11 октября, 2015 
Pokraskin D. Croccolo * M. De Agostinis, G. Olmi and P. Mauri
University of Bologna — DIN, Bologna, Italy
Press fitted and adhesively bonded joints give several advantages: among others, they lead to smaller constructions for a given load capacity, or to stronger constructions for a given size. For that reason, the static and fatigue strength properties of these joints have been studied extensively. Studies dealing with the influence of the Engagement Ratio (i. e. the coupling length over the coupling diameter) are still limited. This work aims at filling the gap, limited to the case of a single component anaerobic adhesive. Coupling and decoupling tests have been performed both on press-fitted and adhesively bonded specimens and on pin-collar samples, considering four different levels for the engagement ratio. The study shows that the engagement ratio has a negligible effect on the shear strength of the adhesive and also on the relationship between the decoupling and the coupling forces. Moreover, the obtained results show that a too high interference level in press-fitted and adhesively bonded joints may have a detrimental effect on the adhesive strength.
Correspondence information: dario. croccolo@unibo. it.
Keywords: Anaerobic Adhesives, Engagement Ratio, Hybrid-Joints, Slip — Fitted Joints, Shear Strength
A Coupling surface [mm2]
Dext, H External hub diameter [mm]
Dint, H Internal hub diameter [mm]
DC Coupling diameter [mm]
Dext, S External shaft diameter [mm]
E Young’s modulus of the shaft and hub material [MPa]
Fint Interference contribution to the axial release force of the joint [N]
Fad Adhesive contribution to the axial release force of the joint [N]
Fcoupi Pushing in force during shaft-hub coupling (peak value) [N]
Ftot Pushing out (or release) force during shaft-hub decoupling (peak
value) [N]
Kj Ratio between the maximum release force and the maximum
pushing in load LC Coupling length [mm]
pC Coupling pressure [MPa]
QH Ratio between the internal and external hub diameters R2 Linear Correlation coefficient for linear regressions
RaH Arithmetic mean surface roughness, considering the internal
surface of the hub [pm]
RaH Arithmetic mean surface roughness, considering the external
surface of the shaft [pm]
U Nominal interference [mm]
Z Actual interference [mm]
ee Strain measured (by a strain gage) on the external surface of the
hub in the tangential direction jxA Axial static coefficient of friction
v Poisson’s ratio
nad Adhesive static shear strength [MPa]
ANOVA Analysis of Variance ER Engagement Ratio
HJ Hybrid Joint (i. e.: press-fitted and adhesively bonded joint)
PFJ Dry Press-Fitted Joint (i. e.: coupled with mere interference joint) SFJ Slip-Fitted Joint (i. e.: Pin-Collar joint)
The research in the field of joint design is nowadays more and more focused on efficient solutions, which are able to ensure the required connection safety and at the same time the reduction of overall structure weight. A possible option consists in interference joints, where the required interfacing pressure and friction are generated by a proper choice of the coupling tolerances. However, achieving strict tolerances usually implies an increase of manufacturing cost. Moreover, an additional drawback of a friction connection between a shaft and a hub consists in the generation of a not negligible tensile stress state in the hub, especially when the interference level is high. For this reason, the adoption of press-fitted and adhesively bonded joints, usually regarded as Hybrid Joints (HJs), is getting more and more frequent in industrial engineering, especially when the transmission of great powers and torques is required. The addition of a suitable adhesive makes it possible to reduce the interference level at the interface, by taking advantage of the bonding layer strength. As a consequence, coarse tolerances may be used, with good outcomes from the point of view of manufacturing costs. Moreover, the tensile stress acting on the hub is strongly decreased. More traditional joining techniques (e. g., keys, pins, interference-fits, bolted joints…) may also be replaced by adhesives, otherwise they can be used in combination. As a matter of fact, dry press-fitted couplings (without adhesive) achieve only 20-30% of nominal contact surface, in the case of a metal to metal joint. On the other hand, the adhesive is able to fill the micro spaces between the crests of surface roughness, so that the contact is extended over the entire area of the mating surfaces [1-5]. A possible drawback of bonded joints stands in the strength of HJs being quite difficult to estimate, since it depends on several factors, such as the coupling pressure [6], the type of materials in contact [1, 7], the surface roughness [8], the curing type and the curing methodology [9], the operating temperature [10] and the loading type [11]. The study in [12] deals with the effect of the type and the way of assembling on the resistance of interference fitted and adhesively bonded joints. In particular, the shear strength of the adhesive is compared, considering press-fitted, shrunk-fitted and cryogenic-fitted bonded specimens. For this purpose, pin-collar specimens were manufactured according to ISO 10123 [13] and involved in release tests after press-fit coupling, whereas, bigger sized specimens were used for shrink-fit and cryogenic-fit, so that the thermal dimensional variations could be more significant. The outcomes of that research indicate that the sample dimensions may affect the adhesive response. In particular, the release strength could depend on the engagement ratio (ER=LC/DC), namely the ratio between the axial length LC of the shaft — hub sample and the coupling diameter DC. In the case of press-fitted and adhesively bonded joints it is also possible to relate the maximum coupling force, namely the ultimate force measured during the pushing in operation, to the maximum pushing out load. In particular, in [12, 14], it was retrieved that the two peak forces are directly proportional and that the ratio between the release and the coupling maximum loads is about 1.2 for an ER equal to 1. This ratio will be referenced as Kf. as a further point of discussion, it is important to investigate if this result is maintained even for different ER values. the main motivation arises from some discrepancies observed when the ER becomes lower than 1 [12, 14]. To the authors’ best knowledge, no results are available in the literature about the influence of the ER on the static strength of the joint and on the ratio between the release and the coupling loads. These two issues have been tackled in the present paper, by running tests on press-fitted and adhesively bonded specimens and on pin-collar samples. The statistical tools of Design of Experiment have been applied for the full randomization of the test orders and for the comparisons of the statistical distributions.
Опубликовано в рубрике 