A tapered-sleeve pin joint for gap-free damper connections

Viscous dampers are widely used in building structures to reduce dynamic response under earthquake and wind excitation, and their effectiveness depends not only on the constitutive behavior of the damper itself but also on the reliability of the connections through which structural deformation is transmitted. In practical applications, these dampers are commonly connected to the main structural system by shaft-pin joints so that the device can work primarily in axial action. Although this arrangement is mechanically simple and widely accepted in engineering practice, it contains a detail that is usually treated as unavoidable during fabrication and installation: a small gap between the shaft pin and the corresponding hole in the connected plate. Previous studies had already shown, in different structural and mechanical settings, that gaps or clearances can alter stiffness, modify load transfer, and degrade cyclic or dynamic performance. Work on viscous damper systems had also pointed to the detrimental effect of imperfect engagement and the zero-force platform that may appear in hysteretic response when a connection gap is present. Still, those studies largely clarified the consequences of the problem rather than offering a direct mechanical remedy for the shaft-pin joint itself. That left an unresolved question in practical damper design: whether the unavoidable clearance in a pin connection could be removed in a workable way without sacrificing the simplicity and functionality that make such joints attractive in the first place. In their paper published in Engineering Structures, Dr. Yi-Qiong Cui, Professor Yang Xiang, Dr. Bo Yang, Dr. Shi-Li Guo, and Professor Guo-Qiang Li from Tongji University address this issue by proposing a pin-joint configuration that incorporates paired tapered sleeves and thrust flanges to eliminate clearance between the shaft pin and the ear-plate holes. On that basis, they examine the connection behavior of the proposed joint under cyclic loading and then assess, through simplified structural modeling, how the removal of joint gap influences the efficiency of viscous dampers in frame systems subjected to seismic and wind actions.

The authors began with a direct comparison between two joint types: a conventional shaft-pin joint and the proposed tightknit version with tapered sleeves. The conventional specimen had a 20 mm pin and a 23 mm hole, giving a 3 mm assembly clearance chosen deliberately to make the mechanical consequence of the gap visible in testing. The tightknit specimen kept the same basic connection logic but inserted fourteen pairs of tapered sleeves and thrust flanges, with lubrication and machined contact surfaces used to support controlled assembly and sliding during tightening. The internal ear plate also retained an annular partition, which helped divide the sleeves into symmetric groups and improved alignment during installation. That design choice matters because the sleeve system is only useful if the radial expansion that removes the gap can be introduced in a balanced and assembly-friendly way. Under cyclic force-controlled loading, the contrast between the two specimens was immediate. The conventional joint developed a horizontal plateau near load reversal, especially as force passed through the vicinity of zero. Mechanically, that plateau corresponds to relative movement without effective force transmission while the pin traverses the internal clearance. The tightknit joint did not show that feature. Its load-deformation response remained continuous, with a nearly linear slope through reversal, which is exactly the behavior one would expect if the internal gap had been removed and compressive and tensile load transfer could proceed without slack.

The conventional joint showed an initial stiffness of 250 kN/mm, whereas the tightknit joint reached 583.3 kN/mm. Also, in the conventional specimen, plastic deformation began at about 150 kN with a relative displacement of 0.6 mm between the upper and lower end plates. In the tightknit specimen, yielding began at about 175 kN and only 0.3 mm. At 300 kN, the conventional joint reached 3.1 mm in that same displacement measure, while the tightknit joint reached 1.6 mm and the authors attributed this to the tapered sleeve assembly improving contact conditions, restricting shaft deformation, and producing a more uniform stress state along the pin-ear interface. The second displacement measure, taken between the shaft pin and the lower end plate, gave larger values for both specimens, but the qualitative picture did not change. That larger reading captured bending-induced geometric distortion of the pin as the upper and lower portions of the joint moved relative to one another under tension. The post-test pin shape confirmed that combined shear and bending had deformed the shaft. Even there, the tightknit joint retained a mechanical advantage, and the authors also note that by limiting severe deformation of the shaft, the proposed configuration helps maintain the integrity of the connection after heavy loading

Afterward, the researchers used OpenSees, to represent the frame, damper, linkage, and joint within a reduced damper-linkage-joint assembly. For conventional joints, they used bidirectional gap elements built from paired ElasticPPGap materials to reproduce the clearance behavior in both loading directions. For the tightknit case, the gap was taken as negligible. When sinusoidal dynamic loading was applied, the zero-gap assembly produced an ideal elliptical hysteresis loop. Once clearance was introduced, the loops broke into semi-elliptical branches separated by a horizontal no-force segment. As the gap increased from 0.5 mm to 1.5 mm, peak displacement rose and output force fell. Over two seconds of loading at the lower excitation amplitude, cumulative energy dissipation dropped from 19107.4 J in the zero-gap case to 12771.8 J at 1.5 mm, a 26% reduction. Even when the loading amplitude increased and the relative importance of the gap weakened somewhat, the adverse effect remained visible.

The study is important in connection mechanics, and Professor Yang Xiang and colleagues carry the joint model into structural response analyses under both earthquake and wind loading, and that move clarifies why the gap problem matters. In the four-story seismic model, dampers improved performance relative to the uncontrolled structure in every case, but the benefit was consistently strongest when the connection was tightknit. Under the El Centro record, the zero-gap configuration reduced drift, velocity, acceleration, and base shear more than the 1.5 mm-gap case. Across a broader set of 46 ground motions, the same pattern persisted: compared with the idealized tightknit condition, a 1.5 mm joint gap increased average peak inter-story drift by 7.6%, peak velocity by 6.4%, peak acceleration by 4.8%, and base shear by 7.9%. The wind analysis is particularly informative and because wind-induced structural deformations are smaller, a gap that might seem modest from a fabrication perspective can occupy a large fraction of the motion available to activate the damper. That is exactly what the twenty-story model showed. When total wind response was considered, the mean component dominated overall drift, so the difference among models was less dramatic. But once the fluctuating component was isolated, the sensitivity to clearance became unmistakable. A 0.3 mm gap brought the controlled structure much closer to the uncontrolled one, indicating a substantial loss in effective damping contribution. Under 24 synthetic wind histories, the tightknit case gave much stronger reductions in response than the conventional gap cases, including reductions relative to the 0.3 mm-gap model of 23.6% in peak inter-story velocity and 33.8% in roof acceleration. That distinction matters because it shifts the design logic. The issue is not simply that connection clearance is undesirable in an abstract sense. It is that clearance can erase a meaningful portion of the working deformation range of a supplemental damping device, and it does so most severely when the demand itself is small. The paper therefore sharpens the engineering interpretation of damper efficiency: performance depends not only on the constitutive behavior of the damper body, but also on whether the connection transmits motion without loss. In that respect, the tapered-sleeve joint is not treated as a secondary hardware refinement but becomes part of the force-transfer mechanism that strongly influences how effectively the damping system can function.