PSI - Issue 17

Felix Kresinsky et al. / Procedia Structural Integrity 17 (2019) 162–169 Felix Kresinsky / Structural Integrity Procedia 00 (2019) 000 – 000

164

3

components, based on local approaches (via FEA), makes impossible the strength calculation of keyed shaft-hub connections (FKM (2012), Wendler/Kresinsky (2016)). Furthermore, there are international studies (Pedersen (2010), Fessler (1983), Eissa (1983)) on the stress distributions at the shaft keyways. Important mechanical tension correlations to the load-bearing behavior of keyed connections are established but the transferability of the results to the whole connection is unsatisfactory. The transferability of the results is rather qualitative, as some of them are based on other key cross-sections.

2. Experimental Procedure

2.1. Test Bench

A keyed shaft-hub connection is a complex tribo-mechanical system in which three components in contact are subjected to high dynamic stresses. Therefore, the described factors and correlations (analogous to fatigue strength) can only be determined experimentally, whereby a reasonable statistical validation was aimed at. The experimental investigations focused exclusively on torsion loads as well as their effects. A test bench with a hydraulic rotary actuator was available to carry out the tests (Fig. 1 left). The maximum transferable static torsional moment is M t,stat = 8000 Nm and M t,dyn = ±5600 Nm can be transmitted dynamically. Based on a keyed shaft-hub connection, a shaft diameter of d = 40 mm thus results in a maximum possible load of τ stat = 636 MPa (static) or τ dyn = 445 MPa (dynamic). The test frequency was 20 Hz. Five million load cycles were selected as the number of limit load cycles. The torsional moment was purely swelling (Fig. 1 right), typical for keyed shaft-hub connections. The torsional moment was transmitted from the rotating cylinder via the flexible coupling into the shaft of the specimen and derived via the hub which is flanged to the torque sensor. All experiments were carried out in oil bath.

Fig. 1. Servo hydraulic test bench with applied torque load.

2.2. Material and Geometry Data

The shafts material with their specific material data as well as the hubs material are shown in Table 1. The known influence of size (Lätzer (2016)) reduced the strength of the hub materials, but this has no influence on the expected failure between shaft keyway and key. The key was made of C45+QT in all tests because its material has no influence on the strength of keyed shaft- hub connections according to Hofmann (2015). Keyed shaft-hub connections can fail via two different criterions: On the one hand, the maximum permissible surface pressure at the contact surfaces between shaft and key or hub and key. Excessive surface pressure can lead to an inadmissible keyway expansion and consequently to a deflection of the keyway. The permissible surface pressure can be calculated on the basis of DIN 6892 (2012). On the other hand, a keyed shaft-hub connection can fail via a fatigue crack of the shaft. The fatigue strength of the shaft or shaft-hub connection can be calculated with the DIN 743 (2012) or FKM (2012). The notch effect is verified only on the basis of the tensile strength of the shaft (see also Bruzek (2012)). The strength verification for all keyed shaft-hub connection under torque load has to be performed according to both criteria (Tab.1)!