CA1191292A - Silicate treatment of impure silica sands - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE

A process is provided for treating impure silica-containing sands to make them more suitable for foundry use. The sands are treated with aqueous alkali metal silicate solutions and the resulting mixture is heated before the sands are coated with a resin binder.
Foundry cores and molds prepared with these treated sands show improved hot and cold tensile strengths.

Description

i~ ~

3 SILICAT~ TREAT~JENT OF
4 ïMPURE SILICA SA~DS

7 FIELD OF THE I~VEI~TIO~

9 This invention relates to silica~-containing foundry sand alld to a process for t;reating silica.-contaill:LnF~:roundry 11 Salld W:i.t}l all alk,ll~. metal S~ i.C.lte to l~llprove the ~ensi~
12 strength of foundry cores or molds Inade from the sand.

14 BACKGROUND OF THE IhVENTION
16 In the foundry art, cores or ~nolcls for makin~ rnetal 17 castings are normally prepared frorn a r)-Lxture of an aggregate 1~ rnaterial, such as sand; and a bi.ndi.ng amoullt of a binder or 19 binder system. Typically, a.fter the ~ggregate rnateri.al and binder have been mixed, the resulting mixture is rammed, 21 blown or otherwise ~ormed to the desired shape or pattern 22 and then cured with the use of catalysts and/or heat to a 23 solid, cured state.

A variety of different processes for forming molds 26 and cores have been developed in the fo~ndry industry. One 27 type of process kno~n as the shell moldin~ process, is ~ell 28 known in the art. While there are man~ variations of this ~1~'32 1 process, the process essentially comprises depositing a

2 corr~bination of sand and potentially thermosetting resin

3 against a heated pattern such tha~ the resin melts and

4 cures to form a rigid shell mold or core section for use in the casting of metals. The combination of resin and sand used in the process can be a mixture of powdered resin 7 and sand, or a free-flowing coated sand in which each grain 8 is coated with a nontacky layer of resin.

The product:ion Or a core or rnold hy the shell 11 proces'i :invo].ves t~o basic st;ep~" I}~e lnves~ arld the e~ ne 12 step. In the first step, the resin-coated sand is dulnped 13 onto or blown against the heated metal pattern. T~e 14 resin-coated sand is held against the pattern (invested) until the shell is thick enough to hold ~netal in a given 16 application. In the second step, the resin-coated sand is 17 dumped or dropped away from the shell of bonded coated 18 particles o~ sand and the resulting shell is cured. After 19 the shell is cured, it is removed from the hot meta~ pat1;ern and is ready for use.

22 Another process, known to the art as the t'no-bake"
23 process, is also used in I`orming resin cores. This process 24 requires no external heating. Instead, curing is accomplished by means of a catalyst added just beI`ore the sand and resin 26 components are introduced into the core box. Base-cured 27 resin components used in the no-bake process are generally 3~ -2-f 1 m tures o î poly~ls and polylsocy2nat~s. Solutions o r these 2 components are usually coated on ~he sand immediately before 4 use.

A third process for making cores and molds employs 6 sands treated with core oil mixes. These mixes contain drying 7 oils and cereal binders. Cores and molds made with such core 8 oil mixes are cured by baking them in an oven.
.In all of thec;e processes, t}l(:' bl11d~r W~ Ch }laS be~n 1l n~ ed wi1;h sand ac~s, whel1 cured, to b:ind tlle particle!~ of s,ancl 12 in the form of the pattern. The core or mold mus~ be strong 13 enough to contain the rnolten metal until it solidifies. For 14 this reason, a core or mold with high tensile strength is required.

17 One factor influencing t;he tensile strength of the lo cores and molds -is the quality of the sand usecl in the:ir l9 preparation. When a silica sand is ernployed, it is generally necessary to use a sand Or high purity. In the past~ when 21 silica sands of lower purity were used, it was necessary to 22 add large amounts of binder to ensure structural integrity 23 of the mold. This was not only costly but led to other 24 undesirable results when gaseous decomposition products of the excess resin penetrated into the molten or solidifying 26 m t l resu~ting in yinholes lnd scarring of the metal shspe.

9~

_ ll Impure silic2 sands, such 2S lake and bank sands, 2 ¦ are readily available in many areas of he United States.
3 These impure sands are sometimes beneficiated by various 4 processes such as water washing. However, it is still necessary to use excess binder with the washed sands to 6 obtain t;he desired tensile strength Or the cores and molds 7 made from them. It is therefore desirahle to develop a 8 process whereby these inexpensive sands can be used to make 9 foundry cores alld molds wi.thout the necd to use excess b:inder 1~ ~ tl~ s~ d.
:1.1 12 Bushey described a method for treating zircon--13 containing sands, U.S. Patent 4,115,345~ and olivine sands, lLI U.S. Patent 4,154,894, with an alkali metal silicate to improve the tensile strengths o~ resin shell molds or 16 cores made from the sands. Hcwever, he reported that when 17 this method was used with si.lica and chromite sands, no 18 improvement in the tensi].e str7ength Or t;he cores and Inolcls 19 was observed.

