Because of the time-dependentprestress loss, the value of fpe decreases gradually throughoutthe service life.When the structure is put to use, e.g., a simply supportedprestressed concrete girder, elongation of the lower prestressingstrands occurs, which leads to an increase in stress until fps isreached in the inner part of the beam (Fig. 1). By increasingthe tensile stress of the prestressing reinforcement, the diame-ter of the strand is reduced because of Poisson’s effect, whichdecreases the contact surface between the strand and the con-crete. Therefore, bond stresses are lower in lfb compared to lt.Because of this phenomenon, the structural concrete standardsconsider a bilinear law for the prestressing strand stress alongthe development length with a higher slope along the transferlength than that of the flexural bond length, as shown inFig. 1[1,4–6].Different experimental techniques exist for characterizing thebond behavior of prestressing steel strands:  Tests for determining the transfer length by manufacturingand analyzing prestressed concretemembers with centeredprestressing steel, called prisms [7–12].  Tests to estimate the development length through themanufacture and flexural testing of scaled prestressedconcrete beams with eccentric prestressing reinforcement[13–17].  Tests on non-prestressed concrete specimens to character-ize the bond behaviors of prestressing steel strands[7,11,17–19].In recent years, extensive experimental
studies have been per-formed in several US laboratories to compare three types of tests:the large block pull-out test (formerly called the Moustafa pull-outtest) [18], the Post-Tensioning Institute (PTI) bond test [19] and theNorth American Strand Producers (NASP) bond test [17]. The re-sults obtained from different laboratories were compared to assessthe reproducibility of each test. In addition, the results of each trialwere compared with the transfer and development lengths ob-tained from the manufacturing and testing of pre-tensioned rect-angular concrete beams and AASHTO-type beams. The studyconcluded that the most reproducible test is the NASP bond testbecause it is most representative of the bond capacity betweenthe prestressing strands and the prestressed, prefabricated con-crete members [17].Fig. 2 shows the typical cross sections of pre-tensioned pre-stressed box type girders produced at a precast factory in Spain. These girders weigh up to 200 metric tons and have a maximumlength of approximately 40 m. Substituting lightweight concretefor conventional concrete would allow an increase in their spanlength and, consequently, a reduction in the number of supportsnecessary for a given bridge deck length, which would significantlyreduce both the costs and environmental impacts of such struc-tures [16,20,21].2. Research significance (93 words)In this research, the three tests described in the previous sectionwere performed to characterize the bond behavior of prestressingstrands with a diameter of 15.2 mm in prestressed concrete gird-ers. The purpose was to study the feasibility of replacing conven-tional concrete with lightweight concrete in the fabrication ofprestressed concrete girders. This replacement would make it pos-sible to reduce the specific weight of girders by 20% and therebyincrease the length of the girders while remaining within the max-imum economically viable transport weight in Spain, which is esti-mated to be 200 tons. 3. Experimental researchThe upper bound of the development length for each concretemix can be estimated during structural testing to determine thebearing capacity of the beams in terms of flexure and shear capac-ity. Therefore, the principal objectives of this study are estimatingan upper bound for the development length of fully developed pre-stressing strands with a diameter of 15.2 mm (0.600) and evaluatingthe ductility of beams made with different mixes of conventionaland lightweight concrete [6].3.1. Variables studied in this investigationThe main variable studied is the type of concrete. The composi-tions of the studied lightweight concrete mixes (LC10-1 andLC10-2), and the typical conventional concrete mix used as at theprecasting plant for girders current production, used as a reference(NC), are described in Table 1.No confining reinforcement was required according to the cal-culations, which followed the analogy of a symmetrical prism[22]. Nevertheless, the first series of beams was prepared usingfour different amounts of confining reinforcement on LC10-1 con-crete beams to analyze the confining reinforcement effect on thepropensity to crack due to splitting. This measure was aimed atpreventing the splitting cracks detected during the fabrication ofall prestressed concrete prisms with this concrete mix that didnot have confining reinforcement [16,20]. Three identical beamswere produced for each confining reinforcement ratio for a totalof twelve beams that were designated V1L10-1 through V12L10-1. The prestress release age for the first series, which were pro-duced during the summer in a prefabrication plant, was 2 days.Therefore, the variable studied during this first series was the con-fining reinforcement ratio.Several days after the prestress release of the first series, split-ting cracks were detected in the lower side of all LC10-1 light-weight concrete beams, regardless of the amount of confiningreinforcement used; as a result, it was decided to modify the light-weight concrete mix to increase its tensile strength. After monthsof research, the LC10-2 mix was adopted, and a second series ofsix beams was produced with the same geometry, prestressingsteel and conventional reinforcement as the first series: three iden-tical beams of both LC10-2 and NC were produced, providing a to-tal of six beams. The maximum amount of confining reinforcementused during the first series of LC10-1 beams was used in all of thesebeams in an attempt to prevent or at least limit splitting. Therefore,the variable studied in the second series was the type of concrete.The prestress release age adopted for this second series, which wasmanufactured in the winter, was 3 days to ensure sufficient con-crete maturity [23].In this paper, the short-term results obtained from the bendingtests of the beams are analyzed. Currently, a long-term study isbeing conducted on the remaining beams to estimate time-depen-dent prestress losses and their influence on both the flexuralcapacity and ductility of these beams. 3.2. Properties of the materialsThe Y1860 S7 prestressing strand was used and had a diameterof 15.2 mm. The surface of the strand was slightly rusty due toweathering in Galicia (Spain), a region of high relative humiditythroughout the whole year. The strands were initially tensionedto 203 kN, which is the standard practice in the prefabricationplant. After the strands were tensioned and before they wereplaced in the concrete, the strands were cleaned with acetone to  Remove any form-release agent that might have stainedthe prestressing strand, which would impede its adhesionto concrete.  Eliminate any residue that had resulted from the wiredrawing of the strands.This procedure was intended to keep the same strand surfaceconditions used in previous studies performed at the universitylaboratory [16,20,21] to compare the experimental results. B400Stype reinforcement consisting of reinforcing rebars with differentdiameters was used [16].The mechanical properties of the quality control specimenstested at the time of release and after 28 days were examined forboth types of beams (age of prestress release was 2 days for theLC10-1 beams and 3 days for the LC10-2 and NC beams); the re-sults are presented in Table 2. The compressive strength at the pre-stress release age, the splitting tensile strength and the directtensile strength showed similar values for the three types of con-crete tested. Relative to the normal weight NC, the modulus ofdeformation at the time of prestress release was 65% for LC10-1concrete and 85% for LC10-2.Fig. 3 shows the compressive stress vs. longitudinal strain rela-tionship observed when testing the longitudinal elastic modulus atthe time of prestress release for the three concrete mixes tested(both tension and elongation are considered positive in this paper).The LC10-1 concrete has a similar compressive strength as the NCthat was used as a reference in this investigation, but its deforma-tion modulus is considerably lower than NC’s. The stress vs. longi-tudinal strain behavior of LC10-1 is fairly linear until failure, whichoccurs at a maximum longitudinal strain of approximately 2800 le(le); this deformation is lower than that of NC, which is 3200 le.The LC10-2 concrete tested after 3 days exhibits a modulus ofdeformation similar to that of NC and a higher compressivestrength. Its behavior is fairly linear until failure, which occurs ata deformation of approximately 2900 le; this value is similar tothat of LC10-1 and 12% lower than that of NC. In short, the com-pressive stress vs. 预应力混凝土梁英文文献和中文翻译(2):http://www.751com.cn/fanyi/lunwen_30871.html