## Abstrakti

The design of railway structures is based on European standards. The specified traffic loads used in dimensioning are presented in EN 1991-2 “Actions on structures Part 2: Traffic loads on bridges” and EN 15528 “Railway applications. Line categories for managing the interface between load limits of vehicles and infrastructure”. The EU Commission regulation No 1299/2014 rules that structures, earthworks and earth pressure effects should be designed according to EN 1991-2 when the structure is new or replaced and according to EN 15528 if the structure if an existing structure is renewed or upgraded. With these regulations, the technical specifications of interoperability (TSI) for loads on railway structures are guaranteed. However, the standard EN 1991-2 is prepared to be applied on bridges and at some cases has been evaluated inappropriate for buried structures.

The aim of the study was to observe the effects of different load models on the vertical stress levels of pile supported embankment slabs under railway structures. The study was conduct-ed with PLAXIS 3D software. Simulations included seven different load models and three dif-ferent embankment thicknesses, or slab installation depths.

The used calculation model has been verified with actual field data. The field data used for verification is taken from a project where a railway embankment located on soft soil area was heavily instrumented and loaded with a special loading car having an axle load of 25 tonnes. The comparison of stress-strain relationship has been conducted only at top parts of the rail-way structure since the field data represents embankment on soft soil area and the models in this study have a stiff base. The measured and modeled results had good correlation and it is safe to assume the model used in this study provides credible results. Three main load mod-els were studied more thoroughly. Load model A (LM71), which TSI determines to be applied on slab structures, though it is intended for bridges; load model F (LM-GEO), which is con-ceptual load model providing an alternative solution to load model A (LM71) (standard EN 1991-2) having line load groups to simulate the actual train load instead of point load group in load model A (LM71); and load model G (E4) (standard EN 15528) which is used for existing infrastructure for 25 ton axle load.

Based on the obtained results load models A (LM71) and F (LM-GEO) are very close to each other in load intensity, or maximum stress increase due train load whereas load model G (E4) provides a smaller load intensity. This is mainly due to distances between load groups. Load model G has always two point loads representing one bogey of a wagon and the distances between bogies are more realistic compared to the actual vehicles. Load model A also has a point load group representing two consecutive bogies but the distance between point loads is fixed. Load model F (LM-GEO), on the other hand, has a replication of the point load group of load model A (LM71) depicted as a line load.

Load model A (LM71) also seems to produce the highest stresses in the most intensively loaded areas. The effect is observable when the installation depth of a pile supported em-bankment slab is less than 5 meters but seems to vanish as the installation depth increases.

The differences between load models seem to contract as the examined section is getting longer. The total traffic loads at the observed maximum section were close to each other. On the other hand, load model A (LM71) results as the highest local load effects due to combina-tion of point loads and line loads and their sequence. Though the total load throughout the slab structure is similar to other load models, the maximum traffic load is roughly half of the live load other load model produce to a single pile.

The obtained results indicate that the load model has a significant effect on the load effects. A load model intended for bridge design is also ruled to be applied on earthworks, which ob-viously leads to uneconomical structures and increasing constructions costs.

The aim of the study was to observe the effects of different load models on the vertical stress levels of pile supported embankment slabs under railway structures. The study was conduct-ed with PLAXIS 3D software. Simulations included seven different load models and three dif-ferent embankment thicknesses, or slab installation depths.

The used calculation model has been verified with actual field data. The field data used for verification is taken from a project where a railway embankment located on soft soil area was heavily instrumented and loaded with a special loading car having an axle load of 25 tonnes. The comparison of stress-strain relationship has been conducted only at top parts of the rail-way structure since the field data represents embankment on soft soil area and the models in this study have a stiff base. The measured and modeled results had good correlation and it is safe to assume the model used in this study provides credible results. Three main load mod-els were studied more thoroughly. Load model A (LM71), which TSI determines to be applied on slab structures, though it is intended for bridges; load model F (LM-GEO), which is con-ceptual load model providing an alternative solution to load model A (LM71) (standard EN 1991-2) having line load groups to simulate the actual train load instead of point load group in load model A (LM71); and load model G (E4) (standard EN 15528) which is used for existing infrastructure for 25 ton axle load.

Based on the obtained results load models A (LM71) and F (LM-GEO) are very close to each other in load intensity, or maximum stress increase due train load whereas load model G (E4) provides a smaller load intensity. This is mainly due to distances between load groups. Load model G has always two point loads representing one bogey of a wagon and the distances between bogies are more realistic compared to the actual vehicles. Load model A also has a point load group representing two consecutive bogies but the distance between point loads is fixed. Load model F (LM-GEO), on the other hand, has a replication of the point load group of load model A (LM71) depicted as a line load.

Load model A (LM71) also seems to produce the highest stresses in the most intensively loaded areas. The effect is observable when the installation depth of a pile supported em-bankment slab is less than 5 meters but seems to vanish as the installation depth increases.

The differences between load models seem to contract as the examined section is getting longer. The total traffic loads at the observed maximum section were close to each other. On the other hand, load model A (LM71) results as the highest local load effects due to combina-tion of point loads and line loads and their sequence. Though the total load throughout the slab structure is similar to other load models, the maximum traffic load is roughly half of the live load other load model produce to a single pile.

The obtained results indicate that the load model has a significant effect on the load effects. A load model intended for bridge design is also ruled to be applied on earthworks, which ob-viously leads to uneconomical structures and increasing constructions costs.

Julkaisun otsikon käännös | Ratarakenteessa oleviin paalulaattoihin kohdistuvat pystyjännitykset |
---|---|

Alkuperäiskieli | Englanti |

Julkaisupaikka | Helsinki |

Kustantaja | Liikennevirasto |

Käyttöönottava elin | Liikennevirasto |

Sivumäärä | 43 |

ISBN (elektroninen) | 978-952-317-416-0 |

Tila | Julkaistu - 2017 |

OKM-julkaisutyyppi | D4 Julkaistu kehittämis- tai tutkimusraportti taikka -selvitys |

### Julkaisusarja

Nimi | Research report of the Finnish Transport Agency |
---|---|

Kustantaja | Finnish Transport Agency |

Numero | 28/2017 |