21 A process has now been discovered which permits 22 the use Or impure sil.ica sands in conjunction with moderate 23 amounts of binder to form foundry cores and molds with 24 improved tensile strength. This process is less expensive thall present beneficiation methods and gives cores and molds -26 with improved tensile strengths.

3 -4- __ ~, . ..

l A further unexpected benefi~ of ~lsing these treated 2 sands .is t~at cores prepared from t}~em by the b2se-curing 3 "no-bake" process are more readily released from the core 4 box. Easy release Or the cores is commercially important, since sticking cores slow down i,he core-making process and 6 often become broken and useless.
8 SUMMARY OF THE INVE~TION
In accordance W:it}l thi.s :imven1.:Jon, therea is ll provi.deA a process for th~:? prepar'.ll;i.on o:f` t~"eal;~i'd ~:Li.i.ca 12 sand which is useI`ul for formillg foundry cores and molds 13 having improved tensile strength. The ~rocess comprises 14 treating an impure silica sand with an aqueous solution of an alkali metal silicate and heating the mixture of sand 16 and silicate.

18 Adclitionally,~in accordance with this invention, 19 there is provided a process for the pr~paration Or a molding composition useful for forming foundry cores and 21 molds having improved tensile strength. The process 22 comprises treating impure silica sand with an aqueous 23 solution of` an alkali metal silicate and heati.ng t~ie 24 mixture of sand and silicate. The treated sand i.s then mixed or coated with an effective bonding amount of a 26 binder selected from the group consisting of shell resins, 27 base-curing "no-bake" resin compounds and core oil mixes. .

' ... .... ..'. ~

( 1 ?urthermore, in accordance with this invention, 2 there is provided silica four!dry sand useful for making 3 foundry cores and molds with i~proved te~sile strength.
~ This is prepared by treating impure sili^a sand with an aqueous solution of an alkali metal sili~ate and heating 6 the mix'ure of sand and silicate.

8 Finally, in accordance with th~s invent:ion, 9 there is pr~ovided a rnolding compositiorl use:ru]. for ~.n p~p~r illK r~O~,n(~ry (~ eS ~ rno~ 3 llav~ r~ vt~ ?llS~
ll strength. Th-is cornpos:i~iorl colllp~lise-s an 9rlL)u:Le sili.ca 12 sand, which has been treated by heating l~ith an aqueous 13 solution of an alkali metal silicate, and an effective 14 bonding amollnt of a binder. The binder is selected from the group consisting o~ shell resin~ base-cur:~ng 16 "no-bake" resin components and core oil ~ixes.

18 DF.TAILF.,D DESCR:r.~:'T:[C)N C)~ L'III', INV:E~lr~':.l-C) Any impure silica sands may be used in the practice 21 of this invention. Examples of such sands are lake and bank 22 sands which generally consist of from a~ut 85% to about 98%
23 by weight of silicon dioxide and small a~ounts of` sllch 24 impurities as aluminum oxide, iron oxide~ alkaline oxides and alkaline earth oxides. The impure s~ica sand can be 26 a naturally--occurring silica sand or a m~ture Or various 27 silica sands. The processes of this inve~tion are useful ~9 ;

¦1 ir the sand or mixture of sands contain ~ess than about 2 ¦ 99~D silicon dio~ide.

4 Commercially available lake and bank silica sands include 20KK Sand, available from the Martin Marietta 6 Corporation, Bridgman, Michigan; Ludington Sand, available 7 from the Sargent Sand Company, Saginaw, ~qichigan; Muskegon 8 Sand No. 850 and Beneficiated Muskegon Sand W~51, available 9 ~rorn the Nugent Sarld Company, Muskegon, ~1-ichigan; and ].0 Vassclr ,',and, available from the Sar¢ent; Sa1ld ~omparly~
11 Sa.g:irla~ M:ichi.~ rl.

13 In the process of this invention, the impure 14 ~ silica sand is treated with an aqueous solution of an alkali metal silicate. Treatment may be carried out ~y 16 stirring a slurry of the sand in a dilute silicate solution.
17 It is often satisractory to treat the sanA with a more 18 concerltrated silicate solution by placin~ the sand in 19 a rnixer and adding the required arnoul1t of s~ icate solution to the sand with mi~ing. Alternatively, the silicare solution 21 may be sprayed onto a thin layer of the sand.

23 Any alkali metal silicate, such as sodiurn and 24 potassium silicate, can be employed in the process of this invention. Solutions of sodium silicate are commerclally 26 available. Such solutions contain varying ratios of 27 sodium oxide to silicon dioxide. These ~eight ratios may 3 -7- __ ~ ,.
. .

