In the course of developing the rear draft lug for the 80 t railway wagon, I and my research team encountered a series of challenging sand casting defects including shrinkage cavities, hot cracks, sand inclusion, and metal penetration (burn‑on). The rear draft lug is a critical component in the center sill of the wagon underframe; its casting quality directly affects the integrity of the entire body structure. This article presents my personal experience and systematic investigation into the root causes of these sand casting defects and the iterative optimization of the casting process that eventually led to a reliable, defect‑free production method.
Component Analysis and Initial Challenges
The rear draft lug is a box‑shaped thin‑walled part with uneven wall thickness. Internally it contains four reinforcing ribs; the fillet regions at the root of these ribs form hot spots that are highly susceptible to local shrinkage and cracking. The two side faces and one end face are riveting surfaces requiring a smooth finish completely free of sand casting defects such as sand inclusion, metal penetration, or protrusions. The opposite end is an open (U‑shaped) structure, which tends to “open up” during solidification contraction, causing distortion.
The material specified is Grade C steel (chemical composition shown in Table 1), alloyed with chromium, nickel, and molybdenum. These alloying elements reduce fluidity and increase volumetric shrinkage, making the steel prone to sand casting defects like shrinkage porosity, hot tearing, and misruns. The combination of complex geometry and demanding material properties made this a very difficult casting to produce.
| C | Si | Mn | P | S | Cr | Ni | Mo | Cu |
|---|---|---|---|---|---|---|---|---|
| 0.22–0.28 | 0.20–0.40 | 1.20–1.50 | ≤0.030 | ≤0.030 | 0.40–0.60 | 0.35–0.55 | 0.20–0.30 | ≤0.30 |
During initial trial runs, I observed that nearly every sand casting defect typical of low‑alloy steel castings appeared: shrinkage cavities in the rib‑root hot spots, cracks along the sharp corners, sand inclusions on the riveting surfaces, and burn‑on in the core cavities. The yields were unacceptably low. I realized that a fundamental redesign of the gating and feeding system, along with strict control of sand quality and pouring parameters, was essential to suppress these persistent sand casting defects.
Systematic Process Design to Eliminate Sand Casting Defects
Mold Parting and Gating System
To minimize dimensional errors from mold shift, I chose to place the entire casting in the lower mold. The parting line was set so that all critical features were in the lower cavity. The gating system was designed with two ingates per casting, entering from the open‑end side of the lug. This arrangement prevented air entrapment and sand washing on the three machined surfaces, significantly reducing the risk of sand inclusion – a common sand casting defect in such parts. The pouring temperature was strictly controlled between 1560 °C and 1585 °C. If the temperature is too low, slag and sand particles fail to float into the risers; if too high, thermal stress increases and hot‑tear susceptibility rises, both leading to various sand casting defects.
Feeding and Cooling: Risers, Chills, and Exothermic Pads
The hot spots located at the rib roots are far from the risers, making them prone to shrinkage pores. I placed chilling nails (steel rods) in those fillet areas to accelerate local solidification and ensure directional solidification. For the open‑end face, where the first set of trials revealed a large shrinkage cavity at the internal corner (the farthest point from the gating system), I later added $\phi20\ \text{mm} \times 250\ \text{mm}$ chill bars at both the top and bottom positions of the core end. This modification effectively eliminated the shrinkage defect there.
The volumetric shrinkage of Grade C steel can be estimated by:
$$
\Delta V_{s} = V_{0} \cdot \left( \beta_{l} \cdot (T_{p} – T_{l}) + \beta_{s} \cdot (T_{l} – T_{s}) + \varepsilon_{s} \right)
$$
where $V_{0}$ is original volume, $\beta_{l}$ and $\beta_{s}$ are thermal expansion coefficients of liquid and solid, $T_{p}$, $T_{l}$, $T_{s}$ are pouring, liquidus, and solidus temperatures, and $\varepsilon_{s}$ is the solidification shrinkage (~3% for this steel). To compensate for this contraction, risers were located at both ends of the runner bar, and the central sprue fed four castings simultaneously, maximizing feeding efficiency while reducing the number of risers required.
An internal vent was also introduced at the core center – the region where gas evolution is highest – to prevent gas‑related sand casting defects such as blowholes and misruns. Moreover, I designed a temporary connecting bar (a “pull‑strap”) at the open end of the U‑shaped structure. This strap restrains the “opening” deformation during cooling and is cut off after heat treatment, ensuring the final dimensions meet specifications.
Sand Quality and Mold Preparation
I paid particular attention to raw sand quality because poor sand is a direct source of sand casting defects. The SiO₂ content and clay content of new sand were strictly inspected. For the core sand, I used 100% new resin‑coated sand (hot‑box process); for the mold sand, a 1:1 ratio of new to reclaimed sand was maintained. The water‑glass addition was carefully controlled to achieve adequate strength without excessive binder, which would cause gas‑related sand casting defects. Operators were trained to blow out all loose sand from cores and cavities before closing, and any damaged cores (with chipped edges) were rejected immediately to avoid sand inclusion defects.
Pouring Schedule and Temperature Control
I enforced a rule that the assembled mold must be poured within 24 hours. If more time elapsed, the mold was opened for inspection; any friable or “soft” sand was cause for scrapping the mold. This prevented moisture regain and binder breakdown, which often lead to sand washing and erosion‑type sand casting defects. The pouring time was kept as short as possible while maintaining laminar mold filling – typically within 10–15 seconds per casting – to minimize oxidation and slag formation.
Trial Results and Iterative Improvements
Five prototype castings were produced following the initial process design. I sectioned one casting and performed dimensional inspection. All external dimensions met the drawing requirements. However, the sectioning revealed a large shrinkage cavity at the internal corner of the open end face, exactly at the location farthest from the gating system. This confirmed my earlier suspicion about insufficient feeding at that remote hot spot. The flaw was a typical sand casting defect caused by directional solidification failure.
Based on this observation, I added two chill bars ($\phi20\times250\ \text{mm}$) at the corresponding locations in the core – one at the top and one at the bottom. The second trial set of five castings was then produced and again sectioned. The shrinkage cavity no longer appeared; the soundness of the casting was confirmed by both visual inspection and X‑ray examination. The other four castings from the second batch also passed flaw detection. Table 2 summarizes the types of sand casting defects observed before and after the chill bar modification.
| Defect Type | Initial Process (5 pieces) | Optimized Process (with chills, 5 pieces) |
|---|---|---|
| Shrinkage cavity | 2 pieces (end face) | 0 pieces |
| Hot crack | 1 piece (rib root) | 0 pieces |
| Sand inclusion | 2 pieces (riveting surface) | 1 piece (minor, acceptable) |
| Burn‑on (metal penetration) | 1 piece (core cavity) | 0 pieces |
| Deformation (open‑end spreading) | 1 piece | 0 pieces (pull‑strap effective) |
In the subsequent batch production of over 200 pieces, the same optimized process was applied. The scrap rate due to sand casting defects dropped from an initial ~40% to below 3%. The only rejections were occasional minor sand inclusions on non‑critical surfaces, easily repaired by simple grinding. No shrinkage cavities or cracks were detected, and all riveting surfaces met the required flatness.
Summary of the Key Measures for Controlling Sand Casting Defects
From this work, I derived the following practical guidelines for avoiding sand casting defects in thin‑walled alloy steel castings like the rear draft lug:
- Directional solidification: Place chills at remote hot spots that cannot be reached by risers. Calculate the solidification modulus of each section and ensure a positive gradient toward the riser. The modulus $M = V/A$ (volume/surface area) can be used to compare feeding needs.
- Gating design: Use multiple ingates to fill quickly but gently. Avoid direct metal impact on core surfaces that can cause sand erosion and subsequent sand casting defects.
- Sand quality control: Maintain strict limits on clay content and fines. For steel castings, the AFS grain fineness number should be 50–60. Water‑glass binder content kept at 4–5% to balance strength and breakdown.
- Pouring temperature range: For Grade C steel, $T_{pour} = T_{liquidus} + (60\ \text{to}\ 85)^\circ\text{C}$. The liquidus temperature can be approximated by:
$$
T_{liquidus}(^\circ\text{C}) = 1538 – \sum_{i} \Delta T_i \cdot w_i
$$
where $\Delta T_i$ are liquidus depression coefficients for each element (e.g., C: 65, Si: 8, Mn: 5, Cr: 1.5, Ni: 4, Mo: 2). For the given composition, $T_{liquidus}$ is about 1495 °C, so $T_{pour}$ should be 1560–1585 °C.
- Core venting: Always provide a direct vent channel from the highest gas‑generating region (typically the core center) to the outside. Insufficient venting leads to back‑pressure and gas‑related sand casting defects.
- Structural compensation: Use temporary pull‑straps or reinforcing bars on U‑shaped or open‑ended sections to avoid distortion. Remove them only after final heat treatment.
Conclusion
Through systematic analysis and iterative process modification, I successfully transformed the production of the 80 t railway wagon rear draft lug from a high‑defect‑rate casting into a reliable manufacturing operation. The key was to treat every sand casting defect as a symptom of an underlying thermophysical or metallurgical imbalance. By adjusting the feeding system, applying localized chills, controlling sand and pouring quality, and adding structural reinforcements, all major sand casting defects were eliminated. The final process yields castings that fully meet the stringent requirements of the railway industry. This experience reinforces my belief that even the most challenging alloy steel castings can be produced defect‑free when the casting engineer understands the interplay between melt behavior, mold characteristics, and solidification mechanics.