The evolution of high-end manufacturing has driven the widespread adoption of stainless steel components produced via the investment casting process. This method offers unparalleled advantages, and is often the only viable choice, for manufacturing small to medium-sized stainless steel parts featuring intricate external geometries and complex internal cavities. However, the design of casting processes and production control for components with sophisticated internal channels present significant technical challenges.
Within the investment casting process, complex internal cavities can be formed using water-soluble wax or salt cores. While feasible, these techniques necessitate additional steps. After dissolving the soluble core from the wax pattern, the standard shell-building procedure of slurry dipping and stuccoing must be performed. This often requires supplementary manual operations or the assembly of multiple wax segments to create the final pattern. A critical limitation arises for deep or narrow features; it is notoriously difficult to produce features where the depth-to-diameter ratio (H/d) exceeds 5. Although larger holes and channels can be successfully shelled, the small volume of refractory material in these areas is prone to sintering and metal penetration during pouring, leading to severe difficulties in post-casting cleanup. Consequently, the use of specialized ceramic cores becomes the optimal and often essential production choice.
Significant research has been dedicated to ceramic cores for superalloy investment casting, with silica-based and alumina-based cores representing mature technologies. However, silica-based cores are prone to deformation at temperatures exceeding 1,550 °C. More critically, silica can react chemically with elements present in stainless steel melts. These reactions not only cause surface defects on the castings but also severely compromise the core’s leachability, making its removal exceptionally difficult. Alumina-based cores exhibit excellent high-temperature chemical stability and inertness, but this very property necessitates removal using strongly alkaline solutions. These aggressive chemicals can corrode the stainless steel casting itself, adversely affecting the final component’s quality and integrity.
This research explores an alternative material system centered on magnesium oxide (MgO) for ceramic cores intended specifically for stainless steel investment casting. Magnesium oxide offers a potentially favorable combination of refractoriness and chemical compatibility. The study investigates the formulation and properties of MgO-based cores, utilizing calcium carbonate (CaCO3) as a sintering aid, or mineralizer, and wood flour as a pore-forming agent. The cores are fabricated via a uniaxial pressing method. The effects of the mineralizer and pore-former on key core properties—including room-temperature strength, porosity, sintering shrinkage, and, most importantly, dissolution behavior in a mild acidic medium—are systematically examined. The ultimate goal is to develop a viable core material that maintains sufficient strength during the investment casting process while being readily removable after casting without damaging the stainless steel component.
1. Experimental Methodology
1.1. Raw Materials
The primary matrix material was commercially available magnesia powder (MgO >98%) with a particle size of 400 mesh. Reagent-grade calcium carbonate (CaCO3), sieved to below 200 mesh, served as the mineralizing agent. Dry wood flour was used as the pore-forming additive.
1.2. Core Fabrication and Processing
The experimental design involved varying the content of CaCO3 and wood flour. The powders were weighed according to the formulations in Table 1 and mixed in a planetary ball mill for 2 hours at 350 rpm using a ball-to-powder ratio of 1:1.
| Designation | MgO (wt.%) | CaCO3 (wt.%) | Wood Flour (wt.%) | Primary Variable Studied |
|---|---|---|---|---|
| M0 | 100 | 0 | 0 | CaCO3 Content |
| M5C | 95 | 5 | 0 | |
| M10C | 90 | 10 | 0 | |
| M15C | 85 | 15 | 0 | |
| M20C | 80 | 20 | 0 | |
| M5C-W1 | 94 | 5 | 1 | Wood Flour Content |
| M5C-W3 | 92 | 5 | 3 | |
| M5C-W5 | 90 | 5 | 5 |
The blended powder was then combined with a 5 wt.% aqueous solution of polyvinyl alcohol (PVA) acting as a binder and plasticizer. The mixture was thoroughly homogenized in a mortar. The resulting granulated feed was pressed in a steel die, pre-lubricated with silicone oil to minimize friction, under a uniaxial pressure of 6 MPa with a 2-minute dwell time. The green compacts had dimensions of 60 mm × 10 mm × 6 mm.