Ij ) 1 ll vary fro~ 1 to 4 ~arts of silicon dio~ide per 1 part of 2 sodium oxide. The amount of water present in the alkali 3 ¦ metal silicate solution is not critical. However, 4 ¦ sufficient water should be present to permit adequate dispersion of the silicate over the surface of the sand 6 grains.

8 The amount of alkali metal silicate used with a 9 given sand should be an arnount that effecti.vely imparts the desi.red strength to the cores OI' mo].ds without .
11 i.nter~Ierintr w.i.th the f.ree I.l.ow:i.nrr properl;ies o.~ the ].2 s:i.]lcate-l;rea.~ed sarltl. l:~ :1s ~ re.t~.rl,7ed to U'iC .E'I'Oln abOllt 13 0.2 to about 1.1 g of silicate on a dry sol:i.ds basis per 14 kg of sand. .
.5 16 hfter the silica sand has been thoroughly rnixed .
17 with the silicate solution, the sand may be iso].ated from 18 the slu.rry by any conventiona]. means such as decantation 19 or filtration. However, whell the more concentrated solutions of silicate are employed, no rnechanical separation 21 of the sand from the silicate solution is requi.red. It is 22 only necessary to heat the sand to about 100C, or above, 23 for a short period of time to evaporate a portion of the 24 water and provide a free-flowing sand for use in the coating process. This simplifies the process by avoiding a 26 decantation or filtration step.

3 -8- _ _ ~ ,, ' . .

Al-ternatively, the sand can be preheated before the silicate solution is added to it. Mixing is -then continued until the water is evaporated.

The silicate--treated silica sands of this i.nvention are used to make foundry molds or cores using the procedures practi.ced with pure silica sand. In generall these processes involve mixincJ the sand w:ith t-~:E:Etect:ive borld~ g amoullts oE
b.Lndters. Usua:l.ly, l:ht-' components oE t.htl. b~ 5tlees aro co,lted on the sand to insure the:ir un:i:Eorm cllstr:ibllt:i.orl, Details of the preparation and use of resin-coated sands in the shell molding process are given in U.S. Patent 3,833,095. Illustrative of "no-bake" processes, using base-curing polyure-thane resin components, are U.S. Patents 3,409,579 and 3,429,848. The use of core oil m:i.xtures as :Eoundry core b:inders is described in U.S. :Patent 2,375,073.

Suitable resins for use in the shell-moldlng process include phenol~formaldehyde novolak resins which become -thermosetting when hea-ted in the presence o:E a curing agen-t. Hexame-thylene-tetramine is a satisEactory curing agent for these resins. Single-stage phenol-:Eormaldehyde shell resins which require no added curi.ng

5' ,. ,~
/'''~
'.;: ,~,.~

- j! agellt can also be used. ~oundr~ sanc, l~hich has been ccateu ~ I or ~rixed ~ith resin is pldced in 2 nlolc ana heeted to cause 3 ¦ the resin to haràen rormin~ a shell of resin-bonded sand.
4 ~-lhen the silicate-treated silica sand of thls invention is used as the sand component in the mold, the resulting mold

6 shows considerably improved tensile strength over the molds prepared using untreated impure silica sand at the same resin 8 loading.

Re;ill cornponent.s useflll in the no-ba~;e~ ~roces~ ilr~
1] polyols aorl pvlyisoc~yilna~es- A var:iety o;t ply~s c;ln i 12 used~ but resole-type phenolic resins are often employed.
13 These are usually dissolved in a solvent mixture and mixed llJ with the sand. Polyisocyanates, either as liquids or in solution are also added- Then a basic catalyst i;s added to 16 the mixture just before it is p~aced in the mold. It cures .
17 without heatirlg. Tertiary amines are commonly useà as the 18 basic catalysts. l~hen t;he sil:icate-treated silica sand 19 of this invention is used -in the ~ase-catalyzed "no-bake"
process, the resulting cores show better tensile strength 21 and better scratch hardness than do cores prepared from 22 untreated impure silica sand- Cores prepared ~rom the 23 treated sand are also ea5ier to remove from the core box.

~4 3 10- --~~
~ , ' . .

It is ol~en tne practice in the foundry ar~ to ~ ¦ include G varietv ol` additi~es in the resins used to 3 ¦ prep2re foundry cores and molds. These additives include 4 such ma~erials as silanes, sources of fluoride, deodorizing agents and the like. Such additives may be used with 6 resins in the present process and do not interfere

7 with the improved tensile strength of the cores and molds

8 obtained from the sands of this invention.

The I`oll owinfr e.xa1-rlples illustrate t,he :inver]t;:i.on.
11 lt is to be understoocl that the e~:alnples are ill.ust:ratlv~
12 only alld do not intend to llrnit the invent:loll in any way.
13 In the examples, all parts and pe-rcentages are by weight 14 and the temperatures are degrees centigrade unless otherwise indicated. All tensile strengths are given in p~unds per 16 square inch (psi).

