In the development of BOG compressors for gas boosting and transportation, a critical challenge arose: the compressor cylinders must operate at cryogenic temperatures as low as -162°C. Traditional compressor cylinders in our experience were designed for a temperature range of 0–300°C, utilizing materials like gray iron or conventional ductile iron, which proved inadequate for these extreme conditions. This necessitated a comprehensive reevaluation of material selection. After considering key factors such as thermal expansion coefficient, we identified austenitic ductile iron, specifically grade QTANi35Cr3, as the optimal material for manufacturing BOG compressor cylinders. This article details our first-hand investigation into the manufacturing process of these low-temperature ductile iron castings, covering material preparation, process optimization, quality control, and numerical validation.
The performance of ductile iron castings at low temperatures is governed by their microstructure and chemical composition. Austenitic ductile irons, with their face-centered cubic crystal structure, retain toughness and resist embrittlement well below freezing points. For the QTANi35Cr3 grade, the high nickel content stabilizes the austenite phase, while chromium enhances specific properties. The fundamental relationship for the carbon equivalent (CE) in high-nickel ductile irons, which influences castability and graphite formation, is given by:
$$ CE = w(C) + 0.3 \times w(Si) + 0.047 \times w(Ni) – [0.0055 \times w(Ni) \times w(Si)] $$
where \( w(X) \) represents the weight percentage of element X. Controlling this value is crucial for achieving sound ductile iron castings.
1. Material Preparation and Chemical Design
The initial step involved defining the precise chemical composition to meet the standard requirements while ensuring economic viability and manufacturability. We referenced GB/T 26648-2011 for the nominal composition of QTANi35Cr3 but tailored the ranges based on practical considerations for ductile iron castings. The target was to obtain a microstructure consisting of spheroidal graphite in an austenitic matrix with minimal carbides and no deleterious graphite forms.
The rationale for each element’s control range is summarized below:
| Element | Function & Influence | Target Range (wt.%) | Rationale |
|---|---|---|---|
| Carbon (C) | Determines graphite amount/morphology, affects fluidity & shrinkage. | 2.1 – 2.2 | Optimized for medium-section thickness to ensure graphitization and feeding. |
| Silicon (Si) | Promotes graphitization, improves corrosion resistance. | 1.6 – 1.8 | Kept at standard lower limit to balance CE and avoid excessive hardening. |
| Nickel (Ni) | Stabilizes austenite, provides low thermal expansion. | 34.0 – 34.5 | Controlled at lower specification limit for cost efficiency without compromising properties. |
| Chromium (Cr) | Enhances oxidation/corrosion resistance, stabilizes austenite. | 2.0 – 2.1 | Minimized to reduce shrinkage porosity tendency in ductile iron castings. |
| Manganese (Mn) | Generally a residual; can segregate. | 1.5 – 1.7 | Kept low to prevent formation of hard phases at grain boundaries. |
| Phosphorus (P) | Impurity, forms brittle phosphides. | < 0.05 | Strictly limited to prevent intergranular corrosion and embrittlement. |
| Sulfur (S) | Impurity, interferes with nodulization. | < 0.03 | Low level essential for effective magnesium treatment. |
The final aimed chemical composition for our ductile iron castings is consolidated in Table 1.
| Element | C | Si | Mn | Ni | Cr | P | S | CE* |
|---|---|---|---|---|---|---|---|---|
| Target (wt.%) | 2.1-2.2 | 1.6-1.8 | 1.5-1.7 | 34.0-34.5 | 2.0-2.1 | <0.05 | <0.03 | 4.1-4.4 |
*CE calculated using the formula above.
2. Melting and Metallurgical Processing
Melting was conducted in a coreless medium-frequency induction furnace. The charge comprised low-sulfur pig iron, clean steel scrap, electrolytic nickel, high-carbon ferrochromium, ferrosilicon, ferromanganese, and returns. Extreme care was taken to avoid trace elements like lead (Pb) and aluminum (Al), as even 0.003% Pb can promote vermicular graphite, drastically reducing mechanical properties. Al may cause pinhole defects. Charge materials were clean, dry, and free of rust to minimize hydrogen pickup, a significant concern for austenitic ductile iron castings.
Nickel, having high gas solubility, was added late in the melt and covered to prevent excessive absorption. Alloying elements like chromium were added after spectroscopic analysis of the base iron for composition adjustment. The high melting point of this alloy required a superheating temperature approximately 50-100°C higher than typical ductile irons to ensure adequate fluidity for casting. The final pouring temperature range was set at 1450-1520°C.
The thermodynamics of gas dissolution in molten iron is relevant. The solubility of hydrogen, for instance, follows Sieverts’ law:
$$ [H] = K_H \sqrt{P_{H_2}} $$
where [H] is the dissolved hydrogen concentration, \( K_H \) is the equilibrium constant (temperature-dependent), and \( P_{H_2} \) is the partial pressure of hydrogen. This underscores the need for a protective atmosphere or rapid processing to minimize gas-related defects in ductile iron castings.
