While it has been demonstrated that high-strength sulfur concrete can be prepared, several detrimental properties have appeared. Sulfur on solidifying undergoes an allotropic transformation from the monoclinic to the orthorhombic form which is more dense and occupies less volume. This results in a concrete that is highly stressed. Both Malhotra (18) and Dale and Ludwig (11) have commented on the low freeze-thaw resistance of sulfur mortars. Temperature variations have caused some sulfur concretes to self-destruct and they are vulnerable to melting at fairly low temperatures and require additives to prevent flammability problems.

The research reported in this paper was aimed at the development of a modified sulfur binder that prevented formation of a highly stressed product and at determining the physical properties of the modified sulfur concrete. The investigation was limited to the use of relatively low-cost modifiers that are commercially available in large quantities. Under these criteria, the olefin polysulfide modifiers used by Duecker would probably be too expensive.

ACKNOWLEDGMENTS

The assistance of Professor Bob Galloway and Dr. Donald Saylak of the Texas Transportation Institute in the freeze-thaw testing of sulfur concretes; metallographic examination and modulus of elasticity determination by Peter Romans and Ronald Lowery of the Albany Metallurgy Research Center, Bureau of Mines; and mineralogical examination of aggregate materials by Howard Heady of the Reno Metallurgy Research Center, Bureau of Mines, are gratefully acknowledged.

EXPERIMENTAL PROCEDURE

Preliminary Tests

Sulfur concrete was first investigated for its suitability in constructing such things as acid leach tanks and containment ponds for waste acid and salt solutions. Previous researchers had reported that sulfur concrete had strength characteristics that should be suitable for such uses (4, 11).

Preliminary tests were made to determine workable mixtures of sulfur with various aggregate materials. The aggregates used were construction sand, silica sand, 1/4- to 3/8-inch volcanic rock, and waste copper mill tailings. The sulfur concrete was made in a small, bucket concrete mixer by heating the mixing bucket and aggregate to 160° C, adding liquid sulfur at 150° C and mixing for 2 minutes. The mixture was poured and tamped into a heated metal mold (135° C) with a demountable center mold. This produced 7-1/2- by 15-1/2- by 7-3/4-inch sulfur concrete boxes with 1-1/2-inch thick bottoms and walls. On cooling, the molds were removed and the boxes were aged for at least 24 hours.

It was found that initially sound boxes of various mixes all developed cracks when filled with water. Figure 1 shows some of the test boxes. Some cracked immediately while others took up to a week to develop cracks large enough for the water to leak through.

These

In the development of sulfur coatings the authors (20) found that two unsaturated hydrocarbons, dicyclopentadiene (DC PD) and dipentene (DP), were useful in modifying sulfur to prevent cracking of the coatings. chemicals are available in commercial quantities at relatively low cost (8 to 12 cents per pound). It was found that reacting the sulfur with 2 or 5 pct dicyclopentadiene for 1 hour before mixing with the aggregate produced sulfur concrete boxes that did not crack when filled with water; however, the use of dipentene for the same purpose was not successful. A summary of the composition and test results is given in table 1. The unmodified boxes all failed while those made with modified sulfur have contained water for 2 years without any signs of cracking.

WO

ere

The failure of the unmodified sulfur concrete boxes was believed to be caused by stresses developed during the allotropic transformation from the monoclinic to the orthorhombic form of sulfur in cooling. In this regard, Dale reported that sulfur has a 13 pct shrinkage from the liquid to the solid state, but that the sulfur concretes shrink only 1 pct on solidification (11). Monoclinic sulfur has a density of 1.96 g/ml at 20° C and an atomic volume of 16.4 ml while the orthorhombic form has a density of 2.07 g/ml at 20° C and an atomic volume of 15.5 ml. Conversion of the sulfur from the monoclinic to the orthorhombic form on cooling and aging induces stresses because the physical nature of the sulfur concrete prevents it from shrinking.

With the successful demonstration of the use of modified sulfur to prevent stress cracking of the sulfur concrete, the testing program was extended to determine the optimum conditions for preparing modified sulfur concretes and to determine the effect that the modified sulfur had on the physical properties of the material as compared with Portland cement concrete and unmodified sulfur concrete.

Materials

The materials used in preparing the sulfur concretes were sulfur, unsaturated hydrocarbons for modifiers, and fine and coarse aggregates. The sulfur was commercial high-purity grade (99.9 min) secondary sulfur. It was in flake form which made it convenient for handling without dusting.

The chemical modifiers were commercial grades of dicyclopentadiene (CH2)--a colorless liquid with a specific gravity of 0.976, a boiling point of 200° C, and a melting point of 33.6° C, and dipentene (C1 H16)--a colorless liquid with a specific gravity of 0.860, a boiling point of 176° C, and a melting point of -40° C. Both of these hydrocarbons have unsaturated double bonds suitable for direct reaction with sulfur. In a previous investigation involving sulfur spray coatings, it was found that reaction of sulfur with 13 pet dicyclopentadiene or 26 pct dipentene would result in a fully plasticized sulfur product (20). Smaller additions of dicyclopentadiene were found to retard or prevent crystallization of the sulfur, while providing comparable physical properties. In this investigation, dicyclopentadiene was chosen as a plasticization agent in the modification of sulfur and the dipentene was used in conjunction with dicyclopentadiene as a viscosity control agent. Small additions (1 pct) of dipentene to sulfur were found to retard the formation of viscous sulfur in the 160° to 200° C range and also improved the ability of the sulfur to wet the aggregate.

