FINAL TECHNICAL REPORT

September 1, 1995, through August 31, 1996

Project Title: EXTRUDED HONEYCOMB CONSTRUCTION MATERIALS FROM FLY ASH

ICCI Project Number: 95-1/3.1A-16

Principal Investigator: Gerald P. Wirtz, Dept. of Materials Science and Engineering, University of Illinois at Urbana-Champaign

Other Investigators: John M. Bukowski, Center for Cement Composite Materials, University of Illinois at Urbana-Champaign

Project Manager: Daniel D. Banerjee, ICCI

ABSTRACT

The objective of the project was to utilize fly ash, produced by burning Illinois basin coal, to produce high strength, lightweight building materials. The desire was to produce a profitable, high value added, environmentally responsible application for large quantities of Illinois fly ash. The approach was to produce thin-walled honeycomb structures by plasticizing the fly ash with the addition of ~20% water and small amounts of water soluble polymer or inorganic plasticizer, and extruding the resultant material through an appropriate die to form the honeycomb. The maximum strength to weight ratio would be obtained by total removal of pores from the solid. For a given strength to density ratio, any density can, in principle, be obtained by varying the ratio of the thickness of the solid walls to the width of the open channels in the honeycomb structure. The extruded honeycomb was solidified or cured by hydration in high pressure steam (autoclaving). Processing additives included Portland cement, lime, and/or silica fume to promote hydrothermal reaction. Based on previous work, a composition of 60% by weight of Class F fly ash and 40% by weight of Type I Portland Cement formed the base composition. Modifications consisted of substituting silica fume for a portion of the fly ash and/or lime or high calcia FBC ash for a portion of the Portland Cement. Most of the mechanical testing was done on one inch square honeycombs with 1/16 inch nominal wall thickness.

Compressive strengths of fly ash based honeycombs have been consistently shown to depend exponentially on the density of the honeycomb walls. Three approaches to decreasing the porosity of the honeycomb walls were pursued in this year's work: a.) decreasing the water content necessary for plasticizing the composition through high shear mixing on the two-roll mill, thereby decreasing the potential pore volume on reacting the water; b.) the addition of fine particle silica fume to improve particle packing in the green state, filling the interstices between fly ash and cement particles; c.) replacing the organic plasticizer with inorganic materials which could be incorporated into the final body, obviating the need for removal of the water soluble polymer and the resultant porosity left behind, and; d.) carbonation of the hydrated body to fill the pores with carbonation reaction product. All of these approaches met with varying degrees of success. The increase in density with the addition of silica fume was particularly striking. Porosities of less that 7 vol. % were attained utilizing silica fume, but little concomitant increase in strength was observed. Scanning electron microscopy of fracture surfaces indicated that the fracture mechanism went from transgranular to intergranular with the addition of more than a few percent silica fume. The available lime in the Portland cement was apparently consumed by the added silica fume, thus decreasing the interparticle bonding and weakening the final structure.

EXECUTIVE SUMMARY

Objectives:

The ultimate goal of this research effort was to utilize ash from coal fired power plants in Illinois to produce strong, lightweight construction components by the extrusion of honeycomb structures from plasticized powders of ash and other additives. This work was restricted to "fly" ash as opposed to bottom ash or other forms of combustion residues. The idea was to utilize those combustion by-products which require a minimum of preliminary processing, such as grinding or ball-milling, to produce a suitable starting material.

Ash characterization:

The Illinois Power (IP) generating station at Wood River Illinois burns pulverized coal (PC) in a conventional burner. The PC combustion results in a typical class F (low calcia) fly ash. Table 1 shows the average chemical analyses, converted to equivalent oxide fractions, of various ashes, as determined by x-ray fluorescence. The column labeled IP(nom.) is the nominal composition of the ash quoted by Illinois Power. Columns labeled IP1 and IP2 are analyses performed as part of this study on the composition of two different batches of ash from the Illinois Power station at Wood River. The column labeled ADM gives the analyses of the ash from the ADM fluidized bed combustor in Decatur, Illinois. Also included in Table 1 is the composition of a typical Class C fly ash, from the White Bluff Steam Electric Station in Arkansas, as measured in this laboratory in a previous project.

