Compressive strength of steel slag stabilized mixes

The compressive strength testing, conducted on stabilisation mixes with steel-furnace slag originating from the Sisak Ironworks, and on ”standard” stabilisation mixes with crushed stone aggregate from ”Velika” Quarry, is presented in the paper. The testing was made to check whether slag can be used as aggregate in pavement base courses, and to determine in what way compressive strength values are influenced by binder content and grading of stabilisation mixes. Compressive strength results obtained during this testing comply with stringent criteria set in technical regulations, and confirm that slag can be used in stabilized base courses of pavement structures.

UDK 666.88+691.215:624.058Ivica Androjić, Sanja Dimter Compressive strenght of steel slag stabilized mixes The compressive strength testing, conducted on stabilisation mixes with steel-furnace slag originating from the Sisak Ironworks, and on "standard" stabilisation mixes with crushed stone aggregate from "Velika" Quarry, is presented in the paper.The testing was made to check whether slag can be used as aggregate in pavement base courses, and to determine in what way compressive strength values are influenced by binder content and grading of stabilisation mixes.Compressive strength results obtained during this testing comply with stringent criteria set in technical regulations, and confirm that slag can be used in stabilized base courses of pavement structures.

Introduction
Great quantities of natural materials are nowadays used in road construction.Such continuous demand for natural materials contributes to the depletion of natural resources, while in areas lacking in quality aggregates the cost of aggregate purchase and supply greatly increases the total cost of construction.Waste materials and industrial by-products can therefore be used as a good substitute to natural aggregate in road construction.These materials, requiring complex disposal and regarded as hazardous to natural environment, can prove to be economically and otherwise quite significant when used as replacement for or addition to standard materials.In the first place, the use of these materials contributes to a more rational utilization of natural aggregate reserves, and is furthermore quite favourable for curbing down environmental problems caused by disposal of waste materials.Slag is one of such materials: it occurs as by product during purification of metal, and during its melting and alloying, and has a significant potential in road construction.It has been used worldwide for many years now, and so numerous studies about its use and properties are now available.Slag is mostly used in asphalt mixes although its good properties have also made it indispensable in other layers of the pavement structure [1].As many as 200 to 240 million tons of blastfurnace slag, and 115 to 180 million tons of steel furnace slag, are produced every year in the world, while annual production in Europe varies around 12 million tons, out of which about 65% of steel-furnace slag is used as aggregate in road construction.
Research undertaken so far in the Republic of Croatia has concentrated on the use of slag as aggregate in concrete [2, 3] and, more recently, on the use of slag as aggregate in asphalt mixes [1].In this respect, test sections were made and slag was incorporated in unbound layers of pavement structures of approach roads, and also in bituminized base courses.Experience gained to this date on test sections and from laboratory testing results, confirm the good quality of this material and give encouragement for further testing work.
Due to heavy traffic load on our roads, pavement structures are in many cases realized with a stabilized base course.While increasing bearing capacity of pavement structures, such stabilisation layers are also a good base for pavement surfacing, and they also increase resistance of pavement structures to harmful frost action [4].Stabilisation mixes used in construction of pavement-structure base courses are mixes composed of granular stone materials (gravel, stone, sand) that are bound by cement.
As the price of stone mix (sand, gravel or chippings) contributes significantly -sometimes with more than 70% -in the price of stabilisation mixes, the idea behind this testing was to reduce the cost of stabilisation mixes by replacing the "standard" stone mix with the crushed steel slag mix 0/31.5 coming from Sisak Ironworks, as in this way we would obtain an economically more acceptable stabilisation mix compliant with current quality requirements [6].The steel slag originating from Sisak Ironworks is a byproduct occurring during production of steel, i.e. during separation of molten steel from impurities in steel manufacturing furnaces.These impurities are the carbon monoxide and silicon, manganese, phosphorus, and some iron in form of liquid oxide.When mixed with lime and dolomitic lime, these impurities form the steel-furnace slag or steel slag.About 150-200 kg of residues or byproducts are generated during production of one ton of steel.The slag stockpiled near the town of Sisak occupies an area of 25 hectares, and is of mixed composition: combined high furnace and electric arc furnace slag.The quantity currently deposited in this area is estimated at 1.5 million tons [5].

Experimental section
The objective of this analysis is to study possibilities for using slag in the stabilized base course of pavement structures, and to determine whether mechanical properties of the stabilized material can actually be realized.This particularly concerns compressive strength, as the possibilities of application are mostly dependent on this property.

