FSsand

7/26/2019 FSsand

1/14

Geoenvironmental behavior of foundry sand amended mixturesfor highway subbases

Yucel Guney a, Ahmet H. Aydilek b,*, M. Melih Demirkan b

a Department of Civil Engineering, 2 Eylul Campus, Anadolu University, Eskisehir 26480, Turkeyb Department of Civil and Environmental Engineering, University of Maryland, 1163 Glenn Martin Hall, College Park, MD 20742, USA

Accepted 16 June 2005

Abstract

The high cost of landfilling and the potential uses of waste foundry sands have prompted research into their beneficial reuse.Roadways have a high potential for large volume usage of the foundry sands. A laboratory testing program was conducted onsoil-foundry sand mixtures amended with cement and lime to assess their applicability as highway subbase materials. The mixtureswere compacted in the laboratory at a variety of moisture contents and compactive efforts and subjected to unconfined compression,California bearing ratio, and hydraulic conductivity tests. The environmental suitability of the prepared mixtures was evaluated byanalyzing the effluent collected during hydraulic conductivity tests. Finally, required subbase thicknesses were calculated using thelaboratory-based strength parameters. The results of the study show that the strength of a mixture is highly dependent on the curingperiod, compactive energy, lime or cement presence, and water content at compaction. The resistance of foundry sand-based spec-imens to winter conditions is generally better than that of a typical subbase reference material. Laboratory leaching tests indicatedthat if these mixtures later come in contact with water that has been discharged directly to the environment (e.g., drainage throughasphalt pavement), the quality of water will not be affected. 2005 Elsevier Ltd. All rights reserved.

1. Introduction

Foundry system sand, which is occasionally referredto as foundry sand, is widely used in the metal castingindustry to create the mold into which molten metal ispoured. The system sand is a blend of silica sand, organ-ic additives, and bentonite as a binder. The properties ofthe foundry sand are determined by the relative amounts

of binder and other additives, and are unique to eachmetal casting process. The continued addition of thesebinders and additives creates an excess volume that can-not be stored in the finite storage volume of the foundrysand system. The excess volume is typically landfilled

even though it may have good engineering properties.For instance, the annual generation of foundry systemsand is approximately 9 million and 12 million metrictons in Europe and the United States, respectively,and most of this sand is disposed in waste containmentfacilities (Abichou et al., 2004). At present, only 32% offoundry sand is beneficially reused in construction. Thisis mainly due to lack of information on its possible ben-

eficial uses or due to regulatory requirements today inforce across some countries that classify foundry sandsamong hazardous wastes (FEAD, 2001; SEPA, 2002).Unless alternative uses of this excess sand are intro-duced, an increase in landfilling costs is inevitable underpresent circumstances.

Significant efforts have been made in recent years touse foundry sand in civil engineering construction. Someof the application areas included highway bases andretaining structures (Kirk, 1998; Mast and Fox, 1998;

0956-053X/$ - see front matter 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.wasman.2005.06.007

* Corresponding author. Tel.: +1 301 314 2692; fax: +1 301 4052585.

E-mail addresses: [email protected] (Y. Guney), [email protected](A.H. Aydilek).

www.elsevier.com/locate/wasman

Waste Management xxx (2005) xxxxxx

ARTICLE IN PRESS
mailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]

7/26/2019 FSsand

2/14

Goodhue et al., 2001), landfill liners (Abichou et al.,1998, 2004), asphalt concrete (Javed and Lovell, 1995),flowable fill (Bhat and Lovell, 1996), and pavementbases (Kleven et al., 2000). Other studies have shownthat the thermal or biological remediation of the foun-dry sands provides an opportunity for their land appli-

cations (Leidel and Novakowski, 1994; Reddi et al.,1996).Existing research has shown that foundry sand can be

effectively used in geotechnical construction due to itscomparable properties with sand-bentonite mixtures(Abichou et al., 2004). However, limited information ex-ists about the use of foundry sand as a component inbase or subbase layers of highway pavements. Roadwayapplications provide an opportunity for high volume re-use of the excess material. Moreover, the effect of differ-ent factors on the mechanical properties of the subbaselayers constructed with foundry sand need to be evalu-ated. These factors are mainly due to differences in con-

structional operations (e.g., compaction conditions),material homogeneity, and the selection of differentmaterials amended with foundry sand.

The objective of this study was to investigate the ben-eficial reuse of foundry sand amended mixtures for sub-base layers in highways. To achieve this objective, abattery of tests was conducted on foundry sand androck-foundry sand mixtures amended with lime or ce-ment. Unconfined compressive strength (qu) and Cali-fornia bearing ratio (CBR) tests as well as scanningelectron microscopy (SEM) analyses were conductedto investigate the effect of cement and lime addition, cur-

ing time, molding water content, and mixture gradationon geotechnical parameters. The effect of winter condi-tions was examined by performing hydraulic conductiv-ity and unconfined compression tests on the specimensafter a series of freezethaw cycles. Finally, the environ-mental suitability of the prepared mixtures was evalu-ated through leaching tests.

2. Materials

The foundry sand used in this study was obtainedfrom Toprak Foundry located in Bilecik, Turkey. Thefoundry sand had approximately 24% particles passingthe US No. 200 sieve (

7/26/2019 FSsand

3/14

AASHTO Classification System. Furthermore, thematerial contained about 1% organic matter. The finescontent of the nonplastic, inorganic subbase materialwas 8%, and it was classified as light brown silty sand

(SM) and A-2-4 according to the USCS and AASHTO,respectively. The final gradations of the materials usedin the testing program are provided inTable 2.

3. Test procedures

The procedure outlined in ASTM D 4318 was fol-lowed to measure the liquid and plastic limits of eachfoundry sand sample. The only deviation from ASTMD 4318 procedure was the hydration period. Based onthe suggestions of Kleven et al. (2000), the hydration

period was extended to one week to ensure the hydra-tion of the thermally degraded bentonite in the foundrysand. Mechanical sieving and hydrometer analyses wereconducted following ASTM D 422. The obtained parti-cle size distributions (PSD) indicated that mixtures, ingeneral, satisfied the AASHTO M 147 subbase PSD lim-its (Table 2).

The specimens were compacted at the optimum mois-ture content as well as at 2% wet and 2% dry of opti-mum, to examine the effect of the molding watercontent on the strength parameters. As with Atterberglimit tests, the foundry sand samples were hydrated forone week prior to compaction. This was necessary, sincedeveloped compaction curves (not presented herein)indicated that the dry unit weight changes significantlypast 3% optimum moisture contents (OMC) dueto moisture-sensitive nature of the foundry sand.

