A few days before paving is scheduled to begin, the contractor mixes a test batch of concrete to verify that the materials at the plant—including Type IS blended cement, fly ash, air-entraining agent and water reducer—produce a mixture with the desired properties. The preconstruction quality control pays off. Paving begins and continues steadily for 10 days with no problems. The mixture performs beautifully.
On the morning of Day 11, though, the mixture begins stiffening prematurely, losing its workability at the construction site. No cause is immediately evident. Subsequent testing reveals another problem: The air content is suddenly off target.
Performance reducers
This scenario has all the markings of materials incompatibility: unexpected changes in a concrete mixture caused by reactions between normally acceptable materials. But preconstruction testing had verified that the materials in stock were compatible. Nothing seems to have changed: Mix proportions, batching sequence and batching process all are being handled the same. Aggregate stockpiles have been consistently monitored for moisture content. Haul times and equipment, even the weather, are the same.
A quick investigation reveals that late on Day 10, the fly ash supply was depleted, and the supply was restocked from a new source. The new fly ash still meets specifications, so batch plant personnel were unconcerned. But, in fact, the new fly ash chemistry is just different enough to cause unexpected and potentially serious changes when it reacts with the sucrose-based water reducer. These changes negatively affect both workability and air content, characteristics that have important implications for pavement quality. The exact mix proportioning needs to be adjusted and revalidated with the new fly ash.
This contractor is reasonably fortunate: The incompatibility was manifested in an immediate, observable change in the concrete. In other cases, unexpected problems related to materials incompatibilities may not be so obvious or may manifest only after the concrete has been placed and it is too late to take mitigating action. In such cases, the problems may end up reducing concrete performance. For example, insufficient air-void spacing increases concrete’s susceptibility to freeze-thaw deterioration, reducing its long-term performance in Snowbelt states.
Chemically enhanced
Contractors need convenient, reliable, cost-effective procedures for flagging and correcting potential materials incompatibilities during mix design, materials selection, preconstruction mix verification, batching and construction. The National Concrete Pavement Technology Center at Iowa State University is helping on two fronts.
First, through the Materials and Construction Optimization for Prevention of Premature Pavement Distress (MCO) pooled-fund project (TPF-5[066]), Iowa State methodically worked with 17 states to identify optimum laboratory and field testing procedures for concrete materials and mixtures. The quick heat generation (or coffee cup) test, which notes changes in as-delivered cementitious materials, would have helped our contractor avoid those bad batches of concrete on Day 11.
A suite of tests identified in the MCO project is described in the Integrated Materials and Construction Practices for Concrete Pavement: A State-of-the-Practice Manual (reprinted November 2007, www.cptechcenter.org/publications/imcp/index .cfm). Three more products—a testing guide, a multimedia instructional hyperdocument for using the air-void analyzer, and an instructional video on the coffee cup test—will be posted at the same site by January 2008. More recently, through another pooled-fund project (Development of Performance Properties of Ternary Mixes, TPF-5[117]), an Iowa State team is focusing specifically on potential incompatibilities in ternary mixtures. By definition, ternary mixes contain two supplementary cementitious materials (SCMs) along with portland cement. The additional chemical complexity may increase ternary mixtures’ susceptibility to potential materials compatibilities.
The ultimate goal of this project is to develop performance-based specifications for ternary concrete mixtures. A first-phase report identifying general incompatibilities in ternary mixtures that can be predicted, and thus avoided, will be online in early 2008 (www.ctre.iastate.edu/research.htm). In summary, it describes incompatibility problems observed in 80 ternary mixtures—findings that are consistent with other recently published literature.
Highlights include the following:
- Mixtures containing ASTM C 618 (1) Class F fly ash require higher dosage rates of air-entraining admixture (AEA);
- Mixtures containing a polycarboxylate, high-range, water-reducing admixture (WRA) achieve air entrainment more readily than mixtures containing a sucrose-based, low-range WRA; and
- A combination of sucrose-based WRA, Class C fly ash and Type I portland cement can severely retard strength gain. Paradoxically, some mixtures incorporating Class C fly ash and surcrose-based WRA demonstrate flash-set tendencies.
Class of 80
The team has studied mortar samples containing natural river sand for the fine aggregate (fineness modulus and absorption of 2.81% and 1.12%, respectively) and various combinations of cementitious materials and chemical admixtures listed in the sidebar above.
First, 10 ternary mixture proportions were selected (Table 1). Based on the annotations listed in the sidebar, the mixture labeled 60TI-30C-10F contains 60% Type I portland cement, 30% Class C fly ash and 10% Class F fly ash by mass.
To study the effects of various admixture combinations on air entrainment, set times and strength development, each mix design was produced with the six possible combinations of air-entraining and water-reducing admixtures, all at the manufacturers’ recommended dosages: AEA1-WR1, AEA1-WR2, AEA2-WR1, AEA2-WR2, etc. This resulted in 60 different mixtures.
