Comparing Fused Silica and Alumino-Silicate Investment Refractories.

Based on an analysis of shell properties such as strength, thermal expansion, hot creep and ease of removal, fused silica investment refractory outperforms alumino-silicate.

The U.S. investment casting industry has used water-based slurries for decades. The two main systems used are fused silica and alumina-silicates. Each refractory has different performance characteristics that influence material selection for building a water-based shell.

When selecting flours or grains for a shell-building operation, both performance and material cost considerations must be taken into account, however, a foundry also must consider its production parameters, including casting type, size and dimensional tolerances. To take this a step further, the total applied cost of a system (including yield losses due to scrap, machine utilization, equipment maintenance, shell knockout characteristics and number of coats) must be factored into any shell material selection decisions.

Based on a presentation at the 48th Annual Investment Casting Institute Technical Meeting & Exhibition held October 16-19 in Dallas, this article discusses testing performed to analyze and compare the performance characteristics of fused silica and alumino-silicate refractories. This article provides investment casters with information that can aid them in their selection of the most appropriate shelling materials for their individual foundry and product mix.

Shell Weight

Fused silica has a specific gravity of 2.20 while alumino-silicate's is 2.70. Thus, an all-fused silica shell will be 18.5% lighter than a 100% alumino-silicate shell, a difference hand-dippers will appreciate. If dipping is done by robot, sometimes more trees can be dipped simultaneously without exceeding weight limitations of the robot. Manual handling of hot shells also is easier at a lower weight.

Slurry Stability

The long-term stability of colloidal silica-based slurries (that utilize zircon or some alumino-silicates) is adversely affected by ions that can leach from the flours used. The soluble ions react with the colloidal silica causing it to micro-gel. This principally is a problem with prime slurries, due to expected longer service lives.

In the case of fused silica, there is nothing to leach out so slurry stability is not a concern. Some foundries have successfully maintained their fused silica slurries for more than 1 year.

Shell Strength and Toughness

Three shell specimens were prepared for testing (Table 1). Specimen A was a fused silica slurry with fused silica 30/50 stucco, Specimen B was an alumino-silicate slurry with an aluminosilicate 16/30 stucco and specimen C was a fused silica slurry with an alumino-silicate 16/30 stucco. The properties being tested were modulus of rupture (MOR), modulus of elasticity (MOE), fracture index (Fl), fracture toughness ([K.sub.1C]) and slow crack growth rates.

Table 2 provides the room temperature "green" results for MOR, MOE, thickness, peak load and fracture index. There were no statistical differences in the results between specimen A and B for any of the parameters. For specimen C, the MOR and fracture index were somewhat lower than A and B.

In a second test, shell specimens were tested after immersion in 195F (91C) water containing a surfactant for 2 min. These conditions were designed to simulate autoclave dewaxing. In this test, the differences between specimens A and B were not statistically significant, except for the normalized fracture index in which the fused silica shell had the edge. As before, specimen Chad a lower MOR, MOE and normalized fracture index.

Another way to evaluate the resistance of a backup shell to cracking is to measure its [K.sub.1C] (fracture toughness). In this test, a pre-cracked shell specimen is loaded in double torsion, causing the crack to propagate the length of the plate. The fracture toughness of specimen A was greater than for the other two, with specimen B being significantly lower. Measurements of slow crack growth rates in the same hot/wet environments yielded similar results.

From these tests, it was concluded that using fused silica--instead of alumino-silicate for the slurry and stucco (or even for just the slurry)--makes for a more autoclave crack resistant shell. This statement assumes the slurry flour particle size distribution and binder are identical and the stuccos are fused silica 30/50 and alumino-silicate 16/30. This is important because both slurry and stucco size distribution can have a major effect on shell strength and toughness.

Casting Dimensions

Data on thermal expansion, hot creep and hot deflection (due to different thermal expansions between prime coat(s) and the backup and due to temperature gradients through the shell wall) illustrate how shell makeup can alter casting dimensions.

Thermal expansion--Fig. 1 shows the thermal expansion behaviors for various refractory blends. A comparison of fused silica and alumino-silicate (test specimens made with only backup shells) shows that when both types of shells are held for 2 hr in a preheat furnace at 2012F (1100C), the same dimensions result. This also has been observed in production foundries using high pre-heat temperatures to fill thin walls. Production foundries (not using high pre-heat temperatures) also have found they can switch from an alumino-silicate shell to a fused silica shell, except for some jobs with tight dimensional tolerances.

