Editors Note: Recent changes in admixture technology are addressing mix-set time. With that in mind, we revisit the basics of mix chemistry in this first of a two-part report.

Concrete producers expect cement to remain versatile, maintaining a consistent and predictable performance with all types of concrete mixes. Concrete workability problems can be costly and affect concrete producers and users alike. Stiffening properties can arise from either false setting or from stiffening due to aluminate control problems (also known as flash setting). Sulfate and aluminate characteristics often are the key to understanding the cause of these stiffening properties.

Premature stiffening of a mix also can result from incompatibility among concrete components. The addition of fly ash will be discussed briefly, since fly ashes containing high amounts of aluminate or alkalies can affect the proper sulfate balance. Chemical admixtures also can disrupt control of the early aluminate hydration by the sulfates, but this topic is too broad to be covered here. Paramount, however, is the fact that many cement parameters can play a role in such incompatibility, including grain size (clinker mineralogy) and amount of tricalcium aluminate (C₃A); the content, chemical form and fineness of sulfate bearing phases; alkali and free lime contents; and, fineness and prehydration of cement.

IMPORTANCE OF SULFATE

Sulfate is added to cement to control aluminate hydration and to enhance tricalcium silicate (C₃S) hydration, promoting improved strength development. The amount, form and fineness of sulfate dictate its solubility and therefore its effect on aluminate hydration. This interaction between the sulfate and clinker phases is important to note, as it influences concrete workability, strength, setting time, drying/shrinkage, and expansion.

What follows are examples of common cement sulfate and aluminate characteristics, descriptions of what happens during hydration, and the resulting paste properties that would be expected. The figures are graphical interpretations meant to demonstrate the reactions and are not necessarily accurate in features such as size and distribution.

CASE 1: Ideal Sulfate Conditions

Condition — “Ideal” sulfate amount, size and form to control the aluminate hydration. The sulfate form in this example consists of both gypsum and plaster. The gypsum added to the finish mill will typically lose some of its water and be converted to plaster, depending on the temperature and relative humidity of the mill and the retention time. It is generally desirable to have some plaster in the cement, as will be described below.

Gypsum: CaSO₄•2H₂O
Plaster (or “hemihydrate”): CaSO₄•½ H₂0
5-10 minutes after hydration

Reactions — Plaster rehydrates and converts to gypsum. Plaster goes into solution quicker than gypsum, providing sulfate ions sooner. Aluminate phase reacts with sulfates in solution, especially those provided by the plaster, forming ettringite.

Paste Properties — With the proper amount of sulfate ions available, the aluminate hydration will be controlled, leading to paste of proper plasticity, normal setting and hardening properties. Calcium sulfate also will enhance the reaction of alite, which is likely a cause of increased early strength with optimum sulfate content. This is the basis for ASTM C 563 (1), which determines “optimum” sulfate solely on strength development at one day. To accelerate the hydration of C₃S, the Ca²⁺ ion concentration must reach super-saturation; CaSO₄ in solution promotes this effect.

CASE II: Low Gypsum/Plaster Ratio

Condition — Low gypsum/plaster ratio (high plaster content). If the finish mill temperature is too hot, the atmosphere too dry or the retention time too long, more gypsum will dehydrate to plaster than is desired. As a rule, high C₃A cements such as Type I will benefit from higher plaster contents, while Types II and V will perform better with more gypsum (or sometimes anhydrite) and less plaster.
5-10 minutes after hydration

Reactions — Plaster rehydrates and converts to secondary gypsum, which is tabular in shape. Large amounts of these tabular gypsum particles interlock. Since plaster goes into solution more quickly than gypsum, abundant calcium and sulfate ions are available to control aluminate reactions, therefore less C₃A is reacted.

Paste Properties — Interlocking tabular gypsum particles cause “false” set, whereupon brief mixing, the cement will appear to have a quick set. Extending the mixing cycle, however, will break up the tabular particles and revive paste plasticity, resulting in more normal setting properties. If a short mix cycle is used in the field and this setting problem occurs, water is often added in an attempt to improve concrete workability, which may result in reduced strength development and durability.

CASE III: Low Sulfate Content or Low Soluable Sulfate

Condition — Insufficient soluble sulfate to account for aluminate hydration. This condition may result from a low total sulfate/clinker aluminate ratio, high gypsum/plaster ratio, or high sulfate in the form of anhydrite (CaSO₄). Anhydrite is much harder than gypsum and has the slowest dissolution rate of the three sulfate forms, in part due to its large size resulting from its hardness.
5-10 minutes after hydration

Reactions — Aluminate starts hydrating. However, with sparse available sulfate ions available for reaction, voluminous hexagonal aluminate hydrates form instead of combining with sulfate to form ettringite.

Paste Properties — Interlocking aluminate hydrates fill up the water-filled space, causing stiffening and a quick set often called “flash set.” Similar reactions would occur if the total amount of sulfate was suitable, but insufficient sulfate ions were available in solution, as when the sulfate source is all anhydrite or coarse gypsum. This situation usually results in poor workability and strength development, possibly because the large hydration plates (mainly aluminate hydrate) may weaken paste microstructure.

CASE IV: Coarse Gypsum Particles

Condition — In most cases, gypsum will be ground finer than clinker, since it is easier to grind. Sometimes, however, clinker and gypsum enter the finish mill with the clinker already crushed. Gypsum will not have enough time in the mill to grind, and therefore, the cement is discharged from the mill with large gypsum particles. Furthermore, mills with high-energy classifiers will often experience high mill airflows with resulting short mill retention times, which may lead to under-grinding of gypsum particles.
5-10 minutes after hydration

Reactions — Large gypsum particles result in slow dissolution and poor distribution of sulfate ions, leading to localized reactions between aluminate and sulfate. Some clinker particles will not have ready access to sulfate ions, and as seen in the previous case, hexagonal aluminate hydrates will form.

Paste Properties — Interlocking voluminous aluminate hydrates will fill up the water-filled space, causing stiffening and flash set. Again, poor workability and strength development will likely result, possibly because the large hydration plates weaken the paste microstructure.

CASE V: Fly Ash with High C₃A Content

Condition — Cement sulfate content and form may be ideal for a pure cement paste, but the balance may be inappropriate in a concrete mix that contains fly ash with high aluminate content.
5-10 minutes after hydration

Reactions — The addition of another aluminate source disrupts proper balance of sulfate to aluminate. The reactions are similar to Case III, in that insufficient available sulfate ions are available to react with the abundance of aluminate. Hexagonal aluminate hydrates form instead of the aluminate combining with sulfate to form ettringite. The sulfate deficiency may also retard the hydration of the silicates and inhibit strength development.

Paste Properties — As in Case III, interlocking voluminous aluminate hydrates will cause stiffening and flash set; the concrete usually demonstrates poor strength development.

Linda Hills, Senior Scientist, and Fulvio Tang, Principal Scientist, both work for Construction Technology Laboratories, Inc., Skokie, Ill.^©2004 IEEE. Reprinted, with permission, from the 2004 IEEE/PCA Cement Industry Conference Technical Conference in Chattanooga, Tenn., April 26-29.

Part 2 of this report will discuss diagnosing reaction during the cement making process, how to optimize sulfate, and what the ideal sulfate is.