Development of a Lost Foam Casting Process for High‑Chromium Impurity Pump Impellers

## Introduction

When I first encountered the challenge of developing impellers for impurity pumps used in hydrometallurgical processes, I realized that this was not simply another casting task. The pumps are required to transport solid–liquid slurries with high chloride ion concentrations, pH values between 1 and 2, and hard particulate contents of 20%–35%. The impeller, as the core rotating component, experiences severe acid corrosion, especially pitting attack induced by chloride ions, combined with abrasive wear, erosive wear, and contact fatigue caused by the hard particles. These combined degradation mechanisms dramatically reduce the service life of the impeller and cause material waste, energy loss, and increased operational costs.

The copper–cobalt mining project in the Democratic Republic of the Congo (DRC) is a typical hydrometallurgical process. For many years, the German company KSB and the Australian company WARMAN dominated this market, strictly controlling their technology and keeping it confidential. Even their impeller life was only 3 to 4 months, and the price was high, with long delivery times. Due to technical limitations in both materials and manufacturing processes, there was no high-quality domestic supplier for this severe acid corrosion and high wear condition. In 2018, our company, with 35 years of professional experience in impurity pump manufacturing and strong technical advantages, signed a trial production contract for impeller spare parts with the DRC copper–cobalt mine, entering the African metallurgical market and enhancing our international reputation.

In this paper, I present my systematic research on the development of a new acid-resistant and wear-resistant FeCr40 ultra-high chromium material and the corresponding lost foam casting process. Through this development, we successfully produced high-quality impellers with a service life of 5 to 7 months, exceeding the lifetime of impellers manufactured by KSB and WARMAN, thereby breaking the international monopoly. Today, our company maintains a long-term cooperation with the DRC copper–cobalt mine, expanding from simple spare part supply to the supply of complete impurity pumps, with annual sales exceeding 100 million RMB.

## Development of the FeCr40 New Material

The first and most critical step was to develop a new metallic material capable of resisting both strong acid corrosion and severe wear, satisfying the hydrometallurgical working conditions and providing a long service life for the components.

### Performance Requirements of the Material

Based on the process parameters of the hydrometallurgical slurry medium, the new material needed to meet the following performance requirements:

– Good resistance to acid corrosion, especially resistance to chloride ion pitting;
– High hardness to resist the scouring and wear of hard solid particles in the slurry;
– Sufficient strength to cope with various stresses since the impeller operates at high rotational speeds;
– Good impact toughness so that the component does not develop cracks when conveying acidic fluids.

Table 1 summarizes these key performance requirements:

### Microstructural Requirements

Using the liquidus surface projection of the Fe–C–Cr ternary system, I determined the solidification structures produced by various alloy compositions. In the Fe–C–Cr system, at the U₁ point of 1275 °C, the four-phase equilibrium peritectic–eutectic transformation occurs:

\[
L + \delta\text{-Fe} \rightleftharpoons \text{M}_{23}\text{C}_6 + \gamma\text{-Fe}
\]

Based on the chemical composition, the carbide volume fraction can be theoretically calculated using the F. Maratray relationship:

\[
\text{Carbide volume fraction} = 12.33 \times \omega(\text{C}) + 0.55 \times \omega(\text{Cr}) – 15.2\%
\]

For the FeCr40 material with low carbon and high chromium content, \(\text{M}_{23}\text{C}_6\) carbides are formed. A high carbide volume fraction of 30%–35% ensures that the material possesses excellent wear resistance.

The eutectic carbon content of the material is determined by:

\[
\text{Eutectic carbon content} = 4.4\% – 0.054 \times \omega(\text{Cr})
\]

For FeCr40 material, the eutectic carbon content lies between 2.3% and 2.8% by mass, placing the alloy in the hypoeutectic high-chromium cast iron category.

