Controlling heat transfer with refractory coating.

Inside This Story

* The cast iron industry is facing higher demands to produce cleaner, thinner-walled castings with closer tolerances.

* Refractory coating manufacturers have been pressed to produce coatings capable of enhancing the performance of resin/sand systems.

* This article details a study to determine the affect of two refractories, diatomaceous earth and graphite, on thermal heat transfer.

Higher demands have been placed on the cast iron industry to produce cleaner, thinner-walled castings with closer tolerances. In order to produce these, it is necessary to make cores that have tighter tolerances and are stable under all pouting conditions. Regardless of the tolerances achieved in a sand core at room temperatures, the distortion of the core upon exposure to molten metal and rapid heating may introduce additional variances to the finished casting.

While new resins and sand systems are being explored to meet these requirements, the metalcasting industry has pressed refractory coating manufacturers to produce coatings capable of enhancing the performance of resin/sand systems. To date, refractory coating's primary function has been to enhance casting surface finish and minimize metal penetration defects.

In most U.S. core applications, the final dry coating deposit typically is limited to 0.004-0.019 in. (0.10-0.25 mm) in order to compensate for dimensional changes associated with coating cores. In Europe, heavier dry coating deposits ranging from 0.10-0.15 in. (0.25-0.4 mm) have demonstrated some anti-veining characteristics, which implies that heavier coating deposits may reduce the thermo-mechanical stress development in resin-bonded sand systems.

One approach to help prevent thermal distortion is to reduce the heat transfer through the coating and into the core. The primary method for this is to lower the thermal conductance of the refractory coating (increase the insulation).

Refractories with insulating characteristics are necessary to reduce heat transfer. The insulating characteristics of the refractory particles in a coating are controlled by multiple factors, including:

* particle chemistry, shape, crystalline structure and density;

* thermal conductivity of the refractory;

* the thickness and alignment of the particles in the deposit.

Investigations were performed to compare diatomaceous earth (DE), a low thermal conductance refractory (high insulation coating), to graphite (G), a high thermal conductance refractory (low insulation coating), at temperatures and pressures that simulated pouring cast iron against a coated phenolic urethane coldbox (PUCB) disc (Fig. 1). The effectiveness of the two types and thicknesses of refractories in preventing heat transfer and thermal expansion to the discs was measured through changes in the thermal distortion curves of the discs and by comparing the heat transfer ([T.sub.transfer]) from the thermal source to the experimental discs.

[FIGURE 1 OMITTED]

Calculating Heat Transfer

Coating thicknesses of 0.004 and 0.008 in. (0.1 and 0.2 mm) were selected as representative of industry practices and because they are the thinnest and heaviest surface deposits that could be applied with the experimental one refractory component coatings. As seen in Fig. 2, refractory layers are not strictly limited to the surface of the discs, but also penetrate into the interstitial space between the sand grains of the disc.

[FIGURE 2 OMITTED]

In addition, the transfer of heat from the contact point at the thermal hot surface used to heat the discs contributes a thermal joint conductance factor to the heat transfer.

As a consequence, no attempt was made to calculate the actual thermal conductance of the two refractories. Instead, the rate of heat loss from the thermal hot surface per unit of time was recorded during each disc heating cycle to determine and compare the amount and rate of the heat transfer. Heat loss from the thermal hot surface ([DELTA]T) was calculated by determining the lowest average temperature the hot surface recorded then subtracting the starting temperature of the host surface from the lowest temperature. The average time (t) to reach the lowest temperature was read directly from the recording, so the rate of thermal heat transfer was calculated as: [T.sub.transfer] = [DELTA]T/[DELTA]t

Laying the Groundwork

In the study, the refractory coating formulas were greatly simplified from production formulas to minimize interactions between typical coating components and the effects to those directly related to the refractories under evaluation. The non-refractory components of the dried coating were less than 1% by weight of the deposit. The median particle sizes of the two refractories were matched as closely as possible, xantham gum was chosen as the suspension/rheological system for both coatings, the surface-active substance used was the same for both coatings, and surface tension of the two coatings was matched as closely as possible (Table 1).

The test cores were 2 x 0.35-in. (5.08 x 0.89-cm) transverse disc specimens made using a common two-part industrial phenolic urethane resin. A commercial grade 4 screen washed lake sand of 48 GFN was used to make the discs (Table 2).

