As a casting engineer working on aerospace fuel systems, I have witnessed a fundamental shift in how we manufacture complex aluminum alloy housings. The internal geometry of fuel control components has become increasingly intricate, and conventional sand molding techniques often struggle to meet the tight tolerances and surface finish requirements demanded by modern aero-engine designs. In our foundry, we have adopted 3d sand printing to produce sand molds directly from CAD data, eliminating the need for expensive pattern tooling and enabling unprecedented design freedom. However, this transition has introduced new challenges related to mold strength, gas evolution, and permeability, especially when combined with low-pressure casting. In this paper, I present a systematic study of the key parameters affecting 3d sand printing mold performance, focusing on resin content and coating process variables. The ultimate goal is to eliminate gas porosity defects in ZL114A aluminum alloy fuel housings produced by low-pressure sand casting using 3d sand printing technology.
1. Introduction
Fuel housings are critical components in aero-engine fuel systems. Their structural integrity directly influences engine reliability and safety. With the continuous push for higher thrust-to-weight ratios and lower fuel consumption, the internal oil passages, valve seats, and mounting interfaces have become more compact and complex. Traditional sand casting methods, while versatile, face limitations when producing thin-walled, intricate cores and cavities. Core shifts, poor surface finish, and gas porosity are frequent problems. 3d sand printing offers a solution by building molds and cores layer-by-layer using a binder jetting process. This method allows me to fabricate integrally cored molds with optimized gating systems that would be impossible to produce with conventional core boxes.
Despite these advantages, the properties of 3d sand printing molds depend heavily on the binder content and the post-processing steps, particularly the application of refractory coatings. The coating serves to protect the mold surface from molten metal erosion, improve surface finish, and control heat transfer. However, it also affects the gas permeability and gas evolution characteristics of the mold. In low-pressure casting, the mold filling is relatively slow and controlled, making proper venting crucial. If the gas generated by the decomposition of resin and coating cannot escape rapidly through the sand mold, the resulting back pressure can hinder filling and lead to gas entrapment in the solidifying metal.
In this work, I have investigated the influence of resin content on the mechanical strength, gas evolution, and permeability of 3d sand printing molds. Furthermore, I examined how coating dilution, immersion time, and coating viscosity (measured in Baumé degrees) affect the gas evolution and permeability. Based on these laboratory results, I optimized the process parameters and verified them in actual production of ZL114A fuel housings. The findings provide practical guidance for foundries that plan to adopt 3d sand printing for low-pressure casting of complex aluminum components.
2. Experimental Procedure
2.1 Materials and Mold Preparation
The base sand used in this study was silica sand with a grain size distribution of 70–140 mesh. Furan resin was used as the binder. Four different resin contents were evaluated: 0.8%, 1.2%, 1.6%, and 2.0% by weight. The sand and resin were mixed in a high-speed mixer and then fed into a commercial 3d sand printing machine. The printer deposited the sand-resin mixture in thin layers and selectively activated the binder by jetting a curing agent onto the areas corresponding to the mold geometry. After printing, the molds were allowed to cure at room temperature for 24 hours before testing. Standard test specimens were printed for strength, gas evolution, and permeability measurements. The test specimens included standard “8” shaped tensile specimens and cylindrical specimens of Φ50 mm × 50 mm for compressive and permeability tests.
2.2 Strength Testing
An XQY-II intelligent sand strength tester was used to measure the tensile and compressive strengths of the printed specimens. For tensile testing, three “8” shaped specimens were tested for each resin content, and the average value was reported. For compressive testing, three cylindrical specimens (Φ50 mm × 50 mm) were used for each condition. The loading rate was maintained at a constant value to ensure consistent results.
2.3 Gas Evolution Measurement
Gas evolution was measured using an FQX-2000 high-precision gas evolution tester. The equipment consists of a high-temperature decomposition furnace and a gas collection and measurement system. For each test, 1.00 g of the sand sample or dried coating material was placed in a ceramic boat. The furnace temperature was set to 710 °C. Once the temperature stabilized, the boat was pushed into the constant-temperature zone of the furnace, and the gas evolution was recorded automatically for 100 seconds. The gas evolution value was expressed in mL/g. Each measurement was repeated three times, and the average value was reported.
