U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-HRT-04-150
Date: July 2006
Retempering is the process of changing the consistency of a concrete mixture by adding water and remixing. As it is common to send the concrete to the placement site with slightly less water than the maximum that may be used, it is expected that a specified amount of water can be added if necessary. The contractor may add the water because the mixture arrives at the site in a condition that would make placement and finishing difficult. These difficult HCCs are often called harsh mixtures. They lack workability. (The only quantitative measure of workability is slump.) (See sections 6.1 and 8.4, Gaynor and Meininger, 1983; Pigeon, Saucier, and Plante, 1990.)
The usual cause of a harsh mixture is sand with a high void content (see appendix D). Sands with a high void content are usually irregular in shape with an abundance of re-entrant angles and internal fractures and voids. Iron-stained clay coatings are common. Other causes of concrete that seems too dry are improper grading (size distribution) of the aggregate and the presence of mud or mud coatings on the aggregate. In addition, a deficiency of fine aggregate or coarse aggregate that is oversized or has a very poor particle shape can create fresh concretes with a difficult texture.
Mixtures with a low w/cm (below 0.45) can be difficult to place unless an effective water reducer is used. A good air-void system or the presence of fly ash as a substitute for part of the cement can help make a mixture with a low w/cm more workable. Apparently the air acts as a fluid and the particles of fly ash are more equant than those of cement and act as ball bearings.
Rims of cement on the aggregate and knots of cement in the paste (see section 8.7) suggest that the following typical scenario may have occurred. When the ready-mix truck arrived at the job site, it was quickly noted that the fresh HCC had a rough texture and looked as if it required more water. If the mixture was designed to have a low w/cm, each of the aggregate particles in the mixture was coated with this very adhesive mixture. Such HCC may be very difficult to place unless a sufficient quantity of a water-reducing admixture was used. If the coarse aggregate is oversized or has a poor shape or the sand is present in an insufficient amount, is unusually angular, contains many cracks, or has many re-entrant angles, the mixture will look stiff and difficult to place (harsh mixtures). It is common under such circumstances for water to be added to the mixture to increase the slump and workability. The additional water must not increase the total water above that designed for the mixture lest the concrete become weakened because of the higher w/cm.
When water is added after hydration of the cement has begun and mixing is restarted, it commonly happens (especially in mixtures with a low w/cm) that the water is not distributed throughout the entire mixture, but is mixed only into the larger spaces between the aggregates. The material already adhering to the aggregates remains as a rim of darker material with a low w/cm around the aggregate particles and in the re-entrant angles. Patches of the original paste (unaltered by the additional water) may remain and can be found to be completely surrounded by the paste with a higher water content. The problems of incomplete mixing are akin to the problems encountered in particular cooking situations. With gravy or white sauce, the thickening agent (such as flour) must be completely mixed with the cool water before the flour is affected by heat and begins to hydrate. If the flour and hot water mixture becomes too coherent, it may be impossible to add more water and create a smooth paste. The added water will mix with only a portion of the paste, and lumps of flour coated with stiff hydrated material will remain no matter how much mixing takes place.
Whenever water is added to the mixture without additional cement being added, the w/cm is raised. The higher the w/cm, the weaker the HCC. When more than the allowable amount of water for a given amount of cement is added to the mixture, the HCC will not have the designed strength. When the rims indicating incomplete mixing are present, a large portion of the cement can be concentrated in the thin bands of very rich paste around the aggregate and in the lumps of the original paste. The remainder of the paste is relatively depleted of cement and is thereby weaker than would be expected from the w/cm calculated from the originally delivered mixture plus the additional water. Thus, it can be seen that areas of HCC with a high w/cm can exist in close proximity to areas with a low w/cm.
It must be remembered that any material is only as strong as its weakest zone. Stress in HCC in service or in a testing apparatus will cause cracking. Cracks will always follow the zones of weakness. In HCCs that have paste areas with different w/cm’s, the cracks are going to develop in the areas of higher w/cm and thus the strength will be dependent on the extent and continuity of those areas.
The skeptic will mention the fact that the bond between the aggregate and the paste in many HCCs is the weakest area and say that the dark rims of high cement content eliminate this problem. Although this is true, the fact that weak bond areas are not as continuous throughout the paste as are the light-colored areas with a high w/cm (low cement content, high water content) obviates the value of rims with a high cement content as bonding agents.
Air-entraining agents are more active in the presence of additional water. When retempering has occurred and the mixing has not been completed, petrographic examination will show that many portions of the paste have a much higher void content than does the HCC of the rims and dark patches. Thus, the weakness of the portion with a high w/cm is compounded by the portion containing more than its proportionate share of air voids. In moderate cases, the spacing factor of the air-void system may change very little because the spacing factor is most dependent on the very small voids. Pigeon, et al. (1990), reported that there was little change in the spacing factor in the mixtures they studied if the retempering did not increase the slump by more than 100 mm.
When remixing takes place after some coalescence of the HCC has occurred, the remixing may occur after the individual integrity of some of the small air voids has formed. In such cases, many of these voids will retain their surface area, but lose their original spherical shape and become ovoid, or squashed, or develop an angular shape. Many angular voids may be seen in figure 74.
Retempering can cause an increase in the size of air voids, the number of air voids, or both. The size of the voids caused by retempering as evidenced by the microscopical examination shows that the larger voids (more than 1 mm across) nearly all occur within the portion with the higher w/cm. In normal, well-proportioned HCC, the percentage of voids whose diameter expressed on the surface examined exceeds 1 mm should be less than 2 percent of the total concrete.
Gaynor, R.D., and Meininger, R.C. 1983. “Evaluating Concrete Sands,” Concrete International, Volume 5, No. 12, pp. 53–60.
Pigeon, M.; Saucier, F.; and Plante, P. 1990. “Air-Void Stability: Part IV, Retempering,” ACI Materials Journal, Volume 87, No. 3, pp. 252-259.