During equilibrium solidification, as the temperature decreases, low-carbon austenite primarily precipitates from the liquid phase, with coarse primary austenite dendrites. With further temperature reduction, the liquid phase diminishes and carbon concentration increases, eventually reaching the eutectic reaction state:

\[
L_E \rightarrow \text{A}_E + \text{Fe}_3\text{C}
\]

The austenite precipitated from the ternary eutectic reaction subsequently precipitates secondary \(\text{M}_7\text{C}_3\) carbides and partially transforms to high-chromium ferrite, while the remaining austenite is retained at room temperature.

The austenite matrix provides excellent toughness, while the ferrite matrix offers high strength and resistance to chloride ion corrosion. Combining both phases achieves excellent corrosion resistance. The FeCr40 material contains a large number of hard alloy carbide phases, satisfying the high wear resistance requirement.

The final microstructure of the FeCr40 material consists of a dual-phase matrix of fine ferrite and austenite, with isolated, dispersed carbides distributed on the matrix.

### Chemical Composition Requirements

Using the JMatProCast software, I analyzed how composition changes affect the phase diagram and continuously optimized the types and contents of elements. The final alloy composition that meets the microstructural requirements is shown in Table 2.

| Element | C | Cr | Mo | Ni | Cu |
|—|—|—|—|—|—|
| Mass Fraction (%) | 1.7–3.2 | 33–45 | 0.5–3.0 | 3.0–6.0 | 1.0–3.5 |

#### Effects of Alloying Elements

**Carbon:** Carbon is the primary carbide-forming element. As a hard phase, increasing the carbon content significantly increases the number of carbides. A high carbide volume fraction is the foundation for ensuring material hardness and wear resistance. The carbon content was selected in the range of 1.7%–3.2%.

**Chromium:** The FeCr40 material is a low-carbon, high-chromium alloy. Part of the chromium atoms replace iron atoms to form \((\text{Fe},\text{Cr})_{23}\text{C}_6\)-type alloy carbides, which have a chromium content as high as 59% and a hardness of HV 1000–1100. This provides the microstructural guarantee for high wear resistance. Most of the chromium dissolves in the matrix. Since a large amount of chromium atoms are dissolved in the austenite, some austenite can be stabilized to room temperature, ensuring good toughness. The high Cr/C ratio of FeCr40 material provides the compositional guarantee for obtaining a ferrite matrix microstructure. When the matrix chromium content is high, chromium combines with oxygen to form a passive film, hindering interphase corrosion; it also raises the electrode potential of the matrix. Both the passive film and increased potential improve the corrosion resistance of the material, protecting the metallic material. Therefore, chromium is the most critical element for FeCr40 alloy to possess corrosion resistance and wear resistance. The chromium content was selected between 33% and 45%.

**Molybdenum:** Molybdenum dissolves in α-iron and γ-iron, strengthening the metal matrix and increasing material strength to prevent cracking. It enriches in carbides, forming high-molybdenum carbide phases, thereby improving the high-temperature chemical stability and enhancing the high-temperature wear resistance of the material. In FeCr40 material, the molybdenum content is 0.5%–3%.

**Nickel:** Nickel stabilizes austenite, preserving the high-temperature stable phase austenite at room temperature. Chromium, molybdenum, and nickel all increase the hardenability of FeCr40 material and improve its wear resistance. Nickel also enhances corrosion resistance and toughness. Furthermore, nickel strengthens the matrix, and a strong supporting matrix provides favorable conditions for carbides to exhibit excellent wear resistance. The nickel content was controlled between 3.0% and 6.0%.

**Copper:** Copper increases the electrode potential of FeCr40 material, enhancing its resistance to chloride ion corrosion in acidic media. The copper content in the FeCr40 material is 1.0%–3.5%.

### Comprehensive Evaluation of FeCr40 Material

The microstructure determines that the material has appropriate strength, high wear and corrosion resistance, and good impact toughness, satisfying the requirements of acid corrosion resistance, high wear resistance, high strength, and good impact toughness for materials in hydrometallurgical media. This provides a microstructural guarantee for the impeller to meet service requirements.