18 E.XhMPL~_l 19 .
An aqueous solution contai.ning 2.8 g/l of sodium 21 silicate was prepared by mi.xing with 10 1 of water 73 g Or 22 a sodium silicate solution available from the Diamond . ..
23 Shamrock Corp., containing 9.1% by weight of Na2O and 29.2%
24 by weight of SiO2. Five kilograms ol 20KK silica sand was added to the silicate solution a.nd the mixture was 26 stirred for 40 minutes. After stirri.ng was stopped, the 27 sand was allowed to settle for 30 minutes before the 3 ~ . . -- .
~, "

213~ ~

I liquid was decanted. The sand was then dried at 121C
2 11 overni~ht. A l-kg sample ol tne trea-ea sand WGS neatea 3 ~o 128~C and added to a Hobart i~i~er. After 30 g Or comrnercial novolak roundry resin was added to the mixer, 1 5 the mixture of resin and sand ~as blended for 90 seconds 6 to melt the resin and coat it onto the sand. Then 14.4 ml 7 of a 27.6% solution o~ hexamethylenetetramine in water 8 was added to the mixer. Blending was continued until the

9 mixture broke up into free-flowing grai,ns of resin-coated sand.

]2 This procedure was repeated using I,udington, 13 Beneficiated Muskegon W/51 and Wedron 7020 silica sands.

Cold tensile and hot tensile strengths of test 16 specimens made from each of the coated sands were measured 17 as fo]lows: , 19 The hot tensile strengths were determilled by use of a Dietert No. 365 Hot Shell Tensile Tester. Tests were 21 run at 232C with a 3-minute cure time.

23 , The cold tensile st:rengths were determined by 24 making 1/4-inch thick "do~-bone" test briquets in a , Dietert No. 363A Heated Shell Curing Accessory. The 26 test briquets were cured for 3 rninutes at,232C and 27 allowed to cool to room temperature. The cold tensile 3 -12- ~=
., , ', ..

~ ll strength of each ~riquet was determined by using a 401 2 I ~ni-~/ersal Sand Strength Tester in the rnann~r set forth 3 by the American Foundryman's Society.

Results of tests using the various silica sands 6 are given in Table I.

8 CO~TR_L TEST I
9 ~ .
The untreated sands used as starting materials 11 in Example 1 were coated with novolal{ resin according to 12 the proceclure o:~ ExalllpIe .1,. Tlle hot an~l co:l.d tensilc 13 st;rerlgths of' cores ma(1e rrc)ln thec;e res:in~-coatecl .~ r-ld~s 14 ~Jere likewi.se tes~ed by the procedure oI' that exarrlple.
. Results of these control tests are given in Table I.

17 CO~TROL TEST 2 18 _ __ __ 19 EaGh of the sands used in Exarnple 1 was washed aild dried using tlle same general procedure o~ Exalnple 1 21 except that no sodillln si].icate was added to the washwater.
22 The washed sand was coated with novolak resill follo'wing 23 the procedure of Example 1, and hot and cold tensile 24 strengths were determined ~or cores made from these resin-coated sands. Results of these control tests are 26 given in Table I.
27 .

3 _1~_ .
. ' ` . .

~ ~3~3~ , ~ hese results show that impure silica lake sands 2 1l ~ive fourdry cores and rnolds with improved tensile strengths ~ if they are treated with a silicate solution before they 4 are coated with a ~oundry resin. In contrast, cores and molds made from resin-coated, silicate-treated pure silica 6 sand show no improvement in tensile strength over those 8 prepared from untreated pure silica sand.

T.~ 3LE
2 ..
O~re Prc)ptrties 3 Hot Cold Tensile Tensile ~ Sand Type Treatment ~psi~
20KKa) Untreated (Control Test 1) 27~ 400 l~ater washed (Control Test 2) 363 q59 6 Silicate treated 432 525 7 Ludingtonb) Untreated (Control Test 1) -l90 230 Water washed (Control Test 2) 230 250 8 Silicate treated 335 345 9 BeneFiciated Untreated (Control Test 1) 297 353 Mllskegon W/51C) l~ater washed~Control Test 2) 2t34 3~
Silicate treatecl 377 fi50 :ll Wedron 7020~l) Untreated (Cont:rol Test 1) 352 465 ~later ~ashed (Control lest 2) 304 500 12 Silicate treated 300 500 a) A lake sand available from the Martin Mariet~a Corp., Bridgman, 14 Michigan, containing about 94% SiO2 and smaller amounts of A1203 plus alkaline oxides and alkaline earth oxides.

b) A lake sand available from the Sargent Sand ~n., Saginaw, Michigan, 16 containing 96.2% SiO2 and smaller amounts of Fe203 and A1203 plus alkaline oxides and alkaline earth oxides. The untreated sand ]7 contained 7.3 ppm (parts per million) sotl;ul-n; the silica-treated 18 sand contained 94 ppm soclillrn.
c) A washed and dried lake sand available From the Nugent Sand Co., 19 Muskegon, Michigan, containing about 95% SiQ2 and smaller amounts of A1203 plus alkaline oxides and alkaline earth oxides.
d) A pure silica sand available from the Martin Marietta Corp., 21 ~ We r n, Illinois, containing over 99.8% SiO2.

