Large rolling bearings are at the heart of modern wind turbines. They support the rotor, transfer rotor loads into tower and support structure and are expected to work reliably for 20 years or more under highly variable operating conditions.
Yet, in recent years, operators and manufacturers have increasingly observed damage and unexpected behavior in main bearings and their surrounding components. One of the more subtle but highly relevant phenomena is ring creep – the slow relative movement of a bearing ring against its shaft or housing seat.
This article explains what ring creep is, why it occurs in large wind turbine bearings, how recent research helps us to understand it better, and what this means for design, manufacturing, and operation.
What is ring creep?
In an ideal design, a bearing ring is mounted on its shaft or housing with a defined interference fit. The ring is slightly smaller than the seat at the main shaft, creating a contact pressure that ensures the ring and seat move as one under all operating conditions. The only relative motion should be possible via the rolling elements.
In practice, this ideal is not always achieved. Ring creep describes the situation where the bearing ring slowly moves against its seat, although this should be prevented by the interference fit. The tangential movement per load cycle is very small and often intermittent, but over many cycles it can lead to a measurable circumferential displacement of the ring.
The consequences typically become visible on the contact surfaces: fretting marks and local wear on the shaft or housing seat. The fit may gradually degrade, preload and internal clearance can change, and local stress concentrations may promote fatigue. In small machinery, this is usually a local design or assembly issue. In large wind turbine bearings, however, ring creep is a system effect that involves not only the bearing but also the shaft, housing, load distribution and spectra, and deformation of surrounding structures.
Why ring creep matters in wind turbine main bearings
The main bearing in a wind turbine is particularly exposed. It must carry large and rapidly changing bending moments from the rotor, combined with radial and axial loads. At the same time, accessibility is limited, expected service life is long, and the exchange is usually very costly.
Ring creep is especially critical at the interface between the inner ring of the main bearing and the main shaft. The inner ring is usually mounted on a shaft, which itself deflects under load. Any local yielding, geometric deviation, or loss of contact pressure can promote micro-movement between shaft and ring. Over time, this may lead to unexpected damage at the main shaft and increased risk of premature bearing failure due to wear and particle ingress. The phenomenon has become more visible as turbines have grown in size and the mechanical system has become more flexible with hollow cast main shafts and further lightweight design.
What full-scale testing can reveal
Understanding such complex behavior requires more than simplified calculations. Large-scale test benches, such as the modular main bearing assembly test bench used in the BeBen XXL [1] and Gusswelle [2] projects at Fraunhofer IWES, offer a way to reproduce realistic main shaft and bearing configurations under representative load conditions. In BeBen XXL and Gusswelle, full-scale components are subjected to accelerated fatigue testing. The bending loads reflect real turbine operation but are scaled to produce relevant damage in a shorter time. The boundary conditions – such as alignment, support stiffness, and load introduction – are controlled much more precisely than in the field.
For ring creep, this approach is helpful for several reasons. It allows researchers to systematically vary fits, surface conditions, and assembly methods, and then observe how the system reacts. Displacements, strains, and tangential ring movement can be monitored during operation, while the contact surfaces at the bearing seat can be examined after testing. In this way, the main bearing assembly test bench provides a link between theoretical assumptions and actual component behavior – and it shows clearly that large components under wind turbine loads can behave differently from what traditional design models predict. In the ongoing research project WEA-RiWa [3], full-scale main bearing assembly tests will be conducted to validate advanced ring creep simulation models.
Ring creep and the role of cast components
Casting is attractive for large drivetrain parts such as the main shaft because it offers cost and design advantages. However, cast components differ from machined steel parts in dimensional accuracy, stiffness and deformation behavior. These differences directly affect the bearing seat. Local deformations as well as local surface quality and residual stresses, influence how contact pressure is distributed around the circumference.
