The temperature effects on an experimental, prestressed concrete, segmental, box-girder bridge were evaluated. Field measurements, including vertical deflections and thermocouple readings for temperature, were collected over a 1-year period under various weather conditions. Critical ambient temperature conditions were gleaned from field observations. With the field observations (surface temperatures) as input, three-, two-, and one-dimensional heat-flow analyses were conducted. The three-dimensional heatf low analysis was a statistical comparison of temperature readings at various longitudinal locations in the experimental bridge. The two-dimensional analysis involved the collection of hourly temperature readings from 24 thermocouples at the midspan of the test bridge for 18 diurnal cycles within a 1-year period. In addition an alternating-direction implicit finite difference analysis was performed, and the results compared well with the field observations. Finally, a one-dimensional approach was formulated with the application of an initial-value method. This was found to be in agreement with the alternating-direction implicit finite difference model and the collected field data.

In the past, the thermal environment was ignored in the design process except for accounting for the longitudinal movement that could occur and the influence such movement could have on joint integrity. However, relatively recent incidents of bridge distress, including the Newmarket Viaduct in New Zealand<!>, the fourth Danube Bridge in Vienna, and the development of a crack 56 ft (17 m) long and 0. 20-0. 24 in.(5-6 mm) wide in the web of the Jag st Bridge in Untergreisheim (2) have prodded the profession to address the complete structural response to the thermal environment. A field study was conducted to determine the complexity and magnitude of the temperature problem