|PDF Printable Version
of This Issue
Concrete placements for the arch were made at night.
The Colorado River Bridge at Hoover Dam—Concrete Production and Placement
Jeff St. John, Obayashi/PSM Joint VentureIn early-to-mid 2005, the joint venture of Obayashi Corporation and PSM Construction USA began assembling on the project site of the Colorado River Bridge. One of the chief topics of discussion amongst the team was the example of public works adjacent to the site—the Hoover Dam, one of the twentieth century’s greatest engineering and construction feats. Discussions hit on many of the usual topics regarding that structure; how did they handle the intense heat of a site, which can approach 130°F (54°C), and how did they get the workers, equipment, and materials to the work site? The same challenges would need to be faced 70 years later.
It was readily apparent that a single challenge dwarfed all others; how to build the concrete arches. The main concern was the concrete itself. This was a multi-faceted concern that involved the mix design, thermal control, concrete delivery and placement, consolidation, and possibly the chief concern of all, quality control.
Work started on developing a mix design 2 years prior to casting the first arch segment. Many of the requirements had been established by the Federal Highway Administration and the designer (T.Y. Lin International). Among these parameters were the required concrete compressive strength of 10,000 psi (69 MPa) at 56 days, aggregate selection to ensure long-term durability, and thermal control requirements to minimize cracking and, again, ensure long-term durability. To these design requirements, the construction team added several others to overcome delivery and placement challenges (pumpability, flowability, and long set time) and schedule challenges (rapid strength gain to minimize the form traveler cycle time). These included compressive strengths of 4000 psi (28 MPa) for launching the form travelers and 6000 psi (41 MPa) for stressing the temporary stays used to support the arch during construction.
The first step was to call in the cavalry: in this case international experts. Dr. Ryuichi Chikamatsu from the Obayashi’s Technical Research Institute in Japan was brought on-board to primarily consult on the mix design. Paul Jordan of Sika Corporation (the admixture supplier) lent his advice and helped with innumerable trial batches. Dr. Wilbert Langley from Halifax, Nova Scotia, Canada, also consulted on the mix design as well as the thermal control requirements. The mix design used was a direct development of Dr. Chikamatsu and Sika, and the thermal control plan was by Dr. Langley.
The mix design met all of the criteria set for it. Short-term and long-term strengths were achieved using the following concrete mix proportions:
This mix typically achieved strengths of 4000 psi (28 MPa) in just over a day and over 12,000 psi (83 MPa) in 56 days. Pumpability and flowability were addressed by the use of a high-range water-reducing admixture, which resulted in concrete slump ranges that neared those of self-consolidating concrete. Set times in excess of 2-1/2 hours were achieved using a retarder.
However, the high cement content had a less desirable effect. The concrete in its natural curing condition would reach temperatures in excess of 190°F (88°C). This was far above the 155°F (68°C) limit of the contract specifications. Most typical mitigation methods, such as using chilled batch water or ice chips, shading the aggregate stockpiles, and casting at night, couldn’t come close to reducing the maximum curing temperature to the target range. Only two realistic options remained; circulation of cold water through pipes embedded in the concrete or the use of liquid nitrogen to precool the concrete to a temperature such that its maximum peak curing temperature would be less than 155°F (68°C). Many miles of cooling tubes had been used to control curing temperatures during construction of the Hoover Dam. The location, cycle time, installation, and maintenance issues involved with cooling tubes ruled them out for the bridge. Only the liquid nitrogen option remained.
The use of liquid nitrogen allowed the temperature of the concrete during the summer to be lowered from a batched temperature of 85°F (29°C) to a predelivery temperature of 40°F (4°C). In turn, this kept the temperature at point of placement in the 60°F (16°C) range resulting in peak curing temperatures of less than 150°F (66°C). The cost of using liquid nitrogen to cool concrete is very high. During the heat of a southern Nevada summer, the cost of the nitrogen required for cooling often exceeds $100/yd3 ($131/m3). However, the high cost was mitigated by the minimal effort needed in other activities. No maintenance for water supply and form insulation and no mitigation efforts such as grouting of cooling tubes and leaving forms in place for an extended duration were required. The precooling resulted in a product that did not require any further thermal control measures and, with the unique bridge structure and location, offered the only viable option.
