Uses of Waste Foundry Sands in Ci vii Engineering

Size: px
Start display at page:

Download "Uses of Waste Foundry Sands in Ci vii Engineering"

Transcription

1 TRANSPORTATON RESEARCH RECORD Uses of Waste Fondry Sands in Ci vii Engineering SAYEED JAVED AND C. W. LOVELL Molds and cores for metal castings are normally sands with chemical or clay binders and other additives. After the casting is complete, the sand is disaggregated, and it is jdged whether it is sitable for rese. However, after several ses, the sand is no longer sitable and is designated as a waste fondry sand (WFS). Sch waste has been disposed in landfills, pblic and private. Althogh some WFSs may contain excessive amonts of harmfl heavy metals, those discarded after ferros castings contain only iron, and this concentration is controlled by magnetic extraction from the waste. At this point, only WFSs from ferros castings are thoght to be environmentally acceptable. Throgh laboratory stdy of the mechanical and physical properties of WFSs from the greensand, shell, and chemically bonded processes for ferros castings, a nmber of civil engineering ses have been identified: embankment fill, sbgrade, flowable fill, and fine aggregate replacement in asphaltic concrete. The index properties of the WFSs and the appropriate test parameters for its varios acceptable ses are described. The next step is to bild demonstration and implementation projects to prove the practicality of the se and to provide long-term evidence of the absence of significant environmental effects. Recent legislation in ndiana has intensified stdy of the se of indstrial and domestic waste prodcts in civil engineering, particlarly in highways. The wastes given primary attention are scrap tires, coal combstion byprodcts, destrcted pavement materials and bilding demolition prodcts, and spent fondry sands. This paper smmarizes properties and potential ses of spent (waste) fondry sand. The desire of ndiana fondries to redce disposal costs led to the sponsorship of a 2-year stdy of waste fondry sand (WFS) in the School of Civil Engineering at Prde University. The athors of this paper wrote the research report reslting from the stdy (J). The potential ses given greatest attention were WFS as embankment and sbgrade material, fine aggregate in controlled low strength material (CLSM), (also called flowable fill), and fine aggregate in asphaltic concrete (3). BACKGROUND Sands have long been sed for metal casting. They are chosen for several important reasons (3): they are readily available, inexpensive, highly refractory, and readily bonded by clays or other inorganic and organic material. There are three types of molding processes: greensand, chemically bonded, and shell molded. n ndiana, the most commonly sed process is greensand molding. The greensand mixtre contains clay, combstible additives like seacoal, and water. The predomi- S. Javed, Geotest Engineering, nc., 5600 Bintliff Drive, Hoston, Tex C.W. Lovell, School of Civil Engineering, Prde University, West Lafayette, nd nant metal cast is gray iron. After a nmber of ses, the molding sands lose the desired qalities and are discarded. The fondry may operate a monofill for these wastes or may transport them to a mnicipal solid waste landfill. n either case, the disposal costs are significant, and the fondries are anxios to redce them. The grain size distribtion of WFS is niform, with most sizes within a narrow range between the No. 50 and No. 100 sieves. This means that they are finer than the materials generally specified for fine aggregates. The greensands, with their additives of clay, demonstrate cohesive type behavior and the chemically bonded and shell-molded sands show cohesionless response. Conseqently, greensand compaction crves show a definite dependence on compaction moistre content. The compacted nit weights of others is almost independent of that moistre content. The reslts of standard heavy metal leaching tests, like the EPTOX and the Toxicity Characteristics Leaching Procedre (TCLP), depend on the metals cast becase the WFSs are commonly contaminated by these metals. Mch of the metal cast in ndiana is ferros, and althogh limits are seldom exceeded, iron is occasionally present in concentrations above those stated. Chemically bonded and shell-molded WFSs contain chemicals and organics that may reqire special environmental assessment before rotine se. CHARACTERZATON OF WFS Experimental work on WFSs in the School of Civil Engineering, Prde University, involved seven samples from greensand processes, two from chemically bonded processes, and one from a shell molding process. All of these involved ferros castings. Compactability of these sands is of considerable interest for embankment and sbgrade applications. The WFSs from chemically bonded and shell-molded processes are cohesionless and respond best to vibratory compaction. The waste. greens ands are compacted best by impact (Pro.ctor) type processes: Figre 1 shows the moistre-density and moistre-california Bearing Ratio (CBR) relations for the raw sand (Rl) that comprises the casting sands. As is common for sch materials, density varies little with compaction water content, althogh it is somewhat higher for a flshed condition. The soaked CBR is also maximized at this higher water content. The WFSs from greensand casting for sample G 1 are illstrated in Figre 2. Unit weight trongly depends on moistre content, and soaked CBR is maximized at abot the optimm moistre content. Note that the swell pon soaking is small arid that the soaked CBR vales compare favorably with compacted natral soils. Shear strength parameters were determined by performing direct shear tests on dry samples at varios densities. The data are sm-

