|
KSC Komponent® — REPORTS AND NEWS
Type-K Shrinkage-Compensating Cement in Bridge Deck Concrete
Paul W. Grunner
G. Alan Plain
Concrete International, October 1993.
Cracking of newly placed concrete bridge decks is becoming an increasing problem in the United States. Use of Type-K Cement concrete is one solution to the problem. Our goal is that this report will encourage others to use Type-K Cement concrete in the construction of bridge decks in view of its proven performance in reducing shrinkage cracks. As more agencies use Type-K Cement in bridge concrete, more data will be available to evaluate this cement’s performance in other areas such as reduced permeability.
The problem – cracking
Today, many states use modified mix designs with higher cement content when specifying concrete for construction projects. In some cases, these modified mixes result from problems with “D-cracking,” or the disintegration-cracking of concrete. Since it has been determined that large coarse aggregates are the primary cause of D-cracking due to their poor freeze-thaw durability, the maximum size of coarse aggregates has been reduced to meet freeze-thaw testing requirements. However, reducing the coarse aggregate size results in decreased unit coarse aggregate content. In order to maintain yield, the coarse aggregate is replaced by mortar that has a higher shrinkage rate than coarse aggregate. Also, maintaining a given slump with the same aggregate shape requires more water when the coarse aggregate size is decreased. This demands a higher cement content for equal water/cement ratios. At the same time, higher cement content and additional water both increase shrinkage. In many cases, cement contents have been increased to provide higher compressive strengths, denser mixes, and/or greater durability.
The higher cement content and higher water content lead to increased drying shrinkage, which causes full-depth transverse cracks in bridge decks. In most cases, these cracks occur shortly after initial curing. Transverse cracks are not only unsightly but also lead to increased water penetration to the levels of the reinforcing steel and to the undersides of the decks where efflorescence at the cracks often becomes apparent. Much debate has arisen about the impact of the cracking on the overall durability of the deck, and the possibility exists that cracks accelerate deterioration. In fact, research from Japan(1) indicates that penetrating cracks cause bridge deck deterioration. Certainly most owners, designers, suppliers, and contractors prefer to have a final concrete product with very few or no cracks at all.
An additional detrimental effect of full-depth cracks is that they permit increased water and chloride leakage through the deck to the supporting superstructure members and substructure units below. A crack-free, low permeability deck combined with joint-free or minimum-joint design can provide valuable protection for all parts of the bridge lying below the deck.
In addition to cracks caused by drying shrinkage, another factor in bridge deck deterioration is the permeability of the deck. Bridge decks that are highly permeable allow water and salt to penetrate deeper into the concrete to the levels of reinforcing steel, thus causing corrosion of the steel and deterioration of the entire deck.
One solution to cracking and permeability
One method of reducing or eliminating drying shrinkage cracking is to use a shrinkage-compensating cement such as ASTM C 845 Type E-1(K), otherwise known as Type-K Cement. Type-K Cement contains a sulfoaluminate (C4A3S) compound that makes it different from other cements. During the hydration process the sulfoaluminate forms ettringite, an expansive material. This ettringite formation creates an initial expansion in the concrete during the first few days of curing. After maximum expansion is reached, the Type-K Cement concrete shrinks at approximately the same rate as portland cement concrete, and the eventual net volume change is near zero. Consequently, very few or no drying shrinkage cracks occur.
It is important that the concrete be elastically restrained, such as by reinforcing steel within the limits suggested by ACI 223-93,(2) to eliminate shrinkage cracking. The reinforcing steel doesn’t resist the shrinkage of the concrete that would result in tension in the concrete and compression in the steel. Instead, tension is built up in the reinforcing steel during expansion. This tension then dissipates during shrinkage. The length-change characteristics of shrinkage-compensating concrete versus portland cement concrete are shown in Fig. 1.
To investigate the performance of its deck overlay and sealing systems, the Ohio Department of Transportation (ODOT) performed a series of tests on a variety of bridge decks. Even though the study was not directed toward the study of Type-K Concrete decks, results indicated that shrinkage-compensating cement concrete is, in fact, less permeable than portland cement concrete with a similar mix design. (3)
During one particular permeability test for a bridge deck in Brown County, Ohio, half of the deck was placed with portland cement concrete, and half with Type-K Cement concrete. After the deck had cured, cores were taken and subjected to the AASHTO rapid permeability test to measure the average through the cores. A higher charge passed would indicate a higher penetration of water and chlorides, which conduct electric current through the concrete. The half of the deck with the shrinkage-compensating cement concrete showed an average charge passed of 2,286 coulombs, while the half placed with portland cement concrete passed 9,015 coulombs.(3) These results may indicate that Type-K Concrete reduces a bridge deck’s permeability.
