Practice Periodical on Structural Design and Construction

Bridge deck rehabilitation via overlaying has been widely used in the colder regions of the United States but not in Alabama due to a warmer climate and nonuse of deicing salts on bridge decks. However, Alabama does have some experience with bonded deck overlays, and the purpose of this work is to document the performance of the overlays that were placed in Alabama prior to the Summer 2000. The service life performance of the Alabama Department of Transportation’s 19 bonded bridge deck overlays to date are as follows:  • Thin (6.4 mm or 1/4 in.) urethane polymer concrete (four overlays) provided a service life of 3 years and left much to be desired before the 3 years.  • Thin (9.5 mm or 3/8 in.) polyester polymer concrete (12 overlays) has provided highly variable performances. Four of the overlays had service lifes of less than 1 year. The remaining seven overlays are approximately 10 years old and continuing in service; however, most are near the end of their service life.  • Thin (9.5 mm or 3/8 in.) epoxy co-polymer concrete (Flexogrid) (two overlays) has provided excellent performance. Both overlays are 8 years old and located on I-20 bridges and remain in excellent condition.  • Thin (12.7–19.1 mm or 1/2–3/4 in.) asphaltic based NOVACHIP (one overlay) has been in service for 3 years. During this short period, it has and is continuing to perform in an excellent manner.

Combined footings on piles are commonly used in bridges and buildings. However, there are very few references available regarding the structural analysis of these footings. This paper discusses three methods for analyzing combined footings on piles. The first approach consists of modeling the combined footing as a beam simply supported by the columns with pile loads represented by concentrated loads. The second approach is similar to the first one with the only difference being that column moments are applied at the supports. The third approach analyzes the footing as a statically determinate beam where the pile loads and moments, shears and axial loads from the columns are applied on the beam. The last approach is more accurate than the first two approaches.

Section K1 of the AISC allowable stress design and load and resistance factor design specifications prescribes a set of requirements by which the strength of the web of a flexural member or column must be evaluated when the web is subjected to a concentrated load. All of the work upon which these requirements are based consider only single point loads. Loading conditions may exist in which a beam is loaded by a pair of compression loads that are spaced in relatively close proximity. Application of the Section K1 provisions without consideration of the spacing of these loads may yield unconservative results. This paper derives means of evaluating when two compressive loads of such a pair may be considered to act independently and how they may be treated if they are close enough to interact with each other with respect to the strength of the web.

The paper recounts the writers’ experience in conducting an 890 kN (100 t) rock bolt pullout test, which was done to confirm the data for the design of the foundation to support a 54.860 m (180 ft) tall steel lattice self-supported communications tower to be located on South Mountain, Phoenix, Ariz. This was the writers’ first experience with a pullout test; a test of 890 kN (100 t) seemed too great a task, even when the design calculations indicated a factor of safety of 2.6. The tower foundation is to be designed with three independent reinforced concrete blocks, each of which must be tied down to the underlying rock by four rock bolts designed to withstand an uplift of 320 kN (36 t). During construction, the bolts will be proof tested to 445 kN (50 t) and then tied off at 355 kN (40 t), 110% of the design load. A 10.67 m (35 ft) rock bolt was placed in a hole with a depth of 9.45 m (31 ft) and a 75 mm (3 in.) diameter and grouted in place. The test was stopped at 890 kN (100 t) to prevent any possibility of a catastrophic failure. The conduct of the test is described, the calculations are explained, and the test equipment is shown. The paper discusses the procedures followed in conducting the test and the difficulty the writers encountered in placing complete faith in the engineering calculations, which were not supported by any prior pullout test experience. The writers recommend that any structural engineer who has a choice begin with a test smaller than 890 kN (100 t) for a first experience.

