Recent studies comparing coremaking processes.

Research on core binders reveals the wide variety of choices available today as well as critical factors that must be considered.

The demands made on coremaking processes have risen sharply during the past few years, and in many cases have undergone significant changes. Any technical and economic assessment of coremaking processes, today, must consider the entire manufacturing cycle, beginning with raw material sources and availability, up to the disposal of production wastes.

Published results from the Foundry Institute of the RWTH (Rheinisch-Westfalische Technische Hocschule) Aachen[1, 6, 7], a comparative study of the INASMET Institute of San Sebastian, Spain[2], assessments of the competition in foundry chemistry[3] and work by two of the authors[4,5] are considered in this paper.

Coremaking Variety

The metalcaster is sometimes overwhelmed by the wide variety of coremaking processes. Whichever method is used, however, the foundry should expect its cores to have good dimensional stability. At the same time, they should exhibit high strength levels during intermediate handling and must temporarily withstand high temperatures during the pouring operation. The binder should also exhibit complementary compatibility with modifying agents, be capable of forming an acceptable level of carbon under pyrolysis conditions and exhibit a certain chemical reactivity.

Thus, the variety of coremaking processes is also related to the reactivity of the resins (as shown in Table 1 using phenolic resins as an example). Phenolic novolaks may be reacted with hexa, and phenolic resoles can undergo reaction with acids, form polyurethanes with diisocyanates, or participate in precipitation (hardening) reactions with methyl formiate or carbon dioxide. Such transformations already form the basis for various coremaking processes.

Table 1. Phenolic Resins: Reactions and Application in Coremaking Processes

Phenolic Resin       Reacts With      Process
Novolak              Hexa             Shell Molding
Resole, Neutral      Acid             Nobake
Resole, Neutral      Acid, Heat       Hotbox
Resole, Ortho        Diisocyanate     PUR Coldbox
Resole, Alkaline     Ester            Ester Nobake
Resole, Alkaline     Ester Aerosol    MF Coldbox
Resole, Alkaline     Carbon Dioxide   [CO.sub.2] Coldbox

These remarks on phenolic resins apply analogously to furan, urea, epoxy and acrylic resins as well. Reactions of the latter resins also form the basis for diverse coremaking processes. In this article, comparative considerations concentrate mainly on heat- and gas-cured processes (Table 2). These processes include:[8]

Table 2. Highlights of Compared Coremaking Processes

Process                Binder Reaction:
Shell Molding          Phenolic Novolak + Hexa
  (Croning)
Hotbox                 Reactive Phenolic or Furan Resin + Hardener
Warmbox                Reactive Modified Furan Resin + Special
                         Hardener
PUR Coldbox            Resole + Diisocyanate + Amine
Gas Curing Epoxy       Modified Epoxy Resin + Oxidizer + [SO.sub.2]
  (Epoxy [SO.sub.2])
Methyl Formiate (or    Alkaline Resole + Methyl Formiate or
  Resole-[CO.sub.2])     [CO.sub.2]

Process                      Temperature           Abbreviation
Shell Molding (Croning)      392-500F (200-260C)   CR
Hotbox                       356-464F (180-240C)   HB
Warmbox                      302-392F (150-200C)   WB
PUR Coldbox                  Room Temperature      PUR-CB
Gas Curing Epoxy (Epoxy      Room Temperature      EGH
  [SO.sub.2])
Methyl Formiate (or          Room Temperature      MF (or "Resole")
  Resole-[CO.sub.2])

* shell, or Croning (CR);

* hotbox (HB);

* warmbox (WB);

* phenolic-urethane coldbox (PUR CB);

* gas curing epoxy (epoxy-[SO.sub.2] (EGH);

* methyl formiate or Resole-[CO.sub.2] (MF).

Core Sand, Bench Life

Various batch or continuous mixing processes may be employed to condition coremaking sand. The choice of (cold) mixing process may be limited by the binder level, since this varies considerably in the core sends for specific processes. For example, levels may be: 0.8-1.2% in the EGH process; 1.0-1.6% in the PUR-CB method; 1.3-2.0% in the WB/HB methods; and more than 2% in CR. The mixing process should ensure homogeneous distribution of the binder throughout the quartz sand even in the case of high-viscosity, solvent-free liquid systems.

The bench life graph shown in Fig. 1 exemplifies the unlimited bench life of CR molding sand and the relatively lengthy bench life of EGH sand. In the other processes, the bench life is limited due to ongoing transformations such as acid catalysis or the PUR reaction. This limitation is accompanied by correspondingly higher costs for mixing logistics and machine cleaning.

