In the production of high-value, safety-critical steel castings such as railway wheels, the selection and control of molding sand is not merely a procedural step but a fundamental determinant of quality, cost, and environmental footprint. My extensive experience in this field has been shaped by the constant pursuit of robust and economical production processes. A significant chapter in this pursuit involved the evaluation and eventual integration of a new silica sand source—Tongliao sand—into our production line for heavy-section steel castings. This journey from initial trial to stable, high-volume application provided profound insights into the complex interplay between sand morphology, binder demand, thermal physics, and final casting integrity.
The conventional sand supply chain for our operations faced repeated disruptions due to resource scarcity and stringent environmental policies, threatening production stability. This necessitated the exploration of alternative sources. Tongliao sand, a wind-borne (aeolian) sand, presented itself as a viable candidate. Its promise was anchored in a reputation for high chemical purity and a spherical grain shape, but its suitability for our demanding process—involving high-temperature pours (around 1600°C) for large steel castings—was entirely unproven. Our production process employs a graphite-facing technique, where a sand liner is shot against a metal pattern/graphite assembly and hardened with CO2-sodium silicate binder before assembly and pouring. The performance of the sand liner under intense thermal load is absolutely critical.
The initial characterization of Tongliao sand revealed a distinct physical profile compared to our incumbent sand. The most striking difference was its significantly higher bulk density, a direct consequence of its grain morphology. While our standard sand exhibited angular, multi-faceted grains, Tongliao sand grains were predominantly spherical or elliptical. This roundness leads to a tighter, more efficient packing arrangement. The bulk density $\rho_b$ can be conceptually related to the solid density $\rho_s$ and the void fraction (porosity) $\epsilon$ by:
$$\rho_b = \rho_s (1 – \epsilon)$$
For a given silica solid density, a lower void fraction $\epsilon$, enabled by spherical grains, yields a higher $\rho_b$. Comparative measurements consistently showed Tongliao sand’s bulk density to be approximately 11% greater.
| Property | Incumbent Sand (Angular) | Tongliao Sand (Spherical) | Implication for Steel Castings |
|---|---|---|---|
| Average Bulk Density (g/mL) | 1.516 | 1.683 | Higher density, tighter packing. |
| Typical Angularity Factor | 1.40 – 1.45 | 1.16 – 1.20 | Spherical grains reduce interlocking. |
| Typical Air Permeability (Pa) | ~2450 | ~2100 | Lower permeability due to dense packing. |
| Binder (Na-silicate) Demand | ~5.0% | ~4.0% (for full replacement) | Lower specific surface area requires less binder. |
| Sand Resilience (Attrition) | Higher fines generation | Lower fines generation | Better resistance to pneumatic handling. |
This fundamental difference in morphology drove several secondary effects crucial for manufacturing steel castings. First, the reduced surface area of spherical grains directly lowered the sodium silicate binder requirement by about 20% for a full sand system replacement. Second, the tight packing resulted in measurably lower gas permeability, a parameter vital for allowing gases generated during the pour of steel castings to escape. Third, the high-density sand matrix exhibited higher thermal diffusivity, accelerating the cooling of the casting. This was initially observed as a higher incidence of “riser sticking,” where the feeder remained attached to the casting due to a faster solidification skin formation at the sand interface.
The initial production trials using 100% Tongliao sand were immediately problematic. While the anticipated benefit of reduced burn-on (metal penetration) on the steel castings was confirmed, a new and severe defect emerged: a high frequency of “water-mark” or veining defects on the cast wheel surfaces. These defects, appearing as fins or buckles on flat sections like the wheel web, are classic symptoms of sand expansion under heat. The rapid heating of the dense sand liner by the molten steel causes significant thermal expansion. The inability of the rigid, CO2-hardened sand mass to accommodate this stress internally leads to micro-cracking or plastic deformation at the mold-metal interface, into which liquid metal penetrates. The linear expansion of silica sand is non-linear and peaks around the α- to β-quartz phase transition at 573°C, a temperature easily exceeded at the surface of steel castings. The stress $\sigma$ generated can be related to the constrained expansion:
$$\sigma = E \cdot \alpha \cdot \Delta T$$
where $E$ is the elastic modulus of the sand mass, $\alpha$ is the coefficient of thermal expansion (which spikes near 573°C), and $\Delta T$ is the temperature change. The dense, low-permeability Tongliao sand system appeared to exacerbate this mechanism.
| Defect Type (Location) | Incumbent Sand System (Defect Rate %) | 100% Tongliao Sand System (Defect Rate %) | Notes |
|---|---|---|---|
| Water Mark (Outer Web) | 0.01 | 4.63 | Catastrophic increase linked to expansion. |
| Water Mark (Inner Web) | 0.00 | 0.68 | Same expansion mechanism. |
| Inclusions/Blow (Web) | 1.71 | 0.80 | Potentially lower due to cleaner sand. |
| Burn-on (Lettering Area) | High | Low | Qualitative observation of improvement. |
This failure mode necessitated a strategic shift. Simply replacing the entire sand system was not viable for quality steel castings. The solution lay in hybridization. The goal was to blend Tongliao sand to retain its anti-burn-on benefits while diluting its negative expansion characteristics. A series of methodical experiments were conducted, varying the blend ratio and the particle size distribution of the Tongliao sand itself.