2.4 Permeability Measurement
Permeability was measured using a ZTY intelligent permeability tester. Standard cylindrical specimens (Φ50 mm × 50 mm) were placed in an airtight sample holder. Air at a pressure of (981 ± 5) Pa (100 mm H₂O) was applied through the specimen. The permeability coefficient was calculated based on the volume of air passing through the specimen per unit time and expressed in cm⁴/(g·min). For each condition, three specimens were tested, and the average value was computed.
2.5 Coating Application
A water-based zirconium silicate coating was used in this study. The coating was diluted with water to achieve different Baumé degrees (measured with a hydrometer). The coating viscosity was varied by adding water: 0, 100, 200, and 300 mL of water per unit volume of the original coating. The immersion time was varied from 2 s to 8 s. The coating was applied to the standard cylindrical specimens by dipping them into the coating slurry. After dipping, the specimens were removed and allowed to drain for a few seconds, then dried in an oven at 150 °C for 1 hour. After drying, the gas evolution and permeability of the coated specimens were measured using the same procedures described above.
3. Results and Discussion
3.1 Effect of Resin Content on Mold Strength
The mechanical strength of the 3d sand printing mold is a critical factor in preventing deformation and cracking during mold handling, core assembly, and molten metal filling. Figure 1 shows the tensile and compressive strengths as functions of resin content. The results are also summarized in Table 1.
| Resin content (%) | Tensile strength (MPa) | Compressive strength (MPa) |
|---|---|---|
| 0.8 | 0.83 | 2.11 |
| 1.2 | 1.25 | 3.60 |
| 1.6 | 1.58 | 4.28 |
| 2.0 | 1.85 | 4.42 |
As expected, higher resin content leads to higher tensile strength. The tensile strength increased from 0.83 MPa at 0.8% resin to 1.85 MPa at 2.0%. The compressive strength also increased significantly when the resin content was raised from 0.8% to 1.6%, but further increase to 2.0% produced only a marginal improvement, from 4.28 MPa to 4.42 MPa. This behavior is typical of resin-bonded sands: beyond a certain resin content, additional resin does not contribute to additional bond strength because the sand grains are already fully coated and the excess resin only occupies the void space. Moreover, excessive resin can make the mold brittle and reduce its collapsibility, which is undesirable for casting complex internal passages.
For the fuel housing application, I found that 1.2% resin content provides a good balance. The compressive strength of 3.6 MPa is sufficient to withstand the low-pressure filling process (typically 0.05–0.1 MPa), and the tensile strength of 1.25 MPa ensures that the mold can be handled without breaking. Higher resin content would increase gas evolution, which is disadvantageous for minimizing gas porosity.
3.2 Effect of Resin Content on Gas Evolution
Gas evolution is one of the most important properties of 3d sand printing molds, because the thermal decomposition of organic resin generates volatile gases. If these gases cannot escape before the molten metal solidifies, they will be trapped as gas pores. Figure 2 shows the gas evolution curves for different resin contents. The total gas evolution values after 100 s are listed in Table 2.
| Resin content (%) | Gas evolution (mL/g) |
|---|---|
| 0.8 | 2.83 |
| 1.2 | 8.60 |
| 1.6 | 9.80 |
| 2.0 | 10.70 |
It is clear that gas evolution strongly depends on resin content. Increasing the resin content from 0.8% to 1.2% raises the gas evolution from 2.83 mL/g to 8.60 mL/g, a more than threefold increase. Further increases to 1.6% and 2.0% result in 9.80 and 10.70 mL/g, respectively. The trend is not linear: the additional gas per unit resin decreases as the resin content increases, possibly due to incomplete decomposition or different decomposition kinetics. Nevertheless, from a practical standpoint, keeping the resin content as low as possible is beneficial for reducing gas porosity.
I also analyzed the gas evolution rate profile. The initial rate of gas release is rapid during the first 20 s, followed by a gradual plateau. This behavior is important for low-pressure casting because the mold filling time is typically 10–30 s. If the gas evolution peak coincides with the solidification of the metal front, the gas has a higher chance of being entrapped. Therefore, optimizing the resin content to reduce the peak gas evolution rate is crucial.
