The study is a comparative desk study of the application of the vertical earth load, traffic live load and the nominal earth pressure in the design methodology of culverts as outlined in TMH7 – Code of Practice for the design of highways bridges and culverts in South Africa Part 2, AASHTO LRFD Bridge Design specification and the DMRB volume 2 section 2 part 12 - BD31/01. It involves the theoretical design and analysis of five single cell reinforced concrete box culverts ranging from 2.1m to 6.0m under different fill depths ranging from 0 to 8.0m by applying load obtained using the three design manuals. The objective of this study is to analyze the methodology involved in estimating vertical earth load on a culvert as outlined in the design manuals to ascertain relevance of the formulae and procedure in TMH7 or/and to recommend the most effective approach for evaluating the vertical earth load on box culverts specific and appropriate for South Africa. By comparing the load derivation methodology outlined in ASHTO LRFD and BD 31/01 and analyzing the load forces obtained from the analysis. Box culverts are designed as rigid monolithic structures to withstand the maximum bending moment and shear force. The design involves the analysis of the various loads acting on the culvert with the weight of the overhead earth embankment being the most significant. The vertical earth load, live load and the lateral earth pressure acting on the culverts at various fill depth are manually derived from equations as outlined in the three design manuals. The culverts are modelled and analyzed in Prokon as two-dimensional plane frame structures using the frame analysis module by applying this load to determine maximum positive hogging moments, maximum negative sagging moments and maximum positive shear forces for each span for the top slab. The load forces obtained for each span are then plotted against the soil cover depth to illustrate the discrete load effect of the vertical earth load and live load on the culverts at varying fill height and to determine the relationship between the culvert geometry, soil cover depth and the applied load. The result of the analysis shows that an increasing non-linear relationship exists between the load effects, the soil cover depth, and the span length. The dead load effect increases with increasing fill depth and culvert span while the live load effect diminishes with increasing fill height and culvert span i.e., for culverts buried at shallow depths, the traffic live load is the most critical load but as the height of the soil cover increases the dead load becomes more significant until a point is reached where the live load is totally insignificant. The vertical earth loads obtained from TMH7 and BD31/01 are constant at a particular fill depth for each culvert despite the different span length. The vertical earth load for these two manuals is estimated from the soil cover depth and density, the load tabulated clearly is independent of the culvert geometry. The load obtained from AASHTO LFRD is the lowest and less than 20% of the load obtained from the other two design manuals. Unlike TMH7 and BD31/01, AASHTO LFRD considers the effect of the soil-structure interaction to adjust the vertical earth load on the structure which automatically reduces the load value. The vertical earth load values obtained from TMH7 and BD31/01 are generally more conservative as compared to those obtained from AASHTO LFRD.

Thesis (MScEng)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: Different types of live loading due to traffic may act on bridges. The focus of this study is on normal traffic loading according to the South African specification of TMH7. Heavy vehicles are not included in normal traffic loading. TMH7 represents the code of practice for the design of highway bridges and culverts in South Africa. The aim of the study is to provide an insight into the flexural analysis of skew bridges, under the effects of normal traffic loading. The need for the study arose since the specification of TMH7 does not explicitly specify application patterns for normal traffic loading. Only the intensity of normal traffic loading is specified and it should be applied to yield the most adverse effects. For these reasons, a set of so-called standard application patterns are investigated and developed through the course of this study. The envelope of the values from the standard application patterns are compared to the most adverse application pattern for flexural effects in certain design regions of the bridge deck. Flexure, as in the context of this study, translates into the bending and twisting of the bridge deck under loads. A number of numerical experiments are performed for typical single span and multi-span continuous carriageways, where the standard application patterns are compared to the most adverse application patterns. The results from the numerical experiments are documented and compared as the angle of skew of the bridge deck increases in plan-view. For this purpose, the development of effective and specialized software was necessary. It was found that the set of standard application patterns can be used as a preliminary approximation for the most adverse effects of normal traffic loading, for specific flexural resultants in certain design regions of a bridge deck. However, for a large number of secondary flexural effects, the set of standard application patterns did not represent a good approximation for the most adverse values. / AFRIKAANSE OPSOMMING: Verskillende tipes lewendige belasting, as gevolg van verkeer, kan op brûe inwerk. Die fokus van die studie is op normale verkeers-belasting volgens die Suid-Afrikaanse spesifikasie van TMH7. Swaar-voertuie word nie ingesluit by normale verkeers-belasting nie. TMH7 verteenwoordig die kode vir die ontwerp van padbrûe en duikers in Suid-Afrika. Die doel van die studie is om insig te verskaf in die buig-analise van skewe brûe, as gevolg van die werking van normale verkeers-belasting. Die rede vir hierdie studie ontstaan aangesien die spesifikasie van TMH7 nie eksplisiet aanwendingspatrone vir normale verkeers-belasting voorskryf nie. Slegs die intensiteit van normale verkeersbelasting word voorgeskryf en dit moet aangewend word om die negatiefste effekte te verkry. Vir hierdie redes word 'n versameling van sogenaamde standaard aanwendings-patrone deur die loop van die studie ondersoek en ontwikkel. Die omhullings-kurwe van die waardes wat deur die standaard patrone gelewer word, word vergelyk met die waarde van die aanwendings-patroon wat die negatiefste buig-effek in sekere ontwerp-areas van die brugdek veroorsaak. Buig-effekte, soos van toepassing op hierdie studie, verwys na buig en wring van die brugdek as gevolg van belastings. 'n Aantal numeriese eksperimente, vir enkel-span sowel as multi-span deurlopende brugdekke, word uitgevoer en die standaard aanwendings-patrone word vergelyk met die aanwendings-patrone wat die negatiefste waardes lewer. Die resultate van die numeriese eksperimente word gedokumenteer en vergelyk soos die hoek van skeefheid van die brugdek in plan-aansig toeneem. Vir hierdie doel is die ontwikkeling van effektiewe en gespesialiseerde sagteware dus nodig. Daar is gevind dat die standaard aanwendings-patrone, vir spesifieke buig-resultante in sekere ontwerp-areas van die brugdek, as 'n voorlopige benadering vir die negatiefste effekte van normale verkeers-belasting gebruik kan word. Dit was egter verder gevind dat vir 'n groot aantal sekondêre buig-effkte, die versameling standaard aanwendings-patrone nie as 'n goeie benadering vir die negatiefste waardes dien nie.