The nitrogen-cooled concrete was beneficial to the placement system during the very warm summer months, where even temperatures at night occasionally did not fall below 100°F (38°C). During the very hottest portions of the summer, it was necessary to precool the concrete pumping line by filling it with chilled water prior to the placement, wrapping it with burlap, and soaking the burlap with chilled water to reduce heat gain through the placement system.
Planning for concrete delivery and the placement system preoccupied the team for a long time. Two options were apparent to get the concrete to the point of placement; use of a pumping system or delivery by cableway (high-line) concrete bucket. Delivery by bucket to the point of placement (the same methodology used for construction of the Hoover Dam) was discarded for several reasons, not the least of which were tying up a critical resource for several hours nearly every day and the size of buckets required to maintain precise control of discharge into a very small target area of the placement openings in the traveler cover forms. The decision was made to use a concrete pumping system.
Challenges for pumping included the harsh aggregates of the concrete mix, the long pump line, the means to place through the restricted openings, and delivery of concrete to the pump. Trailer pumps, specially modified to handle the harsh local aggregates, were selected due to their ability to fit in the tight areas available for setup. Delivery to the pump was easy on the Nevada side of the gorge. The pump could be set up on the roadside that is very near the arches and concrete delivered by truck. The Arizona side, with its tremendously steep cliffs, was another story. There, the trailer pump was set up on the base of the arch in conjunction with a 5 yd3 (3.8 m3) re-mixer. Concrete was discharged from the delivery truck into 8 yd3 (6 m3) concrete buckets supported by the cableway and lowered to the re-mixer, where the buckets were discharged. Use of the re-mixer allowed the buckets to be re-hoisted nearly immediately to receive the next load of concrete. Tying up the cableways for these placements was a significant issue but no other realistic option was identified.
From the trailer pump, the concrete was pumped up the arch through a 5-in. (125-mm) diameter heavy wall line over a distance of 600 ft (183 m) horizontally and 250 ft (76 m) vertically to a 105-ft (32-m) long placing boom mounted on the arch near the form traveler. This placing boom allowed precise control of discharge. A typical arch segment placement took 4 to 5 hours.
Consolidation of the concrete in the formwork was a major concern. The geometry of the arch, with many segments placed at a steep angle, required the use of top surface forms. Openings were established in the forms, not only for placement, but also to allow the use of high-cycle concrete vibrators. In addition, external vibrators were mounted under the bottom soffit form and along the side forms to help eliminate any issues due to the lack of concrete consolidation. As a tribute to the hard-working concrete placement crew, very little honeycombing was encountered when the forms were removed.
During the planning sessions, an often discussed topic was quality control. One bad load of concrete could plug up the placement system and lead to a half-completed segment, which would need to be removed. Another concern, a load of low quality concrete could fail to achieve strength, which might not be determined until several segments later. The financial implications and loss of schedule and momentum would be staggering. It was obvious that the quality control efforts needed to go above and beyond the usual industry standard.
Experience from the footing construction demonstrated that traditional ready-mix concrete batching would not meet the quality requirements of the arch concrete. Too many aspects of the concrete are subject to variations such as the batching efficiency of the truck’s mixing drum and the drive time. After much discussion and research, a portable batch plant incorporating a 5 yd3 (3.8 m3) pan mixer was purchased and set up on the project site.
Pan mixers use high speed paddles to premix the concrete prior to discharging into the truck. They are most traditionally used in precasting plants but were perfect for the application since quality, not volume per hour, was the most important issue. The batch plant operator was a master at determining the quality of the predischarged concrete and was able to make adjustments such that the slump of the concrete rarely varied more than 1/2 in. (13 mm) during a placement. The Quality Control Manager personally tested every single load of concrete prior to sending it to the pump. The proximity of the plant to the site made it extremely easy to make adjustments throughout the course of any placement.
The success of all of these efforts can be seen in the finished bridge. No delays were encountered during arch construction due to pumping or placement issues, nor were any problems encountered with the quality of the concrete. The arch construction actually went faster than anticipated and resulted in a monument that nearly rivals the beauty and awe of its neighbor.
Further information about the project, including construction photographs, is available at www.hooverdambypass.org.
HPC Bridge Views, Issue 63, Sept/Oct 2010