INTERNATIONAL DESIGN CONFERENCE - DESIGN 2002

Dubrovnik, May 14 - 17, 2002.

DESIGN OF AXIAL PIN CONNECTIONS FOR

TORQUE TRANSMISSION

H. Birkholz

Keywords: Shaft-Hub-Connection, Axial Pin, Calculation Methods,

Design

1. State of the Art

Axial pin connections consist of pins arranged axially around the perimeter as driving pins (Fig. 1).
The pins can be varied with respect to number and arrangement.

dS
da,N

dW

Figure 1. Longitudinal pin connection as driving pin in a shaft-hub-connection

There is no literature concerning the design of longitudinal pin connections with clearance
[Faulwasser 1998]. It appears feasible to use the data for no-clearance connections as reference point.
These design references are however few. The first dimensioning basis for the longitudinal pin
connections with clearance was developed at the Institut für Maschinenwesen within the framework of
a research project sponsored by the DFG.

2. Determination of Stress concentration factors

2.1 Stress concentration factors for Torsion
Due to the lack of knowledge concerning the form and fatigue strength of shafts and hubs weakened
by a semicircular notch in the axial direction [Kollmann 1984], the work here concentrates on the
determination of stress concentration factor diagrams for shafts and hubs with the help of the finite
element method. The highest notch stress determined using the FE calculation is divided by the
analytically calculated stress in the undisturbed cross-section:

maximum stress by FE - Analysis (1)

αK =
nominal stress, analytical

TECHNICAL SYSTEMS 1029

This methodology was carried out for the shaft and hub-sided notch for different types of stress and
variations in geometry:
• shaft: ratio of pin diameter to shaft diameter dS/dW; number of driving pins
• hub: ratio of outer hub diameter to shaft diameter da,N/dW at various dS/dW; number of driving
pins
The stress was determined both analytically and using the FEM for an unattenuated geometry in order
to clarify the mounting and load conditions. The chosen boundary conditions with reference to
[Wesolowski 1994] resulted in a maximum deviation of < 1 % from the analytical calculations, i.e. an
adequately high precision was guaranteed. The position of the pins was arranged so that the seam
radius between the shaft and hub runs through the central point of the pin. The parameter calculations
were then carried out using these boundary conditions. Figure 2 shows the stress concentration factor
diagram of the shaft for torsion.
3.3
1 pin
3.1 2 pins
stress concentration factor αkt

3 pins
2.9
4 pins
2.7 6 pins
8 pins
2.5 12 pins

2.3

2.1

1.9

1.7

1.5
0.1 0.125 0.1875 0.25 0.3125 0.35
ds/dw

Figure 2. Stress concentration factor for the shaft at various dS/dW ratios and n pins
For 8 and 12 pins, no meaningful geometry resulted for dS/dW > 0.25 or 0.1875 so that the stress
concentration factors could not be determined in this range. Equations for the analytical determination
of the stress concentration factors were then derived from the curves.
It should be mentioned that although the notch radius increases with the ratio dS/dW and therefore
reaches good values, an increase in the stress concentration factor results. This is due to the influence
of the reduced load-bearing cross-sectional area which outweighs that of the increased notch radius.
Figure 3 shows the stress concentration factors determined for the hub at a ratio of da,N/dW =2.
1.45
1 pin
stress concentration factor αkt

1.4 2 pins
1.35 3 pins
4 pins
1.3 6 pins
8 pins
1.25 max. stress
12 pins at hub notch
1.2

1.15

1.1 max. stress

outer perimeter
1.05

1
0.1 0.125 0.1875 0.25 0.3125 0.35
ds/dw

Figure 3. Stress concentration factor for the hub, da,N/dW = 2; various dS/dW ratios and n pins
The stress concentration factor diagram was also determined for the ratios da,N/dW = 1.5 and da,N/dW =
2.5. The further away a point in a hollow shaft is from the central axis, the greater the torsion stress at
this point. Therefore, the maximum stress must not necessarily occur in the internal notch. The

1030 TECHNICAL SYSTEMS

transitional range of maximum stress from the interior to the exterior is shown in Figure 3 in addition
to the course of the stress concentration factors. At a ratio of da,N/dW = 1.5, the maximum stress is
independent of the number of pins and the ratio dS/dW is always in the range of the notch. At a ratio of
da,N/dW = 2.5, the maximum stress occurs at the outer diameter.

