PSI - Issue 17

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

163

2

Nomenclature a 0

Position of load derivation Width of key and keyway Width of big shouldered hub

R

Load Ratio

b c d

R e , R p0,2 Yield Strength

R m

Tensile Strength

Shaft diameter

z β ξ τ

Variable starting on the load introduction side

D 1 D 2

Small shouldered hub outer diameter Big shouldered hub outer diameter

Notch effect factor Relative interference

f

Frequency

Torque stress

l tr

Load-bearing length of key

(...) stat static (...) dyn dynamic

M t

Torque Moment

1. Introduction

Due to their versatility, keyed shaft-hub connections are one of the most frequently used shaft-hub connections. Although they have been used successfully for many decades as a form-fit shaft-hub connection, there are still many uncertainties in the design, which are shown by numerous in-field failures. Especially the simple assembly/disassembly, as well as cost advantages in manufacturing, justify the frequent use. Keyed shaft-hub connections are characterized by one or more "drivers", the key, which transmits the torque between shaft and hub in a form-fit manner. The design and dimensioning of the 3 components are based on extensive research projects (see below) which are formed into the database of several standards. The cross section of the key is standardized according to DIN 6885 (1968) as a function of the shaft diameter. International European standards (e.g. UNI6604, CSN22562 …) using equivalent specifications. American standards (ANSI B17.1) specify a slimmer and squarer key cross section in dependency of the shafts diameter which limits the transferability of the results. The length of the key and its material remain as significant design parameters. The design according to the maximum permissible surface pressure is based on 40 years old investigations by Militzer (1975). Militzer developed a complex mechanical model on the basis of photo elastic investigations, in which the surface pressure distribution in the longitudinal direction can be calculated in dependency of load introduction, the stiffness of shaft, key and hub. The failure location is always in the contact between the key and the shaft keyway for thick-walled hubs (d/D 1 < 0.7). A failure between key and hub can occur with thin-walled hubs made of cast iron or aluminum. This has been investigated by Floer (2000). Bruzek (2014) systematically investigated the failure of torsion-loaded keyed shaft-hub connections. Failure criteria for short keyed shaft-hub connections (l tr /d = 0.5) were determined according to their material and strength reserves were identified. Extensive investigations have also been carried out into the fatigue strength of keyed shaft-hub connections. Weigand/Renneisen (1990) conducted an extensive literature search, which reflects the state of research before 1990. The results were divided into two categories: keyed shafts and keyed shaft-hub connections. In summary, it can be stated that the results of keyed shafts cannot be transferred satisfactorily to keyed shaft-hub connections. The notch effect of the keyed shaft-hub connections is significantly higher than the notch effect of the keyed shafts. On the basis of this research, Oldendorf (2000) carried out extensive investigations on fatigue strength and life estimation of keyed shaft-hub connections. It was basically limited to circular bending with and without static torsion. He also investigated keyed shaft-hub connections with hollow shafts and confirmed the transferability of the notch effect factors from solid shafts to these. A certain minimum wall thickness must be prepared depending on the diameter. Furthermore, he observed a significant increase in the permissible torques for the superposition of keyed shaft-hub connections with press fits. As a result, he gives improved life estimates and important design information for keyed shaft-hub connections. Forbrig (2006) quantified, besides others, the positive effect of an interference between shaft and hub on the fatigue strength of a bending loaded keyed shaft-hub connection. Furthermore, superimposed static torsion or coating with hard layers (e.g. DLC) increase the strength. Hofmann (2014) investigated the influence of different case hardening processes on bending loaded keyed shaft-hub-connections. It was observed that the higher material strength did not lead to a higher fatigue strength of keyed shaft-hub connections. In the case of keyed shaft-hub connections subjected to bending loads, friction wear was repeatedly identified as the cause of damage (Oldendorf). A failure in contact between two