Two different compactive efforts were studied: stan-dard Proctor (ASTM D 698) and modified Proctor(ASTM D 1557). Crushed rock was added as 55%and 73% of the total weight to some mixtures todetermine its effect on the engineering performanceof the mixtures. Cement and lime were added as bind-ers to 18 mixtures. These binders are commonly usedin highway subbase or subgrade construction to in-

crease the bearing capacity of soils. Table 3 providesa summary of the 40 mixtures used in this study. Ta-ble 3 also provides the optimum moisture contents(OMC) and maximum dry unit weights (cdm) of the

mixtures based on compaction tests. The nomencla-ture represents the compaction effort, percentages ofrock and binders used and the compaction moistureconditions, respectively (e.g., F-M-R73-C5(+2) is amixture comprised of foundry sand, 73% rock, and5% cement compacted at 2% wet of optimum moisturecontent using a modified Proctor effort). Unconfinedcompression, CBR, and hydraulic conductivity testswere performed on the mixtures to evaluate the effectof selected factors on subbase performance. Triplicatetests were conducted on each mixture as quality con-trol and the averages of these three tests are reported

as results. The coefficient of variation (ratio of stan-dard deviation to the mean) was less than 25% inhydraulic conductivity tests and less than 15% in theremaining tests.

3.1. Geomechanical tests

The unconfined compressive strength (qu) is a com-mon measure of the strength of a mixture design inroadways and is often used to determine the structurallayer coefficients of the subbase layers for designingpavements. Strength testing followed the proceduresoutlined in ASTM D 1633. A strain rate of 1%/minwas maintained during the unconfined compressiontests. The specimens prepared with foundry sand,crushed rock, cement, and lime were compacted at twodifferent compactive efforts. After compaction, the spec-imens were extruded with a hydraulic jack, sealed inplastic wrap, and cured for 1 and 7 days at 100% relativehumidity and controlled temperature (21 2 C) beforetesting. The specimens for CBR testing were preparedfollowing the procedures listed in AASHTO T-193 andASTM D 1883. The specimens were compacted at theoptimum moisture content (OMC) using the standardProctor effort and were cured for 1 and 7 days at

Table 2Particle size distribution of materials and mixtures used in the testing program

Specimen name Particle size (mm)

D10 AASHTO M147 limits D30 AASHTO M147 limits D60 AASHTO M147 limits D85 AASHTO M147 limits

F 0.002 0.060.08 0.15 0.10.4 0.25 0.252 0.4 0.911F-L5 0.002 0.060.08 0.075 0.10.4 0.25 0.252 0.4 0.911

F-C5 0.002 0.060.08 0.075 0.10.4 0.25 0.252 0.4 0.911F-R73 0.1 0.060.2 0.7 0.43 6 3.512 15 925F-R55 0.07 0.060.2 0.2 0.43 6 3.512 11 925F-R73-C5 0.08 0.060.2 0.7 0.43 6 3.512 15 925B 0.2 0.060.2 1.9 0.43 7 3.512 23 925D 0.0035 0.060.08 0.3 0.10.4 1.5 0.252 3 0.911R NA 5.5 NA 8.3 NA 15 NA

Note. L: lime; C: cement; R: crushed rock; F: foundry sand; B: reference subbase; D: reference base; NA: not available.

Y. Guney et al. / Waste Management xxx (2005) xxxxxx 3

ARTICLE IN PRESS

7/26/2019 FSsand

4/14

100% relative humidity and controlled temperature(21 2 C) before testing. The swell during soakingwas measured using a dial gauge. Penetration was con-ducted using a metal piston with a diameter of 4.96 cmper ASTM D 1883. Corrections in the penetrationcurves were made as described in the selected proce-dures. A strain rate of 1.27 mm/min was used to de-scribe the penetration curve.

3.2. Hydraulic tests

The hydraulic conductivity of the reference basematerial was determined using the constant-head meth-od in accordance with ASTM D 2434. The proceduredescribed in ASTM D 5856 was followed for the

hydraulic conductivity test conducted on subbase mate-rial and foundry sand-base specimens. Specimens of101.6 mm in diameter and 116.4 mm in height werecompacted at their OMC in five layers with the standardProctor energy and were then cured for 7 days at 100%relative humidity and at 21 2 C following the com-paction. The specimens were left overnight to achieveequilibrium inside the rigid-wall cells before initiatingthe tests. The influent was prepared from a stock solu-tion that contained rain water. The feed solution hadan electrical conductivity (EC) of 0.1 mS/cm and a pHof 7.2, comparable with the properties of water in thenatural environment (Tuncan et al., 2000; Lee et al.,2004). Table 4 provides a list of chemical constituentspresent in the feed solution. The applied hydraulic

Table 3Legend and the composition for the mixture designs

Specimen name Foundrysand, %

Crushedrock, %

Lime, % Cement, % W/C Compactiveeffort

Water contentduring compaction

cdm (kN/m3) OMC, %

F 100 Standard OMC 17.03 12.3F(+2) 100 Standard OMC+2F(2) 100 Standard OMC2

F-M 100 Modified OMC 18.83 9.0F-M(+2) 100 Modified OMC+2F-M(2) 100 Modified OMC2F-L5 100 5 2.9 Standard OMC 16.09 14.7F-L5(+2) 100 5 3.3 Standard OMC+2F-L5(2) 100 5 2.5 Standard OMC2F-M-L5 100 5 2.5 Modified OMC 17.56 12.5F-M-L5(+2) 100 5 2.9 Modified OMC+2F-M-L5(2) 100 5 2.1 Modified OMC2F-C5 100 5 2.5 Standard OMC 16.96 12.5F-C5(+2) 100 5 2.9 Standard OMC+2F-C5(2) 100 5 2.1 Standard OMC2F-M-C5 100 5 2.0 Modified OMC 17.84 10F-M-C5(+2) 100 5 2.4 Modified OMC+2F-M-C5(2) 100 5 1.6 Modified OMC2