2.2 Stress concentration factors for Bending

The stress concentration factor diagrams for the bending strain are formulated using a procedure
analogous to that for the determination of the stress concentration factors for torsion. Figure 4 shows
the stress concentration factor diagram for the shaft.
1.6
1 pin
2 pins
1.5 3 pins
stress concentration factor αkb

4 pins
6 pins
1.4
8 pins
12 pins
1.3

1.2

1.1

1
0.1 0.125 0.1875 0.25 0.3125 0.35
ds/dw

Figure 4. Bending stress concentration factor for the shaft for various dS/dW ratios and n pins
The interaction between the notch radius and the weakened cross section can also be observed for this
type of stress.
Figure 5 shows the bending stress concentration factor diagram for the hub with a ratio of
da,N/dW = 1.5.
1.3
1 pin
stress concentration factor αkb

2 pins
1.25
3 pins
4 pins
1.2 6 pins
8 pins
12 pins
1.15

1.1

1.05

1
0.1 0.125 0.1875 0.25 0.3125 0.35
ds/dw

Figure 5. Bending stress concentration factor for the hub, da,N/dW = 1.5; various dS/dW ratios and
n pins
The values for da,N/dW = 2 and da,N/dW = 2.5 lie clearly beneath those shown in Figure 5. Already at the
ratio da,N/dW = 2, the maximum stress concentration factors occurring lie at αkb = 1.07, so that the use
of a stress concentration factor αkb = 1.1 above a ratio of da,N/dW = 2 for a stability test represents a
conservative design.

3. Influence of the connection width on the load-bearing capacity

The few sources available give a ratio of connection length to shaft diameter l/dW of 1...1.5 as
approximate value for the dimensioning of a longitudinal pin connection [Niemann 1981] [Fronius
1987] [Wächter 1987]. However, extensive investigations of cog-shaft connections [Wesolowski
1997] show that at a ratio of l/dB > 0.6 (l: connection width, dB: reference diameter for spline shafts),
the maximum stress resulting from the jump in stiffness at the start of the connection cannot be

TECHNICAL SYSTEMS 1031

reduced. The maximum stress drops steeply in the axial direction (higher stress gradient) and then
flattens out to a low level until the end of the connection.
Since the geometry of a longitudinal pin connection suggests an analogy to a spline-shaft, the range
l/dW = 0.5...0.75 was examined more closely. Figure 6 shows the influence of the connection length
on the pin.
l=16mm
l/dw=0.5
l=20mm
l/dw=0.625
l=24mm
l/dw=0.75

shear stress τ

pin length

Figure 6. Influence of the connection width on stress for dS/dW = 0.125; dW = 32mm, 4 pins
A distinct influence of the connection length on the pin can clearly be seen. The shearing stress could
be reduced up to 16 % by using a longer connection. This supports the conclusion that a wider
connection reduces the stress of the entire connection.
Figure 7 shows the course of the equivalent stress in the notch base of the hub. A reduction of the
stress in the hub analogous and in the same order of magnitude as that in the shaft can be observed.
l/dw=0.5
l=16mm
l=20mm
l/dw=0.625
l=24mm
l/dw=0.75
equivalent stress

connection length

Figure 7. Influence of the connection width on the stress in the hub for dS/dW = 0.125,
dW = 32mm, 4 pins
Figure 8 shows the course of the equivalent stress for a shaft.
connection
border l/dw=0.
l=16mm
5l=20mm
l/dw=0.625
l=24mm
l/dw=0.7
5
equivalent stress

shaft length

Figure 8. Influence of the connection width on stress in the shaft for dS/dW = 0.125,
dW = 32mm, 4 pins
A change in the connection width has a significantly greater effect on the shaft than the pin and the
hub. A corresponding course of equivalent stress can be seen for all ratios l/dW in the undisturbed
section of the shaft, in other words before the beginning of the connection. A section with a distinct

1032 TECHNICAL SYSTEMS

stress increase is present at the beginning of the connection, which exhibits an increased value of 60%
opposed to the undisturbed section for the ratio l/dW =0.5. A widening of the connection leads to a
significant reduction of the overstress to 19%. At ratios above l/dW = 0.625, it must be observed that
no significant reduction in the overstressing occurs. Therefore, it can be recommended that a ratio of
the connection width to shaft diameter of l/dW = 0.6...0.7 should be maintained.

