497 Main Road
Tasmania 7010 Australia
Phone: 61 3 62497868 (am)
Phone: 61 3 62713000 (pm)
Fax: 61 3 62730010

Printed in cyberspace on recycled electrons

Keeping you informed about the TecEco cement and kiln projects.  Issue 46, 19th April 2005

First Commercial Tec-Cement Concrete Slab a Success

On the 17th March 2005 we poured the first commercial slab in the world using Tec-Cement concrete with the assistance of one of the larger cement and pre-mix companies.

The house slab, a total of 80 cubic meters, was near Whittlesea in Victoria, Australia.

Our formulation strategy was to adjust a standard 20 MPa high fly ash (36%) mix from the company as a basis of comparison.
Given that there was only 180 Kg of PC per cubic meter, strength development, and in particular early strength development was good. Some 90 days later the slab is still gaining strength at the rate of about 5 MPa a month and is now well over 30MPa.

Also noticeable was the fact that the concrete was not as "sticky" as it normally is with a fly ash mix and that it did not bleed quite as much.

We did some shrinkage tests and were please to see that at 7 days, shrinkage was 133 micro strains, at 14 days, 240 micro strains, at 28 days, 316 micros strains and at 56 days, 470 microstrains - much less than normal. The slab was inspected a couple of weeks after it was poured and there were only a few very small micro cracks that were visible from a position of being on hands and knees. 80 cubic meters is a large slab for no crack control.

The curve for Tec-Cement strength development is quite different to that of ordinary Portland cement concretes as can be seen from the plot above. Noticeable is the high early strength gain even with added fly ash and the straight line development thereafter which from our work so far appears to continue for a considerable period.

Pouring at 100+ Slump


Slab Drying Off After Bull Floating


Pozzolan required for the Pozzolanic Reaction

I am often asked how much fly ash or other pozzolan should be used in a Tec-Cement concrete. Having never felt comfortable with a number somebody else has given me unless I can work it out from first principles I decided to do so for the local brand of cement.

The pozzolanic reaction is similar to the hydration reaction for di and tri calcium silicate; the difference is that the components all come together to form compounds that are hydrated silicates with less calcium and bound water.

1.1Ca(OH)2 + SiO2 + .67H2O ==> Ca1.1SiO3.1.2.1H2O
1.1 X 74.08 (Ca(OH)2) + 60.09 (SiO2) + .67X34 (H2O) - molar masses
81.488 (Ca(OH)2) + 60.09 (SiO2) + 22.78 (H2O) - molar masses

The mass ratio of calcium hydroxide to silica is 1 to .737

Atoms don’t come in fractions so the formula suggest that larger molecules are formed that are probably oligopolymers – and with the pozzolanic reaction they are.

The lime that reacts in the pozzolanic reaction is generally supplied entirely from the hydration reactions of belite and alite in Portland cement. Alite produces three times as much lime as belite on hydration on a molar basis and on a 2.27 times as much on a mass basis.

Modern cements have less and less belite or di calcium silicate. For example the local plant where I live in Tasmania, run by Australian cement, produces cement of the following approximate composition.

Ca3SiO3 - 68%, Ca2SiO4 - 6%, 3CaO.Al2O3 - 6%, 4CaO.Al2O3.Fe2O3 - 10%, CaSO4 - 3.6%, CaCO3 - 5.4%, MgO - 2%, CaO <1%, Na2SO4, K2SO4 <.6%

The belite content of cements around the world is falling because people are demanding faster setting mixes and alite hydrates much more quickly than belite. As a consequence, if the objective in a Tec-Cement concrete is to remove lime, more silica is required to react with the higher proportion of alite in the pozzolanic reaction today than say a decade ago.

Total Belite Alite Others Lime Produced Belite Lime Produced Alite Total Lime Silica Required for Reaction
100 6 68 26 1.29 33.184 34.474 25.407
100 10 65 25 2.15 31.72 33.87 24.962
100 15 60 25 3.225 29.28 32.505 23.956
100 20 55 25 4.3 26.84 31.14 22.950

Often not considered is that fact that fly ash is never pure silica and not all of it will react in the pozzolanic reaction. Unfortunately fly ashes vary and it is impossible to give an average composition. Silica is however usually only 50 - 70% of the total.