[Figure 1 ILLUSTRATION OMITTED]

Strength Levels, Moisture

The strength (flexural or tensile) of the cores provides information on the required level of binder and characteristics during coremaking, core handling, storage, and in the mold before and during the pouring operation. Figure 2 illustrates the initial and 24-hr strength levels as percentages based on an index of 100 (800 N/cm2) for the strength in CR cores. This comparison was based on realistic (material-dependent) binder levels and customary curing conditions. Thus, the initial strength in the case of the hot curing processes represents a hot strength level, and in the cold-curing methods, the strength level after 3060 sec. The dotted lines at an index of 25% and 35% (corresponding to 200 and 280 N/[cm.sup.2]) represent practice-relevant tolerance ranges for stripping and handling of the cores. This comparison was based on realistic (material-dependent) binder levels and customary curing conditions. Thus, the initial strength in the case of the hot curing processes represents a hot strength level, and in the cold-curing methods, the strength level after 30-60 sec. The dotted lines at an index of 25% and 35% (corresponding to 200 and 280N/[cm.sup.2] represent practice-relevant tolerance ranges for stripping and handling of the cores.

[Figure 2 ILLUSTRATION OMITTED]

In the CR process, the strength potential cannot be utilized due to the direction of the cure (delayed from the surface inward). The initial strength levels in the HB, WB and EGH cores are approximately equal, whereas they are slightly lower in the PUR-CB cores.

In the case of the MF process, the strength of the cores when stripped is in a critical range. Thus. it is difficult to use this process for complex cores. A similar constellation, with similar relationships, is obtained when the tensile strength is measured, although the polyurethane-bonded cores (PUR-CB) afford slightly higher values than epoxy resin-bonded systems (EGH) in this case. This result is associated with interfacial and systemic (elasticity) differences.

When the cores were stored at high humidity or coated with a water-based wash, the strength levels decreased by about 10% (CR), 50% (Furan-HB), [is greater than] 60% (WB), 50% (PUR-CB), 30% (EGH) and 20% (MF). The large drop in the case of the WB cores is associated with the curing systems. The figure for the PUR-CB process largely corresponds to current quality standards.

Krapohl[1] specified the percentages of binder required to achieve a strength level of 150 N/[cm.sup.2] in the gas curing processes (binder equivalents), and arrived at the following order of performance (from a low to a high binder demand): EGH [is greater than] PUR-CB [is greater than] FRC (free radical cure) [is greater than] MF [is greater than] acetal process. Boenisch[6] determined the initial strength levels of coremaking sand mixtures containing 1.5% binder and established that the strength levels fall in the following order: EGH (acrylic-modified epoxy resin binder) [is greater than] PUR-CB [is greater than] FRC (acrylic-rich binder) [is greater than] acetal process [is greater than] MF method.

Thermal Behavior

The BCIRA hot distortion test affords a reliable indication of the thermal behavior that may be expected. The results are shown in Fig. 3. The extent to which the curve rises characterizes the thermal rigidity, and the length of the time axis until the core breaks roughly describes the thermal stability. The CR plot may be regarded as an ideal curve. The degree of expansion is acceptable, and a relatively lengthy period of time elapses before the core breaks.

[Figure 3 ILLUSTRATION OMITTED]

An epoxy binder specially developed for enhanced thermal stability affords an EGH plot showing high thermal rigidity. The degree of expansion in the other three sand systems (PUR-CB, HB and MF) is acceptable. The lowest thermal stability is experienced in the PUR-CB system. In laboratory tests, the collapse of molding sand, especially coremaking sand mixes for aluminum casting, was determined at 1112F (600C) using three methods.

The special (high acrylic content) EGH binder for aluminum casting employed in these tests exhibited nearly 100% collapse by all three methods (even without shakeout). HB, CB and MF molding sands collapse when subjected to mechanical agitation. As expected, the CR sand exhibits the poorest (system-related) collapse. In the comparison of EGH and HB systems, however, it is important to know that HB binders can exhibit good collapse properties at considerably lower temperatures (less than 752F (400C)] when specially modified.

Gas Evolution/Burst

Continuous measurements of the amounts of gas evolved at 2012F (1100C) in Fig. 4 shows that it is proportionate to the weight percentage of binder (numbers in brackets). The results indicate the amount of evolved gas is lower in the PUR-CB and EGH systems. In parallel tests, measurements were performed at 1472F (800C) using the Bolz method[5]. Table 3 shows that the time at which the gas burst occurs can differ to a great extent depending on the binder.

Table 3. Mean Gas Burst Times A and B (in see) at 1472F (800C) by the Bolz Methods[5]

                A, sec      (%)      B, sec       (%)
EGH              36         (45)       66         (50)
PUR              54         (90)       --          --
Resole           20         (53)       53         (45)
CR               38         (33)       80         (65)
HB               42         (90)       --          --

[Figure 4 ILLUSTRATION OMITTED]

Gas Composition, Emissions

It's difficult to arrive at reproducible data on the composition of the gases produced during casting from the INASMET work[2]. In order to determine the trends, reference is made to another research project[9], in which cured resin systems were systematically carbonized and the resultant pyrolysis gases analyzed in parallel tests.