_lr_ 2 _Y~M?L_ ~ , 3 An a~ueous solution of sodiu~ silica~e was prepared 4 by adding 12.6 g of the co~ ercially available sodium silicate solution used in Example 1 to 200 g of wateI. A mixture of 6 25.7 g of the silicate solution and 1100 g of 20KK sillca 7 sand (0.53 g sodium silicate per kg sand) W2S mixed in a 8 Hobart MiY.er at, room temperature for 12 minutes before it 9 was dried overnight at 232C. Gne thousalld grarns Or the t;reated sarld was coatecl with 30 g oI' p~-lcllo]ic no~olak :Ll rc~ in .~ L~C, ~nc~ rll o~ ,~ 27.G% }~ y:].~ J~
12 solution was added according to the procedure of Example 1.
13 Hot and cold tensile strengths were determined for cores 14 prepared using the resin-coated,sand.

16 This procedure was repeated using Muskegon 850 17 and Vassar sillca sands.

19 Reslllts of the tests are reported in Table II.

21 For control tests, untreated 2OKK, Muskegon 850 22 and Vassar sands were coated with phenolic novolak resin and 23 he~amethylenetetrarnine solution. Hot and cold tensile 24 strengths were then measured on cores prepared ~rom these coated sands. The results of these control tests are also 26 reported in Table II.

3 ' -16-. ' "

3~
I ' li , - T

_Oo~^e Pr~perties_ 4 Hot Cold Tensil eTensile Sand Type Treatrnent 1~5~;~ (psi) 20KKa) Untreated (Control) 251 278 Silicate treated 373 381 7 Muskegon 850b) Untreated (Control) . 242 299 8 Silicate treated 303 350 9 Vassar SandC) Untreated (Control) 165 215 Silicate treated 213 257 11 a) A lake sand available from the ~artin ~arietta Corp., Bridgman, l~ichigan, containing about 94% SiO2 arlcl slnal l~r alnounts of 12 A1~03 l~lus alkaline oxides and all;aline earl:~ oxicles~
13 b) A bank sall(l conl:ainin9 about 91% SiO~ arl~l srnall~r amOllll~.S oF
A1203, Fe203, and alkaline oxides avc;ilable frorn l:he ~uycrlt 14 Sand Co., Muskegon, Michigan.
c) A bank sand available from Sargent Sand Co., ~;agina!~l, Michigan, containing about 90% SiO2 and smaller amounts of A1203, Fe203, 16 ~ alkaline oxides and alkaline earth oxides.

19 This experilnent demonstrates that sllica sands can be treated wi.th a si].:i.cate solut;1.on -ko give ir~lproved 21 :~oundry sands and that i.t ls unnecessary to sepa:rate the 22 silicate solution mechanically ~rom the ~reated sand.

3 . -17.- _ . ,.
. .

3~

3 Sand mixtures were prepared l~sin~ various proportions 4 of Wedron 720, a pure silica sand, and 20KK~ a lake sand containing about 94~ silicon dioxide. The mixtures, which 6 contained from 96.4 to 99.~% silicon dioxide, were trea~ed 7 with sodium silicate solutions by the procedure of Example 2.
8 Both treated and untreated sands were coated with novolak 9 resin according to the procedure of E~amp].e 1. Hot and cold tensile strengths ~ere measured on cores prepared ~`rom these 11 coated sands. Results of these tes-ts sho~Jed that sillcate 12 treatlllent i.s er~ective ln irnprovlng terlsi.le propert.ies Or 13 cores made ~rorn sancls containinf-~ less than about 99% s~ll.con 14 dioxide.
:L6 EXAMPLE 4 18 The general p.rocedure of Example 2 was repeated 19 with 20KK silica sand using amounts of sodium silicate varying from 0.11 to 1.79 ~ of sodium si]icate per kg of sand.
21 Hot and cold tensile strenet}ls were obtained .ror cores 22 prepared from silicate-treated sarlds which had been coated 23 with novolak resin. These tests showed that the impure sillca 24 lake sand gave foundry cores with improved tensile strengths if the sand was first treated with between about 0.2 g and 26 ~ ~.1 g of s ium silicate per kg Or sand.