Four types of aggregate materials were investigated.

These were desert

blow sand, commercial construction sand, volcanic rock, and limestone.

The fine desert blow sand consisted of subrounded to subangular grains 0.2 to 0.5 mm in diameter. Most grains consisted of one mineral with about equal amounts of plagioclase, feldspar, quartz, and microline feldspar. About 25 pct of the desert sand consisted of very fine-grained lithic fragments of granitic rock, cryptocrystalline silicified rocks, limestone, and volcanic rocks. The construction sand was 1- to 3-mm angular grains of medium to

fine-grained granitic rock. The volcanic rock was sized 3/8-inch rock of intermediate to basic intermediate composition and fell on either side of the andesite-basalt composition boundary and was quite vesicular as shown by its lower specific gravity. The 3/8-inch limestone was principally calcite with minor amounts of dolomite, quartz, and feldspar. The physical properties of the aggregates are shown in table 2.

A standard procedure was developed for preparing test specimens for strength measurements. This was necessary to insure sound samples for comparing the properties of various sulfur-concrete mixtures. Sulfur concrete cast into standard concrete compression molds develops a porous shrinkage area in the top of samples even with good tamping and refilling of the mold as it cools. Malhotra has reported on the differences he obtained in compressive strengths with 4- by 8-inch and 6- by 12-inch samples of sulfur concrete (19). The following procedure was used to obtain sound samples and minimize the difference between samples.

Standard cylindrical compression molds 3 by 6 and 6 by 12 inches long were fitted with 3-inch extensions on the top of the molds. Both modified and unmodified sulfur concretes were prepared by mixing together sulfur (140°-150° C) and aggregate (160°-170° C) in a heated mixing bowl (150° C) for 2 minutes with a Hobart or Triumph mixer." When modified sulfur concrete was prepared, the sulfur was first reacted with the modifier before mixing with the aggregate. Sulfur and dicyclopentadiene were reacted for 2 hours at 140° C. Sulfur and dipentene were reacted for 2 hours at 170° C. Sulfur and a mixture of the two were made by reacting with dipentene first for one-half hour at 170° C and then adding the dicyclopentadiene and reacting for 2 hours at 150° C. The The sulfur concrete after mixing was tamped into the molds. molds and tamper were preheated to 120° C. The tamper was a 3/4-inch steel rod with a hemispherical tip. After tamping, the samples were cooled to room temperature and removed from the molds. The top 3 inches of the samples were cut off and discarded. The compression samples were capped with a standard sulfur capping compound before testing. The results of all test measurements reported are for samples aged 1 day unless otherwise specified.

Sulfur concrete beam samples were prepared in a similar manner.

Beams

3 by 3 by 14 and 3 by 4 by 16 inches were cast in vertical molds with tamping.

Reference to specific brands is made for identification only and does not imply endorsement of the Bureau of Mines.

All molds and tampers for beam samples were preheated to 150° c. The molds were 3 inches longer than the desired length and the top 3 inches were cut off before testing. Larger beam samples, 6 by 6 by 30 inches, were cast into horizontal molds with tamping. Some of the larger beams were cast in one pouring and others in four lifts allowing each lift to solidify before adding the next lfit. On the large beam samples additional material was added to the top of the beam surface as shrinkage occurred to obtain a level surface.

The purpose of preheating the sample molds and tamping rods was to prevent segregation of sulfur in the test specimens. When sulfur concrete is poured into a cold steel mold, the mold acts as a heat sink and freezes the sulfur at the surface of the specimens. This action tends to draw sulfur from the core of the sample. The use of heated molds and extensions on the molds resulted in sound test specimens. The various molds used in preparing the test samples are shown in figure 2.

Sulfur Concrete Mixture Design

A workable mixture of sulfur with fine and coarse aggregates is desired in preparing sulfur concretes that have good strength properties. A workable mixture may be described as one that is fluid enough to pour and tamp easily into forms or molds without separation of the sulfur on solidification of the concrete. The strength of sulfur concrete is principally dependent on the strength of the aggregates used, the strength of the sulfur, and the bonding developed between the two on the thermosetting of the sulfur. It would seem feasible that the highest compressive strength sulfur concretes would be obtained by completely filling the voids in the large aggregate with a mixture of sulfur and fine aggregate.

No standard methods of mix design have been developed for sulfur concrete. The most extensive work on characterizing sulfur concrete mixtures is that of Dale and Ludwig (11). They investigated the blending of standard concrete aggregates with fine aggregate and determined the amount of sulfur needed to obtain workable mixes that gave products with acceptable strengths. They also investigated how the type of aggregate used affected strength characteristics and developed a maximum density gradation to obtain highest comprehensive strength. Crow and Bates (4), in their investigation using basalt aggregate to prepare high compressive strength sulfur concretes, utilized the average grain size of the mixture and the coefficient of uniformity in determining their best mixtures. Malhotra used variations of standard aggregate blends in determining his sulfur concrete mixtures and also made use of added silica flour as a workability aid in holding the sulfur in suspension in the mix (18).

In this investigation, 10 aggregate blends were used and the mixture of each blend with the optimum amounts of both sulfur and modified sulfur was found. This was based on the compressive strength of the workable mixtures containing varying amounts of sulfur mixed with the aggregates. The void content of each aggregate blend was determined and used as a guide for the amounts of sulfur needed for each blend to obtain maximum compressive strength