3.3 Effect of Resin Content on Permeability
Permeability is a measure of how easily gas can flow through the porous sand mold. It depends on the void fraction, the connectivity of pores, and the resistance offered by the binder films. Table 3 presents the permeability results for different resin contents.
| Resin content (%) | Permeability (cm⁴/(g·min)) |
|---|---|
| 0.8 | 38.5 |
| 1.2 | 35.2 |
| 1.6 | 33.8 |
| 2.0 | 32.1 |
The permeability decreases slightly as the resin content increases, but the overall change is small, from 38.5 cm⁴/(g·min) at 0.8% to 32.1 cm⁴/(g·min) at 2.0%. This is because the resin volume fraction is relatively low, and the sand grain packing is not significantly altered by the resin content in the range studied. The dominant factor for permeability is the sand grain size distribution and shape, which were kept constant. Thus, for the purpose of this work, the permeability is not highly sensitive to resin content within the practical range. However, when a coating is applied, the permeability can be drastically reduced, as I will discuss in the next section.
3.4 Effect of Coating Process on Gas Evolution
Applying a refractory coating is essential for achieving good surface finish and preventing sand erosion. However, the coating introduces additional organic materials that decompose at high temperatures and contribute to gas evolution. I investigated the effect of coating dilution (amount of water added) and immersion time on the gas evolution of the coated sand specimens. The resin content of the base sand was fixed at 1.2%.
Figure 4 presents the gas evolution results as a function of water addition and immersion time. Table 4 lists the gas evolution values for all conditions.
| Water added (mL) | Immersion time (s) | Gas evolution (mL/g) |
|---|---|---|
| 0 | 2 | 10.2 |
| 0 | 4 | 10.5 |
| 0 | 8 | 11.1 |
| 100 | 2 | 9.4 |
| 100 | 4 | 9.2 |
| 100 | 8 | 10.0 |
| 200 | 2 | 8.9 |
| 200 | 4 | 9.1 |
| 200 | 8 | 9.8 |
| 300 | 2 | 9.7 |
| 300 | 4 | 9.6 |
| 300 | 8 | 10.6 |
From the data, I observed that the gas evolution of coated specimens is generally higher than the uncoated base sand (8.6 mL/g for 1.2% resin). This is due to the additional organic binders and additives in the coating. However, the gas evolution is not monotonically related to water dilution. As the water addition increases from 0 to 200 mL, the gas evolution tends to decrease slightly, likely because the coating becomes thinner and less solid material is deposited on the sand surface. When the water addition reaches 300 mL, the gas evolution increases again. This may be due to the coating slurry becoming non-homogeneous or the binder components concentrating differently during drying. The immersion time also matters: longer immersion (8 s) generally leads to higher gas evolution because more coating is absorbed into the open pores of the sand specimen.
For the best combination, I found that 200 mL water addition and 4 s immersion gives a gas evolution of 9.1 mL/g, which is only slightly above the uncoated value. This condition also provides adequate coating thickness for surface protection, as will be confirmed in production trials.
3.5 Effect of Coating Process on Permeability
The coating, when dried, forms a relatively dense layer on the surface of the sand mold. This layer can significantly reduce the permeability of the mold. Table 5 shows the permeability measurements for the same set of coating conditions.
| Water added (mL) | Immersion time (s) | Permeability (cm⁴/(g·min)) |
|---|---|---|
| 0 | 2 | 12.4 |
| 0 | 4 | 11.8 |
| 0 | 8 | 10.2 |
| 100 | 2 | 18.5 |
| 100 | 4 | 17.9 |
| 100 | 8 | 15.6 |
| 200 | 2 | 33.6 |
| 200 | 4 | 32.0 |
| 200 | 8 | 28.4 |
| 300 | 2 | 36.8 |
| 300 | 4 | 35.0 |
| 300 | 8 | 31.2 |
The results clearly demonstrate that the coating dramatically reduces permeability. For the undiluted coating (0 mL water), the permeability drops to about 12 cm⁴/(g·min), which is only one-third of the uncoated value. Increasing the water dilution improves permeability, because the coating layer becomes thinner and more porous after drying. At 200 mL water, the permeability reaches about 32–34 cm⁴/(g·min), which is close to the uncoated base sand (35 cm⁴/(g·min)). At 300 mL water, the permeability is even slightly higher, but the coating may be too thin to provide adequate surface protection.