1 EXA;~I`L_ 5 3 In this e~periment, 45.5 kg of 2GKK bank sand was 4 placed in a cement mixer. To the mixing sand was added an aqueous solution of sod-i.um silicate prepared by mixing 6 63 g of the commercially available sodium silicate solution 7 used in ~Y.ample 1 with 1000 g water. Mixing was continued 8 at room temperature for 90 seconds before a gas flarne was 9 applied to the mixture. Heating was continued unti] the temperature of the mixture reached 166C. The hot treated 11 sand was transferred to a Mull.er Mixe:r and coatecl with ~2 phel1olic novo:l.ak resin at 1 28C u~:,i.ng the ,sarne rel.atlv~
:1.3 l~oporl;ior]s oI':resl.n, hexalnet;hyl.enetetrarlli.ne and ~;rlrld ~s 14 used in E~ample 2.
16 In a control experiment, un~reated 2 OKK ~ank sand 17 was heated to 180C, transferred to.a Muller Mixer and 18 coated with phenolic no.volak resin by the same procedure 19 used to coat the treated sand.
21 Cores were prepared frorn the t:reated c:oated sand as 22 well as from untreated coated sand which was used as a 23 control. Hot and cold tensile strengths of t~e cores were 24 measured by the standard p:rocedures. Silicate-treated coated sand gave cores which sho~ed a hot tensile stren~;th 26 of 468 psi and a cold tensile strength of 471 psi. These 27 values compared ~ith a hot tensile strength of 336 psi and 28 a cold tensile strength of 362 psi for cores prepared from 29 the untreated coated sand~
-19- _ ' '.....

: i l, ¦ Tnis e~periment sho~^~s that the procedure of this ¦ invention is readily scaled up to a commercially acceptable 3 process without the need for rnechanical separation Or the 4 silicate solution from the treated sand.

8 An aqueous solution of sodiurn silicate was prepared 9 by di].uti.ng 17 g of a sodi.urn silicate soluti.on available :from LO thf' D:i.alnol-ld '~hamrock Corp. corJtaini~g G. 7% N~20 al-ld 25.~3%
.1.1. Si.02 with 196 gr c~ 'lat~.?:r'. 'rh:i.S 5011.1t;:i.0n ~t:l'i li,S~.`d tC~ t;l='l~?al;
12 Vassar sand according to the procedure O:r Example 2 and test 13 cores were evaluated as described in that example.

Foundry cores prepared with silicate-treated sand 16 showed a hot tensile strength Or 253 psi and a cold tensile 17 strength of 270 psi. These values compared with a hot 18 tellsile strength of 165 psi. and a col.d tensi.le s~rengt}l of 19 215 psi for the control sand which had not been txeated w:i.th silicate solution.

22 _XAMPLE 7 - ..
23 .
24 A mixture of 1000 g of 20KK silica sand, treated with sodium silicate solution as in Fxample 2 and 30 g of 26 701 Liquid Shell Resin (a single-stage shell resin solution 27 available from the Acme Resin Corporation Forest Park 3 -20~ = .
.. ..

` ll Illlnois, havingr a pH of 3.5 ~o 4.5, a viscosity a~ 25~C of 2 3500-4500 cps and a solids conten~ of 72% to 75% by weight) 3 ~as mixed in a Hobart Mixer for 3 minuves at 149C. Then 4 14 ml of water was added .,o cool the coated sand and cause the sand to break up into individually coated grains.
After the individual grains had forrned, 1.2 g of calcium 7 stearate was added and mixing ~as continued for 1 minute.
8 Hot and cold tensile strengths of test specimens prepared 9 from the sand were determined by the procedures described in Exarllple 1. The hot tensile strerlgth of the .specirnerls 11 was 1ll0 psi and the cold tellsiLe streng~ was lllO ps~.
.1.~
13 Control tests performed using untreated 20KK sand 14 gave specimens showing 100 psi cold tenslle and 270 psi hot tensile strengths.

17 These results show that co:res prepared using 18 silicate-treated sand coated with sing]e-stage shell ]9 resins have improved hot and cold tensile strerlgth over cores prepared from untreated impure sand.

22 _AMPLE 8 24 This is an example of a "no-bake" foulldry process.
Silicate-treated 20KK bank sand was prepared as in Example 5.-26 TQ 2500 g of the silicate-treated sand in a K-45 Kitchen 27 Aid Mixer was added 17.2 g of Acme Bond 5022 polyol, 14.1 g 2~

.~ ,.

