A THEORETICAL MODEL OF BALLAST FOULING de Paiva, Cassio E. L.; PhD, Professor Peixoto, C. F.; MsC, Professor UNICAMP - State University of Campinas Brazil INTRODUCTION • Oldest Brazilian railways are frequently located near or on the river edges, where flat terrains are observed but subjected to frequent flooding. • The ballast built on these areas probably makes more demands on cleaning ballast work, to remove soil particles clogging voids which modifies track behaviour and reduces permeability coefficient. INTRODUCTION • The stresses and strains developed from track loads and spread to ballast, sub-ballast and to the sub-grade can be affected by the clogging action, changing layers system behaviour. • The main objective of this recherche is to show and to discuss the used math models to study fouled ballast under hydraulic and subterranean theorems. • It is proposed a special model to predict water flux inside ballast, the McLelland Box modified according to this distinguished medium. BALLAST FOULING SOURCES • Ballast fouling occurs under a group of effects. • There is an internal group of effects which produces small particles from the standard ballast: – the dynamic traffic efforts and tamping ballast. – the external events can be grouped into seepage action and wind as a carrier of soil particles to the interior of the ballast. • The main railway features which contribute to ballast fouling are: sleeper wear, particle migration, ballast tamping as a function of repeated load, slurry and hydraulic action. Seepage or superficial drainage also carries solid particles to the ballast voids Platform cross section Ballast and subgrade interface carrying soil particles to the crushed stone Subterranean drainage system subgrade seepage network Oldest railway tracks Railways crossing desert areas are subjected to wind carrying fine soil particles Wind carrying fine sand particles Clogging path by wind The real situation of the freight railways in Brazil • In the nineties, by a governmental decision, almost all Brazilian freight railways are under private concession, because of the absence of resources to keep railways working and also to minimize management difficulties. • Until today, Brazilian Railway Dealers did not achieve budget surplus to start a massive maintenance program. The real situation of the freight railways in Brazil • The Brazilian Concession Model does not guarantee enough funds for the railway maintenance. • So, it is common to see just ballast maintenance instead of rails and sleepers changing. • Putting more crushed stone on the ballast to level the track is one of the most frequent maintenance activity applied on them. • Dealers focus on the waggons and locomotives purchase than investing money on the track maintenance because if they decide to break up the contract they will have to leave all equipments applied on the track. IMPORTANCE OF BALLAST CLOGGING TO THE RAILWAY MAINTENANCE • A high extension of them crosses large areas subjected to high level of precipitation, resulting in higher ballast moisture and fouling through seepage. • It is necessary to develop new maintenance techniques under more precise math and physics approach in order to understand better the relationship between ballast clogging and track behaviour. IMPORTANCE OF BALLAST CLOGGING TO THE RAILWAY MAINTENANCE • The ballast cleaning service can be planned according to the forecasted time to clog ballast voids. • The speed of water at the seepage is an important quantity for service management. • The difference between the forecasted water speed and the real one is subjected to the seepage adopted regime. THEORETICAL MODELS OF CLOGGING BALLAST WATER FLOW • This recherche is in its first phase. • Its beginning was based in the work of Peixoto’s Master Degree dissertation (1988); • The seepage through porous media needs, to be considered as a proportion of hydraulic gradient, and validating the permeability coefficient must analyzed by Equations A or B, according to the researches of Peixoto. THEORETICAL MODELS OF CLOGGING BALLAST WATER FLOW 0.42 ici 2 0.182 D10 (A) Equation A gives the inferior limit of hydraulic gradient, for compacted samples, guarantying the flux as a laminar one. The granular material researched presents effective diametre from 0.1 to 5 mm, i.e., the diametre of grading curve at 10 % passing, possibly shown at fouled ballast. In this case, it is proposed to enlarge this research to bigger diameters, as a cleaned ballast in general shows. THEORETICAL