperature and the rate of increase in the strength and the elastic modulus was higher when the moisture content was larger. On the four mixes of concrete which had different water-to-cement ratios and air contents, tests were made, under the temperatures of down to -70°C, for compressive, tensile, flexural and bond strength and the modulus of elasticity. It was found that the rates of increase in these strengths were higher for the concrete with higher water-to-cement ratios and larger air contents. Also found by tests was the decrease in the strength; of concrete that received the very low temperature shocks between +20 and -196 C. The coefficients of thermal expansion of concrete were calculated from the measurements of the changes in the specimen lengths under the low and the very low temperatures, and were found to be smaller com-pared with those under room temperature. Further discussion is made on the influence of freezing of water in comparatively large pores under low temperatures and of the drop of freezing point of capillary water in smaller pores and accompanying freezing of water in these pores.

This paper presents details of an analysis for strength at failure of prestressed beams subjected to a complex system of applied loads consisting of combined torsion, shear and bending. It is based on a modified skew bending approach incorporating the use of strain compatibility over the beam cross-section to permit recognition of a "non-flat" yield region typical for cold drawn reinforcement. A significant feature of this analysis is the use of a biaxial strain criterion to recognize that the magnitude of the limiting strain in the compressed concrete at failure varies with different combinations of torsion, shear and bending. Other contributors working on this problem have used either a constant limiting concrete strain of magnitude 0.003 as for pure flexure, or some constant fraction of this amount throughout all possible load combinations involving torsion, shear and bending. Incorp-orated also in the determination of the ultimate strength is the ' recognition of the presence of shear stresses on the uncracked failure surface. Results of tests made on eighty four beams were used to verify this analysis. An excellent and consistent correlation was obtained between theoretical and test values for bending moments and resisting torques.

Field measurements on two existing reinforced concrete slabs had to show that the chosen strengthening methods were successful. This was done by determining the bending stiffnesses of the two slabs before and after strengthening. The strengthening methods and the measuring equipment are described. The results showed that the sub-sequent strengthening provided an increase of the bending stiffnesses. It was also found that to achieve good quantitatif results a load test is required, whereas the actual floor loading produces only qualitatif results.

An investigation was conducted to develop information on the time-dependent deformation behavior of concrete in the presence of temperature, moisture, and loading conditions similar to those en-' countered in a prestressed concrete reactor vessel (PCRV). Variables were one concrete strength (6000 psi (41 MPa) at 28 days), three 7 aggregate types (chert, limestone, and graywacke), one cement (Type II), two types of specimens (as-cast and air-dried), two levels of tempera-!, ture during test 73 F and 150 F (23 C and 66 C), and four types of "1, loading (uniaxial, hydrostatic, biaxial, and triaxial). There were 66 test conditions for creep tests and 12 test conditions for unloaded or control specimens. Experimental results are presented and discussed. Comparisons are made concerning the effect of the various test conditions on the behavior of concrete and general conclusions are formulated. Research performed under Int eragency Agreement No. AT-(40-1)-4128 for the Oak Ridge National Laborat ory operated by Un ion Carbide Corporation under contract with the Energy' Research and Deve lopment Administration.

The random behavior of reinforced concrete elements at de-formational limit states and at the ultimate limit state is analyzed by a second moment approximation. The deterministic and stochastic parameters involved, their functional and stochastic dependences, and their experimentally based statistics are discussed. The resulting variances of ultimate carrying capacity and ductility and of crack development are shown. It should be noted that ultimate ductility has an especially strong statistical variation independent of the amount of compressive reinforcement. A two-span beam is analytically modeled to elaborate first order approximations of the means and variances of simultaneously acting live loads and temperature effects which cause different limit states. Imposed deformations do not greatly influence ultimate load-carrying capacity provided that there is sufficient ultimate ductility. However, load-carrying capacity with respect to a limit state of allowable crack width is substan-tially reduced by simultaneously acting imposed deformations.