~ r3~

1 of ~cme ~ond 506~ polyisocyanâte and 0.63 g Or ~cme Bond 5082 2 basic catalyst. The Acme Bond componenls are available~ from 3 the Acme Resin Corporation ~orest Par~ Illinois Sand and 4 resin comDonents were mixed for 1 minute and discharge-d into a Dietert No. 623-50 pyramid core box. The sand was jolted 6 4 times using a Dietert No. 623 core box jolter. A thermometer 7 was inserted about 6 inches into the core. The stripping 8 time is the time it takes to cure the core so hard that the 9 thermometer can no longer be pushed by hand deeper into the core. Strip time was determined to be 5 minutes 15 seconds.
:1.1 ., 12 A s~corld ide?-lt;ical -lnd-re;:in lllix ~as prep~r-ed an~l 13 discharged i.lltO a Dietert No. 696, l2-frJlrl~ tensilc COl'e box 14 to prepare 12 standard Arnerican Foundrymen's Society l-inch dog bone tensile briquets. The cores were cured at room 16 temperature and broken after 1 hour and 24 hours. Humidity 17 testing was carried out by placing tensile briquets in `80~1O
18 and 90% relative humidlty (r.h.) chambers for 24 hours before ]9 determining tensile strengths. The tensile strengths were Jneasured usirlg a Detroit Testirlg Machine Co. Model SCT Te~ter, 21 and scratch hardness was determined using a Dietert; No. 674 22 scratch hardness tester. Results of the tests are summarized 2 5 , 1 n Ta D ` I I .

30 ~ -22- ~

3;~
.

1 .r~S a coi1trol, the above procedure ~as repeated using 2 untreated 2GKK lake sand with tne s~me a~ount of resin 3 components except that 0.75 ~ of the Acme Bond 5032 c?~talyst 4 ~1as used. In this case, a strip time of 5 minutes 30 seconds was obtained. Results o~ the other control tests are given 6 in Table III.

8 TA~LE III
9 _ Tensile, ps~_ and IScratch_Hardness~_ lOCores Prepared 24 hrs 24 hrs From 1 hr 24 hrs~ ~O',Or.h.~ 90% r.h.
l.l1l~ea-t,ed S~nd1~7 (64)267 (72)2l7 (71).157 (61) 12Untre~te~ Sand120 (62)183 (70)20t) (70)123 (64) l3(Control) These results show that cores prepared rrom the 16 silicate-treated sand by a base-catalyzed "no-bake" process 17 generally give irnproved tensile strength and better scratch l8 hardness than the co:res,prepared from untreated impure sand.
].9 Cores prepared from t.reated sand also gave :i.mproved release ~rom the core box. This property .is beneficial 21 because stickin~ to the core box slows production in a 22 foundry and can result in core or mold damage during 23 ~ remov rrom the pattern.

2~

i3~ 1 ll E~ PL~ 9 3 .~ mixture of 4000 g of 20KK silica sand, treated 4 with sodium silicate solution as in Example 2, and 40 g of powdered corn cereal was mulled in a Simpson Mix-Muller 6 (18-inch model) for 1 minute. Then 80 g of water was 7 added and mulling W2S continued for an additional 4 minutes.
8 Mulling was stopped and 20 g of foundry core oil~ obtained 9 from the Archer--Daniels-Midland Cornpany, Minneapolis, Minnesota, was added. The mixture ~as mulled for 1 mi.nute 11 and collected in a polyetllylene ba~. I'he bag was seal~d 12 :ilmnecliately t;o minimi.ze COll~;clC~ with t,he tl i.r~.
:l3 14 Green compression strength of the coated sand was determined by placing 168 g of the material in a Dietert 16 r~etroit No. 315-9 specimen tube. The specimen was rammed 17 three times with a Dietert Detroit No. 315 sand rammer.
18 The resultin~ 2-inch x 2-inch test cylinder was compressed 19 in a Dietert Detroi.t No. 465 compression instrllmellt to deterrn:ine the green cornpression strength.

22 Baked tensile strength specimens were prepared from the coated sand by placing the sand in a tensile specimen 24 mold and ramming it four times with the Dietert Detroit No.
315 sand rammer. Specimens were placed in a tray in a 26 circulating air oven at 224~C. Specirnens were remo~Ted from 27 the oven at varying times. After the specimens had cooled g~

l, l ~ I ~o room temperature, their tensile str~ngths were measured Z ¦ usin~ a Detroit Testing Mzchi~e, Model CST, ~ensile tester. 1, 3 ¦ Each value reported is the average of the strengths measured 4 using three specimen~.

6 For comparative tests, spec~mens were prepared from 7 coated 20KK sand that had not been trea~ed with s,ilicate solutions.

The results given in Table IV show tha-t cores made 11 ~rom silicate-treated sand coated w:i.th a core oil mix exhibit 12 about 25% greater tensile strength tharl ~o cores rnade f'rorn 13 uncoated sand whell ~he cores a:re ba.ked S'or 30 rn.i.nu~es.
11~
TABLE IV

~6 Tests on Specirnens From Control 18 Silicate-lreated 5and Tests ( P~
Green Compress;on 0.5 0.45 Baked Tensile Strength 21 (Baking Time, mirl) 22 30 225 l80 23 45 22l5 -l807 `25 28 _ -Z5- _ .~ ,.

