Friction connection in tubular towers for a wind turbine: Fid-Bau Portal

Friction connection in tubular towers for a wind turbine

Veljkovic, Milan / Feldmann, Markus / Naumes, Johannes / Pak, Daniel / Rebelo, Carlos / Silva, Luis Simoes da

Im Rahmen des europäischen RFCS (Research Fund for Coal and Steel) Projektes HISTWIN wird seit 2005 eine innovative Reibverbindung für Verbindungsstöße von Windenergieanlagen entwickelt und untersucht. Im Vergleich zur konventionellen Flanschverbindung ist die untersuchte Lösung sowohl einfacher als auch kostengünstiger herzustellen, wobei mit Einsparungen von bis zu 80 % für die Verbindung gerechnet wird. Der problemlose Zusammenbau vor Ort wurde durch eine im Werk durchgeführte Machbarkeitsstudie überprüft. Weiterhin wurden experimentelle Untersuchungen zur im Vergleich zu konventionellen Flanschverbindungen höheren Ermüdungsbeanspruchbarkeit der Verbindungsmittel durchgeführt. Der Vorspannkraftverlust in der Langlochverbindung wurde auf Basis von statischen Kurzzeit- und Langzeitversuchen, Ermüdungsversuchen sowie skalierten 4-Punkt Biegeversuchen untersucht. Die sich aus der neuen Verbindung ergebenden Kosteneinsparungen belaufen sich auf ca. 10 bis 15 %.

Loss of the pretension force in a bolt of the friction connection is essential for the safe design and for the economical maintenance of the tower. The estimated loss of the pretension obtained in the segment test is estimated in the range of about 20 % for a period of a tower life time, 20 years. The following effects are accounted for: short term loss of the pretension due to pre-tensioning procedure and due to the bolt stiffness itself, long-term losses due to creep of the finishing and the bolt, long term losses due to the fatigue loading and the meridional stresses. The total loss of the pretension in the bolt for other types of bolts and finishing than presented in the paper should be taken into account with awareness for possible differences in behaviour. Behaviour of the new friction joint during the quasi-static 4-point bending tests is completely elastic until the slip resistance is reached. All test specimens exceeded the expected ultimate slip load significantly. One possible explanation may be that the real friction coefficient of a particular primer used at the time of testing is higher. The slip factor is then higher than 0.56 taken as the lower 5 %-fractile from the preliminary product tests. This issue is dependent on ageing effects of the primer used in the testing. The performance of the connection after the first slip-page is ductile for all specimens and the failure is significantly higher, by about 25 %. The variation of bolt forces, due to micro settlements and a variation of the meridional stresses in the tower, has to be considered for fatigue design of bolts. This local effect, on 95 % of the measured bolt forces in the down-scaled 4-point bending tests, is lower than strain variation ranges predicted by a hand-calculation method proposed. Therefore a correction factor beta = 1.25 is proposed to the hand-calculation formula to obtain safe results. The feasibility tests of the friction connection show that all tasks are solved with rather small tolerances and that the assembling of the friction test specimens is rather easy. Furthermore, large potential for further improvements by using a more industrialized process become obvious.