Parametric modelling of integral bridge soil spring reactions

Date
2022-04
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH ABSTRACT: The integral bridge is a bridge structure in which the deck, piers and abutments are all made integral without the use of expansion joints. The joints would normally allow the expansion and contraction of the bridge deck, however now the flexibility inherent in the piers and abutments is utilized. The main design issue in an integral bridge is the effect of the temperature fluctuations and the corresponding movements in the bridge deck. By its nature, the integral bridge has ends that interact with the embankment soil, as the bridge is subjected to cycles of movement that result in changes in soil pressure from the active to the passive pressure state. Forces inherent in conventional bridge systems also act longitudinally against the backfill soil, however the magnitude of the soil-structure interaction is negligible since the cyclic forces that are created are minor in comparison with those acting on integral bridges. This makes the integral bridge design both a structural and a geotechnical problem of interest. The interaction between the bridge structure and the surrounding soil is a relatively intricate, three dimensional scenario that is simplified in analysis modelling. Engineers are interested in two different (but related) issues. The first is the deformation of the soil as load is applied to it by the structural system, the second is the load carrying capacity of the soil. These two phenomena can be described as soil stiffness and soil strength. Clearly, a greater soil stiffness should lead to larger axial forces and bending moments in the deck due to the longitudinal expansion or contraction of the bridge. Finite Element (FE) approaches to Soil-structure interaction (SSI) usually fall into approaches that characterize the soil using continuum elements and those that represent the soil through springs. In the spring method of analysis, the resistance of the soil lying adjacent to the piles and to the abutment is represented by springs which can be linear, compression only or non-linear in character. The theories used for the calculation of the spring stiffnesses are quite different between the abutments and the piles. In this thesis, the influence on SSI reactions of pertinent bridge geometry and load parameters was investigated. To this end, a series of parametrically varied 2D and 3D bridge models with soil springs were created and subsequently analysed. The parameters of span length, abutment height and soil condition for different percentages of live load and thermal expansion/contraction/gradient were investigated and their influence on SSI reactions were revealed through the series of parametric model testing. Pile lateral loading vs deflection relationships tend to be non-linear, however the spring reaction relationship for increasing span and load (once a pile is incorporated into an integral bridge system) was unknown prior to the analysis work captured in this thesis. The basic hypothesis for the thesis is that 3D models with realistic properties assigned to the springs will be required to capture the true behaviour of the integral bridge springs (which are suspected to be non-linear in nature), and that simplified 2D models can nonetheless provide some basic understanding of their behaviour and characteristics, but will not be able to completely model the bridge spring reaction behaviour. The model test results showed that the maximum spring reactions (for abutments, piles and footings) followed either a linear relationship with increasing span or tended towards more non-linear relationships. The results showed distinct and significant differences between the spring reaction vs span relationships for the abutments and the piles (for the same bridge types). Further interpretation of the test results also showed that as spans increase (irrespective of the abutment height), the maximum spring reaction ratio (abutment/pile reaction ratio) tends towards unity under live loading. In summary, the spring reaction results clearly demonstrated the influence of soil conditions, the bridge geometry and the applied loads on the spring reactions. It was also noted that significant increases or changes in spring reaction often occurred after the span length of 20m was exceeded. It is hoped that the model testing and analysis has added valuable knowledge to the study of integral bridges. Further model testing is recommended to determine the characteristics of the spring reactions for much longer lengths of span, typically found in multi-span bridges.
AFRIKAANSE OPSOMMING: n Integrale se brugdek, kolomme en landhoofde vorm 'n geheel sonder die gebruik van uitsettingsvoë wat gewoonlik uitsetting en inkrimping van 'n brugdek toelaat. 'n Integrale brug struktuur benut die inherente buigsaamheid van die kolomme en landhoofde. Die hoof ontwerpskwessie in 'n integrale brug, is die effek van temperatuur skommelinge en ooreenstemmende bewegings in die brugdek. Die integrale brug se eindpunte is in wisselwerking met die grondwalle wanneer die brug onderwerp word aan die bewegingssiklusse wat gronddruk veranderinge vanaf die aktiewe na die passiewe druk toestand, teweeg bring. Kragte inherent aan die konvensionele brug sisteme, werk ook longitudinaal teen die hervul grond. Die omvang van die grondstruktuur interaksie, is egter weglaatbaar aangesien die sikliese kragte wat gekep word, gering is in vergelyking met die kragte wat op die integrale brug inwerk. Dit maak die integrale brug ontwerp beide 'n strukturele en geotegniese probleem van belang. Die interaksie tussen die brugstruktuur en die omliggende grond is 'n relatiewe ingewikkelde, drie dimensionele scenario wat vereenvoudig kan word deur analise-modellering. Ingenieurs stel belang in twee verskillende (maar verwante) verskynsels. Een aspek is die vervorming van die grond soos wat lading en druk van die struktuur daarop toegepas word. 'n Tweede aspek is die ladingskapasiteit van die grond. Hierdie twee aspekte kan as grondstyfheid en grondsterkte beskryf word. 'n Hoër grondstyfheid behoort duidelik tot groter aksiale kragte en buigmomente in die brugdek te lei, as gevolg van die longitudinale-uitsetting of -inkrimping van die brug. Finite Element (FE) benaderings tot Grondstruktuur Interaksie (SSI), verwys gewoonlik na die benutting van grondelemente en die gebruik van veerstelsels. Met die veermetode van analise, verteenwoordig vere (lineêr of nie- lineêr van aard, of bied slegs kompressie) die weerstand van die grond aangrensend tot die heipale en landhoofde. Die teorieë wat gebruik word vir die berekening van die veerstyfheid, verskil vir die landhoofde en die heipale. In hierdie proefskrif is n reeks parametriese 2D en 3D brugmodelle vir grondvere geskep. Die parameters vir spanlengte, landhoofhoogte en grondtoestand vir verskillende persentasies van die lewendige lading, asook termiese uitsetting/inkrimping/gradiënt, is ondersoek, en die veergedrag is waargeneem en ontleed deur 'n reeks van parametriese model toetsings. Die veerreaksies vir enkelportaal (enkelspan) integrale brûe is oorweeg, met 'n hipotese dat vir 2D modelle, lineêre gedrag waargeneem sal word as gevolg van die vereenvoudigde model aannames. Met die 3D modelle vir heipale, sal dan 'n nie-lineêre gedrag waargeneem word as gevolg van die nie-lineêre laterale heipaal gedrag karakteristieke, en die gebruik van nie-lineêre vere.
Description
Thesis (MEng)--Stellenbosch University, 2022.
Keywords
Bridges -- Structure, Integral bridge, Bridge decks (Floors), UCTD, Bridges -- Design and construction
Citation