Testing suitability of steel slag for use in stabilized base course
The suitability of materials for use in base courses stabilized by hydraulic binder is defined in General Technical Requirements for Road works (GTC), Volume III; Section 5-02 [6].The material to be incorporated in stabilized base courses must meet requirements set with respect to grain size distribution and physicomechanical properties of grains.Partial laboratory tests of slag were conducted in order to establish whether this material is suitable for use in base courses stabilized by hydraulic binder.The grain size distribution of slag is presented in Figure 1, while results obtained by testing physicomechanical properties of slag are given in Table 1.
It can be seen from curve established by analyzing grain size distribution of samples that the slag mix sample (0-31.5 mm) is situated in the grain size distribution range that is considered favourable for use in base courses.In addition to the grain size distribution curve, the following requirements were also checked: uniformity coefficient is U = 32.05(which is within the specified 15-50 range), maximum grain diameter is 31 mm (which is compliant with the specified maximum of 31.50 mm), content of voids smaller than 0.02 mm is 1.95 (which is less than the allowed 3% maximum).shows that slag is suitable for use in base courses stabilized with hydraulic binder (Table 1).
Compressive strength of steel slag stabilized mixes Following laboratory testing aimed at determining suitability of this material for use in base courses stabilized with hydraulic binder, it can be concluded that slug complies with required criteria as related to grain size distribution and physicomechanical properties, with the exception of water absorption where maximum water absorption is 1.6% (5.8% for slag), and that the material is therefore suitable of use in stabilisation mixes.

Properties of stabilisation mix components 2.2.1. Aggregate and binder properties
The objective of the planned testing was to compare compressive strength of stabilisation mixes containing various aggregates.
In one group, the aggregate was the slag originating from Sisak Ironworks (0/31.5),while in the second group the aggregate was the standard crushed aggregate mix from Velika Quarry (0/63) which is normally used in the construction of base courses for pavement structures.The grain size distribution of the

Grain diameter D �mm�
Ivica Androjić, Sanja Dimter Samples measuring 15.00 cm in diameter and 11.80 cm in height were prepared in the standard Proctor mould, with the modified Proctor compaction energy of E=2,66 MJ/m 3 .After preparation, samples were pressed out of the mould by hydraulic jack, and were placed on the plastic surface at the crushed stone mix from Velika Quarry is shown in Figure 2. The grain size distribution curve for the crushed stone mix shows that the grain size ranges from 0 to 63.0 mm.The uniformity coefficient amounts to U=17.06; the maximum measured grain is 70 mm; and the content of particles smaller than 0.02 mm is 0.53.All these results confirm that the stone material is suitable for use in base courses stabilized with hydraulic binder [6,8].
According to the crushed stone grading, this material contains a greater quantity of coarse grains.Lesser quantity of hydraulic binder is necessary as coarser stone mixes are characterized by spot binding of aggregate.Properties of slag used as aggregate are explained in detail in section 2.1.The grain size distribution of slag shows that the grain size is 0/31.5 mm, with considerable quantity of fine grains and with lack of coarser fraction, which is why greater quantities of cement are needed to ensure compliance with strength requirements.The steel slag is porous in structure and has greater water absorption capabilities, while its optimum moisture is by 3-5 percent greater when compared to the crushed stone mix.The cement SPECIJAL, CEM II/A-M (S-V) 42,5N was used to stabilize both mix groups.The cement "Specijal" contains no less than 80 percent of Portland cement clinker, up to 20% of the mixture of siliceous fly ash and high furnace slag (S), and up to 5% of secondary additive (pupil) and binder regulator (natural gypsum).

Preparation and cure of samples
The following five stabilisation mixes were prepared for the testing: 1. steel slag with 0.5% cement (by weight), 2. steel slag with 0.7% cement (by weight), 3. steel slag with 2.0% cement (by weight), 4. crushed stone mix with 0,5% cement (by weight), 5. crushed stone mix with 0,7% cement (by weight).The samples were cured for 3, 7 and 28 days, and were then subjected to compressive strength testing.

Compressive strength test results
The compressive strength of stabilisation mixes depends both on material properties and on curing conditions, and is defined as an average stress in the sample exposed to uniaxial compression at the force that causes failure (HRN U.B1.030 [9]).The sample destined for compressive strength testing is placed in the jack and the pressure is increased at constant rate until sample failure.The force at failure is then registered, and the compressive strength of stabilized mixes is then calculated using the following formula:    where: Results obtained by testing compressive strength of stabilisation mixes are presented in tabular form (Table 2) and in form of a diagram (Figures 4-6).
Compressive strength values for stabilisation mixes with slag were compared at 28 days with compressive strength values specified in General Technical Requirements for Road Works [6], i.e. with HRN U.E9.024 [8].These requirements are presented in Table 3. Ivica Androjić, Sanja Dimter