The sintering cycle was carefully designed to allow for binder burnout and controlled densification. The compacts were heated in an air atmosphere at a rate of 2 °C/min, with intermediate holds at 100, 200, 400, 600, 800, 1000, and 1200 °C for 30 minutes each to ensure uniform thermal treatment. The final sintering was conducted at 1340 °C for 30 minutes, followed by furnace cooling. This profile is critical in the investment casting process for developing the necessary core strength before wax pattern assembly and shell building.
1.3. Property Characterization
1.3.1. Physical and Mechanical Properties:
The room-temperature flexural strength was determined via the three-point bending test on a universal testing machine, using a span of 30 mm and a crosshead speed of 0.5 mm/min. The apparent porosity was measured using the Archimedes method (water immersion). The linear sintering shrinkage was calculated from the dimensional change of the core before and after firing.
1.3.2. Dissolution and Collapsibility Behavior:
A critical performance metric for cores in the investment casting process is their post-casting removability. A dissolution test was devised to evaluate this property. The mass loss of sintered cores after immersion in a leaching solution was used as a quantitative indicator of collapsibility. The cores were immersed in a 40 wt.% acetic acid solution maintained at 80 °C for 10 hours. After leaching, the samples were ultrasonically cleaned in deionized water, dried at 110 °C for 2 hours, and weighed. The dissolution mass loss ratio $K$ is defined as:
$$ K = \frac{m_0 – m_1}{m_0} \times 100\% $$
where $m_0$ is the initial mass and $m_1$ is the mass after leaching.
2. Results and Discussion: Influence of Mineralizer (CaCO3)
The initial phase of the study focused on understanding the role of CaCO3 as a mineralizer in the pure MgO system (samples M0 to M20C).
2.1. Flexural Strength
The addition of CaCO3 dramatically altered the sintering behavior of MgO. As shown in Table 2, the flexural strength increased sharply from 7.57 MPa for pure MgO (M0) to 24.03 MPa with 5% CaCO3 addition (M5C). This significant enhancement is attributed to the mineralizer effect. During heating, CaCO3 decomposes to calcium oxide (CaO) and carbon dioxide (CO2) around 900 °C. The nascent CaO, being highly reactive and having a smaller ionic radius, can diffuse into the MgO lattice, forming a solid solution. This introduces lattice strain, which increases the driving force for solid-state diffusion and sintering, as described by the general diffusion equation for sintering:
$$ \frac{dx}{dt} = \frac{D \gamma \Omega}{k_B T} \cdot \frac{1}{x^n} $$
where $x$ is the neck radius, $D$ is the diffusion coefficient, $\gamma$ is the surface energy, $\Omega$ is the atomic volume, $k_B$ is Boltzmann’s constant, $T$ is temperature, and $n$ is an exponent dependent on the diffusion mechanism. The presence of Ca2+ ions likely enhances the diffusion coefficient $D$. Furthermore, the fine CaO particles fill interstitial voids between larger MgO grains, promoting densification. However, beyond the 5% optimum, strength progressively decreased with further CaCO3 addition, falling to 11.87 MPa for M20C. This decline is likely due to the increasing volume of pores left by the evolving CO2 gas and the potential formation of secondary phases at grain boundaries that may weaken the structure.
2.2. Apparent Porosity and Sintering Shrinkage
The trends in apparent porosity and linear shrinkage complement the strength data (Table 2). The pure MgO core exhibited high porosity (32.15%) and negligible shrinkage (0.22%), confirming its poor sinterability at 1340 °C. The addition of 5% CaCO3 reduced porosity to 28.07% and increased shrinkage to 0.97%, evidence of enhanced densification. As the CaCO3 content increased further, porosity began to rise again (reaching 31.45% for M20C) due to the dominant effect of gas evolution, while shrinkage remained relatively constant between 0.88% and 0.97%.
