553 ACI Structural Journal/September-October 2015 ACI STRUCTURAL JOURNAL TECHNICAL PAPER An experimental study was conducted in which 30 full-scale, precast, pretensioned bridge girders were constructed and instru- mented with the intention of investigating prestress loss. Several different precast beam fabrication plants were used to investi- gate the infuence of different concrete materials and construc- tion techniques. The constructed girders were conditioned in several different climates for up to 3 years. During this period, prestress loss was measured by using vibrating wire gauges (VWG) embedded in test specimens. Following the conditioning period, the girders were fexural service load-tested to quantify the prestress loss at the time of testing and in turn verify the losses measured using VWGs. Prestress losses were found to be heavily infuenced by concrete stiffness, which was dependent on coarse aggregate type and quantity. The measured short- and long-term prestress losses were compared to those determined using several different estimation procedures, suggested by ACI Committee 423. Keywords: prestress loss; prestressed bridge girder; pretensioned concrete. INTRODUCTION Prestressed concrete construction relies on the application of compressive stress to concrete elements; the goal of this compressive stress being to reduce the maximum concrete tensile stresses and thus preventing cracking. The pre-com- pression stress (that is, prestress) is applied to the element using tendons either stressed prior to concrete placement (pretensioning) or after the concrete is allowed to harden (post-tensioning). Over time, the prestress will fuctuate due to concrete behavioral mechanisms (for example, creep and shrinkage) and external events affecting the member (such as deck placement). Any decrease in the prestress is considered prestress loss and any increase is stress gain. Stress gains are caused by elongation of the strand, typi- cally a result of a positive moment being placed on the beam either by an external load or by the differential shrinkage of the deck; many sources elaborate on this phenomenon. 1-4 For the purpose of this paper, losses are positive and gains are negative. The stress in the strand must be properly estimated throughout the life of the concrete member to ensure proper crack prevention (that is, extreme fber concrete stress is always below the prescribed tensile stress limits), if such is desired. The stress in the prestressing strand immedi- ately prior to transfer can be determined by monitoring the applied stress and verifed by measurement of strand elonga- tion. After the stress is transferred to the member, however, a method for estimating the prestress loss and gain is required for the designer to estimate the strand stress at different points in the life of the member. The main factors affecting prestress loss are either related to concrete deformations (elastic shortening, creep, and shrinkage) or relaxation of the prestressing steel. As with typical deformation-related problems, the problem of concrete deformations can be approached by looking at the stresses applied on the system and the effective stiffness of the system, the system being any reinforced concrete member or structure under some type of sustained loading. Stresses acting on concrete are primarily a result of mois- ture movement and externally applied stress. The movement of moisture will drive the concrete behavioral mechanism of shrinkage. Concrete creep will be driven by the exter- nally applied stress but will also be greatly infuenced by the movement of moisture. Moisture movement (primarily in relation to shrinkage) is thought to be driven by several different phenomena. While the beam is being moist- or steam-cured, the initial free water present will be partially “lost” as it is used to hydrate the cement. Once the beam is exposed to the environment, additional water will be lost through further cement hydration and as the internal system’s relative humidity equilibrates with the external ambient rela- tive humidity. Many factors (such as cement particle size, water-cement ratio, aggregate porosity, and supplemental cementitious material type and quantity) affect the ease of the free water to move through the concrete matrix. Prior to initial set, any free water loss will not cause any stresses to develop within the system. After cement hydra- tion has caused the formation of the concrete matrix, further free water loss will cause the development of stress on the system, with the magnitude of these stresses primarily dependent on the pore system and the water content within the matrix. A further explanation of the shrinkage mecha- nism is beyond the scope of this introduction, but is fully explored by several other investigations. 1,5-9 While a detailed explanation of the creep mechanism is also considered to be beyond the scope of this paper, it is important to recognize that creep is also infuenced by the moisture content. The mechanisms driving creep are discussed in detail by several other investigations. 1,8,10-13 The stresses experienced by the system are mainly resisted by the system stiffness, which is primarily provided by the hardened cement paste and coarse aggregate. The coarse Title No. 112-S45 Experimental Investigation of Prestress Losses in Full- Scale Bridge Girders by David B. Garber, José M. Gallardo, Dean J. Deschenes, and Oguzhan Bayrak ACI Structural Journal, V. 112, No. 5, September-October 2015. MS No. S-2014-097.R3, doi: 10.14359/51687909, received September 22, 2014, and reviewed under Institute publication policies. Copyright © 2015, American Concrete Institute. All rights reserved, including the making of copies unless permission is obtained from the copyright proprietors. Pertinent discussion including author’s closure, if any, will be published ten months from this journal’s date if the discussion is received within four months of the paper’s print publication.