However, there are some inherent disadvantages of this material. The casting performance is poor, mainly in the following aspects:

– The material belongs to the hypoeutectic composition; the molten metal has poor fluidity and low filling capacity;
– The material has a wide solidification range and large volume shrinkage, which makes it prone to coarse structures, shrinkage cavities, and shrinkage porosity;
– The material has a high carbide content, making it brittle and prone to cracking defects;
– The material has a high alloy content, making the molten metal prone to slag inclusion and gas hole defects.

To address these drawbacks, I implemented various measures in the lost foam casting process engineering, which are discussed in the following sections.

## Lost Foam Casting Process Design for the Impeller

The impeller was manufactured using the lost foam casting process. The overall layout used a flat-molding, vertical-pouring approach. By adopting a bottom-gating system with large flow channels, balanced solidification risers for shrinkage and cracking prevention, pre-embedded sleeves for grain refinement, and other technical strategies, I successfully eliminated casting defects and ensured the production of high-quality impeller products.

The lost foam casting process for the impeller is shown in the following schematic layout:

– Two top risers for venting and slag removal;
– One impeller casting;
– Two middle risers aligned with two blades for impeller feeding;
– A bottom sprue connected to two lower blades;
– Filter mesh for flow rectification and slag entrapment.

### Gating System Design

The design of the gating system for the impeller lost foam casting was based on the large mouth flow theory. I adopted a bottom-gating, closed, multi-channel, dispersed feeding method, combined with filter meshes for flow rectification and slag prevention, to achieve steady-state pouring and rapid mold filling.

During the lost foam casting process, the molten metal front contacts the foam pattern, forming a displacement region. The metal absorbs heat during this process, and the displacement process impedes the forward movement of the molten metal, reducing its filling capacity. Since the material belongs to the hypoeutectic composition, the molten metal has poor fluidity.

In the process design, I used a bottom-gating dual-runner system with large outlet flow. The cross-sectional areas of each part of the gating system were enlarged to 1.3–2.5 times that of conventional resin sand cavity pouring. On the vertical runner below the sprue, two wide and thick horizontal runners were connected in parallel. At the ends of the horizontal runners, protruding structures were added to accumulate light impurities and reduce the amount carried into the casting.

The ingates were distributed at the junction of the two blades and the front and back shrouds. Ceramic slag-removal filter meshes were set on the ingates through which the molten metal enters the casting. The bottom-gating, dispersed, dual-runner, large-mouth flow gating system enables the molten metal to fill the mold quickly from bottom to top. The thermal field is distributed uniformly, weakening stress concentration and reducing the potential risk of brittle cracking of the impeller casting. The slag-prevention and rectification effects of the filter mesh purify the molten metal and adjust the flow pattern from turbulent to laminar, making the molten metal flow stable and significantly improving slag prevention.

### Riser Design

According to the balanced solidification theory, I conducted comprehensive riser design to achieve the dual goals of shrinkage prevention and crack suppression. The FeCr40 material has ultra-high chromium content, causing large alloy shrinkage and a wide crystallization range. The hub root of the impeller has obvious hot spots, and the junctions of the blades with the front and back shrouds form five continuous structural hot spots.

The large shrinkage attribute of the material and the hot spot structure of the part both require an enhanced riser feeding strategy to prevent shrinkage cavities and porosity. The material contains many hard and brittle alloy carbide phases, while the lost foam negative-pressure pouring process accelerates the solidification of the metal and concentrates thermal stress.

Enlarging the riser can cause double stress superimposition on the brittle material, potentially leading to brittle cracking. Therefore, the riser design must balance the need for larger risers to prevent shrinkage defects against avoiding the risk of cracking due to excessive stress concentration. This is a critical process step to ensure the production of high-quality impellers.