1 Thus, it is apparent that the~e has been provided, 2 in accoraânce with the invention, a process for the 3 preparation of resin-coated silica. sanQs that fully 4 satisfies the objects, aims and advantâGes set forth above. While the invention has been described in 6 conjunction with specific ernbodiments Ihereof, it is 7 evident that many alternatives, modifications, and 8 variations will be apparent to those skilled in the 9 art in ].ight of the foregoing description. Accordingly, it`is intended to inc]ude all such alte~Anatives, 11 modifications, and variations as s~t ror~h ~li.th:i.rl th~

13 Spl t ~nd soope of lhe app~ ed cla:~me.

' 16 3 ~ -26- . .

Claims (23)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A process for the preparation of treated silica sand useful for the preparation of foundry cores and molds having improved tensile strength which comprises treating impure silica sand containing from about 85% to less than about 99%
of weight of silicon dioxide with a solution consisting of an alkali metal silicate and water and heating the mixture of sand and silicate to give a treated silica sand containing from about 0.2 g to about 1.1 g of silicate per kg of sand on a dry solids basis.

2. The process of claim 1 wherein the alkali metal silicate is sodium silicate.

3. The process of claim 1 wherein the impure silica sand is selected from the group consisting of 20KK lake sand, Ludington lake sand, Muskegon bank and lake sands, and Vassar bank sand.

4. The process of claim 1 wherein the silica sand is separated from the aqueous solution of an alkali metal silicate before the mixture of sand and silicate is heated.

5. A process for the preparation of a molding composition useful for forming foundry cores and molds having improved tensile strength which comprises treating impure silica sand containing from about 85% to less than about 99% by weight of silicon dioxide with a solution consisting of an alkali metal silicate and water heating the mixture of sand and silicate to give a treated silica sand containing from about 0.2 g to about 1.1 g of silicate per kg of sand on a dry solids basis, and mixing or coating the treated sand with an effective bonding amount of a binder selected from the group consisting of shell resins, base-curing "no-bake"
resin components and core oil mixes.

6. The process of claim 5 wherein the alkali metal silicate is sodium silicate.

7. The process of claim 5 wherein the impure silica sand is selected from the group consisting of 20KK lake sands, Ludington lake sand, Muskegon bank and lake sands, and Vassar bank sand.

8. The process of claim 5 wherein the silica sand is separated from the aqueous solution of an alkali metal silicate before the mixture of sand and silicate is heated.

9. The process of claim 5 wherein the binder is a shell resin which further comprises the curing agent hexamethylenetetramine.

10. The process of claim 5 wherein the binder consists of "no-bake" resin components which comprise a polyol and a polyisocyanate.

11. The process of claim 10 wherein the resin components further comprise a tertiary amine.

12. The process of claim 5 wherein the binder is a core oil mix comprising a drying oil and a cereal binder.

13. A silica foundry sand useful for the preparation of foundry cores and molds having improved tensile strength prepared by treating impure silica sand containing from about 85% to less than about 99% by weight of silicon dioxide with a solution consisting of an alkali metal silicate and water and heating the mixture of sand and silicate to give a product containing from about 0.2 g to about 1.1 g of silicate per kg of sand on a dry solids basis.

14. The product of claim 13 wherein the alkali metal silicate is sodium silicate.

15. The product of claim 13 wherein the impure silica sand is selected from the group consisting of 20KK lake sand, Ludington lake sand, Muskegon bank and lake sands, and Vassar bank sand.

16. The product of claim 13 wherein the silica sand is separated from the aqueous solution of an alkali metal silicate before the mixture of sand and silicate is heated.

17. A molding composition useful for the preparation of foundry cores and molds having improved tensile strength comprising an impure silica sand containing from about 85% to less than about 99% by weight of silicon dioxide, previously treated by heating with a solution consisting of an alkali metal silicate and water to give a treated sand containing from about 0.2 g to about 1.1 g of silicate per kg of sand on a dry solids basis, and an effective bonding amount of a binder selected from the group consisting of shell resins, base-curing "no-bake" resin components and core oil mixes.

18. The composition of claim 17 wherein the alkali metal silicate is sodium silicate.

19. The composition of claim 17 wherein the impure silica sand is selected from the group consisting of 20KK lake sand, Ludington lake sand, Muskegon bank and lake sands, and Vassar bank sand.

20. The composition of claim 17 wherein the binder is a shell resin which further comprises the curing agent hexamethylenetetramine.

21. The composition of claim 17 wherein the binder consists of base-curing "no-bake" resin components which comprise a polyol and a polyisocyanate.

22. The composition of claim 21 wherein the resin components further comprise a tertiary amine.

23. The composition of claim 17 wherein the binder is a core oil mix comprising a drying oil and a cereal binder.