F-R73 27 73 Standard OMC 20.95 7.5F-R73(+2) 27 73 Standard OMC+2F-R73(2) 27 73 Standard OMC2F-M-R73 27 73 Modified OMC 21.34 5.6F-M-R73(+2) 27 73 Modified OMC+2F-M-R73(2) 27 73 Modified OMC2F-R55 45 55 Standard OMC 20.22 7.7F-R55(+2) 45 55 Standard OMC+2F-R55(2) 45 55 Standard OMC2F-M-R55 45 55 Modified OMC 20.65 6.1F-M-R55(+2) 45 55 Modified OMC+2F-M-R55(2) 45 55 Modified OMC2F-R73-C5 27 73 5 1.5 Standard OMC 20.98 7.7F-R73-C5(+2) 27 73 5 1.9 Standard OMC+2F-R73-C5(2) 27 73 5 1.1 Standard OMC2F-M-R73-C5 27 73 5 1.1 Modified OMC 21.40 5.7F-M-R73-C5(+2) 27 73 5 1.5 Modified OMC+2F-M-R73-C5(2) 27 73 5 0.7 Modified OMC2B Standard OMC 22.25 8.3B-M Modified OMC 23.22 6.5D Standard OMC 20.77 7.0D-M Modified OMC 21.88 6.0

Note. L: lime; C: cement; R: crushed rock; F: foundry sand; M: modified Proctor energy; B: reference subbase; D: reference base; OMC: optimummoisture content; cdm: maximum dry unit weight; W/C: water-to-cementing agent ratio.

4 Y. Guney et al. / Waste Management xxx (2005) xxxxxx

ARTICLE IN PRESS

7/26/2019 FSsand

5/14

gradient was 3 for the crushed rock amended specimens,while a hydraulic gradient of approximately 45 was se-lected for the specimens without crushed rock. Each testwas terminated after ensuring the stabilization of flow

and following the criteria given in ASTM D 5084. Theeffluent from the column test (leachate) was collectedon a regular basis and the samples were stored for chem-ical analysis. US EPA (Environmental ProtectionAgency) standard methods were used to store and ana-lyze the leachates.

3.3. Climatic tests

To observe the effect of winter conditions on geotech-nical engineering parameters, some of the mixtures weresubjected to strength and hydraulic conductivity tests

after a series of freezethaw cycles. Specimens with vary-ing cement, lime, and crushed rock contents were com-pacted at their optimum moisture contents usingprocedures outlined in ASTM D 698. After 7 days ofcuring, the specimens were frozen in a temperaturechamber at 23 1 C for 24 h and then thawed in ahumidity chamber for 23 h per ASTM D 560. Specimenswere frozen and thawed at zero overburden stress. Theweight loss, water content, hydraulic conductivity, andunconfined compressive strength were measured at theend of 1, 4, and 8 freezethaw cycles.

3.4. Environmental tests

The pH of selected mixtures was measured using theCole Parmer 39000-50 pH meter following the proce-dures given in U.S. EPA Method 9045. Electrical con-ductivity (EC) measurements were performed usingOmega CDB-70 conductivity meter. Each specimenwas left for 1 h and 24 h at room temperature(20 3 C) before conducting the pH and EC measure-ments, respectively. Cation concentration analyses wereconducted using atomic absorption spectroscopy (AAS)and following the procedure outlined in SW 846 EPAMethod 6010B.

3.5. Microscopy analysis

Undisturbed specimens of 7-day cured specimenswere prepared for scanning electron microscopy (SEM)analysis by the critical point drying technique as outlinedin Bennett et al. (1977). The specimens were initially trea-

ted with acetone and a critical point drying apparatuswasutilized to replace the acetone with CO2. The specimenswere held on an aluminum sample holder with adhesivetape. Later, they were coated with gold to minimize anycharge build-up. The microstructure and chemical com-position of the samples were examined under LEO 440Model SEM using the energy dispersive X-ray (EDX)technique.

4. Results and discussion

4.1. Influence of crushed rock content and molding water

content

It is a common practice to prepare highway subbaseswith crushed rock or rock amended mixtures. As part ofthis study, two different percentages by weight ofcrushed rock, 55% and 73%, were selected such thatthe PSD of each mixture satisfied the AASHTO M147 gradation limits.Fig. 1shows the unconfined com-pressive strength (qu) of the specimens compacted attheir optimum, 2% wet of optimum, and 2% dry of opti-mum moisture contents. Unconfined compression testswere not conducted on the reference subbase material

(B) since it was cohesionless. The results presented inFig. 1indicate that the strength of specimens compactedat optimum or 2% dry of optimum moisture content in-creases with addition of crushed rock; however, thestrength seems to be insensitive for rock contents above

Table 4Chemical properties of feed solution used in column tests

Chemical composition Value (mg/L)

Calcium (as CaO) 8.0Chloride 7.4Potassium (as K2O) 3.5Sulphate 2.6

Magnesium (as MgO) 1.2Nickel NDChromium NDLead NDZinc NDCopper NDCadmium ND

Note. ND: not detected.

Fig. 1. Effect of compaction moisture content and crushed rockcontent on strength (Note. F: foundry sand; M: modified Proctor; R55and R73 designate the specimens with 55% and 73% crushed rock,respectively. Each specimen was cured for 1-day before the test).


ARTICLE IN PRESS

7/26/2019 FSsand

6/14

55% when the standard Proctor compaction effort wasutilized. Additionally, high rock contents are needed toobserve an increase in qu when the modified Proctorcompaction is performed. On the other hand, the mea-sured strength values of crush rock-amended mixturesare higher than those observed for foundry sand. This

trend is clearer for the specimens compacted at dry ofoptimum. For instance, qu is 313 kPa for the mixtureincluding 55% crushed rock (F-R55), which is almosttwice the strength determined for the foundry sand (F)when standard Proctor effort compaction and dry-of-optimum conditions are considered.

Fig. 1also suggests that the molding water content atcompaction can significantly affect thequof the mixturedesign. An increase in molding water content generallyleads to a decrease in strength of the specimens com-pacted with the same effort. The effect of water contenton strength can be explained on the basis of suction the-ory. Specifically, the decreased pore water pressures due

to the lack of water on the dry-of-optimum side of thecompaction curve led to higher strength values. Similarconclusions were made by previous researchers that thedry-of-optimum moisture conditions generally increasesthe strength of highway bases and compacted soils (Ben-son and Daniel, 1990; Aydilek and Arora, 2004).

4.2. Influence of compactive effort, curing period, and

binder addition

The effect of compaction energies onquand CBR arepresented in Fig. 2. For the specimens cured for the

same period of time, higher qu and CBR are observedwith modified Proctor compaction effort. This is mainlydue to higher maximum dry unit weights of the speci-mens that result from the increased compaction energy.For instance, the cdmof foundry sand (F) increased from17.03 to 18.83 kN/m3 with increasing compactive en-ergy. A similar increase in maximum dry unit weightsis evident for all mixtures with varying degrees of changein their qu and CBR (Table 3andFig. 2). The uncon-fined compressive strengths of specimens compactedwith modified Proctor effort are 96% to 675% higherthan those compacted using the standard Proctor effortwhen a 1-day curing period is considered. An increase of75% to 1050% in CBR can also be observed for the spec-imens compacted at a higher compaction effort (i.e.,modified Proctor effort).

