Calculating the large storage tank shell thickness and designing of user interface

Shell thicknesses of vertical cylindrical tanks from the most common equipment used in industrial installations has great importance according to safety of construction, material costs, labor costs, life expectancy, etc.. The increase in thickness increases the weight of the construction as well as weakens the design in terms of buckling. For this reason thin shell thicknesses should be obtained as thin as possible. The calculation of shell thicknesses in pressure and non-pressure tanks is based on the calculation of the shell thickness at which the internal pressure can carry the load. If liquid is stored in the tank, the effect of liquid static height must also be considered. Internal pressure, liquid static pressure and safety factors must be considered together to make calculations. The tank has a cylindrical, spherical or a different geometry. There are two methods used for common in tank shell thickness calculations. One foot method (1FM) and variable design point method (VDPM) recommended by API 650 standard are used for shell thickness calculation in industrial storage tanks. In the industrial applications of storage tanks, tank shell thickness calculations differ in the above mentioned baselines and thicknesses in terms of tank manufacturing and material cost. Thicknesses are the largest at the bottom of the cylindrical storage tanks and the thickness at the top should be the smallest since there is no load. This makes the tank manufacturing impossible. For this reason, it is necessary to use graded thicknesses. Every step is called a course. The thickness of the course is calculated so that it will bear the liquid pressure which affects it. Each course has different thicknesses so that welded construction could be manufactured. In such a design, it seems that there will be excesses in thickness. In this case, the upper part of the lower unit could carry some amount of the upper unit's load. Thus, the fluid height could be reduced somewhat in the calculation of the course thickness. This value is obtained empirically and used as 1 ft (0,304 m). The value of the liquid height is subtracted by one foot (1 ft) from the value of the course thickness. Thus, the thicknesses calculated are thinner and the total cost is reduced. The second method, VDPM, takes the distance taken in one foot method by calculating it with an empirical expression. This value, which is generally higher than 0,304m, enables thinner shell thicknesses to be obtained. Tanks constructed with VDPM are subject to occasional extreme stresses. For this reason, designs should be inspected with finite element analysis. 1FM and VDPM methods could cause heavy designs or unsafe designs. In this thesis study, the equation that Timoshenko wrote for storage tanks under hydrostatic pressure of vertical cylindrical tanks was solved with finite difference method (FDM). Deformation and stress values were examined and compared with the currently used two methods. By creating an interface for the obtained solution code, it has become a useful in house design program. In cases where the functions are given analytically, it is easy to calculate the function values at the desired points, to find the derivatives of the desired order at certain points of the function, or to calculate the integrals at certain intervals of the function. While some discrete values of the functions are known, such calculations could be approximated using finite difference method (FDM). Even where analytical solutions are evident, higher-order differential equations could be preferred in terms of ease of solution by use of finite differences. Finite difference method, solution of heat propagation equations, electromagnetic force equations, and preferred method with ease of use in nonlinear systems and solution of high order differential equations in places where structural analysis is performed. If the general equation for vertical cylindrical storage tanks is applied to the finite difference method, the fourth-order differential equation is transformed into a set of linear equations. In order to be able to control the accuracy of this solution by means of an experiment, it is necessary to adapt the thickness of a certain tank to a numerical method. The numerical solution of the design of the tank designed for the solution and the stress values on the tank courses must be found by numerical method. It is seen that the maximum deviation of the calculated error between the model and the experiment is -6,23%. The model we construct by the finite difference method characteristically overlaps with the experimental results although there are deviations. The solution algorithm which was validated with the experimental results was compared with the other methods used in the literature on eighteen tanks. The first six tanks are recommended diameter values in the API 650 standard. The next six tanks are the tanks previously examined in the literature. The last six tanks are the tanks that are outside the standard used dimensions. 2.8m was selected for the tank course height. The stored liquid water is considered as open tanks to the atmosphere. The thicknesses obtained are shown in charts and graphs. Corrosion allowance is not included. Welding efficiency (JE) is assumed as 1,0. As a result of the obtained interface program, tank shell thicknesses were calculated according to eighteen tanks 1FM, VDPM and FDM methods. FDM and 1FM are very close to each other in the first course and intermediate courses. The differences of thicknesses are range from 1% to 4%. The FDM results are thinner than 1FM. In the top course, the deviation between the thicknesses is large because the thicknesses that are too small. In other words, if the top course is close to the lower courses' heights, the deviations will be in the same order again. It is possible to make economic designs by changing the heights of the courses. FDM and VDPM are close to each other with 6% to 8% difference in initial and intermediate courses. However, there is still a big deviation as it is in the 1FM. In tanks which tank diameters greater than 100 m, the differences between the VDPM and FDM solutions are out of character by increasing tanks' heights. A deviation generalization is not possible. As a result, the analysis with FDM gives thinner courses' thicknesses result than 1FM methods in the literature. In this case, a more economical solution has been obtained for the risk of buckling in design, manufacturing and construction processes, such as thinner shell thickness, less material use, field work, labor costs, logistics. The work done is turned into a program and made available continuously. As a result, in this study, to calculate the shell thicknesses of large storage tanks, the shell deformation equation for vertical cylindrical tanks under hydrostatic pressure of Timoshenko is solved using the finite difference method. The resulting solution code, interface and program have been prepared. Thus, a useful in house design program has been obtained.