Title: Quay walls in urban areas: A method to predict the remaining life span
Abstract: Without doubt the need to give a founded estimation of the remaining life span is needed. In the NEN 8700 and 8701 and the guideline of Rijkswaterstaat (RBK) a way to determine the structural reliability is given. The corresponding estimate of the remaining life span is based on but not equal to the suggested reference period of the loads. These can vary from 1 to 50 years. The main objective is to calculate the remaining life span. The idea is that the reliability index decreases and the failure probability increases in time. Deteriorating materials and changing loads are the main reason for this. To find the reliability index full probabilistic models are a possibility. For these calculations a lot of parameter input is required and the calculations are often difficult, time consuming or need specialised software. As an alternative the Eurocode and also SBRCUR 186 give partial safety factors to make reliability-calculations without the need of actual probabilistic computations. As a start the quay wall is assessed by simplified hand calculations to find the most critical failure mechanisms and other aspects, such as input parameters. Information that can be found in the archive of the municipality of Dordrecht is used as first input. The second step is to find a more detailed way to model the structural reaction of the loads. It is chosen to use Plaxis. A strutting wall and the large amount of piles in combination with a small retaining screen make it necessary to use a 3D model. The computation time is extensive and therefore a 2D model was made. This model is calibrated on the 3D model by means of the critical aspects in the construction, which were found in the hand calculations. With the 2D model it is possible to make a large amount of computations to estimate the sensitivity on certain scenarios within a limited time span. The fact that there are many unknowns makes this useful and even inevitable. The reliability indices follow from a given set of partial safety factors for which the construction passes all unity checks. By analysing the fault tree the reliability index is found in more detail. Each failure mechanism is investigated separately to calculate the overall reliability index without making any probabilistic computations. By extrapolating the indices from the time of construction via the present condition to the crossing with the required reliability the remaining life span is found. At the time of computing there were no inspection results available. The information that was used resulted in an unsafe structure. By improving the soil retaining screen and a complete inspection the remaining life span could be 40 years and 33 years for the unfavourable scenarios. The scenarios are based on soil mechanical reactions on an overly dredged port basin, an extreme ground water level and pile geometry. Some main remarks are placed with this method: *The consequence class is assumed to be 1. *The extrapolation of the reliability indices is assumed to be linear. *The Eurocode is used to assess the strength and some of the old building materials that are not used anymore (smooth reinforcement steel) are not treated in detail in the codes. *It is assumed that the partial safety factors in SBRCUR 186 are calculated in a similar way as CUR 211 and CUR 166, on which it was based. This is the way how the importance of the different failure mechanisms, as parts of the total (6.25), is determined. *The use of different guidelines that were not meant to be used together in this manner makes that the error of the method is estimated to be 50%. The partial safety factors have an error of 0.05 to 0.10 in 1, resulting in an error of the reliability index of approximately 10%. Due to the extrapolation the error increases to 50% and even 100% taking into account modelling errors. This is shown by a grey band in the graph. As a direct result the main recommendations are given to be: *The retaining screen has to be improved and the relieving floor has to be inspected or the remaining life span is 0 and the quay should be replaced. *Redo the calculations using inspection results and make detailed calculation after inspections. *In a few years the quay wall should be assessed again to find a new reliability index. A third point on the line to extrapolate should give more insight to the distribution. Alternatively one could predict the material properties to predict a future reliability index. Challenges with the reference period could arise. *The Eurocode, or a part in the specialised NEN 8700 series should give calculation rules for common old building materials. *SBRCUR 186 should give the fault tree and parts of the total reliability index for each of the failure mechanisms to make calculations for all the given main types of quay walls possible. As a result the background of the given partial safety factors, which are sometimes counterintuitive, is also known in more detail. *An experienced engineer could be able to fit the uncertainties into scenarios. The large errors and scenarios give a wide range of the remaining life span. A detailed inspection to the general outline, state and the material properties of the structure should be done beforehand. *More quay walls should be assessed to validate the usefulness of the presented method.
Publication Year: 2016
Publication Date: 2016-03-31
Language: en
Type: article
Access and Citation
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot