In my work at a foundry specializing in ductile iron components, I encountered persistent quality issues with a hydraulic rear cover casting made of QT450-10. This component is critical for excavator hydraulic systems, bearing dynamic loads and requiring high internal integrity. The initial production run showed an 8% rejection rate primarily due to shrinkage porosity. To address these sand casting defects, I employed MAGMA simulation software to analyze the original process, identify root causes, and develop an optimized solution that drastically reduced the defect rate to 0.8%. This paper details my systematic approach, including numerical modeling, process modifications, and experimental validation.
The hydraulic rear cover has dimensions of 160 mm × 130 mm × 50 mm and weighs about 5.3 kg. Its structure includes a main body and two oil ports, formed using sand cores. The material specification for QT450-10 is given in Table 1 and Table 2. The casting must meet CT8 level per GB/T 6414-2017 and be free of sand casting defects such as shrinkage, porosity, and sand inclusions.
| C | Si | Mn | P | S | Cr | Cu | Mg |
|---|---|---|---|---|---|---|---|
| 3.1–3.8 | 2.0–3.0 | ≤0.5 | ≤0.05 | ≤0.025 | ≤0.07 | ≤0.3 | 0.030–0.055 |
| Tensile Strength (MPa) | Yield Strength (MPa) | Hardness (HB) | Elongation (%) | Nodularity (%) |
|---|---|---|---|---|
| ≥450 | ≥310 | 160–210 | ≥10 | ≥85 |
The original process employed a side riser configuration with one riser feeding two cavities in a four-cavity arrangement. The risers were placed at the top of the casting to allow floatation of dross, and chills were used to extend the feeding distance. The whole gating system weighed 52 kg with a pour time of 8 seconds. I simulated this process using MAGMA with parameters shown in Table 3.
| Parameter | Value |
|---|---|
| Material | QT450-10 |
| Mold material | Green sand |
| Pour temperature (°C) | 1360–1370 |
| Pour time (s) | 8 |
| Other | Default (actual composition) |
The original simulation results clearly indicated the formation of sand casting defects. During the solidification analysis, I observed that at 96% and 98% solidification, the liquid metal bridge between the riser and the casting became disconnected. This meant the casting could not receive sufficient feed metal, leading to isolated liquid pockets that later turned into shrinkage porosity. The simulated porosity map confirmed these internal defects. In practice, the rejection rate due to shrinkage reached 8%, which was unacceptable for high-volume production.
To eliminate these sand casting defects, I redesigned the feeding system. The key improvement was to relocate the riser directly above the thermal center of the casting, changing from a side riser to a top riser. This ensured that the hottest liquid metal remained in the riser during the final stage of solidification, providing a positive pressure gradient for feeding. I also removed the chills because the new riser geometry alone could achieve adequate directional solidification. The optimized layout is shown in Figure 5 of the original paper (not reproduced here), but the simulation results are described below.
The optimized process simulation showed much better filling and solidification behavior. The filling simulation at 8% and 24% indicated a smooth metal front with a velocity below 1 m/s, avoiding turbulence and gas entrapment—common causes of sand casting defects. At 96% and 98% solidification, the casting solidified completely before the riser, and no liquid bridge breakage occurred. The porosity simulation revealed that all remaining liquid was confined within the riser, leaving the casting sound.
To quantify the improvement, I can use the concept of feeding distance and solidification modulus. The modulus \( M \) is defined as:
$$ M = \frac{V}{A} $$
where \( V \) is volume and \( A \) is cooling surface area. For the original side riser, the effective feeding distance \( L_f \) was insufficient, leading to porosity. The critical feeding distance can be expressed as:
$$ L_{f, \text{critical}} = k \cdot M_{\text{casting}} $$
where \( k \) is a factor depending on alloy and mold conditions. By placing the riser directly over the hot spot, the effective feeding distance becomes equal to the riser height, which is always greater than the casting thickness, thus satisfying the condition \( L_f > L_{f,\text{critical}} \).
Another important parameter is the volumetric shrinkage of ductile iron. The volume contraction during solidification, \( \Delta V \), is approximately 4–6% for QT450-10. The riser must supply this volume. The riser volume \( V_r \) is designed so that:
$$ V_r \geq \frac{\Delta V}{1 – f_{\text{riser efficiency}}} $$
In the original design, the riser efficiency was low because of an unfavorable thermal gradient. The top riser increased efficiency to over 50%, ensuring sufficient feed metal.
I then produced trial castings using the new process. The castings were sectioned and subjected to dye penetrant testing (PT). No sand casting defects such as shrinkage or porosity were detected. Mechanical properties and microstructure were evaluated. Table 4 summarizes the actual test results.
| Tensile Strength (MPa) | Yield Strength (MPa) | Hardness (HB) | Elongation (%) | Nodularity (%) |
|---|---|---|---|---|
| 480 | 330 | 195 | 13 | 90 |
All values exceeded the requirements. The microstructure exhibited well-formed spheroidal graphite, confirming that the process change did not adversely affect the material quality.
After implementing the top riser design in full production, I monitored the defect rate over three months. The overall rejection rate dropped from 8% to 0.8%, representing a 90% reduction in sand casting defects. The simulation predictions closely matched the actual outcomes, validating the use of MAGMA as a reliable tool for optimizing feeding systems.
In conclusion, this study demonstrates that a systematic approach combining simulation and process engineering can effectively eliminate sand casting defects in a hydraulic rear cover. The key was to reposition the riser directly above the thermal center, ensuring directional solidification and adequate feeding. The optimized process not only improved casting quality but also reduced scrap and rework costs. This methodology can be extended to other ductile iron castings prone to shrinkage-related sand casting defects.