Fig. 2also shows that the quand CBR of cement orlime amended mixtures increase with increasing curingtime despite the fact that only two curing periods wereevaluated in this study. On the other hand, a significantincrease is not evident for the remaining mixtures. Thepostulated mechanism is that the delayed release of cal-cium hydroxide (Ca(OH)2) by Portland cement or freelime (CaO) by quicklime caused these increases andthe temperature of the curing chamber and availability

of 100% relative humidity also enhanced these cementi-tious reactions. It is believed that the specimens did notsignificantly hydrate after 1 day of curing and, thus, theincrease is relatively small. On the other hand, thestrength increase is more evident for the 7-day curedspecimens. For instance, the 1-day strength of F-C5 in-creased almost five times after 7 days of curing from 150to 720 kPa when it was compacted at OMC. Similar in-creases in strength with increasing curing period were re-ported by previous researchers (Vishwanathan et al.,1997).

It can be observed inFig. 2that the CBR of foundrysand-based specimens is either comparable or higherthan that of the reference subbase material when similarcuring period or compaction effort is considered. TheCBR generally increases with addition of cement or limeand this increase is clearer when cement was the binder.The CBR of specimens compacted with modified

Fig. 2. Effect of compaction effort, curing period, and cement or limeaddition on (a) strength; and (b) CBR (Note. F: foundry sand; B:reference subbase; R55 and R73 designate the specimens with 55% and73% crushed rock, respectively; L5 and C5 designate the specimenswith 5% lime and cement, respectively).


ARTICLE IN PRESS

7/26/2019 FSsand

7/14

Proctor effort exceeds 50, a generally accepted limit forsubbase applications (Asphalt Institute, 2003). Further-more, the cement or lime amended specimens (L5, C5,and R73-C5) compacted with standard Proctor effortexhibits CBR values higher than 50 after 7 days of cur-ing. The SEM photographs inFig. 3show the agglom-

eration of foundry sand as a result of cement or limeaddition, which may be an indicator of an increase inCBR. Relatively higher amounts of calcium are evidentas a result of the addition of lime, as shown in the EDXplots ofFig. 4. The same figure also suggests that mostof the silica in the foundry sand was consumed to formcalcium silicate hydrates, which in turn hardened thespecimen. Additionally, oxygen (i.e., air) present in thepores of the foundry sand decreased, possibly due to adecrease in the hydraulic conductivity.

Fig. 5 shows that the compaction moisture contentplays a major role in the qu and CBR of the cementor lime amended mixtures. Wet-of-optimum condi-

tions reduce the qu and CBR significantly, and thisis the case when the specimens are compacted with

standard or modified Proctor effort. The effect ofwater content on strength can be explained by thecharacteristics of cementitious reactions. The water-to-cementing agent ratio (W/C) is important in thesereactions, even though it cannot always be optimizedin solidification/stabilization work. Cement or lime

doubles its volume upon hydration, which creates anetwork of small gel pores (Conner, 1990). WhenW/C reaches the range of 0.220.25, the binder startsto fully hydrate which leaves free water (pore water),gel water, and air voids. Very high W/C ratios leavebleed water, which is water that appears as standingwater on the surface of the solid mass. This starts tooccur at a W/C ratio of about 0.48 or greater (Con-ner, 1990). The observed decrease in unconfined com-pressive strength with increasing molding watercontent may be attributed to relatively high W/C ra-tios for the specimens used in the current study (theratio W/C was greater than 0.48, even for the mix-

tures compacted at optimum moisture content usingthe standard Proctor effort).

Fig. 3. SEM photograph of (a) foundry sand prior to compaction, (b) compacted foundry sand, (c) lime amended foundry sand, and (d) cementamended foundry sand (specimens were compacted with standard Proctor effort and cured 7 days).


ARTICLE IN PRESS

7/26/2019 FSsand

8/14

A significant amount of swell was not observed whenthe specimens were kept submerged for four days; hence,the mixtures do not appear to have long-term swellpotential. As seen in Fig. 6, a maximum swell value of0.5% was obtained for foundry sand when it wascompacted at optimum water content using standard

Proctor effort. The other mixtures resulted in swellvalues ranging from 0.05% to 0.4%. Compacting thespecimens at wet of optimum would have resulted inlower swell values (Kleven et al., 2000) even though thiswas not studied herein.

4.3. Influence of winter conditions

In any stabilization application, the stabilized mate-rial should be able to withstand climatic stresses, partic-ularly freezethaw cycles (TFHRC, 2002). Subjectingthe specimens to strength tests after freezethaw (FT)cycles and recording the change in weight have been re-ported as indicators of durability; however, it has beennoted byKalankamary and Donald (1963)that the eval-uation of durability by weight loss as a result of freezethaw cycles (ASTM D 560) is overly severe, and this testprocedure does not totally simulate field conditions.Therefore, some of the U.S. highway agencies currentlyrequire unconfined compression tests in lieu of durabil-ity testing. Previous research indicates that 812 cyclesof freezing and thawing could be considered adequatein investigating the effect of F-T cycles on various engi-neering parameters including strength (Zaman and Naji,2003).

Fig. 6. Swell test results (Note. F: foundry sand; R55 and R73designate the specimens with 55% and 73% crushed rock, respectively;L5 and C5 designate the specimens with 5% lime and cement,respectively. All specimens were compacted with standard Proctoreffort).

Fig. 5. Effect of moisture content on (a) strength; and (b) CBR ofcement or lime amended foundry sand mixtures (Note. F: foundry

sand; M: modified Proctor effort; R73 designate the specimens with73% crushed rock; L5 and C5 designate the specimens with 5% limeand cement, respectively. Each specimen was cured for 1-day beforethe test).

Fig. 4. EDX plot of the SEM photograph of (a) compacted foundrysand (F) and (b) 7-day cured lime amended foundry sand (F-L5).