Considering the material characteristics, part structure, and the triple factors of the lost foam process, I performed comprehensive riser design according to the balanced solidification theory. The riser design was based on the principle of providing sufficient feed metal for shrinkage compensation while using multiple risers distributed discretely to reduce stress concentration and prevent brittle cracking. For local hot spots, dedicated riser intervention was applied to feed the hot spots and prevent shrinkage porosity, thereby achieving the dual effects of shrinkage prevention and crack prevention.

The quantitative design of the risers is as follows. The risers are arranged in two layers, with two risers in each layer. The lower risers are located near the hot spots on both sides of the hub, aligning with two blades respectively. Their primary function is to feed the hub hot spot. The risers also feed the front shroud and the impeller inlet edge through the blade passages.

For the riser quantitative design, based on the part structure and hot spot size:

\[
R_{\text{neck}} = 1.05 \sim 1.25 \times R_{\text{hot spot}}
\]

\[
R_{\text{riser}} = 3.5 \sim 5 \times R_{\text{hot spot}}
\]

\[
H_{\text{riser}} = 1.2 \sim 1.5 \times R_{\text{riser}}
\]

The two upper risers are positioned at the highest points of the blade–shroud junctions. The top risers mainly serve to weaken the adverse effects of the lost foam wall adherence effect, unblock gas venting channels, and prevent slag inclusion and gas hole defects. The top risers also provide gravity feeding from top to bottom through the front and back shroud channels.

The upper risers are slightly smaller than the lower risers:

\[
R_{\text{neck, upper}} = 1.05 \sim 1.15 \times R_{\text{hot spot}}
\]

\[
R_{\text{riser, upper}} = 1.3 \sim 1.5 \times R_{\text{hot spot}}
\]

\[
H_{\text{riser, upper}} = 1.1 \sim 1.3 \times R_{\text{riser, upper}}
\]

For riser removal, I used the hammering method. When striking, I paid attention to the force direction to avoid excessive force that might damage the casting and cause cracking.

The five blades of the impeller correspond to five heat channels: two blades correspond to ingates and three blades correspond to risers. The thermal field is uniformly distributed, the solidification of the molten metal is balanced, hot spots are fully fed, and stress concentration is weakened, thereby achieving the dual goals of shrinkage and crack prevention.

### Pre-embedded Sleeve Design

By setting a pre-embedded sleeve in the impeller shaft hole, I achieved the purpose of chill-induced grain refinement, strengthening the parent material structure, and facilitating thread machining.

Under working conditions, the torque load of the impeller is mainly transmitted through the threads in the shaft hole. Ensuring the thread strength also ensures the normal operation of the impeller during its service life. Since the FeCr40 material has high hardness, drilling and threading during machining are difficult to perform directly.

Therefore, I adopted the technical measure of installing a pre-embedded sleeve in the hub hole of the impeller. The pre-embedded sleeve acts as an internal chill, chilling the shaft hole, refining the structure, increasing material strength, and enabling the drilling and threading process to be completed on the pre-embedded sleeve.

The pre-embedded sleeve was designed with a quenched 45# steel material. If the wall thickness of the pre-embedded sleeve as an internal chill is too large, micro-cracks may form in the parent material of the impeller. If the wall thickness is too thin, the sleeve cannot withstand the torque load during impeller operation, potentially causing loosening or even detachment.

The reasonable wall thickness of the pre-embedded sleeve was designed to fully accommodate the thread structure. The sleeve has a four-layer annular band structure. The bottom is the largest and thickest, preventing axial loosening and pull-out from the parent material. The annular band surface is milled with flat edges and concave grooves arranged in a dovetail pattern to increase the fusion with the parent material and prevent radial rotation.

The pre-embedded sleeve also serves to chill the hot spot, assist the directional solidification of the riser, and improve the overall casting quality of the impeller.

Table 3 summarizes the pre-embedded sleeve design parameters:

### Casting Process Virtual Simulation

Through the simulation of the impeller lost foam casting process, I continuously adjusted process parameters and optimized the process plan, achieving the goal of short-cycle, low-cost, and high-quality physical production of the impeller.

Using the JMatProCast software, I simulated the liquid fraction during solidification. The simulation results showed that the liquidus temperature of the FeCr40 material is 1360 °C. By fully considering the impeller part size, structure, heat absorption from metal-foam displacement, deep fusion of the pre-embedded sleeve with the parent metal, and the multiple factors of riser feeding and crack prevention, I determined the pouring temperature for the FeCr40 impeller lost foam casting process to be:

\[
T_{\text{pouring}} = 1440 \pm 10 \,^\circ\text{C}
\]

Using the same method, I analyzed the thermophysical property parameters of the FeCr40 material, including solidus temperature, density, viscosity, thermal conductivity, and specific heat capacity, using JMatProCast software. These parameters were imported into the user database of the CASTSoft casting process simulation software, providing fundamental theoretical data for subsequent CASTSoft simulation of the lost foam casting filling and solidification processes.

The lost foam casting simulation process consists of pre-processing and post-processing. In the pre-processing stage, I defined the materials of the casting (impeller), chill (pre-embedded sleeve), and mold (ceramsite sand), as well as the heat transfer coefficients at the boundaries between them. Then the filling and solidification processes were simulated.

The post-processing stage outputs the simulation results. The filling simulation outputs the flow field temperature field, flow field velocity field, and flow field pressure field, presented in the form of time and percentage. The solidification simulation outputs the temperature field, cooling rate, solidification process, and shrinkage porosity, also presented in the form of time and percentage.

### Simulation Results, Optimization, and Evaluation

From the filling simulation results, in terms of the temperature field, velocity field, and pressure field, the molten metal filled the mold steadily with an appropriate filling speed. The total filling time was 20 seconds. The filling pressure remained relatively stable, and the molten metal entered the mold in a laminar flow state. This indicates that the gating system design was reasonable and could meet the requirements for molten metal filling.

From the solidification simulation results, in terms of the solidification temperature field, cooling rate, solid–liquid two-phase, and shrinkage cavities/porosity, the total solidification time was 4320 seconds. The original process showed a slight shrinkage porosity zone at the top riser A. This indicated that the riser feeding capacity was insufficient. I optimized the riser to design B by increasing the riser modulus and feeding capacity, which eliminated the shrinkage porosity defect at that location, achieving the balanced solidification goal and ensuring the quality of the impeller lost foam casting.

Table 4 presents the simulation results and the corresponding design modifications:

### Determination of Pouring Parameters

Based on the combined simulation results and material characteristics, I determined the key pouring parameters. Table 5 summarizes the final optimized process parameters:

The filling velocity during the lost foam casting process can be expressed as:

\[
v = \frac{Q}{A}
\]

where \(Q\) is the volumetric flow rate and \(A\) is the cross-sectional area of the flow channel. By designing the cross-sectional areas of the gating system components appropriately, the desired filling velocity was achieved.

## Promotion and Application of FeCr40 Material in Impurity Pumps

Starting from the successful production of the Φ500 impeller using the lost foam casting process, I promoted the industrial application of the FeCr40 material across eight pump models with diameters Φ100, Φ200, Φ300, Φ400, Φ500, Φ600, Φ800, and Φ1000. This included 29 types of flow-pass components such as impellers, volutes, and front/rear liners produced using the FeCr40 material with lost foam casting.

Our company progressed from trial production of individual spare parts for the DRC client to batch supply of complete pumps, successfully entering the African hydrometallurgical market and breaking the foreign monopoly. This demonstrated our company’s professional technical capability. Order volumes have increased year by year, creating considerable economic benefits and generating positive international social effects.

