ASME STP-PT-027:2009 pdf free download.EXTEND LOW CHROMESTEEL FATIGUE RULES.
3 CREEP-FATIGUE DATA
Creep-fatigue data have been developed in tests utilizing many combinations of strain rdnge. hold time. temperature and load measurement. For the most part, creep-fatigue tests arc run with loads that cycle between compressive and tensile and with tensile hold periols ranging from seconds to times exceeding a few minutes, but rarely more than an hour. Plastic strain amplitudes typically do not exceed 1-2 percent and are usually only a fraction of 1 percent. The total number of cycles applied before failure or a specific load reduction is reached may extend into the thousands, but because of the high cyclic frequency, the total time of exposure may be only tens or, at most, a few hundreds of hours.
For the purpose of this study, creep-fatigue data on several of the strain softening alloys were gathered from many sources. The data from which the plastic strain range may be estimated are shown in Figure 3. Included in the plot are some data from tests that show the effects of tensile hold times. Most of the data are from relatively high frequency tests where the accumulated time at creep temperature is very short. It appears that longer hold time tests result in fewer numbers of cycles. i.e. there is a creep-fatigue interaction. A line provided by a producer of 2 ¼ Cr-I Mo-V alloy, shown in Figure 3, did not include significant hold time effects.
The most widely scattered points in the figure are for hold time tests of one brittle heat of an alloy for which the “no hold time” tests were also mainly outside the scatter band. It is not expected that pressure vessel alloys of interest in this project will behave in a creep brittle manner when tested in uniaxial tension.
The curves presented in Figure 10 start with a hold time of 15.000 hours for the maximum plastic strain amplitude covered (2%). The hold time was reduced by the one third power of the plastic strain to permit more cycles at low strain amplitudes wherein less damage is done per cycle. The strain damage factor, fi. was kept at the benchmarked value of 2 observed for the I Cr Mo V alloy as noted above.
For specified values of plastic strain amplitude, the number of design cycles is calculated using the above equation for several conservatively calculated rupture lives. The stress tbr each value of plastic strain was computed using a simple work hardening law based on minimum specified nxm temperature properties reduced to the values at X50F. Work hardening was based on a plastic strain of lxl0 at 30 ksi (about 2/3 of yield) and 0.002 at 48 ksi, about the 0.2% offset stress. The strain hardening coefficient was then calculated to be 0.0618.
The design curves bear a striking oftet relationship from the test line for 2 ¼ Cr-lMo-V. This is remarkable since the ‘no hold limeTM test results were not used in the modeling. The appearaiice derives perhaps from the flict that hold time data on a similar Cr-Mo-V alloy was used in henchmarking the proposed model.