Immersion time also affects permeability. Longer immersion times result in lower permeability, especially for the less diluted coatings. This is because the coating slurry penetrates deeper into the sand pores and, after drying, fills more of the void space. For a given dilution, an immersion time of 2–4 s is preferred to balance coating pick-up and permeability.
3.6 Effect of Coating Viscosity (Baumé Degree) on Permeability
In production, coating viscosity is often measured by Baumé degree. I prepared coatings with different Baumé degrees by diluting the original coating with water. The permeability of coated specimens was measured, and the results are shown in Table 6 and visually in Figure 6.
| Baumé degree (°Bé) | Permeability (cm⁴/(g·min)) | Visual observation |
|---|---|---|
| 14 | 40.2 | Thin coating, poor coverage |
| 16 | 37.8 | Thin coating, some surface exposed |
| 20 | 32.0 | Uniform coating, good coverage |
| 24 | 26.5 | Thick coating, good coverage |
| 30 | 18.3 | Very thick coating, cracking after drying |
When the Baumé degree is below 20°Bé, the coating is too thin; it does not provide sufficient refractory layer to prevent sand burning and metal penetration. When the Baumé degree is above 24°Bé, the coating becomes too thick, leading to reduced permeability and also cracking and peeling during drying. I observed that at 30°Bé, the dried coating layer exhibited obvious cracks and detachment, which could cause defects such as sand inclusion and cold shut. Therefore, I recommend a Baumé degree in the range of 20–24°Bé. Within this range, the permeability is still acceptable, and the coating provides the necessary protective function.
3.7 Combined Optimization and Mathematical Model
To quantitatively describe the trade-off between gas evolution and permeability, I introduce a performance index \(P\) that combines both factors. A common approach is to define a venting capability index \(V\) as follows:
$$ V = \frac{k}{\dot{G}_g} $$
where \(k\) is the permeability (cm⁴/(g·min)) and \(\dot{G}_g\) is the gas evolution rate (mL/(g·min)) measured at the peak. The higher the \(V\) value, the better the mold’s ability to vent gases. Based on my measurements, I calculated \(V\) for several key conditions, as shown in Table 7.
| Condition | Permeability (cm⁴/(g·min)) | Gas evolution (mL/g) | Peak gas rate (mL/(g·min)) | V (min·cm⁴/(g·mL)) |
|---|---|---|---|---|
| 1.2% resin, no coating | 35.2 | 8.6 | 12.8 | 2.75 |
| 1.2% resin, coating 20°Bé, 2 s dip | 33.6 | 8.9 | 13.5 | 2.49 |
| 1.2% resin, coating 20°Bé, 4 s dip | 32.0 | 9.1 | 14.0 | 2.29 |
| 1.2% resin, coating 20°Bé, 8 s dip | 28.4 | 9.8 | 15.2 | 1.87 |
| 1.2% resin, coating 24°Bé, 4 s dip | 26.5 | 9.3 | 14.5 | 1.83 |
From Table 7, the uncoated specimen naturally has the highest venting capability. Among the coated conditions, the 20°Bé coating with a 2 s dip provides the best \(V\) value of 2.49, followed closely by the 4 s dip with 2.29. Although the 2 s dip is slightly better in venting, the coating layer may be too thin for complex internal oil passages where erosion resistance is critical. Therefore, I selected the 4 s dip as the compromise for production.