3. Nodulization and Inoculation Treatment
The heart of producing high-integrity ductile iron castings lies in the successful transformation of graphite into spheroidal form. We evaluated two nickel-magnesium master alloys for nodulization: Ni70Mg30 and Ni85Mg15. The former resulted in inferior nodularity (~82%) and lower elongation, while the latter yielded excellent nodularity (~94%) and met all mechanical specifications. Consequently, Ni85Mg15 was selected as the nodulizing agent, added at 1.0 wt.% of the treated iron.
The treatment was performed using the sandwich method (a type of pour-over technique) within a preheated treatment ladle equipped with a fume extraction system. The procedure was:
- Place nodulizer (1.0%) in the well of the ladle, cover uniformly with inoculant (1.0% FeSi alloy).
- Tap approximately two-thirds of the total melt at 1530°C into the ladle, directing the stream away from the well to initiate controlled reaction.
- After reaction subsidence, tap the remaining one-third to dilute and homogenize.
- Skim slag, take a wedge test sample for quick visual assessment of nodulization.
- Proceed to pouring if successful; otherwise, perform corrective treatment.
Inoculation was critical for graphite nucleation and preventing chilling. A dual inoculation practice was adopted: primary (ladle) inoculation with the nodulizer, and secondary (late) inoculation via stream inoculation during pouring to combat fading.
The kinetics of nodulization can be described by a first-order rate equation for magnesium fade:
$$ [Mg]_t = [Mg]_0 \cdot e^{-kt} $$
where \( [Mg]_t \) is the residual magnesium at time \( t \), \( [Mg]_0 \) is the initial magnesium after treatment, and \( k \) is the fade rate constant, which is higher for austenitic ductile irons due to their higher melting temperature, necessitating rapid pouring.
The microstructure of successfully produced ductile iron castings showed spherical graphite (Type I & II ≥90%) in an austenitic matrix with minimal carbides.
4. Material Properties: Room and Cryogenic Temperature
Test castings were produced to evaluate the mechanical and physical properties of the developed material. The results at room temperature are presented in Table 2.
| Property | Tensile Strength (MPa) | Yield Strength (0.2% Offset, MPa) | Elongation (%) | Hardness (HBW) |
|---|---|---|---|---|
| Value Range | 370 – 450 | 220 – 240 | 7 – 10 | 150 – 180 |
Cryogenic testing was conducted at 103 K (-170°C). The results, shown in Tables 3 and 4, confirm the material’s suitability for low-temperature service.
| Sample | Elongation at Break (%) | Tensile Strength (MPa) | Elastic Modulus (GPa) |
|---|---|---|---|
| 1 | 7.2 | 372 | 145 |
| 2 | 7.3 | 373 | 149 |
| Sample | Length Change ΔL/L303K (%) |
|---|---|
| 1 | 0.0909 |
| 2 | 0.1150 |
The low thermal expansion coefficient is a key advantage for dimensional stability in cryogenic ductile iron castings. The average coefficient (\( \alpha \)) over the temperature range can be calculated from the data:
$$ \alpha_{avg} = \frac{\Delta L / L_0}{\Delta T} $$
where \( \Delta T = 200 \, K \). For Sample 1, \( \alpha_{avg} \approx 4.55 \times 10^{-6} \, K^{-1} \), confirming its low expansion nature.
5. Casting Process Design for the Compressor Cylinder
The compressor cylinder is a pressure-retaining component with complex geometry. The casting design prioritized uniform wall thickness to minimize thermal gradients and shrinkage defects, which are pronounced in high-nickel ductile iron castings due to their mushy solidification mode. The designed cylinder had a nominal weight of 1930 kg.
Given the significant volumetric shrinkage (comparable to steel), a rigorous feeding system was mandatory. The casting process employed a top riser on the cylinder body, accounting for approximately 30% of the casting weight (580 kg). The total poured weight was about 2500 kg. The gating system was designed for rapid, turbulent-free filling to avoid oxide entrapment and gas pickup.
To validate the process, numerical simulation using ProCAST software was performed. The simulations analyzed filling patterns, temperature gradients, solidification sequences, and defect prediction (shrinkage porosity). The results confirmed that the designed riser provided adequate feed metal to compensate for shrinkage throughout the solidification period, ensuring sound ductile iron castings. The solidification time (\( t_s \)) for a section can be estimated using Chvorinov’s rule:
$$ t_s = k \left( \frac{V}{A} \right)^n $$
where \( V \) is volume, \( A \) is surface area, \( k \) is a mold constant, and \( n \) is an exponent (often ~2). Simulations helped optimize the modulus \( (V/A) \) of different sections to achieve directional solidification towards the riser.
6. Process Optimization and Quality Control Metrics
Manufacturing high-quality ductile iron castings for cryogenic service requires tight control over every process variable. We established a multivariate control system. Key relationships between process inputs and casting quality were modeled. For instance, the nodularity percentage (\( N \)) as a function of residual magnesium (\( Mg_{res} \)) and cooling rate (\( \dot{T} \)) can be expressed empirically:
$$ N = a \cdot Mg_{res} + b \cdot \ln(\dot{T}) + c $$
where \( a, b, c \) are material-specific constants. Maintaining \( Mg_{res} \) between 0.04-0.06% and optimizing cooling through mold design was crucial.