The resin level was 1.25% based on sand (BOS) with a 55:45 ratio of part 1 to part 2, and technical grade TEA was used as a catalyst. A four-off corebox was used with a blow pressure of 50 psi (nitrogen) for 0.5 sec. and the catalyst gas purge cycle (nitrogen) of 40 sec. at 20 psi.

The experimental discs were randomized by corebox activity and by coating. They were held at. a 30-degree angle and dipped 20 min. after stripping. Only one side of each disc was coated with a refractory coating. The wet coated discs were allowed to drain for 15-20 sec. before being transferred to a forced air oven for drying. Once cooled to lab temperature, the dry coated discs were weighed and then transferred to Western Mich. Univ. for final testing.

The net change in dry weight caused by refractory coating was considered indicative of increased coating deposit. The change in dry weight was calculated for every disc by subtracting the strip weight from the final dry weight of the coated disc (Table 3).

Reading the TDCs

In this investigation, PUCB specimens were coated with DE or G to two thickness levels while control cores were left uncoated. All specimens were tested at a molten cast iron temperature of 2,320F (1,2880 and a weight of 1.46 lbs. (663 g) for 3 min.

The thermal distortion curves (TDCs) for all systems showed undulations indicating thermo-mechanical and thermo-chemical changes in the binder system at an elevated temperature. During the first 2-5 sec., the control cores and the [G.sub.0.008], [DE.sub.0.004] and [DE.sub.0.008] thermal distortion curves all demonstrated a slight downward deflection between 0.0002-0.0004 in. (0.005-0.01 mm). After this initial deflection, the [DE.sub.0.008] held constant, while all other test groups demonstrated an upward deflection between 5-60 sec. The order of upward deflection, from smallest to largest, was [DE.sub.0.004], control, [G.sub.0.004], [G.sub.0.008]. After reaching their peak, these four samples then began a downward trend out to the 180 sec. tested. After 180 sec., the order of least to most downward deflection of the test specimens was:

1. [DE.sub.0.008];

2. [DE.sub.0.004];

3. [G.sub.0.004];

4. Control;

5. [G.sub.0.008].

Ultimately, both G refractory coated systems demonstrated significant expansion during the first 65 sec., followed by plastic deformation for the remainder of the test period. The uncoated control system had a similar TDC (Fig. 3). The G refractory systems increased the thermal distortion of the discs by 38.2% at the 0.004-in. (0.01-mm) deposit and by 57.8% at the 0.008-in. (0.02-mm) deposit.

[FIGURE 3 OMITTED]

The DE systems were thermally stable. After the first 5 sec., the 0.004-in. (0.01-mm) DE demonstrated slight expansion for 110 sec. with a slight plastic deformation at the end of the test. [DE.sub.0.008] showed only slight plastic deformation. The thermal distortion was reduced by 42.2% at the 0.004 in. (0.01-mm) deposit and by 62.2% at the 0.008 in. (0.02-mm) deposit.

With respect to systems tested, there was a significant difference in thermal distortion (Table 4, Fig. 3). The G specimens had more thermal distortion than the controls, and the DE specimens had less thermal distortion than the control discs.

Comparing Hot Surface Temperature

Certain trends were observed while comparing the data for average hot surface temperature. The [G.sub.0.004], [G.sub.0.008] and control samples caused a thermal gradient at the hot surface and, after 20 sec., were reduced to 2,039, 2,075 and 2,102F (1,115, 1,135 and 1,150C). The temperatures then began the return to the steady state condition (Fig. 4). The G refractory increased the heat transfer (relative to the control) by 10.6% at the 0.004-in. (0.01-mm) deposit and by 12% at the 0.008-in. (0.02-mm) deposit. Both G samples demonstrated a maximum heat loss at 19 sec. with a change in temperature of -276 and -245F (-171C and -154 C). The average calculated heat transfer for graphite was -8.55C[s.sup.-1]. The graphite systems demonstrated at 12% increase in heat transfer from the hot surface when compared to the uncoated control disc.