Table 4 shows the resultant trends in the case of phenolic and epoxy resin systems. All systems yield monomeric aromatic hydrocarbons and phenolic compounds in small amounts. A certain tendency to produce higher polycyclic aromatic hydrocarbons is present in the case of the novolakhexa (CR) binder.

Table 4. Trends in Compositions of Pyrolysis Gases from Binders

                   Monomeric aromatic     Phenolic       Dinuclear
                   hyrocarbons            Compounds      Compounds
                       (%)                  (%)            (%)
Novolak-Hexa           0.1                 5-10             1
Resole                 0.1                 3-8              1
Epoxy                  0.05                3-6              0.4
Epoxy-Acrylate         0.03                2-4              0.3
Polyurethane(*)        0.1                 5-10             0.5

                   Higher Polycyclic             Benz(a)pyrene
                   Automatic Hydrocarbons (%)       (%)
Novolak-Hexa                     0.3              < 0.01
Resole                           0.1              < 0.01
Epoxy                          < 0.01             < 0.01
Epoxy-Acrylate                 < 0.01             < 0.01
Polyurethane(*)                < 0.01             < 0.01

However, the levels of any benz(a)pyrene produced were below the limit of detection. As expected, nitrogen-containing compounds may be observed in the case of the PUR-CB binder. In epoxy resins, production of aromatic compounds is considerably reduced for structural reasons. Thus, measurements of the emissions developed during the pouring operation have shown that the benzene emissions from an EGH binder range down to as little as one-tenth of those produced in a HB system. A comparison of the emissions produced by the PUR-CB and EGH systems affords a similar result.

Casting Defects

Table 5 shows the trends in casting defects for various systems[2]; the data refers to uncoated molds containing no additives. Practice shows that inorganic additives hinder recycling of old sand and that organic additives can increase gas burst on pouring. This listing, in which formation of lustrous carbon and the tendency to erosion could not be included for reasons of test design, may be interpreted as a ranking of casting quality (with an index of 100 for the CR system) which falls in the following order: CR [is greater than] EGH [is greater than] MF-CB [is greater than] PUR-CB [is greater than] HB. Table 5. Trends in Casting Defects in Uncoated Molds[2] (Scoring: 1 marked 10 = none)

               PURCB    EGH    Resole-CB    CR    HB

Gas              8       9         6         4     5
Finning          4       6         8         8     7
Burnout          5       4         3         9     5
Penetration      8       7         6         8     6
Sum             23      26        23        29    23
Index           80      90        80       100    80

Despite all such considerations of casting quality, however, it should be kept in mind that there is also an economic side to good casting, which in the case of hot curing processes such as CR can include some other important considerations.

Used Sand and Recycling

The cited references[1, 6, 7] provide detailed information on specific systems. One noteworthy mention is the effect of various fresh coremaking sands on the strength levels of green sand described in one of the cited studies[1] end illustrated in Fig. 5. According to this data, molding sands derived from the CR, EGH and PUR-CB processes offer advantages that contribute to improved green sand properties without reducing green strength.

[Figure 5 ILLUSTRATION OMITTED]

Our own experience shows that removal of the fines or at least holding their level constant represents a key factor in problem-free use of recycled sand, particularly thermally regenerated sand, and that particular attention must be paid to the alkali level.

The Furan-[SO.sub.2] Process

The furan-[SO.sub.2] method[11] is the predecessor of the epoxy-[SO.sub.2] method with respect to the type of process--gassing with [SO.sub.2] in closed systems--and is based on the same principle (Table 2). Specially condensed furan resins are used in this process. Subsequent inclusion of this method in the previous comparisons results in the following analysis.

The furan-[SO.sub.2] method has offered significant advantages, such as high thermal stability, good core collapse and low levels of glossy carbon formation. These were earlier contrasted by the disadvantages experienced in corebox deposits (which can accumulate to such an extent that dimensions are distorted), exceeded tolerances, corebox sticking and plugging of the vents.

Conventional binders for the FGH (gas-curing furan resin) process contain up to 50% free furfuryl alcohol, which on the one hand produced the high reactivity of the binder, but on the other is also partially responsible for the corebox deposits. A marked improvement has been achieved by development of furan resins condensed in a completely different manner[12]. These are nearly free of monomer, but still exhibit adequate reactivity when cured by gassing. These new binders reduce corebox deposits and cleaning times to tolerable levels comparable to those experienced in other gassing processes.

The following remarks survey the properties of the furan-[SO.sub.2] molding sands compared to those of the previously cited methods.