Hot creep--Hot creep of the shell occurs when one or more of the following applies: shell preheat temperature is high; molds are poorly supported during preheat or pour; molds are stacked in the preheat furnace; the pour temperature is high and the molds are large; and slow cooling is used to obtain a desired grain morphology. Hot creep was studied using a four-point bending approach only on specimens of backup shell. Figure 2 compares hot creep behavior of an all fused silica specimen with another where the slurry was of the low viscosity fused silica type and the stucco alumino-silicate. The choice of slurry had little effect on hot creep. It is clear that the use of fused silica stucco vs. alumino-silicate stucco provides for a shell that is more resistant to hot creep. The difference increases at higher temperatures and decreases at lower temperatures.

Hot deflection--In castings with areas of more than 35 sq in. and without features such as ribs, flanges, bosses, etc., the deflection of the mold wall due to differences in thermal expansion becomes a factor with respect to dimensions. The effect of a high-expansion prime coat and a low-expansion backup were examined under isothermal conditions. The mold wall bends toward the side of the higher expansion layer. It was shown that multiple prime coats increase the extent of this hot deflection. This effect can be balanced by using one or more primes as the final coat(s). In a second case, a shell with a moderate or high thermal expansion coefficient experiences a thermal gradient. This occurs when the metal is poured into the mold and a temperature gradient from inside to outside of several hundred degrees centigrade develops. This causes the mold to flex toward the metal (while it is still molten), resulting in a thin wall after solidification. Shortly after metal solidification, the inside to outside temper ature difference is less, and the shell wall moves toward its original position. This opens a gap between the hot solid metal and the shell. Such gaps have a major effect on the metal/shell interface heat transfer coefficient, so the metal in the center portion of this featureless area cools more slowly than the surrounding area. This can affect the grain size and morphology in certain alloys. In severe cases it can result in undesirable metallic phases that degrade mechanical properties.

Shell Removal

One of the main reasons fused silica is used in building investment shells is the ease of shell removal. One way to quantify this effect is to perform room temperature crush tests on shell specimens exposed to temperature conditions approximating what a production shell might experience during pre-heat and pour/cooling. The compression test results for various fused silica and alumino-silicate shell combinations are shown in Table 3.

The data suggests that an all fused silica shell will be much easier to mechanically remove after casting than one based on all alumino-silicate. If a fused silica slurry/alumino-silicate stucco system is used, ease of shell removal should fall in-between, but closer to the all alumino-silicate system.

The crushing data above was for all backup specimens. Other approaches to improve shell removal characteristics include using mixtures of alumino-silicate and fused silica for the backup stucco with fused silica flour for the slurry; and using one or two intermediate dips of fused silica slurry/fused silica stucco and alumino-silicate sluny/alumino-silicate stucco for the backup dips.

Shell Leaching in Caustic

Shells that cannot be removed mechanically are dissolved in a caustic solution or molten caustic. For an alumino-silicate slurry/alumino-silicate stucco system, the caustic solution dissolved out the colloidal silica binder, but left the flour and stucco grains virtually untouched. When this happened, the specimen disintegrated. For the fused silica slurry/alumino-silicate stucco system, the caustic dissolved the colloidal binder and the fused silica flour, but left the alumino-silicate stucco grains largely unaffected. For the fused silica slurry/fused silica stucco system, the shell specimens dissolved uniformly rather than disintegrating into individual stucco grains.

Feeding

When casting into a fused silica shell, it is possible to watch the metal rise through the shell wall as it fills the cavity. This is because fused silica is transparent in the visible spectrum and in the near infrared. Transfer of heat by radiation, in addition to conduction and convection, means castings cool more rapidly. (The size and amount of shell porosity as well as the size of the fused silica stucco grains are factors.) Consequently, feeding distances are greater in a fused silica shell than in other types. For a given part and gating system, a sounder casting would be anticipated if the shell were all-fused silica.

The thermal conductivity of backup shell specimens is shown in Fig. 3r. At low temperatures, the fused silica shell has a lower thermal conductivity, since fused silica is a poorer thermal conductor than alumino-silicate and the shell porosity is higher. At higher temperatures, fused silica becomes substantially more thermally conductive than alumino-silicate because of radiation transport through the stucco grains. At the highest temperatures, the actual thermal conductivity of the fused silica specimens is understated due to problems with the test method. The small increase in conductivity of the alumino-silicate backup shell at 2192F (1200C) is the result of porosity reduction due to sintering.

