Page 215 - Biaxial Multiaxial Fatigue and Fracture

Estimation of the Fatigue LiJe of High Strength Steel Under ... 199

Table 3. Calculated fatigue lives Nb,eaf [blocks] under uniaxial variable-amplitude tension-
compression according to the Palmgren-Miner (PM) and Serensen-Kogayev (SK) hypothesis
for different coefficients a

ca max PM PM SK SK SK

[MPa] a=l a=0.25 a=l a=0.81 a=0.64
675 * 1137 276 160 109 65

675** 1137 313 160 109 65
736** 42 1 180 59 41 32
ER 942% 240% 48% 18% 3 8%
* - course CARLOS-fl (13568 extrema), ** - course CARLOS42 (46656 extrema)

Table 4. Results of fatigue life calculations under variable-amplitude loading according to
the Serensen-Kogayev hypothesis with coefficient a = 0.8 1

Tension- compression Tension with torsion Torsion

uamax Nb,cal %a1 Nb,cal %a1 Tamax Nb,cal "tal
[MPal [blocks] ["I [MPal [blocks] ["I [MPal [block] ["I

675* 109 0 595" 150 18 488* 208 45
675** 109 0 595** 150 18 488** 208 45
736** 41 0 525** 62 45
* - course CARLOS-fl (13568 extrema), ** - course CARLOS42 (46656 extrema)

CONCLUSIONS
The normal strain energy density parameter in the critical plane is an efficient quantity
describing the fatigue life under variable amplitude tension-compression, torsion and
proportional tension with torsion.
Under constant amplitude torsion where the stress gradient occurs, we can determine the
equivalent stress amplitude and accumulate the damage with use of the Wohler curve for
tension-compression.
Application of the linear hypothesis of damage accumulation, formulated by Palmgren-
Miner, leads to too high fatigue life as compared with the experimental life. Application
of the Serensen-Kogayev hypothesis, including amplitudes of the analysed energy pa-
rameter higher than 81% of the fatigue limit (expressed in energy), leads to correct esti-
mation of the fatigue life.
Application of the energy parameter distinguishing tension (positive value of the energy
parameter) and compression (negative value of the energy parameter) for fatigue life es-