| Sample | Flexural Strength (MPa) | Apparent Porosity (%) | Linear Shrinkage (%) | Dissolution Loss, K (%) |
|---|---|---|---|---|
| M0 | 7.57 ± 0.8 | 32.15 ± 1.2 | 0.22 ± 0.05 | 100.00 |
| M5C | 24.03 ± 1.5 | 28.07 ± 1.0 | 0.97 ± 0.07 | 50.22 ± 3.1 |
| M10C | 19.45 ± 1.3 | 29.83 ± 1.1 | 0.92 ± 0.06 | 45.17 ± 2.8 |
| M15C | 15.21 ± 1.1 | 30.76 ± 1.3 | 0.88 ± 0.08 | 52.49 ± 3.5 |
| M20C | 11.87 ± 1.0 | 31.45 ± 1.4 | 0.95 ± 0.07 | 55.34 ± 3.7 |
2.3. Dissolution Behavior
The dissolution behavior presented a clear trade-off. The pure MgO core (M0), being highly porous and poorly sintered, disintegrated completely in the acetic acid solution within 10 hours (K=100%). In contrast, the mineralizer-containing cores, which were more densely sintered, only experienced surface erosion, maintaining their overall shape. Their mass loss values were significantly lower, ranging from ~45% to ~55%. While increased CaCO3 content slightly improved K due to higher residual porosity, the improvement was marginal compared to the substantial strength penalty. The core with 5% CaCO3 (M5C) presented the best compromise, offering high strength while retaining a moderate dissolution rate. This balance is crucial for surviving the rigors of the investment casting process, including wax injection, shell handling, and metal pouring, while still being removable afterwards.
3. Results and Discussion: Influence of Pore-Former (Wood Flour)
To improve the dissolution characteristics of the optimized M5C composition without resorting to excessive mineralizer, wood flour was introduced as a sacrificial pore-former (samples M5C-W1 to M5C-W5).
3.1. Flexural Strength and Porosity
The incorporation of wood flour had a pronounced and predictable effect, as summarized in Table 3. The flexural strength decreased markedly with increasing wood flour content. Compared to the baseline M5C core (22.19 MPa), strength dropped by approximately 55% to 10.56 MPa with just 1% wood flour (M5C-W1). It further decreased to 7.03 MPa (3% wood flour) and 4.73 MPa (5% wood flour). This severe reduction is a direct consequence of the increased porosity generated when the organic particles burn out during sintering, leaving behind voids that act as strength-limiting defects. Concurrently, the apparent porosity increased systematically from 28.07% (M5C) to 39.77% (M5C-W5). The relationship between porosity ($P$) and strength ($\sigma$) often follows an exponential decay model:
$$ \sigma = \sigma_0 \cdot \exp(-bP) $$
where $\sigma_0$ is the theoretical strength at zero porosity and $b$ is a material constant. The data obtained aligns well with this model, indicating that wood flour is an effective agent for introducing interconnected porosity.
3.2. Sintering Shrinkage and Dissolution Behavior
The sintering shrinkage showed a non-monotonic trend, initially decreasing with 1% wood flour addition before rising significantly at 5% content. The initial decrease may be due to the temporary skeletal support provided by the wood particles during the early stages of sintering. The subsequent increase at higher wood flour loadings likely results from the greater volumetric fraction of voids, which allows surrounding ceramic material to consolidate more freely into the empty space, leading to net macroscopic shrinkage.
The dissolution performance improved dramatically. The mass loss ratio $K$ after 10 hours in acetic acid increased from 50.22% for M5C to 70.12% for M5C-W5, an improvement of nearly 40%. The cores with higher wood flour content disintegrated into finer debris. The enhanced collapsibility is directly linked to the higher open porosity, which allows the leaching solution to penetrate the core’s interior more rapidly and extensively, attacking a much larger surface area of the MgO matrix. The dissolution process can be conceptually modeled as a surface reaction coupled with diffusion into the pore network. The increased porosity reduces the tortuosity of the pore path, accelerating the overall process.
| Sample | Flexural Strength (MPa) | Apparent Porosity (%) | Linear Shrinkage (%) | Dissolution Loss, K (%) |
|---|---|---|---|---|
| M5C (Baseline) | 22.19 ± 1.4 | 28.07 ± 1.0 | 0.97 ± 0.07 | 50.22 ± 3.1 |
| M5C-W1 | 10.56 ± 0.9 | 31.31 ± 1.2 | 0.46 ± 0.06 | 57.71 ± 3.0 |
| M5C-W3 | 7.03 ± 0.7 | 35.18 ± 1.5 | 0.77 ± 0.07 | 63.47 ± 3.4 |
| M5C-W5 | 4.73 ± 0.6 | 39.77 ± 1.7 | 1.58 ± 0.10 | 70.12 ± 3.8 |
This creates a fundamental design compromise for the investment casting process: higher porosity improves post-casting removability but reduces the core’s green and sintered strength, which must be sufficient to withstand handling, wax injection pressures, and the metallostatic head of the molten metal.
