The Hydro-Support

Abstract

The "classic" solution for the support of a translating lockgate, the wheelonrail support, has relatively high construction as well as inspection and maintenance costs. An alternative support which has previously been developed for use in the Prins WillemAlexander lock is the socalled "hydrosupport", a hydrostatic thrust bearing that slides on an elastic track and is connected to the lockgate by an elastic support. After a runningin period, this support shows low friction. In this thesis several methods to further improve this type of support have been studied. The direction of these improvements has been guided by the following observation: The typical dimensions and manufacturing standards of the bearing and the track are in conflict. On the large scale of a lockgate, a hydrostatic bearing typically requires sliding surfaces with a surface waviness smaller than 0.1 mm/m. However, the track can not be manufactured easily with a surface waviness smaller than typically 0.5 mm/m. This means that contact between the bearing and track will be inevitable. In this thesis, methods have been studied to use this contact in order to improve the performance (namely reduced flow rate and pumping power) of the hydrosupport. A mathematical model has been developed, incorporating the elastic deformation of the track, bearing and support, and the partial contact and hydrostatic lubricating film between the bearing and track. Several track waviness models have been developed, among others a random periodic surface waviness. Furthermore the concept of an "ideal" support has been introduced, which under compression exhibits a reaction pressure equal to the hydrostatic pressure in a lubricating film with a constant height. Not only has this mathematical model been developed in this thesis, it has also been implemented in a numerical program and used to test the influence of a number of design parameters on the performance of a hydrosupport. It has been shown that, using the contact between the sliding surfaces, the tilting stiffness of a bearing with 1 small recess is comparable to that of the 4-recess bearing. Additionally, a 1 recess bearing requires a smaller or even no restrictor and therefore a smaller supply pump. Furthermore, it has been shown that the hydrofender with its large length/width ratio has comparable or even better performance than the circular hydrofoot, while requiring a narrower track. In addition, it has been shown that, for a given load, a hydrofender with a small bearing thickness and with a standard elastic support design exhibits a smaller flow rate and larger bearing coefficient than a bearing with an ideal support design. Finally, using the results of these parametric studies, a procedure has been developed for the design of hydrosupports. This procedure has been used in two examples.