| This dissertation presents the results of a series of tests aimed at quantifying self-equilibrating thermal stresses caused by AASHTO design nonlinear thermal gradients in segmental concrete bridges. Negative gradients (deck cooler than web) can cause significant tensile stresses to develop in the top few inches of bridge decks, leading to requirements for large prestressing forces to counteract this tension.Design gradients are based on field measurement of temperature variations on both a seasonal and diurnal basis. There is, however, little data in which actual stresses have been measured during these peak gradients to verify that the stresses are indeed as high as predicted by analysis. One reason for this is the difficulty of stress measurement in concrete. Stress is generally estimated by measuring strain, which is then converted to stress by applying an elastic modulus. This works well for homogeneous elastic materials but there is less confidence in this procedure when applied to concrete due to material variability at the scale of the strain gauge, temperature compensation of strain gauges, creep, and shrinkage.A 20 ft-long 3 ft-deep segmental T beam was constructed and tested in the laboratory for the purpose of quantifying self-equilibrating thermal stresses caused by the AASHTO design nonlinear thermal gradients. The beam was made of four 5 ft segments externally post-tensioned together with four high-strength steel bars. By embedding rows of copper tubing into two of the beam segments, and passing heated water through the tubes, the desired thermal gradients were imposed on the heated segments.Two independent methods were used to measure stresses at the dry joint between the heated segments. The first was to convert measured stress-inducing thermal strains to stresses using the elastic modulus. Stresses determined using this method were referred to as elastic modulus derived stresses (E stresses). The second method was a more direct measure of stress using the known stress state at incipient opening of the joint. This method of determining stresses was referred to as joint opening derived stresses (J stresses). Stresses determined using both methods are compared with AASHTO predicted self-equilibrating thermal stresses and discussed. |