[FIGURE 4 OMITTED]

The thermal gradient at the hot surface after 20 sec. for the DE specimens showed an insulative property and only dropped to 2,246 and 2,210 F (1,230 and 1,210C) before beginning their return to the steady state condition. The DE refractory reduced the heat transfer (relative to the control) by 4.2% at the 0.004-in. (0.01-mm) deposit and by 5.6% at the 0.008-in. (0.02-mm) deposit. The average calculated heat transfer for diatomaceous earth was--2.56C[s.sup.-1]. The DE systems demonstrated a 66% decrease in heat transfer from the hot surface when compared to the uncoated control disc.

Future Opportunities

The investigation results suggest that refractory coatings have a more significant impact on core distortion and, consequently, on casting tolerance than had been thought previously. Those actively exploring thin wall iron castings and precision iron castings may want to investigate the contributions that refractory coatings can make to their efforts.

Future investigations are planned to evaluate the impact of refractory coating heat transfer on total core distortion and the subsequent potential for vein formation. Research into the effects that refractory coating penetration between sand grains has on thermal distortion and thermal conductance also is planned.

There are numerous other chemical binder systems from which additional data could be gathered to learn about thermal properties, and additional work could be performed at different loads and different temperatures simulating other alloys and pressures representative of larger or smaller castings.

For More Information

"Evaluating Refractory Coatings: A Practical Approach," S.G. Baker, MODERN CASTING, Oct. 2002, p. 21-23.

Orville Guyer is a lab manager and Robert Emptage is a product development chemist at HA International, Toledo, Ohio. Sam Ramrattan is an FEF Key Professor at Western Michigan Univ., Kalamazoo, Mich.

Table 1. Refractory Coating Properties Relative to Coated Discs.

                                Brookfield   Thixotropic
        Coating      % Solids   Visc. (cP)      Index

A    [G.sub.0.004]     22.13        330          1.74
B    [G.sub.0.008]     24.72        609          1.74
C   [DE.sub.0.004]     23.32        259          1.76
D   [DE.sub.0.008]     25.07        460          1.76

                     Median Particle    Surface Tension
        Coating      Size ([micro]m)   (dyne/[cm.sup.2])

A    [G.sub.0.004]         9.38              31.64
B    [G.sub.0.008]         9.38              30.94
C   [DE.sub.0.004]         9.28              34.32
D   [DE.sub.0.008]         9.28              34.58

Table 2. Properties of Sand.

Source   AFS/gfn      Shape     Screens   % Resin

  MI        48     subangular      4       1.25%

              Roundness/
Source   Sphericity/Krumbein      pH

  MI           0.5/0.7         7.2-8.4

Table 3. Changes in Disc Weight Caused by Dry Coating.

         Refractory Coating                        G
                                       High Thermal Conductance

     Dry coating thickness (in.)            0.004    0.008
      Mean coated disc wt. (g)              25.967   26.123
Refractory penetration (sand grains)          5         2

         Refractory Coating                       DE
                                       Low Thermal Conductance

     Dry coating thickness (in.)           0.004     0.008
      Mean coated disc wt. (g)             26.151   26.260
Refractory penetration (sand grains)         5         2

         Refractory Coating             Control
                                       (Uncoated)

     Dry coating thickness (in.)            0
      Mean coated disc wt. (g)            25.616
Refractory penetration (sand grains)        0

Table 4. Physical and Thermo-Mechanical Properties of the
Refractory Coated and Uncoated Discs Systems.

    System            Initial Temp          Lowest Recorded
                 [degrees]C([T.sub.0])   Temp [degrees]C(T1))

 [G.sub.0.004]           1,288                   1,117
 [G.sub.0.008]           1,288                   1,134
[DE.sub.0.004]           1,288                   1,234
[DE.sub.0.008]           1,288                   1,209
    Control              1,288                   1,151

    System          Change in Temp          Time to Lowest
                 [degrees]C([DELTA]T)   Temp (sec., [DELTA]t)

 [G.sub.0.004]           -171                     19
 [G.sub.0.008]           -154                     19
[DE.sub.0.004]            -54                     26
[DE.sub.0.008]            -79                     30
    Control              -137                     18

    System       [T.sub.transfer]   Thermal Distortion
                 ([degrees]C/sec.       Range (in.)

 [G.sub.0.004]          -9                   0
 [G.sub.0.008]         -8.11               0.01
[DE.sub.0.004]         -2.08               0.02
[DE.sub.0.008]         -2.04                 0
    Control            -7.61                 0