Some of the other compounds in fly ash will however participate in other silicification reactions such as surface hydrolysis and bonding.

As the theory behind TecEco cements is to consume all the lime it is recommended that excess fly ash is added. 30% may be a good figure to aim for.

Water Consumption by Magnesia

One characteristic we are very interested in with MgO is the extent to which, when it reacts with water, it actually forms a "gel" containing hydrated complexes – as by doing so it would use a lot more water than calculated in a simple hydration reaction from magnesia to Brucite.

Magnesium Complexes

Both magnesium and calcium show an ability to form complexes with water.
Magnesium as a hydroxide or carbonate appears to have a very strong affinity for water. In solution Mg++ complexes with water more readily than Ca++ forming ions of the general form [Mg(H2O)N]2+. Mg++ can also hydroxylate forming H3O+ and Mg+OH and hydrated forms of Mg+OH.

According to Skalmowski 1#, in the first stage of magnesium oxide hydration Brucite is not formed; instead meta stable magnesium hydroxide [Mg(OH)2•nH20] is formed and only after some time does it re crystallizes to Brucite which is the stable.

Investigations done in Japan proved that the structure of meta stable magnesium hydroxide is crystalline. It differs from Brucite in that it holds mono molecular layers of water between neighbouring Mg(OH)2 packets. Intra packet water is removed during re-crystallisation and Brucite is formed as hexagonal lamellae. Increasing the temperature increases the rate of transformation. According to Skalmowski the properties of meta stable Mg(OH)2 are different from Brucite. It is four times more soluble in water, but still less soluble than Portlandite.

We suspect that the hydrate complexes form what amounts to a meta stable “gel” holding water for slow release right through the matrix of the concrete resulting in more complete hydration of PC. In ordinary PC concrete up to 15 - 20% may remain un hydrated. The hydration of more of this fraction would add to efficiency of PC as a binder and result in greater strength.

1Skalmowski, W, Chemistry of Building Materials, Warszawa, Budownictwo I Architectura

Water Consumed in Tec-Cements without taking Complex Formation into Consideration

Unfortunately as scientists without much money we only easily see the end results. Furthermore we would have to make a lot of guesses as to the extent of complex formation.

Without taking the as yet unknown amount of water in hydrated magnesium hydroxide gels into account and using the cement produced in Tasmania by Australian cement of the following composition:

Ca3SiO3 - 68%, Ca2SiO4 - 6%, 3CaO.Al2O3 - 6%, 4CaO.Al2O3.Fe2O3 - 10%, CaSO4 - 3.6%, CaCO3 - 5.4%, MgO - 2%, CaO <1%, Na2SO4, K2SO4 <.6%

The hydration stoichiometry is as follows:

MgO + H2O ==> Mg(OH)2
40.31 (MgO) + 18 (H2O) ==> 58.31 Mg(OH)2 – molar mass
11.20 (MgO) + 18 (H2O) ==> 24.30 Mg(OH)2 – molar volume

The mass ratio of magnesia to water (taken as 1) is 2.23 to 1
The volume ratio is a minimum of .62 to 1

The stoichiometric water demand of the mix with 10% added magnesia using Tasmanian cement rises to 28% (an probably a lot more due to complex formation) hence our claim that magnesia as it hydrates consumes water. By calculation a minimum of 7.6 % more water is consumed internally although this could be a lot more because of hydroxide hydrate formation (see above). As magnesia does not hydrate all that quickly in at least the first half hour or so this water consumption is a potentially a very useful property. There are a number of potential consequences.

• Bleeding and the introduction of associated problems such as efflorescence due to lime, freezing of bleed water and weaknesses such as interconnected pore structures and high permeability do not appear to occur as much. Denser concretes without interconnected pour structures are more durable.

• Tec-cements concretes tend to dry out from the inside due to the water demand of magnesia as it hydrates and combined with a lower long term pH, density and the low solubility and reactivity of Brucite, improved durability results. Brucite protects against sulfates, chlorides and other aggressive salts and delayed reactions do not occur.

• Drying shrinkage does not occur as the water consumed is converted expansively to solidus.

• There is no loss of alkalis in bleed water and the early pH is forced up by water removal. A higher pH during the early plastic stage results in better pozzolanic reactions

The advantages of using Portland cement such as ambient temperature setting, easy placement and strength are not diminished.