ICF10B

1500 oK at the surface to within 1oK of ambient (300oK) at about 1 mm below the surface, for the few ms of convective heating typical of cannon firing. Temperature dependent material properties [6-8] of the 0.1 mm thick electro-plated chromium bore coating and the Ni-Cr-Mo steel substrate were used for the analysis, in the form δ(T) = C0 + C1 T , with values of C0 and C1 as in Table 1. The inputs to the finite difference calculations, in addition to the chromium and steel properties, were: the initial ambient temperature, Ti , 300 oK; the duration of the convective heating pulse at the tube surface, 0.008 s; the convection coefficient, h, of the heating pulse, 193,000 W/m2 oK; and the mean gas temperature of the pulse, with values as discussed in the upcoming results. TABLE 1 – TEMPERATURE DEPENDENT PROPERTIES OF CHROMIUM AND STEEL Chromium Ni-Cr-Mo Steel property units range; oK C0 C1 C0 C1 thermal diffusivity, δ m2/s 300 - 2000 29.6 E-6 -12.6 E-9 11.7 E-6 -5.3 E-9 thermal conductivity, k W/m oK 300 - 2000 97.2 -0.0266 43.6 -0.0097 elastic modulus, E GPa 300 - 1000 -- -- 248 -0.0968 thermal expansion, α 1/oK 300 - 1000 -- -- 13.5 E-6 0 yield strength, SY MPa 300 - 1000 -- -- 1740 -1.62 Expressions for the near-bore, transient, in-plane, biaxial compressive thermal stress, ST, and the tensile residual stress, SR, produced when the transient stress exceeds the material yield strength are: ST = - E α [ T{x,t} - Ti ] / [1 - ν] [6] where ν is Poisson’s ratio; the transient temperature, T is for a given depth, x below the bore surface and time, t after the start of a heating pulse; the term [1 - ν] accounts for the biaxial nature of the temperature and stress distributions. Then SR is determined directly using the linear unloading concept: SR = - ST - SY. Fatigue Intensity Factor versus Life The effects of autofrettage and thermal residual stresses on fatigue life will be described in the work here using the fatigue intensity factor (FIF) concept [5]. In brief summary, the FIF concept adds a quantitative measure of material yield strength and initial crack size to the stress parameter in a log stress vs log life description of fatigue life behavior of a structure. By adding material strength and initial crack size to S-N life descriptions, significant improvements in life modeling of cannon pressure vessels have been shown. The key expressions for an FIF description of ID-initiated fatigue life of a pressurized, autofrettaged cannon pressure vessel are: fatigue intensity factor = ∆SLOCAL × (SY-NOM / SY) × (ai / ai-NOM) 1/6 [7] ∆SLOCAL = kt SAPPLIED + SRESIDUAL + p [8] Note that the local stress range for ID-initiated cracking includes: the familiar Lamé applied ID stress for the appropriate p and b/a, with stress concentration factor, kt if one is present; the ID hoop residual stress, accounting for the Bauschinger reduced strength, using Eqs. 1-4, but no kt even if present because it would be negated by the Bauschinger effect; and an added stress equal in magnitude to p, to account for the effects of pressure on the crack faces [5]. The second term in Eq. 7 accounts for the yield strength of a given vessel, when different from the nominal yield strength of the vessels under comparison. The third term accounts for the initial crack size of a given vessel, if different from the nominal initial crack size of the vessels under comparison. As will be seen in the forthcoming results, the effect of thermally induced tensile residual stresses in cannon vessels can be accounted for by using ai values associated with the depth of the tensile residual stresses in cannons critically subject to hydrogen cracking. RESULTS Mechanical Residual Stress Effects on Fatigue Life A baseline description of the effect of mechanical autofrettage residual stresses on ID initiated fatigue life for cannon tubes is available from prior results of full-scale cannon fatigue tests [5]. Four groups of the prior fatigue tests that were ID initiated and had well documented initial crack sizes will be used for a baseline fatigue life, see Table 2. Two of the groups were rifled cannons, with a kt of 1.7 to account for the radius

RkJQdWJsaXNoZXIy MjM0NDE=