A Fraunhofer IWES study [4] shows that even when nominal tolerances are respected, the resulting contact conditions are not equally distributed around the bearing seat. Under cyclic loads, zones with lower contact pressure are particularly susceptible to micro-movement between the inner ring and the shaft. At first, this manifests as tiny relative movements, but over many cycles and load cases they can accumulate into measurable ring displacement.
An important conclusion is that classical design is not always sufficient to guarantee a robust, creep-free fit in operation. The actual deformation of the main shaft, the load direction and magnitude, and the stiffness of the surrounding structure must be considered together with geometric tolerances.
Looking inside the fit: numerical simulation of ring creep
Field observations and test bench results show that ring creep occurs – but they do not reveal exactly how the contact behaves inside the fit. For this, detailed and validated numerical simulations are needed.
Fraunhofer IWES researchers have developed sophisticated models to describe ring creep in a wind turbine main bearing [5]. The model combines realistic shaft and bearing geometries, material properties, an interference fit, and time-dependent load histories that represent turbine operation. The goal is to simulate how contact pressures and relative movements evolve over several load cycles.
The results provide a more nuanced picture of ring creep. Rather than a simple transition from “stick” to “slip”, the process appears as a progressive, path-dependent mechanism. In some load states, sections of the contact zone may relax and undergo micro-sliding; in others, contact pressure increases and the ring sticks again. The net effect is a small incremental rotation or tangential displacement that builds up over time.
By varying parameters in the model – for example, interference fit, geometry, material behavior or wall thickness – researchers can identify combinations that are particularly sensitive to creeping. The simulations also highlight the importance of non-linear elastic behavior in large components.
Design and manufacturing lessons
First, the traditional design methods are not sufficient for large and more flexible components of modern wind turbine drivetrains. The bearing seat must be seen in the context of shaft stiffness, load spectra and deformation. It is the distribution of contact pressure that matters: local under-pressurized regions are starting points for creeping.
Second, material behavior and structural interaction must be included in design calculations. Integrating nonlinear-elastic models and realistic load conditions into the design process allows more robust fits and more realistic safety margins to be defined.
Outlook
The research discussed here shows how full-scale testing, targeted investigations on inner ring creep and advanced numerical simulation together build a more complete picture of ring creep in large wind turbine bearings. As turbines continue to grow in size and complexity, such interface phenomena will become even more important for reliable operation.
Future work will focus on refining models, validating them against the full-scale test data, and translating the findings into practical design and assembly guidelines. Ultimately, avoiding ring creep is part of a broader effort to understand the real behavior of large wind turbine components under real loads. The better we understand these details, the more reliably and economically the turbines will perform over their lifetime.
References
[1] Herrmann, J., Holz, K., Kyling, H., Zeise, B.:
BeBen XXL – Beschleunigter Experimenteller Betriebsfestigkeitsnachweis von Windenergieanlagen – Großkomponenten am Beispiel der Hauptwelle. https://doi.org/10.2314/KXP:1666359084
[2] Kirsch, J., Kyling, H., & Niewiadomski, J.:
Projekt Gusswelle – Werkstoff- und Bauteiloptimierung für leistungsfähigere Gusskomponenten im Antriebsstrang von Windenergieanlagen durch den Einsatz von Kokillenguss: Schlussbericht
[3] Kirsch, J., Kyling, H.:
Optimized cast components in the drive train of wind turbines and inner ring creep in the main bearing seat. Forschung im Ingenieurwesen 85(2), 199–210 (2021). https://doi.org/10.1007/s10010-021-00458-x
[4] Grosse, P., Kyling, H.:
Numerical simulation of ring creep on a wind turbine main shaft. Forschung im Ingenieurwesen 89, 2087–2112 (2025). https://doi.org/10.1007/s10010-025-00805-2
You may also be interested in:
The step to the digital test procedures: hybrid testing of wind turbines
Wind propulsion as a sustainable solution for shipping
A new approach to data management: NEXUS meets complex requirements