Konvektiivinen lämmönsiirtyminen ratapenkereessä

Modern track evenness demands are high which is why frost heave causes
considerable problems in Finland. Nowadays railway ballast and sub-ballast are often
of crushed rock whose thermal performance differs from that of the traditional gravel.
Many studies around the world have found that convective heat transfer may occur in
very coarse crushed rock aggregate. This can lead to increased heat transfer from the
embankment which allows frost to penetrate deeper than normal calculation methods
predict.
Many studies have been conducted on free convection internationally, and there is a
calculation model for estimating its probability. However, the calculation involves
problems because the calculation model considers intrinsic permeability, which is
difficult to define for coarse materials. Intrinsic permeability can be calculated from
water permeability. Unfortunately, water permeability is difficult to determine for
coarse materials. The results of research conducted abroad are not as such applicable
in determining convection in Finnish railway embankments because of different grain
sizes. Therefore, new laboratory test apparatus was built. Three materials of different
gradation were tested in the lab: 31.5/63 mm ballast aggregate, 5/16 mm crushed
rock aggregate, and 0/63 mm sub-ballast material.
Strong natural convective heat transfer via air occurred in the railway ballast tested in
the lab. If the medium was moist air or water, convection could occur at smaller
temperature differences. No significant convection occurred in the crushed rock and
sub-ballast material via convection, but when water was added to the sample, thermal
conductivity of the material increased significantly due to convection. The same
phenomenon was observed in the case of 5/16 mm crushed rock, but adding of water
did not increase thermal conductivity significantly.
Two actual railway embankments on which temperature data was available were also
tested. The data allowed estimating the temperature differences in the embankments.
Based on the results from the test embankments in Seinäjoki, Finland, the thermal
performance of the gravel embankment and crushed rock embankment differed, but
the possibility of natural convection could not be proved nor completely excluded.
The study showed that crushed rock ballast and sub-ballast material may allow
natural convection to occur if its grain size distribution is close to the most coarseand
even-grained materials allowed by the current guidelines Sorting of the material
in the building phase can increase this susceptibility to convection.