In addition, to determine the effects of higher dosages of low- and high-range water-reducing admixtures, mixtures containing AEA1 were produced with twice the recommended dosage of either WR1 or WR2. This resulted in a total test matrix of 80 mixtures.
Watered-down versions
Initial and final set times were determined for the various mortar mixtures using the Vicat testing procedure (ASTM C 191) (3). Early-age strength gains were determined using 2-in. cubes of mortar cast according to ASTM C 109 (4) and tested at seven days.
Air-void systems were tested using the air-void analyzer (AVA), a relatively new procedure and equipment that measures the size and spacing of entrained air bubbles in fresh concrete. The AVA is included in the suite of tests investigated by the MCO project, but the value of its results is still being proven in the concrete paving community.
The AVA procedure involves placing water in a riser column, filling the bottom of the column with glycerol, then injecting a mortar sample into the glycerol and stirring it for 30 seconds to release air from the mortar. Released into the thick glycerol, the air bubbles retain their quantity and size as they rise through it and then up through the column of water.
Large bubbles rise faster than small bubbles. All are collected on an inverted glass dish at the top of the riser column, and information about their size and number is automatically recorded as changes in the dish’s buoyancy over time. Within 30 minutes, software algorithms compute, and a monitor displays, the cumulative distribution of air voids, a histogram of air voids and numerical values for spacing factor and specific surface.
Samples for the AVA in this project were prepared by mixing mortar (ASTM C 305) (5) with a water-to-cementitious materials ratio (w/cm) of 0.45 and a sand-to-cement ratio of 2.75:1. The fresh mortar was then hand-packed into syringes and struck off at the desired 20-cc sample size. This method deviates from that recommended by AASHTO TIG (6) only in that it uses mortar samples prepared specifically for the test instead of samples obtained from fresh concrete.
Setting limits Set times
Table 2 shows the set times for all 80 mortar mixtures, including mixes with sucrose-based, low-range water reducers (WR1) or polycarboxylate-based, high-range water reducers (WR2) at recommended and double dosages. The acceptable range for initial set (ASTM C 191) is 45 to 375 minutes.
Certain results are especially noteworthy:
- Mixtures containing double dosages of WR1 generally have significantly shorter initial and final set times, indicating an incompatibility. To improve workability, change water reducers or other aspects of the mix design;
- Mortar mixes containing Class C fly ash generally set more quickly than those containing Class F ash;
- Increasing a mixture’s dose of WR2 generally delays initial set to a greater degree than it delays final set. This can extend mix workability; and
- Some mixtures containing Class C fly ash and WR1 display very rapid or even flash set. The set-time incompatibility may be eliminated by switching to WR2.
Air-void systems
Air-void systems were compared among mixtures containing high levels of Class C or Class F fly ash, respectively (Figures 3 and 4). The results generally confirm other published findings.
These results indicate that mortar mixtures containing blended cements (TIP, TISM, TIPM) generally produce better air-void systems. Conversely, mixtures containing large amounts of high loss-on-ignition (LOI) (that is, high in carbon) Class F fly ash generally produce poorer air-void systems.
Note that mixtures containing each of the three AEAs, using the same water reducer, produce similar spacing factors. However, as shown in Figure 4, mixtures with the two different water reducers generally produce significantly different spacing factors. Mixtures containing high-LOI Class F fly ash and WR2 may have more adequate air-void systems for freeze-thaw durability.
The majority of mortar mixtures in Figure 3 do not exceed the maximum limit for air spacing factor (0.2 mm [0.008 in.]). The mixtures in Figure 4, however, exceed the threshold unless they include WR2. In other words, a greater dosage of AEA is needed in ternary mixtures that include a high-LOI Class F fly ash.
Strength gain
Similarly, strength gains (average seven-day compressive strength) were compared among mixtures containing high levels of Class C or Class F fly ash, respectively (Figures 5 and 6).
Figure 5 clearly indicates reduced compressive strengths for mixtures containing 30% Class C fly ash, Type I portland cement and WR1.
This incompatibility is not evident in similar mixtures using polycarboxylate-based, high-range water reducer (WR2) or using blended cement instead of Type I portland cement. If development of compressive strength is unacceptably delayed in the field, simply changing water reducer, switching to a blended cement or adding fly ash may solve the problem.
In general, mixtures containing Class F fly ash do not experience strength-gain incompatibility issues (Figure 6). Strength gains are about the same for all admixture combinations.
Turning to ternary
Ternary concrete mixtures—those containing two supplementary cementitious materials along with portland cement—can be used effectively in pavement applications. However, before construction begins it is important to use actual project materials to verify the mix.
In addition, it is useful to test for incompatibilities as appropriate throughout the project. New project materials that meet mix design specifications may vary enough to cause incompatibility problems. Testing the air-void structure, setting times and early strength gain allow the contractor to take remedial action—like substituting cement from another source or changing the type of chemical admixture—if incompatible material combinations occur.
