API TR 6AF2
Technical Report on Capabilities of API Flanges Under Combinations of Loading - Phase II
|Publication Date:||1 April 2013|
The evaluation of the load carrying capacity of API 6A integral flanges is the objective of this work. The applied loading includes the end tension and bending moment in addition to the conventional rated pressure and makeup forces. The effect of a temperature difference corresponding to 250 °F on the inside and 30 °F on the outside was also evaluated.
Three-dimensional finite element meshes were generated for the
Type 6B, and Type 6BX flanges. The bending moment load case
required a model of one quarter of the flange which was built up
from the smaller segments and the half-bolt superelements. The
computer program SESAM was used to obtain the stresses at selected
critical flange and hub sections and to determine the gasket
reaction due to each of the four unit load cases and the
temperature difference load case. Leakage criterion was defined as
the load combination which reduces the initial makeup compressive
forces in the gasket to zero. The stresses in each defined section
were linearized in accordance with the ASME Section VIII, Division
2, procedure to determine the membrane and membrane-plus-bendin
The results of the analysis carried out indicate that the leakage criterion governs the capacity of the smaller flanges in the Type 6B flanges. Leakage was governing for up to 9 in. size flanges in both the 52.5 ksi and 40 ksi makeups for the 2000 psi pressure. Leakage was governing the 5 1/8 in. for the higher pressures. Leakage was also found to be governing all Type 6BX flanges for working pressures of up to 5,000 psi. For the 10,000 psi and 15,000 psi flanges, leakage governed only in the larger size range greater than 2 9/16 in. Leakage was governing in all the 20,000 psi API 6BX flanges. The leakage model adopted in this study employs several approximations that have not yet been evaluated. Therefore, the actual leakage forces, i.e. load combinations leading to leakage, may be considerably higher than assumed herein. In reality, the gasket only leaks when its energized capacity is exceeded.
The state of stress at the stress governing hub section under the combined loading of makeup, pressure, tension and bending moment is considered to be "secondary." However, when pressure, tension, and bending moments are applied together with the necessary makeup to resist these actions without leakage, the state of stress is rendered "primary" and, therefore, the allowable stress intensities are halved. This does not seem to be consistent, and it may by far exceed the intention of the code. However, the oversight subcommittee preferred to adopt the conservative route, which may be overly conservative pending further evaluation. Therefore, it may be concluded that when the hub stresses are treated as primary, most flanges do not possess significant reserve strength beyond the leakage condition. In fact, if the leakage condition was somewhat conservative, the stress condition may become governing for most flanges.
The temperature difference of 250 °F internal and 30 °F external leads to increases in the load-carrying capacity of the flanges. This condition is caused by the compressive forces generated in the gasket due to this temperature difference, and the increase in the allowable stresses when the self-limiting temperature load condition is included. It is recommended that a 3-D finite element, nonlinear material and geometric models of approximately eight flanges be carried out to determine the actual failure mechanism that governs the behavior of these flanges. This includes the prediction of the response of the gasket under increasing load and a more accurate definition of the leakage mechanism. The elimination of the raised face does not significantly reduce the stresses in the hub which caused six Type 6B flanges to fail to meet the ASME criterion for makeup load only (52.5 ksi for 105 ksi bolting). The stress intensities were reduced only by about 5 % when the raised face was eliminated, increasing the thickness of the flange by about 10 %. The hub thickness for these flanges had to be increased by up to about 27 % of their existing thicknesses together with the elimination of the raised face.
The bolt stresses did not govern for any of the flanges analyzed. Bolt stresses are typically within approximately 67 % of their yield strength due to makeup, pressure, tension, and bending moment loads. The bolts are expected to be made up to half their yield. The stresses in the bolts due to temperature differences increase by about 5 ksi to 7 ksi, which is about 6 % to 8 % of the bolt yield stress. The other load conditions (pressure, tension, and bending moments) increase bolt stress by twice the increase due to the temperature difference. Therefore, it is concluded that the bolts will not approach their limiting criterion under the investigated load conditions.