The Influence of Load Misalignment During Uniaxial Low‐cycle Fatigue Testing—ii: Applications

Abstract

Models for predicting scatter bands due to bending have been applied to four alloys, namely AISI 316L, Nimonic 101, 9 Cr-l Mo, and IN 718. The alloys were tested extensively by 26 laboratories in an international round robin exercise sponsored by the Community Bureau of Reference (BCR) of the EC. After initially selecting data for analysis on the basis of their confirmed conformance to the ASTM bending criterion, it has been shown that in all four materials a major fraction of the data scatter could be attributed to bending. Furthermore, at the lowest strain range the predicted bending component represents the highest proportion of the experimental interlaboratory scatter. Many laboratories did not report any measure of bending and so could not be used in the initial analysis. However, a further interesting deduction from the models is that the entire BCR data-set can be encompassed within a scatter band based upon a bending criterion that is twice the allowable ASTM limit. Differences in the extent of scatter between materials at a given total strain range can be attributed to the gradient of the logarithmic plot of total strain range as a function of lifetime. NOMENCLATURE c = distance from neutral axis C, C' = material constants e = lateral offset E = Young's modulus E, = cyclic modulus ( = Ao/A€) f= extensometer factor = 1 for dual extensometer = 2 for single extensometer F = total axial force through testpiece I = moment of inertia of the cross-section K = material constant 1 = testpiece length L = total length of machine's cross-beam M = bending moment Nrl = repeatability minimum-cycles-to-failure Er2 = repretability maximum-cycles-to-failure IVr, = reproducibility minimum-cycles-to-failure Nr2 = reproducibility maximum-cycles-to-failure NF = number of cycles to failure under no bending A F f = range of repeatibility scatter-band (= Nr2;Nrl 1 AN, = range of reproducibility scatter-band (= Nr2-NfI ) r = testpiece radius R = radius of curvature z = distance CL, a' = exponents = slope of log A6 and log N f curve 529 530 F. A. KANDIL and B. F. DYWN y = angular offset; angle of deflection tb = maximum bending strain tbx = bending strain at point x tk = maximum bending strain at axial compressive peak strain cb, = maximum bending strain at axial tensile peak strain c, = control-strain amplitude t, = strain amplitude at point x to = strain amplitude at the centre of testpiece t , = maximum strain amplitude t2 = minimum strain amplitude At = strain range Ac,, = bending strain range ( = t b , -tb) Atc = control-strain range Ate = elastic strain range Acp = plastic strain range Au = stress range u,, = bending stress 0 = angle in the circular cross-section between the intersections of the maximum bending plane and the extensometer plane

The Influence of Load Misalignment During Uniaxial Low‐cycle Fatigue Testing—i: Modelling

“…The reproducibility limits predicted in Fig. 8 will be discussed further in Part I1 of this paper [3] in connection with the extremes of scatter found in the BCR round robin exercise [1,2]. Numerical prediction for the limits of the scatter band according to equations such as (30) require the determination of the bending error (ACb/Atc) at conditions corresponding to the (tensile and compressive) peaks of stress/strain in the fatigue test.…”

“…It will be shown that the magnitude and bias of the data-scatter depends upon both the type of extensometer and the material-dependent fatigue crack initiation mechanism. In Part I1 of this paper [3], the predictions of this modelling will be compared with lifetime data obtained on the four materials used in the BCR exercise.…”

Uncertainty in fatigue loading: Consequences on statistical evaluation of reliability in service

“…Kandil and Dyson [8,9], by analyzing inter-laboratory low cycle fatigue datasets [10], also showed that, even if a test is performed under the stringent conditions imposed by the ASTM standard E 606-80 [11] concerning the maximum allowable bending stress in axial fatigue tests (smaller than 5%), the ratio of the reproducibility limits of the fatigue life reaches the value of six suggesting a reappraisal of the British Standard BS 7270:1990 [12] and ASTM standards E 606-80 [11] and E 1012-89 [13]. The above cited results were confirmed in a further research supported by the Versailles Project on Advanced Materials and Standards [14,15].…”

Influence of bending on low cycle fatigue life of cylindrical test pieces at elevated temperature

“…With fatigue cyclic-through-zero-testing, Kandil and Dyson [1,2] differentiate between misalignment due to ''non-reversed'' bending and ''reversed'' bending ( Figure 2). Non-reversed bending misalignment is caused by angular or lateral offset of the test piece axis with respect to the grips.…”

“…Here, bending strains change sign in a tension-compression cycle and therefore contribute to cyclic strain loading. Load -strain offset is believed to be more damaging to fatigue testing than angular or lateral offset [1,2].…”