The first approach was to blend standard-grade Tongliao sand with reclaimed sand (previously used sand, thermally cycled). The hypothesis was that the “dead,” calcined reclaimed sand would have a lower expansion coefficient. A 2:1 blend (Tongliao:Reclaimed) was tested. While water-mark defects reduced slightly compared to the 100% trial, the rate remained unacceptably high at nearly 1%, indicating that the expansion contribution from the fresh Tongliao fraction was still dominant. Subsequent trials focused on blending with fresh, angular sand. A key finding emerged: a blend of only 25% standard Tongliao sand with 75% angular sand brought the overall defect rate for steel castings down to a level comparable with the baseline process, while still conferring some burn-on reduction.
To increase the utilization of Tongliao sand beyond 25%, a different variable was manipulated: the grain fineness. The original Tongliao sand had an Average Grain Fineness (AFS GFN) similar to our standard sand (around 46). We procived coarser grades. The logic was two-fold: 1) Coarser grains naturally create a more open, higher-porosity structure, increasing permeability and potentially providing micro-space to accommodate expansion. 2) Coarser grains have a lower specific surface area, further optimizing binder usage. The void fraction $\epsilon$ for a packed bed of spheres is relatively constant, but the mean pore throat size $d_t$ scales with the grain diameter $d_p$:
$$d_t \propto d_p$$
A larger $d_t$ improves gas escape and may reduce back-pressure during metal infiltration into sand cracks.
We tested coarse Tongliao sand (AFS GFN 41-47) and a very coarse grade (AFS GFN 38-40) at a 50% blend ratio. The results were illuminating and formed the cornerstone of the final production solution.
| Sand Blend Composition | Avg. Fineness of Tongliao Sand | Total Relevant Defect Rate (%) | Water Mark Defect Rate (%) | Severe Burn-on (Lettering) Rate (%) |
|---|---|---|---|---|
| 50% Tongliao / 50% Angular | 44-50 (Standard) | 2.30 | 0.47 | 1.25 |
| 50% Tongliao / 50% Angular | 41-47 (Coarse) | 2.11 | 0.30 | 0.96 |
| 50% Tongliao / 50% Angular | 38-40 (Very Coarse) | 2.32 | 0.16 | 1.63 |
The data revealed an optimal point. The very coarse sand (38-40 GFN) virtually eliminated water-mark defects, proving the expansion was successfully mitigated. However, it caused an increase in severe burn-on, likely because the excessively large pores in the sand matrix were more easily penetrated by molten metal. The standard-grade blend still carried a moderate water-mark risk. The coarse-grade blend (41-47 GFN) struck the perfect balance: it reduced water-mark defects significantly compared to the standard-grade blend, maintained a low overall defect rate, and kept burn-on at a manageable minimum. This blend, with a sodium silicate percentage stabilized at 4.8% (slightly lower than the original 5.0%), was subsequently adopted for full-scale production.
The long-term performance of this optimized sand system for steel castings has been exceptional. Over a period encompassing hundreds of thousands of castings, the process has demonstrated remarkable stability. The targeted benefits have been fully realized:
| Metric | Original Sand System | Optimized System (50% Coarse Tongliao) | Impact |
|---|---|---|---|
| Overall Relevant Defect Rate | 2.78% | 2.51% | Comparable quality, reliable production. |
| Binder Consumption | 5.0% | 4.8% | ~4% reduction, leading to annual cost savings and improved sand reclamation. |
| Burn-on Severity (Area Index) | 2.56% | 1.96% | ~23% reduction, drastically lowering grinding/fettling costs and abrasive dust generation. |
| Supply Chain Risk | High | Mitigated | Dual-source strategy ensures production continuity. |
The reduction in burn-on is particularly significant for the economics and environmental impact of producing steel castings. Less metal penetration means less manual grinding is required post-casting, saving hundreds of thousands annually in labor and consumable costs. Furthermore, it dramatically reduces the amount of metallic dust released into the workshop air, easing the burden on filtration systems and improving the workplace environment. The lower binder demand also contributes to easier reclamation of the sand, closing the loop on material usage.
In conclusion, the successful integration of Tongliao sand into our process for manufacturing high-grade steel castings was not a simple substitution but a systematic re-engineering of the sand system. It underscored that the properties of foundry sand are a holistic system: morphology dictates packing, which influences permeability, binder demand, and thermal transfer. The spherical, dense Tongliao sand offered superb refractoriness and low burn-on tendencies but introduced a severe thermal expansion challenge under the conditions specific to large steel castings. By intelligently blending it with angular sand and carefully coarsening its grain size, we created a composite system that retained the advantages while suppressing the disadvantages. The final 50% coarse Tongliao sand blend delivers consistent quality, reduces consumable costs, minimizes finishing labor, and enhances environmental performance. This case study exemplifies the meticulous, science-informed approach required to advance the art and efficiency of producing demanding steel castings.