ARTICLE IN PRESS

7/26/2019 FSsand

9/14

In the current study, the specimens were cured for 7days as normally practiced in pavement construction.Unconfined compression and hydraulic conductivitytests were conducted on the mixtures after each freezethaw cycle and the results are summarized in Figs. 7and 8. The unconfined compressive strength ratio

(qur=qun/qui) and hydraulic conductivity ratio(Kr= Kn/Ki) inFigs. 7 and 8are ratios of hydraulic con-ductivity and unconfined compressive strength after nfreezethaw cycles (Kn or qun) to the initial hydraulicconductivity and strength (Ki or qui). Fig. 7 indicatesthat the qur of all foundry sand-based specimens staysnearly constant between the first and the eighth freezethaw cycles even though they lost 4050% of their initialstrength after the first cycle. The effect of freezethaw onstrength can be explained in terms of the retardation oracceleration of the cementitious reactions. Freezingaction retards the cementitious reactions, which causesa reduction in strength; conversely, thawing action con-

tributes to the increase in strength via accelerating thecementitious reactions. Evidently, the freezing tempera-ture and period are dominant in the first cycle of themixtures. It is believed that between cycles of 1 and 8,both freezing and thawing compensated each otherand hence the observed variation in qur is minimal. Alarger change can be observed in the strength of speci-mens including crushed rock, probably due to their rel-atively higher porosity and susceptibility to frost action.

Fig. 8 shows the effect of freezethaw cycles on thehydraulic conductivity of the mixtures. The resistanceof foundry sand-based specimens to winter conditionsis generally better than that of the two reference materi-als. For instance, the Kr stays in a range 632 for thetwo reference materials whereas the same ratio rangesfrom 2 to 24 for the foundry sand-based mixtures. Theonly exception to that phenomenon is lime amendedspecimens, which experience more than a three order

0

0.2

0.4

0.6

0.8

1

0 2 4 6 8 10

FF-L5F-C5

qur=

qun

/qui

Freeze-thaw cycles

No crushed rock

0

0.2

0.4

0.6

0.8

1

0 2 4 6 8 10

F-R55

F-R73

F-R73-C5

qur=

qun

/qui

Freeze-thaw cycles

Specimens including

crushed rock

a

b

Fig. 7. Effect of winter conditions on unconfined compressive strength(Note. F: foundry sand; R55 and R73 designate the specimens with55% and 73% crushed rock, respectively; L5 and C5 designate thespecimens with 5% lime and cement, respectively).

1

10

100

1000

104

0 2 4 6 8 10

BDFF-L5F-C5

Kr=

Kn

/Ki

Freeze-thaw cycles

No crushed rock

1

10

100

1000

0 2 4 6 8 10

BDF-R55F-R73F-R73-C5

Kr=

Kn

/Ki

Freeze-thaw cycles

Specimens including

crushed rock

a

b

Fig. 8. Effect of winter conditions on hydraulic conductivity (Note. F:foundry sand; B: reference subbase; D: reference base; R55 and R73designate the specimens with 55% and 73% crushed rock, respectively;L5 and C5 designate the specimens with 5% lime and cement,respectively).


ARTICLE IN PRESS

7/26/2019 FSsand

10/14

of magnitude increase in their hydraulic conductivities,i.e., Kr> 1000. Visual observations after the eighth cyclealso indicated that these specimens were at the verge ofdisintegration. The measured weight loss for all of themixtures was in a range of 2% to 30% after eight cyclesof freezethaw (not shown herein), and lime amended

specimens and the foundry sand experienced the greatestweight loss. On the other hand, the weight losses for thecement amended specimens were less than 14%, a gener-ally accepted limit for cement-treated bases. Visualobservations of these mixtures did not reveal cracks ordegradation. Therefore, it was assumed that the volumeof pores was capable of accommodating the formationof ice lenses during freezing without causing noticeabledamages. However, it is important to note that the lab-oratory simulated freezing and thawing conditions areharsher than the normally encountered winter condi-tions (e.g., including the conditions in Turkey), andmost of the mixtures are likely to exhibit better perfor-

mance in the field.

4.4. Environmental suitability

To evaluate the characteristics of leachate from foun-dry sand used in this study, effluent was collected forchemical analyses during the hydraulic conductivitytests. The pH and electrical conductivity (EC) measure-ments conducted on the specimens are given in Fig. 9.As expected, the pH increases when lime or cement isadded. The release of Ca(OH)2or CaO in these bindersincreases the pH, and the measured values are compara-

ble at different times. Furthermore, the encapsulationprocess, a common phenomenon observed during ce-ment stabilization, decreases the EC of the cement-amended mixtures (Tuncan et al., 2000). The encapsula-tion process usually occurs in the first day of the mixingprocess due to the fast hydration of Portland cement,which may be a reason for the relatively higher ECobserved at 24 h. Effluent collected from the specimensduring the tests was initially grayish and cloudy; how-ever, the color and cloudy appearance of the effluent be-gan to diminish after about 8 h. This improvement inclarity is consistent with the small decrease in electricalconductivity observed between 24 and 72 h, as showninFig. 9(i.e., from 0.520 to 0.317 mS/cm).

One drawback of using compacted foundry sand as ahighway subbase material is potential leaching of toxicconstituents. Several studies have been conducted inthe past to evaluate the characteristics of leachate fromfoundry sand (Ham et al., 1981; Lovejoy et al., 1996;Naik et al., 2001). Most of this work showed that foun-dry sand did not cause groundwater or surface watercontamination since the measured concentrations weresignificantly below the US EPA maximum concentra-tion limits. Furthermore, Freber (1996) indicated thatthe concentrations of metals in groundwater underlying

highway embankments are comparable with thoseencountered in the natural environment, i.e., embank-ments built with natural soils. On the other hand, Leeand Benson (2002) and Coz et al. (2004) showed thatthe concentrations of zinc, lead, chromium, and ironleaching from foundry sand may exceed the US EPAlimits; however, they concluded that the difference isonly 10%, which may be considered acceptable.

In the current study, measurements were conducted onthe effluent specimens for selected metals only (nickel,chromium, lead, copper, zinc, and cadmium) due to theloworganiccontentofthefoundrysandusedinthisstudy.The results of the chemical analyses given inFig. 10andTable 5indicate that the leachate is not significantly con-taminated with metals. The metals were likely leachedfrom either fresh constituents added to the system sandcircuit in the foundry or cooling of the molten metal inthe mold. For all metals, the concentration decreasesgradually and relatively lower concentrations are mea-sured at later stages (i.e., 48 and 72 h), which indicatesthat the construction stage of the subbase may be the mostcritical one and mass loading of the groundwater may

6

8

10

12

14

20 30 40 50 60 70 80

F

F-C5F-L5B

pH

Time (hours)

0

5

10

15

20

25

20 30 40 50 60 70 80

FF-C5F-L5B

Electricalconductivity(mS/cm)

Time (hours)

a

b

Fig. 9. (a) pH; and (b) electrical conductivity of foundry sand-basedmixtures (Note. F: foundry sand; B: reference subbase; L5 and C5designate the specimens with 5% lime and cement, respectively).