2 110 TRANSPORTATON RESEARCH RECORD >-- ry CD 8-4- As Compacted CSR Soaked CSR , E... O' 1.65 "iii c <U 0 >-. a io Water Content, % FGURE 1 Moistre density-cbr relationship of Rl. marized in Table 1, along with typical vales for natral sands reported in works by Terzaghi and Peck ( 4) and Peck et al. (5). The symbol G represents greensand, C chemically bonded,. S shell molded, and R raw sand. The <P vales represent peak strengths for dense samples and ltimate strengths for loose samples: (Dr) is relative density. To determine the sitability ofwfss for sbgrade,resilient modls (Mr) tests were condcted. These are reported in detail elsewhere (1). t was fond that the laboratory-compacted greensands had vales comparable with or higher than soils typically sed for sbgrades in ndiana (6). Use of WFS as a fine aggregate spplement in asphalt concrete has been reported elsewhere (2). Based on simple testing, it was determined that as mch as 15 percent of the conventional sand content of asphalt concrete cold be replaced by WFS. A promising new engineering material is a mixtre called controlled low strength materials (CLSM). These mixtres of cementitios materials, fly ash, sand, and water can flow into hard-to-access locations and, within a few hors, can set p to prodce strengths comparable to compacted soils. Becase it is often desirable to be able to remove or replace these materials (e.g., in tility cts), their strengths shold be relatively low. The CLSM mix will then have parameters of flow, rate of strength gain, and final strength, which vary with mix constitents and percentages. n this stdy, the constitents were Type 1 cement, Class F fly ash, water, and WFSs. The se of WFS in CLSM mixes is relatively new, and therefore comparisons with mixes sing conventional sand appeared appropriate. Table 2 shows these comparisons with vales reported in works by Amon (7) and Nantng (8).

3 Javed and Lovell As Compacted CSR Soaked CBR Swell -. '\ \!R 16 / Ck'. 0.2 CD Q; (/) 12!R 8 4 i \ \ SJ O , E O" 1.65 v; c:: QJ Q > a Water Content, sg FGURE2 Moistre density-cbr relationship of G 1. The WFSs are assmed to be free of charge and, within a short distance of sorce, prodce CLSMs that are economical. Mix 3 is preferred within this set, bt the se of WFS raises qestions of porosity and permeability, as well as setting rate. These isses are being addressed in crrent research (9). ENVRONMENTAL CONCERNS All of the WFSs tested in this stdy were from ferros castings. Therefore, one wold expect that the only elements present in the waste sand wold be the additives and a small amont of iron. Becase the principal additives in greensand are clay and seacoal, these materials wold be expected to pass all reasonable environmental tests. The greensands pass the varios leaching tests. Bioassay testing (a waste water techniqe) is also planned for these materials, bt they are expected to pass. There is no specific environmental testing for the chemicals and other inorganic or organic sbstances sed in chemically bonded and shell-molded WFSs; therefore, more potential risk is involved. One also expects that greensand casting of other metals (e.g., alminm), will yield a nonhazardos waste. Conversely, other castings, like brass, wold be highly sspect. Two se scenarios appear well-sited for demonstrating the se of greensand from ferros casting: Well-drained sitations like sbgrades and embankmentsleachate in sch sitations is small and is frther minimized by clayencasing layers on the sides. Some sort of growing layer wold need to be placed at the sides, even withot the encasement reqirement.

