Powerful tsunami attacked to the northeast coast of Japan due to the Great East Japan Earthquake occurred on March 11, 2011. Some aboveground oil storage tanks (ASTs) were subjected to external pressure due to the tsunami. Wind girders are installed to the shell plates in AST for the prevention of buckling by wind pressure. Although many researches on the buckling of ASTs or cylindrical shells concerning tsunami were made, the effect as reinforced rings of the wind guarders was not taken into consideration in them. This paper presents the bifurcation buckling of shell plates in90, 000m³ AST under both internal and external liquid pressure using the axisymmetric shell finite element method. The buckling external liquid height of AST with wind girder is compared with that without wind girder. As a result, the effect of the wind girder on buckling under external liquid pressure is negligibly small. Furthermore, in order to increase the buckling strength in case the internal liquid is empty, one more ring stiffener is added to the shell plate below the lowest wind girder. The result shows that it is effective to attach a ring stiffener to the low position of the shell plate.

View full abstract

The authors study on the technology of an induction bending process and attempt to develop the accurate finite element model for its numerical simulation in which the electromagnetic field-heat transfer-large displacement coupled analysis is required. In the present paper, as part of the research works, a large displacement analysis to bend the STPG370 steel pipe whose diameter is 8 inches, gage is Sch40 by 2DR has been carried out alone and the effects of the temperature distribution given as one of the conditions to the simulation result are discussed. Here, the temperature dependency and the strain rate dependency of the material are taken into consideration. The FE results suggest that wrinkling, a kind of plastic buckling, occurs on the compression side during the whole bending process when the length of the high temperature region over 700°C is too long, for an example, its length which is presently designed around21mm may be shorter than at least 29mm and must not be larger than at least 50mm in this simulation. The mechanical properties of the employed material depend on temperature and greatly change around 700°C. The finding is applicable to the design and management of operational sites and provides guidelines for the required accuracy of the heat transfer analysis which should be coupled with the large-displacement analysis.

View full abstract

Experiments of induction bending at the actual processing site have been performed on the STPG370-E steel pipe from which the systematic data on profile, temperature, strain and dimensions with such parameters as diameter, thickness and bending radius. The obtained data show that the processing is a good success with all the parameter cases, and that some characteristics of induction bending, including appearance of real deviation such as asymmetry from the ideal state and actual situation as mentioned below, were observed. Thermocouples measured the temperature distributions around the induction coil whose peak was just below 1000°C, width along the x-axis was about160mm to 300mm, and the graph curve resembled the shape of a smooth mountain. Two axial strain gages measured the strain distributions along the straight area of the pipes which were simple, and suggested the bending moment acts the area to generate 0. 034% elastic strain. We used our own tracing instrument newly created to draw the profiles of the bent pipes well which proved that springback can be ignored. An ultrasonic thicknessmeter measured the thickness along the bent area to calculate the rates of variations which were in the range of 20% to 30%, satisfied the requirement and showed that in the harsher conditions like 2DR case they fluctuated wider in accordance with time. Vernier caliper measured the diameters along the bent area to calculate the rate of variation in ovality which was almost less than 1% to 3% and a diameter tape measured the circumferential length. These results satisfy the requirement while they show some interesting features of the process. The obtained data with various contrivance can be available for verifying the accuracy of coming FE analyses.

View full abstract

In previous studies, we developed an inverse method to estimate the thickness of refractory in two-layered structure composed of refractory and steel plate. In the proposed method, outer temperature change on the steel plate was measured when periodic temperature fluctuation was applied on the inner refractory side. Then, obtained delay time of the outer temperature change against inner temperature change due to thermal conduction was inversely analyzed to estimate the thickness of refractory. In the previous report, we investigated the feasibility of the inverse analysis was demonstrated for the hot-blast stove in steel mills. In this paper, the authors investigated the estimation errors of the thickness of refractory by the inverse analyses and factors influencing on them, based on the experimental studies using small two-layered combined plane specimens and finite element analysis. It was found that the refractory thickness estimated by inverse analyses were influenced by several factors such as in-plane thermal conduction, waveform of the temperature fluctuation on the inner surface. However, effects of these factors are not significant in the case of thickness estimation of refractories of the large scale actual hot-blast stove, and the proposed method is found to be effective for actual applications.

View full abstract