Fatigue of steel lock gates

More Info
expand_more

Abstract

The Netherlands contains more than 400 hydraulic structures, of which 120 are locks. Inspections carried out since 2010 by Rijkswaterstaat revealed that a number of cracks per lock gate have been found. Based on experience and an evaluation of the photos made during the inspections it is assumed that the cracks occur due to fatigue. At this moment there is no good insight in the fatigue calculation of steel lock gates. The standard EUROCODE 3 used at the moment, has a part dedicated to fatigue strength, but describes that it is only applicable to structures in atmospheric conditions. As lock gates function in (sea) water conditions, it is questionable whether this standard can be used. Another aspect is that there is no guideline how to determine a load spectrum of forces on lock gates. The following topics are covered: a description of the forces on lock gates, a summation of the observations where cracks have been found, which part of the lock gates should be investigated and the standard that should be used and it’s limitations. Two calculations of fatigue damage based on a simplified model of the forces on lock gates are made. The first investigated case is the West lock of Sambeek and the second is the East lock of Terneuzen. For these lock gates the waterlevel spectrum is determined based on the waterlevel measurements of Rijkswaterstaat. From this spectrum a fatigue damage calculation is made. The results of the fatigue damage calculation based on the simplified schematisation of forces on lock gates did not match the fatigue cracks that have been found during the investigations. Some aspects causing this disparity might be that the schematisation of the lock gate is too simplified and the influence of waves is neglected. Another aspect that could influence the fatigue damage is the maintenance. A poor state of the wooden sealing at the back post, the coating layer and/or the cathodic protection system, could result in lower fatigue damage values. All these aspects have a negative influence on the S-N curve. The fatigue damage calculated with a simplified model cannot determine correct fatigue damage values. Therefore a FEM-model should be used to determine the stresses in the lock gate. The stresses in the lock gate itself should be monitored. A lock gate could be fitted with strain gauges to determine the real stresses in the lock gate. These real stresses could be compared with the result of the FEM-model and confirm the reliability of the FEM-model. Research on the influence of the salt and fresh water condition with or without cathodic protection should be done and rules included in the EUROCODE 3 on the influence of salt and fresh water.