Table 1. Chemical analysis of "as received" fly ash

Oxide	IP(nom.)	IP1	IP2	ADM	Class C
Silicon Dioxide, SiO2	57.90	45.68	50.89	17.62	32.99
Aluminum Oxide, Al2O3	29.68	26.31	24.66	5.65	19.10
Iron Oxide, Fe2O3	4.87	7.54	11.89	5.60	8.01
Calcium Oxide, CaO	1.75	7.89	1.92	36.69	25.74
Magnesium Oxide, MgO	0.81	2.07	1.04	13.89	5.04
Potassium Oxide, K2O	1.75	0.93	2.40	0.91	0.46
Sodium Oxide, Na2O	0.65	1.45	2.31	0.50	2.08
Titanium Dioxide, TiO2	1.67	1.43	1.42	0.31	1.47
Phosphorus Pentoxide, P2O5	0.29	1.66	0.28	0.14	1.16
Manganese Oxide, MnO	0.07	0.02	0.02	0.069	0.02
Strontium Oxide, SrO	0.23	0.27		0.04	0.28
Barium Oxide, BaO	0.25	0.49		0.04	0.73
Sulfur Trioxide, SO3	0.08	0.66	0.34	18.55	2.13
Loss on Ignition (1000C)		3.34	2.34		0.20
Total	100.00	99.74	99.51		99.41
H2O (110C)		<0.1	0.06		0.08

Note: weight % ignited basis

The major compositional difference between the two lots of IP ash in Table 1 was the factor of 4 difference in the calcia content. The other notable difference is that there was 50% more iron in IP2 than in IP1. The iron content is notably high in both. The acidity of the two IP batches of fly ash differed markedly. The initial pH of batch IP1 with 100 g of fly ash added to 100 ml of deionized water was 12.26 and relatively constant out to 7 days. The corresponding initial pH of batch IP2 was as low as 3.32 and increased to 8.41 after 7 days in the deionized water. For comparison, the ADM fly ash exhibited a pH very close to that of the IP1 fly ash. The class C fly ash from the White Bluff steam electric station also exhibited a pH of nearly 12, virtually constant over several days.

Forming processes:

Since a major part of the load which a building must support is its own weight, the reduction in weight of the building material would be reflected in less massive, less expensive construction. Also, since a major part of the cost of construction is labor, reducing the weight of the material of construction would reduce the labor of construction and another major savings would be realized. Decreasing the density of a ceramic component necessarily involves incorporating open volume into the body. This might be accomplished by increasing the porosity of the ceramic body. The problem is that ceramic materials invariably exhibit an exponential decrease in strength with porosity when the porosity is randomly distributed throughout the material. The extrusion of a thin-walled ceramic honeycomb, similar to the exhaust catalyst carrier in the automotive industry, provides an ideal structure for high strength lightweight materials with superior thermal and acoustic insulation values. The density of the building component will depend upon the relative widths of solid walls and open channels in the honeycomb structure, which are controlled by the die design Ð the thinner the walls, the lighter the weight. The strength of the honeycomb will scale linearly with this bulk density, but will depend upon maximizing the density of the honeycomb walls. The lower limit of the honeycomb wall thickness which can be designed into the die is determined by the maximum particle size in the powder to be extruded.

Plasticizers:

The plasticizers and/or binders added to the starting composition imparted the plasticity and green strength necessary to extrude complex shapes from fly ash powders. The plasticizer consisted of 1-5% (by weight) of water soluble polymer such as methyl cellulose, or up to 2.5% by weight of bentonite (the mineral montmorillonite), added to the powder. The polymers were also observed to act as water reducers and retardants in fly ash compositions containing Portland Cement. Plasticizer compositions based on hydroxyethyl cellulose (HEC) were clearly less retarding to the hydration reaction than earlier methyl cellulose based composition. Extrusion of thinner walled honeycombs was more difficult with the HEC compositions, however, because of the lower stiffness of the green body. Fly ash-Portland cement compositions were readily extruded using bentonite, but the thinner walled configuration was not attempted.

