A Method for Applying Automatic Temperature Control to the Transient Finite Element Analysis of a Pressure Vessel Undergoing Postweld Heat Treatment

For complex finite element problems it is often desirable to prescribe boundary conditions that are difficult to quantify. The analysis of a pressure vessel undergoing postweld heat treatment (PWHT) is an example of such a problem. The PWHT process is governed by Code rules, but the temperature and gradient requirements they impose are not sufficient to precisely describe the complete vessel temperature profile. The imposition of such a profile in the analysis results in uncertainty and errors. A suitable but difficult approach is to specify heater power instead of temperatures, letting the solver determine the temperature profile. Unfortunately, the individual heater power levels necessary to meet the Code requirements are usually not known in advance. Determining the power levels necessary is particularly difficult if a transient solution is required. A means of actively controlling the heaters during the FEA solution is requirement for this approach. A simple and adaptive control algorithm was incorporated into the FEA solver via its scripting capability. Heat flux boundary conditions (heater power) were applied instead of transient temperature boundary conditions. Heater power levels were optimized to achieve predetermined time/temperature goals as the solution proceeded. The algorithm described was successfully applied to a pressure vessel PWHT with 14 zones of control. The approach may be adapted to other problems and boundary conditions.

A Thin Shell Theoretical Solution for Two Intersecting Cylindrical Shells Due to External Branch Pipe Moments

Two intersecting cylindrical shells subjected to internal pressure and external moment are of common occurrence in pressure vessel and piping industry. The highest stress intensity occurring in the vicinity of junction, which is a complex space curve when the diameter ratio d/D increases. As the new process of theoretical solution and design criteria research developed by the authors, the stress analysis based on the theory of thin shell is carried out for cylindrical shells with normally intersecting nozzles subjected to three kinds of external branch pipe moments. The thin shell theoretical solution for the main shell with cutout, on which a moment is applied, is obtained by superposing a particular solution on the homogeneous solution. The double trigonometric series solution of cylindrical shell subjected to arbitrary distributed normal and tangential forces based on Timoshenko equation is used for the particular solution and the Xue et al.’s solution, for the homogeneous solution based on the modified Morley equation instead of the Donnell shallow shell equation. The displacement function solution for the nozzle with a nonplanar end is obtained on the basis of the Goldenveizer equation instead of Timoshenko’s. The presented results are in good agreement with those obtained by experiments and by three-dimensional finite element method. The present analytical results are in good agreement with WRC Bulletin 297 when d/D is small. The theoretical solution can be applied to d/D ≤ 0.8, λ = d/DT ≤ 8 and d/D ≤ t/T ≤ 2 successfully.

The Effect of Reduced Design Margin on the Fire Survivability of ASME Code Propane Tanks

The design margin on certain unfired pressure vessels has recently been reduced from 4.5 to 4.0 to 3.5. This has resulted in the manufacture of propane and LPG tanks with thinner walls. For example, some 500 gallon ASME code propane tanks have had the wall thickness reduced from 7.7 mm in 2001 to 7.1 mm in 2002 and now to 6.5 mm in 2004. This change significantly affects the fire survivability of these tanks. This paper presents both experimental and computational results that show the effect of this design change on tank fire survivability to fire impingement. The results show that for the same pressure relief valve setting, the thinner wall tanks are more likely to fail in a given fire situation. In severe fires, the thinner walled tanks will fail earlier. An earlier failure usually means the tank will fail with a higher fill level, because the pressure relief system has had less time to vent material from the tank. A higher liquid fill level at failure also means more energy is in the tank and this means the failure will be more violent. The worst failure scenario is known as a boiling liquid expanding vapour explosion (BLEVE) and this mode of failure is also more likely with the thinner walled tanks. The results of this work suggest that certain applications of pressure vessels such as propane transport and storage may require higher design margins than required by the ASME.

Classification of Thermal Piping Loads Using Limit Load Analysis

Equipment nozzle loads essentially originate from sustained (gravity) sources and restraint of the free thermal displacement of the attached piping. A common practice has been to assume that these thermal piping loads develop only secondary stresses. That is, a 1.5Sm [2] check on membrane stress intensities arising from thermal piping loads is typically not performed. The key assumption used in support of this approach has been that these loads decay appreciably with local shell deformation such that the associated stresses are truly self-limiting in nature. This paper illustrates that this assumption may not be appropriate in all instances. A typical pressure vessel and piping configuration is examined. In this example, the associated stresses and deformations developed due to thermal piping loads resulted in significant deformation of the shell arrangement. In static evaluations of local stresses in shells, the ASME Code only offers two classifications that may be applied to stresses resulting from thermal piping loads: primary or secondary. Given these results it may be more reasonable to treat thermal piping load membrane stresses as being primary.

An Iterative Approach for the Treatment of Nonlinear Supports in the Dynamic Analysis of Piping Systems

An iterative approach for the dynamic analysis of piping systems with nonlinearities in the support system is presented. This approach of analysis is based on the use of describing functions technique to obtain quasilinear approximators for the nonlinearities. The quasilinear approximator form and parameters depend on the form and parameters of the input excitation to the system. The describing functions representation is used iteratively in a linear dynamic analysis of the system. This iterative approach can be applied to different types of dynamic analyses employed in the design of piping systems. The iterative method of analysis is presented for the analysis of piping systems with gap and dry friction nonlinearities present in the supports. This method of analysis, which can be integrated into linear elastic analysis programs, is used in a case study of the frequency response analysis of a pipeline model.