Table 5 outlines the critical process control parameters and their monitored ranges for producing these specialized ductile iron castings.
| Process Stage | Parameter | Control Range | Target / Rationale |
|---|---|---|---|
| Melting | Final C Content | 2.15 ± 0.05% | Ensure correct composition, minimize gas. |
| Melt Superheat | ≥ 1550°C | ||
| Hold Time at Temp | < 10 min | ||
| Treatment | Nodulizer Type/Amount | Ni85Mg15, 1.0±0.1% | Achieve nodularity >90%, effective inoculation. |
| Inoculant Amount | 1.0±0.1% (FeSi75) | ||
| Pouring | Temperature | 1475 ± 25°C | Balance fluidity and gas evolution. |
| Pouring Time (for 2500kg) | < 60 seconds | ||
| Solidification | Riser Modulus Ratio | > 1.2 (Riser/Casting) | Adequate feeding for shrinkage. |
Statistical process control (SPC) charts were maintained for key chemical and mechanical properties to ensure consistency across multiple heats of ductile iron castings.
7. Advanced Analysis: Thermodynamic and Kinetic Modeling
To deepen understanding, we applied thermodynamic software (e.g., using CALPHAD databases) to predict phase stability in the QTANi35Cr3 system. The driving force for graphite nucleation (\( \Delta G_{nuc} \)) during inoculation is a critical factor. A simplified expression is:
$$ \Delta G_{nuc} = \frac{16 \pi \gamma^3}{3 (\Delta G_v)^2} f(\theta) $$
where \( \gamma \) is the interfacial energy, \( \Delta G_v \) is the volume free energy change, and \( f(\theta) \) is a function of the contact angle. Effective inoculation increases \( f(\theta) \), reducing the energy barrier for graphite spheroid formation in ductile iron castings.
Furthermore, the solidification path and fraction solid (\( f_s \)) vs. temperature were simulated. The Scheil-Gulliver model (assuming no diffusion in solid) provides insight into microsegregation:
$$ C_s^* = k C_0 (1 – f_s)^{k-1} $$
where \( C_s^* \) is the composition of the solid at the interface, \( k \) is the partition coefficient, and \( C_0 \) is the initial liquid composition. This helps predict the formation of carbides at the end of solidification, guiding heat treatment needs for ductile iron castings.
8. Comparison with Alternative Materials and Economic Assessment
The choice of austenitic ductile iron castings over alternatives like austenitic stainless steel castings or aluminum alloys was justified through a multi-attribute analysis. Key comparison factors are shown in Table 6.
| Material | Approx. Yield Strength @ -170°C (MPa) | Thermal Expansion Coeff. (10-6/K, ~200K) | Relative Cost Index (Raw Material + Processing) | Castability Rating (1=poor, 5=excellent) |
|---|---|---|---|---|
| Austenitic Ductile Iron QTANi35Cr3 | ~370 | ~5-8 | 1.0 (Baseline) | 4 |
| Austenitic Stainless Steel CF8M | ~450 | ~15-18 | 2.5 – 3.0 | 3 |
| Aluminum Alloy A356-T6 | ~280 | ~22-24 | 1.5 | 5 |
The ductile iron castings offered the best compromise of adequate strength, very low thermal expansion (critical for clearances), good castability, and lower cost than stainless steel. The total cost model for producing one cylinder casting (\( C_{total} \)) included material (\( C_m \)), energy (\( C_e \)), processing (\( C_p \)), and yield loss (\( Y \)):
$$ C_{total} = \frac{(C_m + C_e + C_p)}{Y} $$
Where yield \( Y \) was improved from an initial ~65% to over 85% through process optimization, significantly impacting economics.
9. Summary and Key Findings
The successful manufacture of BOG compressor cylinders from QTANi35Cr3 austenitic ductile iron castings hinged on several pivotal factors. First, meticulous control of chemistry, particularly keeping nickel and chromium at the lower specification limits while maintaining strict limits on impurities like phosphorus and sulfur, was fundamental. Second, the melting practice required vigilance against gas absorption and precise sequencing of alloy additions. Third, the choice of nodulizing agent (Ni85Mg15) over alternatives was decisive in achieving the required high nodularity and mechanical properties, especially elongation at cryogenic temperatures. Fourth, the casting process itself had to accommodate the substantial shrinkage characteristics of this alloy, necessitating a steel-like approach with large, properly sized risers. Numerical simulation proved an invaluable tool for validating the feeding system before committing to expensive tooling and melt.
The project demonstrated that robust, cryogenic-grade ductile iron castings are viable for demanding pressure vessel applications. The knowledge gained extends beyond compressor cylinders to other components requiring dimensional stability and toughness at low temperatures, such as valves, pumps, and cryogenic storage equipment. Future work may explore optimized heat treatments to further enhance ductility or the development of modified compositions with reduced nickel content for cost-sensitive applications without sacrificing low-temperature performance. The continuous improvement in the manufacturing process for these advanced ductile iron castings remains a key engineering endeavor.