To further illustrate, thermal conductivity data for fused silica refractories (from the literature) is shown in Fig. 31. The contribution of radiation to the total conductivity is seen for the "dense fused silica." The porous samples are of a fine grain size so the amount of porosity dominates the conductivity. A fused silica shell with 25% porosity and large stucco grains

would fall between the 12% porosity curve and that for the dense specimen.

Thin-Wall Fill

Fused silica has a lower heat capacity than alumino-silicates and half that of zircon. In thin walls, it is only the quantity of heat transferred to the first few shell layers that controls fill. Thus, greater fill of thin-walled castings occurs when using a fused silica shell.

Most metals do not readily wet either fused silica or zircon. Consequently, the intimacy between the solidifying metal and the interior of the shell can affect interface heat transfer. Most aerospace foundries use fine flours (325 mesh and finer) for their prime. However, some foundries use primes that are predominantly 120 fused silica with its angular flour grains. The resulting rougher prime surface means the metal only will be in intimate contact with the tops of the mountains and not the valleys, meaning a reduced interface heat transfer coefficient with better filling of thin-walls. The as-cast surface may not be as smooth, but the surface finish of most investment castings is the result of the blasting operations.

Core Compatibility

Ceramic cores are usually 75-100% fused silica. With an alumino-silicate shell, it is necessary to make slip joints between the shell and the core or the thermal expansion differences will break the core. Cores with an "L" shape are nearly impossible to "float" so they won't fracture. When a fused silica shell is used, its expansion is very low and matches that of the core, so "slip joints" are not required. This also improves the precise positioning of the core.

Which Is Better?

For every characteristic considered, fused silica, in part or whole, improved shell system performance. Whether these performance advantages justify the higher cost of fused silica is a question each foundry must decide for itself. In general, foundries that produce more precise and higher value castings will have a greater incentive to use fused silica. This is partly because the shell materials will constitute a smaller percentage of the total manufacturing cost. (If the casting is rejected for shell related problems, however, the relative cost of the shell material becomes insignificant.) The other part of the equation is that without fused silica, it may be impossible to meet the specifications for the casting.

This article was adapted from a paper presented at the 48th Annual Investment Casting Institute Technical Meeting & Exhibition. Proceedings are available from ICI at 214/368-8852.

                   Porosity, Permeability and Capillary
                  Adsorption Coefficient of Shell Systems
System                             Porosity (%)     Permeability
                                                ([10.sup.-10] sq Cm)
A--Fused silica slurry with
fused silica 30/50 stucco              24.6             18.3
B--Alumino-silicate slurry
with alumino-silicate 16/30 stucco     23.2             8.6
C--Fused silica slurry with
alumino-silicate 16/30 stucco           24              18.2
System                               Capillary Adsorption
                                          Coefficient
                                   (kg/[sq m[(sec).sup.1/2]]
A--Fused silica slurry with
fused silica 30/50 stucco                   0.0261
B--Alumino-silicate slurry
with alumino-silicate 16/30 stucco           0.023
C--Fused silica slurry with
alumino-silicate 16/30 stucco               0.0286
                   MOR, MOE and Fracture Index for Room
                    Temperature "Green" Shell Specimens
                     A--Fused silica    B--Alumino-silicate
                    slurry with fused  slurry with alumino-
                   silica 30/50 stucco silicate 16/30 stucco
MOR (psi)                  577                  596
MOE (ksi)                  726                  768
Normalized
fracture index           0.0167               0.0166
Specimen
thickness (in.)           0.293                0.285
Fracture load (lb)        7.05                 7.05
                     C--Fused silica
                   slurry with alumino-
                     silicate stucco
MOR (psi)                  489
MOE (ksi)                  612
Normalized
fracture index            0.0155
Specimen
thickness (in.)           0.312
Fracture load (lb)         6.86
                   Compression Tests of Fired Specimens
System                             1900F (1037C) for 2200F (1204C) for
                                       1 hr hold         2 hr hold
Fused silica slurry with
fused silica 30/50 stucco              1717 psi          1947 psi
Alumino-silicate slurry
with alumino-silicate 16/30 stucco     6172 psi          9456 psi
Fused silica slurry with
alumino-silicate 16/30 stucco          4547 psi          5008 psi