Using Finite Element Analysis Combined With Strain Gage Testing to Do Proof Test to Establish MAWP

An amine reboiler was constructed with very large openings in one semi-elliptical head. The openings extended beyond the “spherical” portion of the head into the knuckle region. The vessel was designed to 1998 ASME Section VIII Division 1 (VIII-1). Initially the manufacturer of the amine reboiler vessel chose the proof test after the calculations submitted to the approval agency were not accepted. Non-destructive strain gage proof testing per VIII-1 UG-101(n) was planned, but the minimum proof test pressure to achieve the desired MAWP exceeded the maximum firetube flange test pressure therefore an alternate method was chosen. Finite element analysis (FEA) was done in addition to the strain gage testing. The strain gage results at the maximum hydrotest pressure were used to verify the FEA calculations. The FEA calculated strains were higher than the measured strains. This indicated that the assumptions made in the computer model were conservative. By combining FEA with strain gauge testing, the design was proven to meet Code requirements.

Recent Design Approaches of Large Bore Piping Exeeding 100 of Ratio D/T

Integration and scale up of utility facilities tend to use large bore pipes which are thinner than ever to reach an optimum design. The size range of pipe from 50” to 120” in diameter and ratio of diameter to thickness (D/t) ranges from 100 to 215 are chosen for the optimum design. These size ranges are outside the scope of conventional standards such as ASME piping codes and pipe fitting codes. They are treated in a limited manner in the design guide or manual such as provided by AWWA[1]. However these conventional codes shall be paid much attention when extrapolating the design methods beyond the limitation of conventional methods especially for large bore piping. Also as the detailed analysis methods assure reliable operation, the current design method by using Finite Element Analysis (FEA) has been proposed. It is applied to deal with piping design for the large bore piping in recent overseas projects, such as for the layout of piping, stress concentration on pipe fittings, optimum saddle configurations, and special heavy duty anchor.

Thermal Stresses at Dissimilar Pipe-Stanchion Interfaces

High temperature steam lines in power plant piping systems are often supported by the use of pipe support stanchions welded to the steam pipe. The end of the pipe stanchion has a steel plate welded to it, which typically slides on rack steel. The temperature of the stanchion drops from the process pipe interface along the length of the stanchion. The material for the process pipe carrying high temperature steam can be stainless steel, alloy steel, or carbon steel. The material for the stanchion can also be stainless steel, alloy steel, or carbon steel. It is of course cheaper to use carbon or low alloy steel for the stanchion as there is no steam flow in to the stanchion, when the process pipe is made of stainless steel, or other high alloy steel such A335 Gr. P91. In this paper, finite element thermal analysis is utilized first to obtain steady state temperature distribution due to decay or attenuation from the steam line surface along the stanchion. Conduction of heat from process pipe to stanchion, and convection from stanchion surface are considered. Then finite element structural analysis was performed to obtain steady state thermal stresses at the pipe-stanchion interface utilizing the temperature distribution obtained from thermal analysis as an input. The current industrial practice is to use similar materials for both process pipe and stanchion materials conservatively. Normally encountered pipe materials were considered. The materials studied include 304 & 316 Grade stainless steels, A335 Grades P91, P22, & P11 alloy steels, and A106 Grade B carbon steel. The temperature and stress results are presented. Guidelines are provided for the acceptability of pipe-stanchion dissimilar interfaces.

ASME Section III Design-by-Analysis Criteria Concepts and Stress Limits

The ASME Section III design-by-analysis approach provides stress criteria for the design of nuclear components. Stresses are calculated elastically for the most part, although plastic analysis is recognized. Limits are specified for primary, secondary, and peak stresses. Inherent in these limits are factors of safety against several modes of failure. The purpose of this paper is to explain the design-by-analysis criteria and fundamental concepts behind the approach. With this basis, some of the technical issues that have been identified are discussed.

An Investigation of a Heat Exchanger Tubesheet Crack Caused by a Process Upset

A heat exchanger tubesheet experienced cracks in the ligaments between tube holes. This paper describes the analyses that were performed in order to determine the cause of the crack. The first analysis examined the residual stresses in the tubesheet caused by the tube-rolling. It showed that the residual stresses caused by a typical rolling operation would be negligible. The second analysis examined the steady-state operating condition. It indicated that a tensile stress field existed on the surface of the tubesheet — a potential crack propagation driving force. The third analysis examined a thermal transient caused by a process upset. This transient created high peak stress intensities. However, a fatigue analysis indicated that the stress intensities were inadequate to initiate a crack in the low-cycle fatigue regime. The results of these analyses pinpoint the locations of the cracks accurately, and indicated that while a crack propagation mechanism existed, a crack initiation mechanism did not. Therefore, it was concluded that the crack may have been caused in the manufacturing process.