MODELS OF CLOGGING BALLAST WATER FLOW • Equation B gives the same limit but for non compacted samples, all the other features kept equal. ici 0.098 0.008 2 D10 (B) Ici: hydraulic gradient for tamped samples (m/m); D210: effective diameter (mm) of non compacted samples Proposed Dynamic Drainage Clogging Ballast Model Ballast seepage pumping Dynamic load (hammer effect) Ballast equipotencials and flowlines Gap to the studied displacement Proposed Dynamic Drainage Clogging Ballast Model • During ballast lifetime a huge amount of dynamic loads subjects sleepers like a hammer on the ballast top. • After several load actions some sleepers can show a gap between its bottom surface and ballast top. • During flooding period, this gap can work as a piston due to the trains frequency, pumping water and subgrade soil particles to ballast media voids. • This scenario is represented at previous figure, which pictures the moment when the train acts on the rails, the sleeper settles and the gap diminishes. Then, just some seconds after, the sleepers return to its previous level, pumping water and soil particles. PHYSICAL MODEL AT NATURAL SCALE OF BALLAST SEEPAGE • It was conceived a physical model to reproduce soil water seepage considering transient flux by McClelland (1943), a single model which permits to predict and to evaluate water table in two dimensions in natural scale into a ballasted box. • It is highlighted that this model focuses transient flux study, not permanent one, describing the variation of water table level during the flux changing according to a special rain, taken as a representative one. PHYSICAL MODEL AT NATURAL SCALE OF BALLAST SEEPAGE • McClelland aimed to study the transitional flux, but, in this approach, it is proposed to evaluate also ballast clogging, under the solid and water flux controlling, measuring clogging, permeability coefficient, strain and stresses variation. PHYSICAL MODEL AT NATURAL SCALE OF BALLAST SEEPAGE • It is proposed an adjustment to the McClelland Box, as shown at figure ahead, where the charge of water would be calculated to guarantee the adequacy of Darcy’s Law. • The water tankers of both sides were installed to allow measurements of water density according to Stokes Equation, for indirect estimation of the amount of soil particles left inside the ballast. • This test would evaluate the variation of permeability coefficient and the fouling action to the voids in function of time. PHYSICAL MODEL AT NATURAL SCALE OF BALLAST SEEPAGE Symetrical platform axis 1,35 1,05 Pressures gauges (1 a 5): 1,20 1,20 D siphon W/2 S Flexible connetion D variou de 45 cm a 135 cm W/2 from 120 cm to 240 cm Perforated drain 6,00 (no scale) 1,20 0,45 ballast 1,80 Water source Geotextil screen 0,45 S was abort 10 cm, 15 cm and 20 cm 0,45 0,45 0,60 (m) Adjusted McClelland Box to evaluate ballast transitional flux and clogging PHYSICAL MODEL AT NATURAL SCALE OF BALLAST SEEPAGE • After this first test, it is proposed to install a sleeper part on the ballast top and to start again these tests, with a cleaned ballast tamped and graded the previous one. • In this case it is intended to evaluate the variation of permeability coefficient and clogging voids in function of time, influenced by crushed stone breakdown. PHYSICAL MODEL AT NATURAL SCALE OF BALLAST SEEPAGE • In the next phase of tests, it will be evaluated the influence of tamping and water carrying soil particles. With the resulted curves it is possible to help better maintenance planning. • In the last phase of tests the water source will vary during the time, allowing evaluation of the influence transitional flux ballast fouling. • It is intended to adopt a time which lasts a specific strong rain, to calculate the amount of fines inside the ballast just after a single strong rain. PHYSICAL MODEL AT NATURAL SCALE OF BALLAST SEEPAGE • The ballast fouling behaviour and its lifetime demand more researches to achieve new models to conceive better planning ballast maintenance under low costs. • A natural or greater scale model of ballast and subgrade system subjected to a controlled seepage, according to the ambience conditions helps to achieve this goal. PHYSICAL MODEL AT NATURAL SCALE OF BALLAST SEEPAGE • In this case, it is proposed to apply the McClelland Model Box, permitting to control these quantities and to evaluate the ballast coefficient in relation to the outline structure changing as a function of the amount of solid particles clogging the voids. Thanks for your attention
© Copyright 2024