Wood I-joists have taken a large share of the wood floor framing market and have distinct advantages over solid-sawn joists. Like other wood products, I-joists are intended for dry-use environments. However, in the unusual event that I-joists are exposed to abnormally high or rapidly changing moisture environments, such as occur in certain disaster situations, research shows that the I-joists can undergo greater creep deflection than solid-sawn members exposed to similar environments. Field experience of the writer indicates that creep of I-joists in abnormally high moisture situations can render them unserviceable due to excessive deflection, even though the strength of the I-joists may not have been significantly reduced. Structural engineers and insurance investigators who are involved in natural disasters where I-joists are exposed to water, such as in floods or fires, should take measures to prevent damage to the I-joists from moisture-related creep deflection. Owners and designers of buildings with wood I-joists should also be aware of and take into account this unique performance characteristic of wood I-joists. A brief review of the literature on moisture-related creep deflection of I-joists is provided. Then field observations are discussed to illustrate practical situations in which I-joists can be exposed to abnormally high moisture conditions and experience problems due to excessive creep deflection. These situations include buildings that are waterlogged by fire-fighting operations, are flooded, or have very moist or wet crawl spaces.

Different types of fiber-reinforced concrete have been advocated and used as a replacement for welded-wire reinforcement (WWR) in slabs on ground, leading at times to conflicting opinion and/or unrealistic expectation about the benefits of using either WWR or different types of fiber in slabs. In order to verify some of these opinions, tests have been carried out on model slabs $(2.5m\times 2.5m\times 150\mathrm{mm}$ thick) cast on grade in a test pit and loaded to failure. Results are compared to evaluate the load-carrying capacity for slabs having equivalent amounts of either WWR, fibrillated polypropylene fibers, or steel fibers. Test results show that steel fibers are a suitable alternative to using properly positioned WWR. However, use of polypropylene fibers at a low dosage, which is most common for slab construction (0.9 kg/m³), provides little benefit in terms of increased load-carrying capacity when compared with a plain unreinforced slab. Slab capacity is affected mainly by the postcracking strength of the concrete section, and results from flexure beam tests give a good indication of expected behavior of slabs. The test results also show that a poor subgrade support can be compensated by reinforcing a slab with a moderate amount of either WWR or steel fibers (at a dosage of 30 kg/m³).

Constructability considerations are of importance in segmental bridge construction as the occurrence of failures of bridge superstructures under construction has highlighted. For a safe and economical construction process, the interactions between construction loads and the permanent structural system, depending on the chosen erection method, need to be evaluated. Segmental bridges can be constructed with methods like balanced cantilever construction, where individual spans are counterweighted about their substructure support. Time-dependent material properties like strength of the newly cast concrete, as well as shrinkage, creep, and relaxation influence the structural system resistance. Resulting stresses in the unfinished bridge structure during construction can even exceed the final stresses during service. This paper makes an educational contribution by illustrating these concepts with the case study of the Wilson Creek Bridge in Virginia. This five-span, cast-in-place bridge was constructed using balanced cantilever construction. Two form travelers were used to construct cantilever arms about the pier tables until the full span was finally connected at midspan; casting cycle duration for a single segment was 7 days. The contractor implemented major constructability changes in both the design and the construction of the bridge to facilitate a more economical construction process.

Two long-term loads concrete stress/strain models are introduced. One model, the typical long-term model, uses modified parabolic and linear equations to include creep and shrinkage effects. The other, the realistic long-term model, incorporates a realistic load path for concrete in compression as well as the creep and shrinkage phenomena. The proposed models were used to investigate the influence of their constituents on the ultimate strength of reinforced concrete-column cross sections under long-term loading. An analytic procedure was used to generate the strength-interaction relationships of the column cross sections. A series of calculations was performed for two selected column cross sections, and the short-term and long-term results are compared. The short-term model strengths computed using the same procedure compare well to the strengths computed according to ACI 318. Small differences are discernible between the strengths obtained using the typical long-term model and the short-term model at low reinforcement ratios. The differences increase at large reinforcement ratios. A noticeable reduction in strength can be observed above the balance point when using the realistic long-term model.