While Type-K Concrete may appear to reduce permeability, the elimination of shrinkage cracking is its most important and proven function. If drying shrinkage and plastic shrinkage cracking are eliminated, the only path remaining for chloride ion migration is through the pore structure and capillaries of the concrete. According to a report prepared by Gulen and Sudol for the Indiana Department of Highways in 1984, expansive cement significantly reduced deck cracking.(4) In addition, the expansive cement concrete surface appeared to be as durable as regular cement concrete when placed with normal care and good workmanship.
Gulen and Sudol did publish some conflicting results to other reports. Contrary to Bunke’s findings regarding the lower permeability of shrinkage-compensating cement concrete, Gulen and Sudol found that expansive cement concrete has about the same resistance to chloride penetration as regular cement concrete.(3,4) These results indicate that more permeability testing is required. While most people who have used shrinkage-compensating concrete agree that it may form a more dense concrete matrix, making it a less permeable form of concrete, additional testing of in-place decks is necessary to verify that decreased permeability does exist.
Type-K Cement concrete in bridge decks
Testing of Type-K Cement concrete in bridge decks has been conducted in Ohio since 1968, on U.S. Route 33, in Marysville and since 1974 on Interstate 675 in Dayton. The use of Type-K instead of Type I concrete on Ohio bridges has consistently resulted in excellent performance:
- Fewer cracks have been noted.
- Chlorides have not migrated through the cracks to the bottom of the bridge decks.
- Virtually none of the Type-K Concrete decks have required overlays.
Evidence supporting the improved performance of Type-K Cement concrete is mounting. The Ohio Turnpike Commission (OTC) began a bridge replacement program on the Ohio Turnpike in 1984. Approximately 50 bridges were scheduled to be replaced all that year; all but three had extensive transverse cracking. One of the three employed Type I cement; one deck used stay-in-place forms and was post-tensioned; and the last of the three was the only one of the 50 bridges constructed with Type-K Cement concrete.
The OTC contracted with Construction Technology Laboratories (CTL) to study the problem. CTL issued a report titled “Evaluation of Concrete Deck Cracking for Selected Bridge Deck Structures of the Ohio Turnpike.”(5) The report concluded that the problems were caused by too high a cement content, too low a large aggregate content, and improper curing. It was recommended that the mix design be changed to lower the cement content, increase the course aggregate size, and increase the aggregate content. Alternatively, a shrinkage-compensation cement could be used. In addition, water curing was recommended for hot weather placing, and Type III cement was to be avoided. Wiss, Janney, Elstner Associates, Inc., also studied deck cracking problems in the Hope Memorial Bridge, Cleveland, for ODOT(7) and arrived at some similar conclusions.
However, OTC could not readily change its mix design. Because a preponderance of the concrete produced in Ohio is in accordance with ODOT specifications, aggregate producers charge premiums for special aggregates that require additional screening. OTC therefore, chose the alternative of using Type-K Cement concrete, which uses a standard aggregate. As a result all decks replaced on the Ohio Turnpike since 1985 have been constructed with Type-K Cement concrete and modified water curing methods with great success. Between 1985 and 1990, mor than 300 decks have been replaced with this type of concrete, and no further transverse cracking problems have been observed. The results are shown in detail in Table 1. Sixteen of these decks have also included lightweight aggregate to improve the load-carrying capacity of the bridges.(8)
Many decks with Type-K Cement concrete have been constructed by ODOT, including the I-490 bridge in Cleveland that is 3,462 ft long with 14,000 cu. yd. of deck concrete.(6) Since April, 1991, Type-K Cement concrete is included in all projects as an alternative bid item, and the Department normally pays the additional cost. The I-490 deck is one of a number constructed by ODOT that have included Type-K Cement in the deck concrete, epoxy-coated reinforcing steel in both mats, and a latex-modified concrete overlay. An important practice when using a Type-K Cement concrete deck with a latex-modified concrete is to place the latex-modified concrete overlay a few days after the deck concrete, when the concrete has attained its maximum volume. Montgomery County, Ohio, has also used these techniques on several bridges, and the results for both ODOT and Montgomery County have been high quality, finished decks with virtually no cracks. Clark County, Ohio, recently constructed a bridge with a deck containing silica fume and Type-K Cement.