Analysis of testing results
The comparison of compressive strength results of stabilisation mixes under study shows that the mixes with slag containing 0.7 and 2.0 percent of cement (mixes 2 and 3) meet stringent criteria for base courses of motorways and roads with very heavy traffic load.The stabilisation mix with slag containing 0.5 percent of cement (mix 1) does not meet compressive strength requirements for both groups, heavy and medium traffic load, as the compressive strength of this group amounts to 2.04 MN/m 2 .The following conclusions were made after analysis of compressive strength results for steel slag mix and crushed stone mix, with equal percentage of cement and different moisture: Although the crushed stone mix 0/63 is of coarser composition, with significantly smaller quantity of finer grains, it is still within the range permitted by General Technical Requirements for Roads.When compared to slag mix, this stabilisation mix requires less hydraulic binder to achieve the required strength.
The grading of the steel slag is 0/31.5, and it has a considerable quantity of finer fractions.This mix is also within the range permitted by General Technical Requirements for Roads [6].
As the mix contains a great quantity of fie particles, a greater percentage of cement is needed to achieve the required strength.Considering the differences between the mixes with regard to grain size distribution at similar binder content, the stabilized slag mix has lower compressive strength results (Table 4).
It can be seen from diagrams presented in Figures 4 to 6 that compressive strengths of steel slag stabilisation mixes with 0.50 percent of cement are lower than the strength of crushed stone stabilisation mixes: 98.70% at 7 days and 72.50% at 28 days.A significant decrease in the difference between the compressive strengths can be noted at 28 days.For stabilisation mixes with 0.70% of cement, the difference in compressive strengths at 7 days is 67.29% while it falls to 30.49% at 28 days.
It can be concluded from these results that the difference in compressive strength decreases with an increase in the binder quantity in stabilisation mixes, and with longer sample curing times.
After the curing time of 3 days the compressive strengths amounted to 58.45-81.68% of the total compressive strengths realized at 28 days.The lowest increase in strength at 3 and 7 days with respect to strength at 28 days is registered for the steel slag mix (mix 2) with 0.70% of cement (50.45%), while the highest increase in compressive strength is registered for steel slag mix with 2.0% of cement (mix 3).The lowest increase in strength at 7 days with respect to strength at 3 days is registered for the steel slag stabilisation mix with 2.0% of cement (5.60%), while for other mixes this increase ranges from 13.02 to 16.25% with respect to strength registered at 7 days.
When compared to crushed stone mix, the steel slag is characterized by higher water absorption rate and a more porous structure, and it requires higher quantity of binder.At that, it needs a higher quantity of water to achieve an optimum moisture and maximum dry bulk density at a given compaction energy.Due to higher water requirement (3-4%) when compared to the crushed stone mix, and at equal binder content, the compressive strength of the steel slag stabilisation mix is lower.

Conclusion
Results obtained by testing compressive strength of steel slag stabilisation mixes from Sisak Ironworks, and crushed stone stabilisation mixes from the Velika Quarry, are described in the paper.The testing was made to check whether slag can be used as aggregate in pavement base courses, and to determine in what way compressive strength values of steel slag stabilisation mixes are influenced by binder content and moisture.
The results are compared with a "standard" crushed stone stabilisation mix at curing times of 3, 7 and 28 days, and the following conclusions were made: Based on laboratory testing made to determine suitability of material for use in base courses stabilised with hydraulic binder, it can be concluded that slag meets required criteria with respect to grain size distribution and physicomechanical properties, except for water absorption criterion, where the water absorption maximum is 1.6% (while 5.8% was registered for slag).However, additional testing revealed that the material is suitable for use in stabilisation mixes.
During preparation of stabilisation mixes, water content is considered to be highly responsible for increase in compressive strength.Due to difference in structure between the slag and crushed stone aggregate, the former requires greater quantity of water (3-4%) to achieve maximum density in compacted state, which is why compressive strength results for slag are lower.
During the compressive strength testing, much greater compressive strengths were registered for crushed stone stabilisation mixes when compared to steel slag stabilisation mixes, and this at similar binder content.
Steel slag stabilisation mixes with 0.7 and 2 percent of cement (mixes 2 and 3) meet stringent compressive strength requirements for traffic load categories according to [6 and 8].Compressive strength of steel slag stabilized mixes Steel slag stabilisation mixes with 0.5% of cement (mix 1) do not meet these compressive strength requirements.
Taking into consideration the above test results, it can be concluded that steel slag stabilisation mixes with 0.7 and 2 percent of cement can be used in construction of stabilised base courses in pavement structures for motorways and roads with very heavy traffic load, and for roads with heavy and medium traffic load.The use of steel slag in pavement structure courses would be acceptable from both economic and environmental standpoints: [1] [2] [3] [4] [5] [6] [7] [8] [9] great quantities of highly affordable (waste) material would thus be used and, at the same time, the quantity of slag deposited on stockpiles would be reduced.
The results described in this paper are a contribution to the study of possibilities for using waste materials in pavement structures.Further continuation of this research, and preparation of appropriate test sections, would enable us to gain a better insight into the properties and behaviour of steel slag stabilisation mixes.

Figure 1 .
Figure 1.Grain size distribution curve for slag

Figure 2 .
Figure 2. Grain size distribution diagram for crushed stone mix

Figure 3 .
Figure 3. Stabilisation mix samples (slag on the left, and crushed stone on the right)

Figure 4 .
Figure 4. Compressive strength of stabilisation mixes Figure 6.Compressive strength increase for stabilisation mixes under study Compressive strength ofstabilisation mixes MN/m 2Compressive strength ofstabilisation mixes MN/m 2Compressive strength ofstabilisation mixes MN/m 2