ARTICLE IN PRESS

7/26/2019 FSsand

11/14

happen during this phase. The results also indicate thatthe lime or cement amended mixtures result in lower metalconcentrations, possibly due to decreased solubility ofthese constituents at high pH values or decreased hydrau-lic conductivities as a result of agglomeration capacity ofcement and lime. Nickel concentrations are higher than

those measured for other metals at all times; however,drinking water standards are not exceeded for any ofthe measured constituents (Davis and Cornwell, 1998).From these analyses, it is evident that water passingthrough foundry sand or foundry sand-based mixtureswill not become contaminated with metallic compounds,since themeasured concentrations werelower than the USEPA limits. Thus, if these mixtures later come in contactwith water that has been discharged directly to the envi-ronment (e.g., drainage through asphalt pavement), thequality of water will not be affected with leaching metals.Other elements occasionally present in foundry sands,such as cyanide and fluorides, may also leach into surface

or groundwater (West Virginia Department of Environ-mental Protection, 2000; South Australian EnvironmentProtection Authority, 2003); however, such a study wasnot conducted herein.

5. Subbase thickness design

Unconfined compression and CBR test results can beused to estimate the thickness of the subbase layer in aflexible pavement, and the procedures defined in theAASHTO Guide (1993)were followed herein to design

the subbase thicknesses. First, the structural numbers(SN) were calculated using a design serviceability loss(Change in the Present Serviceability Index, DPSI) of1.9 and a roadbed material effective resilient modulus(MR) of 34.5 MPa, which was selected based on previ-ous literature (Huang, 1993). Equivalent single-axleloads (ESALs or W18) of 5 million and 50 million wereused in the analyses to simulate low traffic (Case I) andhigh traffic (Case II) conditions, respectively. The over-all standard deviation (So) and reliability (ZR) were as-sumed to be 0.35% and 95%, respectively. Eq. (1) was

Fig. 10. Amount of leached constituents from foundry sand-basedmixtures during column tests (Note. F: foundry sand; B: referencesubbase; L5 and C5 designate the specimens with 5% lime and cement,respectively).

Table 5Mass of leached constituents from foundry sand-based mixtures at the end of testing (at 72 h)

Constituent F F-C5 F-L5 B

L/S(L/kg)

Mass leached(mg/kg)

L/S(L/kg)

Mass leached(mg/kg)

L/S(L/kg)

Mass leached(mg/kg)

L/S(L/kg)

Mass leached(mg/kg)

Ni 0.47 0.09 0.009 0.004Cr 0.22 0.16 0.009 0.004Pb 0.55 0.08 0.56 0.04 0.57 ND 0.44 0.004Cu 0.08 0.05 0.04 0.008Zn 0.11 0.08 ND NDCd 0.009 0.005 0.005 ND

Note. F: foundry sand; B: reference subbase; L5 and C5 designate the specimens with 5% lime and cement, respectively; ND: not detected; L/S:volume (in L) of influent percolated/mass (in kg) of dry material.


ARTICLE IN PRESS

7/26/2019 FSsand

12/14

used to back calculate the structural numbers (SN) fortwo traffic conditions:

logW18 ZR S09.36log10SN 1 0.20

log10DPSI=4.2 1.05

0.41094=SN15.19

2.32log10MR 8.07 1

The top layer of asphalt (D1) was fixed as 102 mm forCase I and 152 mm for Case II. The MRof asphalt wasassumed to be 2965 MPa, which corresponds to a struc-tural coefficient ofa1= 0.44 according to theAASHTOGuide (1993). An MR of 207 MPa (corresponding to astructural coefficient of a2 = 0.14) and a thickness of152 mm (D2) were assumed for the base layer for bothcases based on AASHTO road tests (Huang, 1993).The structural coefficient of the subbase layer (a3) wascalculated for its corresponding CBR values of 7-daycured specimens compacted with standard Proctor effort

and 1-day cured specimens compacted with modifiedProctor effort. Additionally, the MR values were esti-mated using the procedure given in theAASHTO Guide(1993)and they were also used to calculate a second ser-ies of a3 values. Finally, the subbase thicknesses weredetermined using the following formula:

D3SN a1D1 a2D2m2

a3m3; 2

where m2 and m3 are the drainage coefficients for thebase and subbase layer, respectively. They were chosenas 1.2 which assumes excellent drainage conditions with-in the pavement system (Huang, 1993).

Table 6 shows the required subbase thicknesses forCases I and II. Layer coefficients of F-R73-C5 couldnot be interpreted from the CBR test results because

the CBR exceeded 100. The calculated thicknesses arerelatively large as compared to generally practiced onesin highway construction, and greatly exceed the mini-mum limit of 152 mm recommended in the AASHTOGuide (1993). Nevertheless, the results clearly indicatethat the use of foundry sand-based mixtures requires a

comparable or lower subbase thickness than those ob-tained for reference subbase material, when same com-paction effort is considered. Thicknesses of the subbaselayers based on CBR andMRare generally comparable.Additionally, an increase in compaction energies and de-crease in traffic load decreased the required thicknesses.

The calculations indicated that the compaction mois-ture content has a significant effect on the calculatedsubbase thicknesses due to the moisture-sensitive natureof the foundry sand.Table 7 is given as an example todemonstrate this phenomenon. An increase in moisturecontent beyond OMC leads to a significant increase inthicknesses, and in most cases measurements were not

possible on the nomographs given in the AASHTOGuide (1993)due to the low CBR. On the other hand,dry-of-optimum conditions required lower subbasethicknesses. As discussed before, winter conditions gen-erally lead to a decrease in strength of cement-treatedmixes. This would indicate larger subbase thicknesses,even though not presented herein. However, it shouldbe noted that the climatic stresses may have unexpectedeffects on the mixtures even in the short-term (i.e., dur-ing construction), and therefore precautions should betaken to protect specimens from in-situ freezing condi-tions in harsh environments (e.g., insulation).