Additionally, I derived a simple empirical relationship to estimate the permeability of coated 3d sand printing molds as a function of the coating Baumé degree \(B\) (in °Bé) and the immersion time \(t\) (in s). Based on the experimental data, the following equation was fitted:
$$ k(B,t) = k_0 – \alpha (B – B_0)^2 – \beta t $$
where \(k_0\) is the permeability at the optimal Baumé degree \(B_0 = 20\) °Bé, \(\alpha = 0.12\) cm⁴/(g·min·°Bé²), and \(\beta = 1.2\) cm⁴/(g·min·s). This model reproduces the observed trends for \(B\) in the range 16–24 °Bé and \(t\) in the range 2–8 s. The parabolic dependence on \(B\) reflects the fact that both too low and too high viscosity reduce the effective open porosity. The linear term in \(t\) accounts for the progressive filling of pores by the coating.
Using this model, I can predict that at \(B = 20\) °Bé and \(t = 4\) s, the permeability is \(k = 35.2 – 0.12 \times 0 – 1.2 \times 4 = 30.4\) cm⁴/(g·min), which is close to the measured value of 32.0 cm⁴/(g·min). The slight discrepancy may be due to the nonlinearity of the absorption process. Nevertheless, the model provides a useful guideline for process design.
3.8 Production Validation
Based on the laboratory results, I implemented the optimized parameters in the foundry: resin content of 1.2%, coating Baumé degree of 20–24°Bé (with 200 mL water addition per unit volume of concentrated coating), and immersion time of 4 s. The 3d sand printing molds were produced for a ZL114A aluminum alloy fuel housing with complex internal channels. Low-pressure casting was performed using a standard low-pressure casting machine with a filling pressure of 0.06–0.08 MPa and a holding time of 300 s.
Before optimization, the fuel housings exhibited a high scrap rate due to gas porosity, especially in thick sections and at the end of the filling path. After applying the optimized process, I examined the castings using X-ray radiography and metallographic analysis. No gas porosity defects were observed in any of the critical areas. The internal channels had a clean surface with minimal sand adhesion, and the dimensional accuracy was within the specified tolerance.
The production validation confirmed that the combination of moderate resin content and an optimized coating process provides an excellent balance of mechanical strength, gas evolution, and permeability for 3d sand printing molds used in low-pressure casting. The elimination of gas porosity can be attributed to the sufficient venting capability of the mold, which allowed the evolved gases to escape through the porous sand before the metal solidified.
4. Conclusions
In this work, I have systematically evaluated the effects of resin content and coating process parameters on the properties of 3d sand printing molds for low-pressure casting of fuel housings. The following conclusions can be drawn:
(1) Resin content significantly influences both the mechanical strength and gas evolution of 3d sand printing molds. Increasing the resin content from 0.8% to 2.0% raises the compressive strength from 2.11 MPa to 4.42 MPa, but also increases the gas evolution from 2.83 mL/g to 10.70 mL/g. The permeability is only slightly affected in this range. A resin content of 1.2% provides a compressive strength of 3.6 MPa and a gas evolution of 8.6 mL/g, which is suitable for low-pressure casting.
(2) Coating application increases gas evolution and decreases permeability of the sand mold. For a given coating, the immersion time is the dominant factor. A short immersion time (2–4 s) with a diluted coating (200 mL water addition) yields the best combination, with gas evolution around 9.1 mL/g and permeability around 32 cm⁴/(g·min). Longer immersion times (8 s) significantly reduce permeability and increase gas evolution.
(3) Coating viscosity, expressed as Baumé degree, must be carefully controlled. Coatings below 20°Bé are too thin to provide adequate surface protection, while coatings above 24°Bé cause excessive permeability loss and drying defects such as cracking. The optimal Baumé degree range is 20–24°Bé. A simple empirical model relating permeability to Baumé degree and immersion time was developed and can be used for process optimization.
(4) The optimized process, consisting of 1.2% resin content and a coating with 20°Bé and 4 s immersion time, was successfully verified in production. ZL114A aluminum alloy fuel housings produced by low-pressure sand casting using 3d sand printing molds showed complete elimination of gas porosity defects. These results demonstrate that 3d sand printing combined with proper coating practice is a robust manufacturing route for complex aerospace fuel system components.
The findings from this study provide a practical reference for foundries using 3d sand printing in low-pressure casting. Future work will focus on predicting gas flow during casting using computational fluid dynamics and integrating the coating model into the mold design phase. I believe that 3d sand printing will play an increasingly important role in the production of high-integrity aluminum castings for aerospace applications.