4. Influence of the pin diameter on the connection stress

A reduction of the shearing stress in the pin is to be expected due to the increase of the pin diameter
and the corresponding increase of the shearing area. This is clearly shown in Figure 9. The shearing
stress in the pin is significantly reduced.

ds/dw=0.12
4mm
5
6mm
ds/dw=0.1875
10mm
ds/dw=0.312
5
shear stress

pin length

Figure 9. Influence of the pin diameter on the pin stress. 4 pins, dW = 32; l/DW = 0.625;
da,N/dW = 2
The level of stress in the notch root is shown in Figure 10.
connection
equivalent stress

ds/dW = 0.125
ds/dW = 0.1875
ds/dW = 0.3125

shaft length

Figure 10. Influence of the pin diameter on the shaft stress. 4 pins, dW = 32; l/DW = 0.625;
da,N/dW = 2
Due to the increase in the pin diameter, the stress peak at the beginning of the connection is reduced
on the one hand and on the other, the cross-section of the shaft in simultaneously weakened causing
the stress before the contact to increase significantly. With a pin diameter of dS = 6 mm, which
corresponds to a ratio of dS/dW = 0.1875, the stress before and in the contact zone are almost equal for
a connection with 4 pins.
In contrast to the shaft, the equivalent stress in the hub sinks, whereby the changes are slight for a pin
diameter of 6 mm, Figure 11. An increase in the stress is to be expected for larger pin diameters dS>10

TECHNICAL SYSTEMS 1033

mm because the load-bearing cross-section is reduced.

dS/dW = 0.125
4mm
dS/dW = 0.1875
6mm
10mm
dS/dW = 0.3125

equivalent stress

connection length

Figure 11. Influence of the pin diameter on the hub stress. 4 pins, dW = 32; l/DW = 0.625;
da,N/dW = 2

5. Conclusion
Extensive parameter studies lead to the development of stress concentration factor diagrams for the
shaft and hub for the torsion and bending stress for a shaft-hub connection with axially arranged pins.
The influence of the different geometry variations could clearly be seen when the entire longitudinal
pin connection system is taken into consideration. A protrusion of the pin at the beginning of the
connection reduces the stress in the pin and shaft, the influence on the hub is only slight. The
advantage over other stress-reducing measures, such as increasing the number of pins or the pin
diameter, lies in the fact that neither the shaft nor the hub are weakened through a higher notch effect
or that the manufacturing expenditure increases through additional notches and pins.

References
Faulwasser, M.: Beanspruchungsverhalten drehmomentbelasteter Längsstiftverbindungen. unveröffentlichte
Studienarbeit, IMW, TU Clausthal, 1998
Kollmann, F.G.: Welle-Nabe-Verbindungen. Springer-Verlag, Berlin, Heidelberg u.a., 1984
Niemann, G.: Maschinenelemente. Erster Band: Konstruktion und Berechnung von Verbindungen, Lagern,
Wellen. 2. Auflage Berlin, Heidelberg, New York: Springer 1981
Fronius, St.: Taschenbuch Maschinenbau in acht Bänden. Dritter Band: Maschinenelemente, Getriebe,
Mechanismen, Hydrostatische und pneumatische Antriebe, Elektrische Antriebe. Berlin. Verlag Technik 1987
Wächter, K.: Konstruktionslehre für Maschineningenieure. Grundlagen, Konstruktions- und Antriebselemente.
Berlin: Verlag Technik 1987
Wesolowski, K.: Dreidimensionale Beanspruchungszustände und Festigkeitsnachweis drehmomentbelasteter
Zahnwellen-Verbindungen unter elastischer und teilplastischer Verformung. Düsseldorf: VDI Verlag 1997
Wesolowski, K.: Ausnutzung von Symmetrieebenen zur Berechnung drehmomentbeanspruchter Bauteile.
Mitteilungen aus dem Institut für Maschinenwesen Nr. 19; Clausthal 1994

Dipl.-Ing. Hagen Birkholz

Institut für Maschinenwesen der TU Clausthal
Robert-Kochstr. 32, 38678 Clausthal-Zellerfeld, Germany
Tel.: +49/5323/72-2683
Fax: +49-5323-723501
Email: [email protected]

1034 TECHNICAL SYSTEMS