Table 6Required subbase thickness for different mixture designs for two trafficconditions (all thickness values are in mm)

Specimen name CBR MR

Case I Case II Case I Case II

F 478 860 477 858F-M 355 639 356 642F-L5 371 668 365 656F-M-L5 365 656 357 642F-C5 327 589 327 589F-M-C5 368 663 368 662F-R73 526 947 546 983F-M-R73 349 629 352 633F-R55 395 712 408 735F-M-R55 361 650 359 646F-R73-C5 787F-M-C5 362 368 525F-R73 369 526 NAF-M-R73 323 349 >787F-R55 372 395 >787F-M-R55 346 361 >787F-R73-C5 388 787B 488 515 NAB-M 411 455 >787

Note. L: lime; C: cement; R: crushed rock; F: foundry sand; M:modified Proctor energy; B: reference subbase. Minimum thicknessrequirement (AASHTO Guide 1993) for ESALs greater than 5,000,000is 152.4 mm. NA: not available since the specimen failed duringtesting.


ARTICLE IN PRESS

7/26/2019 FSsand

13/14

6. Conclusions

Foundry system sand is a blend of silica sand, organicadditives, and bentonite as a binder. Continued additionof binders and additives to the foundry system sand cre-ates an excess volume that cannot be stored in landfills.

Highway subbases are one of the largest applicationareas, and reuse of the foundry sand can provide signif-icant cost savings. A battery of tests including uncon-fined compressive strength (qu) and California bearingratio (CBR) tests, as well as scanning electron micros-copy (SEM) analyses, were conducted to investigate theeffect of cement and lime addition, curing time, moldingwater content, and mixture gradation on the geotechni-cal properties of foundry sand-based highway subbases.The effect of winter conditions were examined by per-forming hydraulic conductivity and unconfined com-pression tests on the specimens after a series of freezethaw cycles. Finally, the environmental suitability of

the prepared mixtures was analyzed through leachingtests. The observations are summarized as follows:

1. An increase in strength can be obtained in the field bycompacting the foundry sand-based mixtures usinghigher compactive efforts (i.e., energy simulatingmodified effort). The test results showed that the watercontent at compaction could affect thequof the mix-ture design considerably. The performance of thefoundry sand-based mixtures can be significantlyincreased by preventing the intrusion of excess waterin the field. It is recommended that the subbase layer

be compacted at dry of optimum for higher strength.Alternatively, compaction may be performed at opti-mum moisture content; however, engineers should becareful about rain or any other addition of unwantedwater at the time of compaction.

2. Lime or cement treatment had a beneficial effect onthe strength of the mixtures. Addition of lime orcement increased the qu and CBR of the fullyhydrated (i.e., cured for 7 days) specimens. Further-more, addition of cement increased the qu and CBRof the rock-foundry sand mixtures.

3. For all mixtures, the qu decreased and the Kr increasedwith an increasing number of freezethaw cycles. Thedecrease in strength with the number of cycles wasmore prominent for mixtures that contained crushedrock owing to their relatively higher porosity and frostsusceptibility. The resistance of foundry sand-basedspecimens to winter conditions was generally betterthan that of the two reference materials as observedin relatively lower increases in their hydraulic conduc-tivity during freezethaw cycles.

4. Laboratory leaching tests indicated that water pass-ing through foundry sand-based mixtures will notbecome contaminated with undesirable compounds.Thus, if these mixtures later come in contact with

water that has been discharged directly to the envi-ronment (e.g., drainage through asphalt pavement),the quality of water will not be affected.

5. Lower subbase thicknesses would be required if foun-dry sand-based mixtures were used in lieu of a refer-ence subbase material. The calculations also indicated

that the compaction moisture content has a signifi-cant effect on the calculated subbase thicknessespartly due to the moisture-sensitive nature of thefoundry sand. Additionally, an increase in compac-tion energies and decrease in traffic load decreasesthe required thicknesses.

6. The results indicated that the foundry sand utilized inthis study satisfies the geomechanical as well environ-mental limits and can be safely used as a component inhighway subbases. The physical and chemical proper-ties of this particular foundry sand are similar to thefoundry sands that exist in many locations; therefore,the reported trends may be considered for future stud-

ies. On the other hand, the effects of organic additivesand other elements (e.g., cyanides, fluorides) on sur-face water or groundwater were not analyzed herein,and such tests should be conducted if the trendsreported in the current study will be adopted.

Acknowledgements

The funding for this project was provided by Anad-olu University, Turkey, through contract No. 03-210.

This support is gratefully acknowledged. Dr. Yucel Gu-ney was the principal investigator (PI) in this project.

References

AASHTO Guide, 1993. Guide for Design of Pavement Structures,American Association of State Highway and TransportationOfficials, Washington, DC.

Abichou, T., Benson, C.H., Edil, T.B., Freber, B.W., 1998. Usingwaste foundry sand for hydraulic barriers. In: Vipulanandan, C.,Elton, D. (Eds.), Recycled Materials in Geotechnical Applications,Geotechnical Special Publication 79. ASCE, Boston, MA, pp. 8699.

Abichou, T., Benson, C.H., Edil, T.B., Tawfiq, K., 2004. Hydraulicconductivity of foundry sands and their use as hydraulic barriers.In: Aydilek, A.H., Wartman, J. (Eds.), Recycled Materials inGeotechnics, Geotechnical Special Publication 127. ASCE, Balti-more, Maryland.

Asphalt Institute, 2003. Thickness Design Highways and Streets,Manual SeriesNo. 1, AsphaltInstitute, Lexington,Kentucky, 110p.

Aydilek, A.H., Arora, S., 2004. Fly ash amended soils as highway basematerials. In: Yegian, M.K., Kavazanjian, E. (Eds.), GeotechnicalEngineering for Transportation Projects, Geotechnical SpecialPublication 126, vol. 1. ASCE, Los Angeles, CA, pp. 10321041.

Bennett, R.H., Bryant, W.R., Keller, G.H., 1977. Clay Fabric andGeotechnical Properties of Selected Submarine Sediment Coresfrom the Mississippi Delta, NOAA, Mississippi, Professionalpaper, pp. 891901.


ARTICLE IN PRESS

7/26/2019 FSsand

14/14

Benson, C.H., Daniel, D., 1990. Influence of clods on the hydraulicconductivity of compacted clay. Journal of Geotechnical Engi-neering 116 (8), 12311248.

Bhat, S.T., Lovell, C.W., 1996. Use of Coal Combustion Residues andWaste Foundry Sands in Flowable Fill, Purdue University-JointHighway Research Project Report, Federal Highway Administra-tion, Washington, DC, 240 p.

Conner, J.R., 1990. Chemical Fixation and Solidification of Hazard-ous Wastes. Van Nostrand Reinfold, New York, 692 p.

Coz, A., Andres, A., Soriano, S., Irabien, A., 2004. Environmentalbehavior of stabilized foundry sludge. Journal of HazardousMaterials 109, 95104.

