| Summary: | After the winter of 1962-1963 the frost damage in the Netherlands to road constructions was enormous. The "Studie Centrum Wegenbouw" made inquiries on road constructors about this damage. A special commision was brought together to look after these inquiries. It proved that many physical aspects about frost penetration were unknown. One of the activities therefore had to be a study on frost penetration into soils. For this reason a new freezing model is derived. With this model values for frost depths and heave can be calculated with it. The new model differs from known models because of different assumptions. Two important conditions are the original temperature distribution and the original water distribution. Both are assumed to be linear functions of place. The interpretation of masstransport and frozen fringe differs from Jones and Miller. The transport is in agreement with Darcy's law. However, the quantity of transported mass is not only depending on the waterpotentialgradient itself, but also on the location of this gradient and on time. The importance of the factor time follows from the heatfluxes at the freezingfront.Model values are compared with experimental results. Three different types of setups are used. A closed system (no free water) was used to investigate the influence of watercontent on frost penetration. Moisted silversand was applied. The volumemetric watercontent, changed to 6%. By the so called T.R.R.L. method an opensystem (free water available) was copied. Moistured morainesand was used. Measured frostdepths and heaves are compared with calculated values. A frost-tube, developed by the "Heidemaatschappij" was used to measure frostdepths in the field.Several tubes were placed at different locations into the soil. Airtemperatures and frostdepths were measured during winterperiods. The measurements of the winter 1979 are used here. Unfortunately there was no knowledge about several thermal parameters. The missing values were replaced by the numbers of morainesand.A simple numerical ...
|
|---|