Experimental validation of a high torque density flange connection
摘要
Prevailing developments of wind turbine drivetrains aim toward the increase of the power density of the turbine. Since this directly leads to higher torque transfer demands on the rotor-side end of the drivetrain, current solutions for torque transfer are reaching their performance limits. This applies, in particular, to connections between components using conventional friction-locked bolted connections.
While one approach for surpassing the present limitations on these frictional bolted connections is to use friction-enhancing technologies, another promising approach is to substitute them with form-fitted pin connections. One alternative to conventional press-fitted pin connections that works with manufactured overclosure is using connection elements, which are only tensioned during assembly, thereby significantly simplifying and accelerating the assembly process. GripCon elements from Schaaf, which have just been introduced for the connection of the main shaft of a wind turbine, provide such a solution. Moreover, the GripCon elements reduce weight, costs, and space requirements and are suitable for multiple installations and removals. A crucial prerequisite for the introduction of the GripCon solution in wind turbines was the experimental validation of the torque capacity and verification of durability. Consequently, Schaaf designed and manufactured an ultra-compact test rig utilizing hydraulic short-stroke cylinders developed in-house to perform the validation of the connection with a maximum torque capacity of 16 MNm. The test rig was subjected to static tests up to 14 MNm providing successful proof of both the overload capacity as well as the durability of the flange connection using GripCon elements.
A particular challenge throughout the tests was the in-situ measurement of the applied torque on the connection during the measurement. It was stipulated that the torque is to be measured on one of the flange parts close to the connection such that friction losses do not affect the results. Therefore, measurements of either the hydraulic cylinder pressures or its forces were not sufficient. Due to the compact design of the test rig, no cross-sections in the flange parts featuring a homogeneous stress/strain distribution on the surface were identified, which excluded the conventional strain gauge measurement in the solution space available for this measurement. Consequently, IME Aachen has adopted a model-based measurement method that combines shear strain measurements with high spatial resolution and an evaluation method based on an advanced finite element deformation model of the entire test rig. The proposed article and presentation will provide insight into the design of the test rig and the model-based measurement method.