4. Validation: Investment Casting Trial
To validate laboratory findings in a real investment casting process, cores with different wood flour contents (M5C-W1, W3, W5) were used to cast CF8 stainless steel test specimens. Standard shelling with silica sol binder was performed. The molds were fired at 1160°C, and steel was poured at 1570-1575°C. After casting, the shells were removed by vibration, and the test pieces were cut from the tree and shot blasted.
Post-blast examination revealed that a significant portion of the core remained inside the cast channels. These cores-in-castings were then subjected to static leaching in 40% acetic acid at 80°C. The depth of core removal from the channel entrance was measured over time. The results, plotted in Figure 1, are critical for process planning.
During the initial 6 hours, the dissolution rates were similar (~3.6-3.8 mm depth), likely due to the shot blasting process inducing surface microcracks that provided an initial rapid leaching pathway. Subsequently, the rates diverged based on the intrinsic core structure. The M5C-W5 core, with the highest porosity, exhibited the fastest dissolution, with a steady-state rate of approximately 0.78 mm/h, completely removing the residual core in about 20 hours. The M5C-W3 core dissolved slowest at 0.49 mm/h, while M5C-W1 showed an intermediate rate of 0.57 mm/h.
This trial confirms the core performance in an actual investment casting process context. The dissolution in a static bath, while effective, points to a clear opportunity for process optimization. In an industrial setting, the leaching step of the investment casting process can be significantly accelerated by employing agitated tanks, ultrasonic assistance, or forced solution flow through the core channels. The dissolution rate $v$ in a convective system can be described by an expression of the form:
$$ v = k \cdot (C_s – C_b) $$
where $k$ is a mass transfer coefficient dependent on fluid flow, $C_s$ is the saturation concentration at the core surface, and $C_b$ is the concentration in the bulk solution. Agitation increases $k$, thereby increasing $v$. Therefore, the core removal time in a production investment casting process can be substantially less than indicated by the static tests.
5. Conclusions
This study successfully developed and characterized a magnesium oxide-based ceramic core system tailored for the stainless steel investment casting process. The key conclusions are:
- Mineralizer Optimization: Calcium carbonate is an effective mineralizer for MgO. An addition of 5 wt.% CaCO3 sintered at 1340°C produces cores with an optimal balance of properties: high flexural strength (~24 MPa) for handling and casting integrity, moderate sintering shrinkage (~0.97%), and acceptable initial dissolution behavior.
- Porosity Engineering for Removal: The incorporation of a sacrificial pore-former, wood flour, is a highly effective strategy to engineer interconnected porosity, thereby dramatically improving the core’s dissolution collapsibility in mild acetic acid. A 5 wt.% addition increased the 10-hour mass loss from ~50% to over 70%.
- The Inevitable Trade-off: The study quantitatively demonstrates the fundamental property trade-off in core design for the investment casting process. Enhancing leachability via increased porosity directly and significantly reduces the mechanical strength of the sintered core. The M5C-W5 composition, for instance, saw its strength drop by about 78% compared to the non-porous counterpart.
- Process Validation: Casting trials confirmed the functionality of these cores. The porous cores (M5C-W5) demonstrated the fastest in-situ dissolution rate (~0.78 mm/h in static acid) after casting. This validates that the designed microstructure translates to improved removability in a real investment casting process sequence.
- Process Integration Outlook: The use of a mild acetic acid solution for core removal is a significant advantage for stainless steel components, avoiding the corrosive damage associated with strong alkali leachants used for alumina cores. Furthermore, the removal process within the investment casting process can be optimized beyond static immersion. Implementing solution agitation, flow, or ultrasonic energy will increase mass transfer, significantly reducing the total core removal time and enhancing overall production efficiency.
This work provides a foundational framework for designing MgO-based ceramic cores. Future work may involve optimizing the particle size distribution of the MgO powder, exploring alternative mineralizers or composite formulations, and testing the high-temperature creep resistance and chemical interactions with various stainless steel grades under investment casting process conditions.