Table 6 lists the pump models and components manufactured with FeCr40 material:

| Pump Model | Impeller Diameter (mm) | Flow Components Produced |
|—|—|—|
| Φ100 | 100 | Impeller, volute, front liner, rear liner |
| Φ200 | 200 | Impeller, volute, front liner, rear liner |
| Φ300 | 300 | Impeller, volute, front liner, rear liner |
| Φ400 | 400 | Impeller, volute, front liner, rear liner |
| Φ500 | 500 | Impeller, volute, front liner, rear liner |
| Φ600 | 600 | Impeller, volute, front liner, rear liner |
| Φ800 | 800 | Impeller, volute, front liner, rear liner |
| Φ1000 | 1000 | Impeller, volute, front liner, rear liner |

### Key Technical Innovations

The success of this development can be attributed to several key technical innovations. The FeCr40 ultra-high chromium alloy impeller has a sound material composition, forming a dual-phase matrix of ferrite and austenite with isolated, dispersed carbides. This microstructure satisfies the service requirements of strong acid corrosion and severe wear in hydrometallurgical slurry media.

The lost foam casting process for impeller production incorporates five key technologies:

1. Large mouth flow for rapid mold filling;
2. Balanced solidification risers for shrinkage and crack prevention;
3. Pre-embedded sleeves for structure refinement;
4. JMatProCast software for process parameter calculation;
5. CASTSoft software for filling and solidification simulation.

These technologies eliminated shrinkage porosity, cracks, gas holes, and slag inclusion defects, enabling successful impeller production.

Table 7 summarizes the key technical measures and their functions:

## Performance Verification of the Developed Impeller

The performance of the FeCr40 ultra-high chromium impeller produced by the lost foam casting process was verified through actual field operation in the DRC copper–cobalt mine hydrometallurgical project.

Table 8 compares the service life of different impeller materials:

The FeCr40 material impeller achieved a service life exceeding that of the foreign competitors, successfully breaking the international technical monopoly. This advancement significantly improved the domestic manufacturing level of impurity pumps and promoted the application of new materials in the hydrometallurgical field.

## Conclusion

From my research and engineering practice, I can draw the following conclusions:

1. The FeCr40 ultra-high chromium alloy impeller has a well-designed material composition. The resulting microstructure of a dual-phase ferrite + austenite matrix with isolated, dispersed carbides satisfies the service requirements of strong acid corrosion and severe wear in hydrometallurgical slurry media.

2. In the lost foam casting process for impeller production, the following five key technologies were implemented:
– Large mouth flow gating system for rapid mold filling,
– Balanced solidification risers for shrinkage and crack prevention,
– Pre-embedded sleeves for structure refinement and easy machining,
– JMatProCast software for accurate calculation of process parameters,
– CASTSoft software for filling and solidification simulation.

3. These technologies successfully eliminated common casting defects including shrinkage porosity, cracks, gas holes, and slag inclusions, enabling the successful production of high-quality impellers.

4. The successful development demonstrates that combining advanced material design with optimized lost foam casting technology can produce components that outperform international competitors, break technological monopolies, and create significant economic value.

The lost foam casting process for the FeCr40 ultra-high chromium impeller not only improved the domestic manufacturing level of impurity pumps but also established a strong foundation for further applications of new materials in the hydrometallurgical field. This development represents a significant step forward in the manufacturing of wear-resistant and corrosion-resistant components for demanding industrial applications.

The experience gained from this research confirms that the lost foam casting process, when properly designed and simulated, is a highly effective manufacturing method for complex impeller geometries in highly alloyed wear-resistant materials. The combination of advanced material development and robust casting process design provides a comprehensive solution to the challenging problem of manufacturing components that must withstand simultaneous acid corrosion and abrasive wear in hydrometallurgical applications.

Furthermore, the successful application of the FeCr40 material in impurity pumps demonstrates its versatility and reliability for various flow-pass components beyond impellers, including volutes and liners. This broad applicability makes the material and the lost foam casting process valuable assets for the industry’s continued advancement toward more durable and cost-effective pump components.