4 TABLE 1 Reslts of Direct Shear Tests Loose Dense Sample# Dr c (kpa) 0 (deg) Dr c (kpa) 0 (deg) G G C Sl Rl Uniform Moderately Very medim sanda dense dense Sandb Loose Dense psi = 6.89 kpa a (4) b (5) TABLE 2 Spread, Density, and Compressive Strength for Different Mixes Mix No ln ES-lb Cement (kg/m 3 ) Fly ash (kg/m 3 ) Water (kg/m 3 ) WFS or Sand (kg/m 3 ) Spread (mm) c 127-i52 Density (kg/m 3 ) c 2147 Compressive strength (kpa) 3 - Day c Day Day pcy = kg/m 3, 1.0 in = 25.4 mm, 1.0 pcf = kg/m 3, 1.0 psi = 6.89 kpa a (7) 1>(8) c Data not available

5 Javed and Lovell 113 Stabilized mixtres like CLSMs-althogh these are wet and poorly drained sitations (trenches, arond pipes and tanks), the WFSs are reasonably well fixed in the mix. Porosity and permeability of the CLSM need to be given greater attention to increase confidence in this sage. CONCLUSONS The WFSs passed all environmental tests and demonstrated good physical and mechanical properties. Greensands from ferros castings appear likely to pose very low environmental risk. Shell-molded and chemically bonded WFSs pass all existing leaching tests bt warrant frther stdy with respect to chemical and organic additives before se. Uses most favored for demonstration projects are (a) welldrained embankments and sbgrades and (b) a fine aggregate in CLSMs. The WFSs are likely to be economically competitive when the project is close to a disposal sorce. ACKNOWLEDGMENT The research reported herein was fnded by the ndiana Cast Metals Association, which has its headqarters in ndianapolis, ndiana. REFERENCES 1. Javed, S., and C. W. Lovell. Use of Waste Fondry Sand in Highway Constrction. Report JHRP/NDOT/FHW A-94/2. School of Civil Engineering, Prde University, West Lafayette, Jly Javed, S., C. W. Lovell, and L. E. Wood. Waste Fondry Sand in Asphalt Concrete. n Transportation Research Record 1437, TRB, National Resarch Concil, Washington, D.C., 1994, pp Parkes, E., G. Westwood, and R. Grigg. Molding and Core Sands. n Applied Science in the Casting of Metals, (K. Strass, ed.), Foesco nternational Limited, Permagon Press, Birmingham, Great Britain, 1970, pp Terzaghi, K., and R. B. Peck. Soil Mechanics in Engineering Practice. John Wiley and Sons, nc., New York, Peck, R. B., W. E. Hanson, and T. H. Thornbrn. Fondation Engineering. John Wiley and Sons, nc., New York, Lee, W. Evalation of n-service Sbgrade Resilient Modls with Con. siderations of Seasonal Effects. Ph.D. thesis, Prde University, West Lafayette, nd Amon, J. A. Controlled Low-Strength Material. Constrction Specifier, Vol. 43, No. 12, Dec. 1990, pp Nantng, T. Design Criteria of Controlled Low Strength Materials. Ph.D. thesis. Prde University, West Lafayette, nd. Ag Bhat, S. T., C. W. Lovell, C. F. Scholer, and T. E. Nantng. Flowable Fill Using Waste Fondry Sand. Proc., 1 lth/nternational Symposim on Use and Management of Coal Combstion By-Prodcts, Vol. 2 (CCBs), EPR TR V2, Project 3176, Orlando, Fla., Jan. 1995, pp to Pblication of this paper sponsored by Committee on Cementitios Stabi /i.zation.