Framtagande av temperaturlaster i broar med hjälp av allmänt tillgängliga klimatdata

To study and evaluate temperature distributions and subsequent load cases in concrete structures it is common to use thermal simulations. Such simulations have been performed in multiple previous studies using weather data measured in the vicinity of the studied object. However, this method has limitations, since important factors such as solar radiation and longwave heat radiation are only measured in a few places in Sweden. This means that thermal simulations can only be performed for structures located close to the measuring locations, if local measurements is to be avoided. This, in turn, limits the possibility of including all parts of the country when e.g. evaluating the load levels for various cases.

An alternative is to use models for estimating solar radiation and longwave heat radiation, such as STRÅNG, provided by SMHI, the Swedish Radiation Safety Authority (Strålsäkerhetsmyndigheten, SSM) and the Swedish Environmental Protection Agency (Naturvårdsverket), for solar radiation and models based on cloud cover and air temperature. With this kind of methodology, it would be possible to simulate the temperature in most places in Sweden without having to perform local measurements, i.e. the simulation can be made with easily accessible public data only. There is however an uncertainty related to the accuracy of STRÅNG and other possible models for estimating radiation.

In this study, the possibility of using easily accessible public data in thermal simulations is evaluated. The temperature has been simulated with easily accessible public data for three reference objects where the temperature in the structures has been measured in previous studies. Comparisons are made between the new simulations and measured temperatures, as well as with simulations using local data as input.

The results show that if the general trend and maximum and minimum values over time is of interest, the simulations using easily accessible public data give adequate results. If, however, the temperature at a specific point in time is sought, the new simulations may be unreliable. If the input data are taken from a location with a different climate than the location of the object, the results may differ significantly which is the result in one of the reference cases. The results presented here creates the possibility of e.g. producing isotherm maps for thermal gradients and differences in temperature between structural parts, since local variations can be included all over Sweden with the use of easily accessible public data.