Davis, M.L., Cornwell, D.A., 1998. Introduction to EnvironmentalEngineering. Mc-Graw Hill, New York, NY.

FEAD, 2001. Position Paper on Classification of Wastes (http://www.fead.be/downloads/ClassificationWastes_055050901.pdf).

Freber, 1996. Beneficial reuse of selected foundry waste material,Proceedings of the 19th International Madison Waste Conference,University of Wisconsin-Madison, pp. 246257.

Goodhue, M., Edil, T.B., Benson, C.H., 2001. Interaction of foundrysand with geosynthetics. Journal of Geotechnical and Geoenviron-mental Engineering 127 (4), 353362.

Ham, R.K., Boyle, W.C., Kunes, T.P., 1981. Leachability of foundryprocess solid wastes. Journal of Environmental Engineering 107(2), 155170.

Huang, Y.H., 1993. Pavement Analysis and Design. Prentice-Hall,Inc., New Jersey, 805 p.

Javed, S., Lovell, C.W., 1995. Uses of waste foundry sand in civilengineering. Transportation Research Record, No. 1486, pp. 109113.

Kalankamary, G.P., Donald, D.T., 1963Development of Freeze-ThawTest for Design of Soil-cement, No. 36. Highway Research Board(HRB), pp. 7796.

Kirk, P.B., 1998. Field Demonstration of Highway EmbankmentConstructed Using Waste Foundry Sand, Ph.D. Dissertation,Purdue University, West Lafayette, IN, 202 p.

Kleven, J.R., Edil, T.B., Benson C.H., 2000. Evaluation of excess

foundry system sands for use as subbase material. Proceedings ofthe 79th Annual Meeting, Transportation Research Board, Wash-ington, DC, CD-Rom, 27 p.

Lee, T., Benson, C.H., 2002. Using waste foundry sands as reactivemedia in permeable reactive barriers, Geo Engineering Report 02-01, University of Wisconsin-Madison, 268 p.

Lee, T., Park, J., Lee, J., 2004. Waste green sands as reactive media forthe removal of zinc from water. Chemosphere 56, 571581.

Leidel, D.S., Novakowski, M., 1994. Beneficial sand reuse: making itwork. Modern Casting 84 (8), 2831.

Lovejoy, M.A., Ham, R.K., Traeger, P.A., Wellander, D., Hippe, J.,Boyle, W.C., 1996. Evaluation of selected foundry wastes for use inhighway construction. Proceedings of the 19th InternationalMadison Waste Conference, University of Wisconsin-Madison,pp. 1931.

Mast, D.G., Fox, P.J., 1998. Geotechnical performance of a highwayembankment constructed using waste foundry sand. In: Vipulan-andan, C., Elton, D. (Eds.), Recycled Materials in GeotechnicalApplications, Geotechnical Special Publication 79. ASCE, Boston,MA, pp. 6685.

Naik, T.R., Singh, S.S., Ramme, B.W., 2001. Performance andleaching assessment of flowable slurry. Journal of EnvironmentalEngineering 127 (4), 359368.

Reddi, L.N., Rieck, G.P., Schwab, A.P., Chou, S.T., Fan, L.T.,1996. Stabilization of phenolics in foundry waste using cemen-titious materials. Journal of Hazardous Materials 4 (23), 89106.

SEPA, 2002. European Waste Catalogue (http://www.sepa.org.uk/pdf/consultation/closed/2002/haz_waste/TGNappendixa.pdf).

South Australian Environment Protection Authority, 2003. UsedFoundry Sand-Classification and Disposal (http://www.epa.sa.go-v.au/pdfs/guide_foundrysand.pdf).

Tuncan, A., Tuncan, M., Koyuncu, H., 2000. Use of petroleumcontaminated drilling wastes as subbase material for road con-structionWaste Management and Research, vol. 18. SAGE Pub-lications, Beverly Hills, CA.

Turner-Fairbanks Highway Research Center (TFHRC), 2002. CoalFly Ash User Guideline Stabilized Base (http://www.tfhrc.gov/hnr20/recycle/waste/cfa55.htm).

West Virginia Department of Environmental Protection, 2000. SpentFoundry Sand Beneficial Reuse Guidelines (http://iofwv.nrcce.w-vu.edu/metal/foundry.pdf).

Vishwanathan, R., Saylak, D., Estakhri, C., 1997. Stabilization of

subgrade soils using fly ash. Ash Utilization Symposium. CAER,Kentucky, pp. 204211.Zaman, M.M., Naji, K.N., 2003. Effect of freezethaw cycles on class

C fly ash stabilized aggregate base. Proceedings of the 82nd AnnualMeeting, Transportation Research Board, Washington, DC, CD-Rom, 22 p.


ARTICLE IN PRESS
http://www.fead.be/downloads/ClassificationWastes_055050901.pdfhttp://www.fead.be/downloads/ClassificationWastes_055050901.pdfhttp://www.sepa.org.uk/pdf/consultation/closed/2002/haz_waste/TGNappendixa.pdfhttp://www.sepa.org.uk/pdf/consultation/closed/2002/haz_waste/TGNappendixa.pdfhttp://www.epa.sa.gov.au/pdfs/guide_foundrysand.pdfhttp://www.epa.sa.gov.au/pdfs/guide_foundrysand.pdfhttp://www.tfhrc.gov/hnr20/recycle/waste/cfa55.htmhttp://www.tfhrc.gov/hnr20/recycle/waste/cfa55.htmhttp://iofwv.nrcce.wvu.edu/metal/foundry.pdfhttp://iofwv.nrcce.wvu.edu/metal/foundry.pdfhttp://iofwv.nrcce.wvu.edu/metal/foundry.pdfhttp://iofwv.nrcce.wvu.edu/metal/foundry.pdfhttp://www.tfhrc.gov/hnr20/recycle/waste/cfa55.htmhttp://www.tfhrc.gov/hnr20/recycle/waste/cfa55.htmhttp://www.epa.sa.gov.au/pdfs/guide_foundrysand.pdfhttp://www.epa.sa.gov.au/pdfs/guide_foundrysand.pdfhttp://www.sepa.org.uk/pdf/consultation/closed/2002/haz_waste/TGNappendixa.pdfhttp://www.sepa.org.uk/pdf/consultation/closed/2002/haz_waste/TGNappendixa.pdfhttp://www.fead.be/downloads/ClassificationWastes_055050901.pdfhttp://www.fead.be/downloads/ClassificationWastes_055050901.pdf

Documents

FSsand