ASME STP-PT-036:2010 pdf free download – BOLTED FLANGED CONNECTIONS IN ELEVATED TEMPERATURE SERVICE
A similar approach is used by Marin [8] to address creep in a loose-ring flange, however, similarly to previous papers, the outlined approach neglects the effects of strain hardening. The second report [9] of the Institute of Mechanical Engineers Pipe Flange Research Committee was released in 1939 and contained substantial additional test results on gaskets, stud bolts and on both actual flanges and on stacked loose rings. The creep tests on a variety of 0.75 in. diameter stud Ni-Cr-Mo and Cr-Mo bolt alloys under constant stress showed that the compression of the 0.125 in. thick mild steel washer was less than 0.001 in. in all cases and that the creep/embedment of the nuts contributed between 10% and 20% of the total bolt creep. The Ni-Cr-Mo bolt materials were also subject to impact test after creep to examine whether the time at temperature had led to embrittlement of the material. The results demonstrated that specimens subject to no stress exhibited as-new fracture toughness, but that stressed specimens subject to creep damage showed significant loss of toughness. In addition to the bolt tests, 0.0156 in. thick and 0.0312 in. thick asbestos gaskets were tested for both creep and leakage. The creep tests showed that there was initial gasket “flow” with significant reduction in gasket load until a temperature of 300°F, but after this temperature, the gasket stabilized and no further significant creep was witnessed.
The flange creep tests were conducted on 8 inch BS10 Table T flanges with alloy bolts. The flanges were tested under similar conditions to the first report tests, but in this case the steam leak rate was measured and “failure” of the joint was indicated when the leak rate reached 150 grams/hour. It should be noted that in these tests some leakage was witnessed from the early stages of the test, so this quantitative measure of “failure” was nominally selected. The tests were conducted at a range of temperatures and the time to leakage “failure” measured for each case. The flanges were retightened and re-tested in order to generate the effect of successive re-tightening on the time to failure. The results of these tests are shown in Figure 4 and it can be seen that the time to failure increases significantly with successive re-tightening of the bolts. In addition, it can be seen that the relationship between life and operating temperature appears adequately represented by a linear (temperature) vs. log(life) relationship, enabling short term tests to be extrapolated to determine the 100,000 hr temperature limit. In this case, the temperature limit for a life of 100,000 hrs was determined to be 835°F.
The tests did not exhibit a significant difference between dismantling the joint versus simply tightening the bolts after leakage for the no gasket and thin gasket cases. In addition, the advantage of the stress relaxation mode of the joint versus a constant stress test (from which allowable stresses in the creep range are determined) is demonstrated by comparison between the tensile creep tests on the bolt material and joint leakage tests (Figure 5). It should be noted that in this test the results are not directly comparable, in that the joint leakage test bolts were initially tightened to a higher stress (21.6 tons/in² vs. 15 tons/in²) and the tests were conducted at 1000°F versus 975°F for the tensile tests. This difference in temperature corresponds to approximately a halving of the creep life in tensile tests. Therefore, the fact that the bolt strain associated with the flange leakage tests is significantly lower than the tensile test demonstrates the advantage of the flange configuration in creep versus the tensile creep case and the overly conservative nature of using allowable stress values based on a tensile creep test. In addition, it can be seen that the successive re-tightening of the flange did not contribute to accelerated creep damage.