Curing Processes:

Ordinary Type I Portland Cement (OPC) was added to fly ash compositions, typically in amounts 40% to promote cementitious or pozzolanic bonding during hydrothermal processing in an autoclave. Honeycombs containing 40% Portland Cement were autoclaved at 180C for times of 1 to 24 hours. Prehydration of honeycombs to be hydrothermally processed in the autoclave made them easier to handle without fracture. It was also shown to enhance the subsequent crushing strength achieved upon autoclaving. Prehydration of 24 hours in a moist atmosphere at 60C before autoclaving was ultimately chosen as part of the standard curing process.

Performance Results:

Table 2 reiterates the comparison previously presented of compressive strength and density measured on extruded honeycombs to values obtained for typical construction materials. The materials are listed in order of increasing strength to density ratio, normalized to unity for construction grade pine lumber.

Table 2. Strength/Density Comparison with Typical Construction Materials

Material	Bulk Density (gm/cc)	Compressive Strength (MPa) (psi)	Normalized Strength/Density Ratio
Aerated Concrete	0.69	4.4 640	0.10
Concrete	2.2	35 5100	0.25
Clay Brick	2.3	45 6500	0.31
Steel	7.8	455 66000	0.91
Pine	0.48	30 4350	1.00
Autoclaved Honeycomb	1.13	95 14000	1.36
Fired Honeycomb	1.32	155 22500	1.89

The autoclaved honeycomb in Table 2 was composed of 60% by weight IP2 fly ash and 40% OPC. It was prehydrated for 24 hours at 60C and autoclaved for 12 hours at 180C. Change of dimension on curing was negligible for the autoclaved honeycomb. Wall porosities, determined from mercury intrusion porosimetry, was about 25% by volume. For a given wall density, the strength would vary linearly with the fractional cross sectional area of the walls, as would the bulk density. To a first approximation, therefore, the strength to density ratio would be independent of bulk density. In principle, any desired bulk density could be achieved while maintaining the same strength to density ratio by simply varying the die design. An autoclaved honeycomb with the same density as pine (0.48 gm/cc) would therefore be expected to have a 36% greater compressive strength than pine (5900 psi) from Table 2. By increasing the wall density, even more impressive strength to weight gains should be achieved.

Densification Results:

Three approaches to decreasing the porosity of the honeycomb walls were pursued: a.) high shear mixing on a twin-roll mill to decrease the water content necessary to plasticize the composition, thereby decreasing the potential pore volume on removing the water; b.) the addition of silica fume to improve particle packing in the green state, filling the interstices between fly ash and cement particles; c.) replacing the organic plasticizer with inorganic materials which could be incorporated into the final body, obviating the need for removal of the water soluble polymer and the resultant porosity left behind, and; d.) carbonation of the hydrated body to fill the pores with carbonation reaction product. Table 3 shows typical results of each of these initiatives. The standard batch corresponds to the composition and processing of the autoclaved sample in Table 2, plasticized with 5% by weight of the standard combo organic binder. The high shear mixing results are for the same composition with the water requirement reduced from 35 to 28% by volume through the high shear mixing operation. The carbonated sample corresponded to the standard batch subjected to 500 psi CO₂ pressure for 24 hours. The silica fume addition consisted of replacing 10% of the fly ash with silica fume combined with the high shear mixing. In the bentonite and silica fume sample the organic binder in the silica fume addition sample was replaced by 2 weight % bentonite.

Table 3. Density Increases Achieved with Various Process Modifications and Additives.

Process Modifications and Additives	Porosity (Volume %)
Standard Batch	24
High Shear Mixing	18
Carbonation	14
Silica Fume Addition	7
Bentonite and Silica Fume	4

The increase in strength predicted by the marked decrease in porosity of the final three samples has not yet been realized. The carbonated samples formed microcracks on carbonation, drastically reducing the observed strength. The dense samples formed by silica fume additions fractured with an intergranular fracture path, indicating that the bonding between fly ash particles was degraded, presumably through competition of the silica for the pozzolanic lime in the Portland cement. Honeycombs plasticized with bentonite also showed extensive cracking after autoclaving. The latter samples were not cracked after prehydration at 60C in moist air, suggesting that process modifications might alleviate this cracking problem.