Bench Life (Fig. 1)--The bench life--8-24 hr, depending on the mix--is slightly shorter than in the EGH method.

Strength Levels (Fig. 2)--The levels are high and, with indices of 35% and 70%, generously exceed the minimal requirements for stripping and handling. The high strength levels are also reflected by the low Krapohl binder equivalent[1] of 0.9-1.0% for 150 N/[m.sup.2].

Thermal Properties (Fig. 3)--The hot distortion curve approximately corresponds to that in the furan HB or MF-CB process. It documents the high hot strength levels, good dimensional stability during pouring, and finally the low levels of casting defects familiar in practice.

Collapsibility--The break-down properties following pouring are excellent, particularly in the case of light metals casting, and are generally better than those achieved using all the molding sand systems.

Gas Evolution (Table 3)--As is typical for furan resins, an early and nearly complete burst of gas is observed. As much as 85-90% of the total gas is evolved in the first 20 sec. The absolute quantity of evolved gas is greater than that experienced in the epoxy [SO.sub.2] method, but less than in the PUR-CB, HB and CR molding processes.

Pyrolysis Trends (Table 4)--The types of furan resins used as furan-[SO.sub.2] binders afford higher carbon yields (about 55%) and lower fractions of volatile decomposition products (less than 45%) than epoxies. The latter circumstance also explains the low tendency of furan-[SO.sub.2] binders to form glossy carbon.

As far as development of pollutants during pyrolysis is concerned, small amounts of phenol, aromatic hydrocarbons (toluene and xylene) and naphthalene may be expected on the basis of analyses performed using straight binder systems. Benzene, benz [a] pyrene and similar compounds couldn't be detected.

Effect of Old Sand on Green Sand (Fig. 5)--Based on the chemical composition of cured furan-[SO.sub.2] coremaking sands, which for example are slightly acidic, a similarly positive effect as that exerted by used EGH sand on green sand (increased strength levels) may be expected.

Trends In Casting Defects (Table 5)--When graded according to the scheme illustrated in Table 5, the furan-[SO.sub.2] process affords an index of around 95, exceeded only by the CR method. This is mainly due to the low levels of gas defects, burnout and penetration.

Summary

The demands made on coremaking processes have changed. Up to the present, the key features were nearly exclusively of a technological nature. Today, an evaluation must consider the total picture of raw materials, energy and ecological factors, as well as opportunities for utilization of by products. Bearing this in mind, the most widely used coremaking processes have been compared on the basis of contemporary methods employed in everyday practice.

Coremaking processes requiring the application of heat, in most cases can be replaced by methods that use gas curing. The furan-[SO.sub.2] process has recently been developed to a point at which its earlier weaknesses (corebox deposits and associated problems) have been effectively eliminated, the cores now exhibit improved thermal properties and a lower tendency to casting defects than ones made by the EGH process. The future of hot and cold coremaking processes will partially depend on raw material availability, but even more on environmental legislation and the utilization of used sand.

For a free copy of this article circle No. 344 on the Reader Action Card.

References

[1.] H.P.Krapohl, "Gashartende Giessereisandkerneein Eigenschaftsvergleich Von DerKernfertigung Bis Zur Regenerierung," Dissertation, RWTH Aachen Foundry Institute (1991).

[2.] L. Cobos, "Vergleichende Betrachtung Verschiedener Kernherstellverfahren," INASMET Centro Technologico de Materiales, San Sebastian, Unpublished Work (Fete 1993).

[3.] J. Urreiztieta, "Arenas Quimicas Para Moldesy Machos," Paper Delivered at the Fourth Congreso Iberico de Fundicion, Sevilla (1991).

[4.] A. Gardziella, "Chemisch Gebundene Formstoffe-Stand Der Technik Und Zukunftsentwicklungen," Paper Delivered at the VDG Seminar Arbeitsstoffe in der Giesserei, Wurzbach/Thuringen (May 24-26,1992).

[5.] A. Gardziella, A. Kwasnick, Unpublished Work of Bakelite AC, Iserlohn (1993).

[6.] D. Boenisch, "Recycling von Kernsandstoffen Teil 1: Die Kaltregenerierung Kunstharzgebundener Altsande," Giesserei, vol 78, number 21, pp 733-744 (1991).

[7.] D. Boenisch, "Recycling von Kernaltstoffen Teil 2: Kernabfall-Ein Vielseitiger Werkstoff," Giesserei, vol 79, number 11, pp 428-435 (1992)

[8.] A. Gardziella, "Glessereihilfsmittel Auf Basis Von Reaktionsharzen," Becker/Braun, Kunststoff Handbuch, Vol 10, pp 931-984.

[9.] M. Belsue, "Synthetic Resins as Carbon Forming Agents," INASMET-Centro Technologico de Materiales, San Sebastian, Partly Published Work (1990-1993).