Several other states have successfully used Type-K Cement recently and are reported as intending to increase their usage. Decks have predominately been supported on steel beam or girder superstructures, but Type-K Cement concrete has also been used in decks on pre-stressed concrete I-beams and in continuous slab superstructures spanning longitudinally. The decks on prestressed concrete I-beams have been less successful, apparently due to the massive prestressed members and large prestressing forces resist the initial expansion. Some reduction in cracking is realized, but not the nearly 100 percent reduction experienced in Ohio.
Type-K Cement specifications
Having been used since the late 1960’s, Type-K Cement is not a new or experimental material. Many specifications are available for materials used in construction with Type-K Cements. Some of these specifications include ASTM C 845; ACI 223-93, Standard Practice for the Use of Shrinkage-Compensating Concrete, and the Department of the Army Corps of Engineers Manual EM1110-2-2000, dated September 25, 1985, Engineering and Design Standard Practice for Concrete.
For approximately five years, ODOT has been specifying Type-K Cement in selected decks by means of a few simple plan notes. The states of Michigan and Indiana also use special provisions in their contract proposals for Type-K Cement on selected projects.
Construction Procedures
No special equipment is required for placing and finishing concrete made with Type-K Cement. Generally speaking, normal, sound practices can be followed for mixing, placing, finishing, and curing the concrete. Even though Type-K Cement concrete is more expensive than portland cement concrete, it has several advantages that can mitigate these costs. For example, it is easier to finish and place Type-K because of its higher slump requirements. Therefore labor costs are reduced, especially for decks of 300 cu. yd. or more.
With a few simple precautions, Type-K Concrete produces higher slump and cohesiveness – or “fat” characteristics. These characteristics provide good finishing qualities. However, because attributes make Type-K Cement concrete different from other concretes, owners and contractors must consider the following factors when using Type-K Cement concrete:
- Fog-spraying should be available, especially for placement during hot weather, to reduce possible plastic shrinkage cracking.
- Delaying placing activities at the jobsite should be avoided since shrinkage compensating cement concrete loses slump faster than normal concrete.
- Concrete made with shrinkage-compensating cement will exhibit little or no bleed water; therefore, care must be taken to begin finishing operations at the proper time.'
- Placing of the concrete should be scheduled early in the day before temperatures rise; the deck should not be placed if the ambient temperature is above 80 F. Evening placement also has advantages in that the greatest exothermic heat dissipation during the cool nighttime ambient temperatures.
- Water curing should always be used with Type-K Concrete decks.
Special uses of Type-K Cement
Type-K Cement has a variety of special uses. It can be employed for widening of bridge decks – since the initial expansion compensates for the shrinkage, cracks usually will not develop at the joint between old and new concrete. Consideration should be given to permitting expansion parallel to the joint. As indicated above, concrete using the special cement is also easily pumped in lightweight as well as normal weight concrete.
In situations where the appearance of abutment walls, wing walls, and other portions of a structure are important, Type-K Cement concrete is effective in preventing unsightly cracks. Care should be given to wall footing details in accordance with Chapter 3 of ACI-223-93. (2) Some limited testing is currently being performed using concrete with other additives or systems, including corrosion inhibitors, latex, silica fume, and special overlays.
ACI member Paul W. Gruner is a graduate of the University of Cincinnati and a registered professional engineer. He is a Partner of Woolpert, responsible for the firm's transporation and structural engineering, and a member of the State of Ohio Board of Building Standards. He is a member of ACI Comittee 223, Expansive Cement Concretes.
G. Alan Plain is a graduate of Cleveland State University and a registered Professional Engineer. He is the deputy Executive Director and Chief Engineer of the Ohio Turnpike Commission and is responsible for the Commission's successful use of Type-K Cement concrete in its bridge deck replacement program. He has served as chairman of the Engineering and Design Committee of the International Bridge, Tunnel, and Turnpike Association.