Gas Porosity Control in ZG1Cr10MoVNbN Castings

In my work with ZG1Cr10MoVNbN supercritical martensitic stainless steel castings, I have found that gas porosity is one of the most persistent and damaging casting defect families. The material is designed for high-pressure, high-temperature service in supercritical and ultra-supercritical steam turbine components, including main regulating valve housings, high-pressure valve bodies, and high-temperature partition castings. Because these parts must combine high strength, toughness, creep resistance, oxidation resistance, and corrosion resistance, the allowed tolerance for any casting defect is extremely small. A gas porosity casting defect can reduce fatigue life, create local stress concentrations, interrupt the continuity of the oxide scale, and provide initiation sites for cracking under thermal cycling. In my experience, the problem is not caused by a single variable but by the interaction of alloy chemistry, melting practice, mold and core gas evolution, gating design, solidification rate, and heat treatment.

The alloy is a Cr-Mo-Ni-Nb-V-N multi-component martensitic heat-resistant cast steel. It contains deliberate nitrogen, and it also contains vanadium and niobium, both of which have a high affinity for oxygen and nitrogen at elevated temperature. This combination makes the steel sensitive to gas absorption, oxide inclusion formation, and gas-driven casting defect formation. I therefore treat the entire route from raw material selection through AOD refining, pouring, solidification, and heat treatment as one integrated control system. The purpose of this article is to explain how I analyze gas porosity in ZG1Cr10MoVNbN castings and how I have applied casting and melting measures to suppress this casting defect.

1. Casting Defect Classification and Relevance to the Alloy

When I classify casting defects in this material, I separate them according to origin, morphology, location, and detection response. The most important distinction is between gas porosity, shrinkage porosity, and gas-shrinkage porosity. A gas porosity casting defect forms when gas bubbles are nucleated and trapped during solidification. A shrinkage casting defect forms when liquid feed metal cannot compensate for volume contraction. A combined gas-shrinkage casting defect occurs when gas precipitation and inadequate feeding act together. This third type is especially common in thick valve housings because the thermal gradient is low, the solidification time is long, and gas solubility drops sharply as temperature falls.

Defect Class Typical Morphology Primary Origin Common Location Detection Response Effect on Service
Gas porosity Rounded or near-spherical cavities Gas dissolved in liquid steel, mold/core gas, air entrapment Upper sections, hot spots, core surfaces Radiographic testing, penetrant testing, ultrasonic testing Stress concentration, reduced fatigue life, leakage path
Blowhole Large smooth internal cavity Mold or core gas evolution, poor venting Near mold wall, core interface, top surface Radiographic testing, visual inspection after machining Local loss of section, pressure boundary risk
Pinhole Small dispersed cavities or surface openings Hydrogen or nitrogen supersaturation, slag-steel reaction Surface layer, subsurface layer Penetrant testing, metallographic examination Surface crack initiation, corrosion attack
Shrinkage porosity Irregular dendritic or interdendritic cavities Insufficient feeding, low thermal gradient Hot spots, thick sections, junction zones Radiographic testing, ultrasonic testing, sectioning Reduced load-bearing area, creep damage
Gas-shrinkage porosity Irregular cavities with rounded internal surfaces Gas precipitation during long freezing plus poor feeding Heavy sections, isolated thick bosses Radiographic testing, ultrasonic testing Combined stress concentration and reduced toughness
Inclusion-related porosity Angular inclusions with attached voids Oxide, nitride, or slag inclusions acting as nucleation sites Gating paths, last solidified regions Metallography, scanning electron microscopy, ultrasonic testing Local embrittlement, corrosion initiation

I use the following general expression for gas pore stability in liquid steel. A bubble can survive only when its internal pressure is sufficient to balance the local pressure and the surface tension resistance:

$$P_g \ge P_m + P_s + P_t$$

where \(P_g\) is the gas pressure inside the bubble, \(P_m\) is the metalostatic pressure, \(P_s\) is the surface tension contribution, and \(P_t\) is any additional pressure from mold gas or trapped atmosphere. For a spherical bubble of radius \(r\),

$$P_s = \frac{2\sigma}{r}$$

where \(\sigma\) is the surface tension of the liquid steel. The critical radius for bubble nucleation is therefore

$$r^* = \frac{2\sigma}{P_g – P_m – P_t}$$

This relation explains why a casting defect of the gas porosity type is more likely to form when the local pressure is low, when the bubble radius is small, or when the gas concentration is high. In thick sections, the metalostatic pressure may remain low in upper regions, while the solidification time is long. This creates favorable conditions for gas bubbles to grow and become trapped.

For a more detailed nucleation view, I also consider the Gibbs free energy change:

$$\Delta G = \frac{4}{3}\pi r^3 \Delta P + 4\pi r^2 \sigma$$

where \(\Delta P\) is the pressure difference driving bubble growth. The barrier to nucleation decreases as \(\Delta P\) increases. In ZG1Cr10MoVNbN, nitrogen and hydrogen are the most important gas-forming species, although carbon monoxide can also participate when carbon and oxygen are not balanced. The solubility of each gas in liquid steel follows a Sieverts-type relation:

$$[H] = K_H \sqrt{P_{H_2}}$$

$$[N] = K_N \sqrt{P_{N_2}}$$

$$[O] = K_O \sqrt{P_{O_2}}$$

During cooling and solidification, the solubility coefficients decrease, so the gas concentration can exceed the equilibrium solubility. The excess gas then precipitates as bubbles or forms a casting defect at dendrite boundaries. I therefore regard gas control as both a melting problem and a solidification problem.

2. Alloying Elements and Internal Chemistry Control

The chemical composition of ZG1Cr10MoVNbN must be controlled very tightly. Carbon provides the martensitic matrix and the carbide-forming potential. Chromium provides corrosion and oxidation resistance. Molybdenum strengthens the matrix and improves creep resistance. Vanadium and niobium form fine carbonitrides that improve creep strength and refine grains. Nitrogen participates in precipitation strengthening and solid solution strengthening. However, nitrogen is also a gas-forming element, and if it is not combined with nitride formers or kept below the solubility limit, it can produce a gas porosity casting defect.

Element Target Internal Range, wt.% Primary Role in ZG1Cr10MoVNbN Effect on Casting Defect Risk
C 0.10–0.13 Martensite matrix, carbide formation, strength and hardness Too low gives insufficient strength; too high reduces toughness and may increase CO gas porosity
Mn 0.50–0.70 Deoxidation, sulfide shape control, solid solution strengthening High levels can raise inclusion load and affect gas evolution
Si 0.40–0.60 Deoxidation, oxidation resistance support High levels can increase slag volume and inclusion-related porosity
S ≤0.005 Residual impurity High sulfur increases inclusion and hot-tearing sensitivity
P ≤0.025 Residual impurity High phosphorus worsens toughness and segregation
Cr 9.20–9.80 Corrosion and oxidation resistance, carbide formation Affects nitrogen solubility and oxide scale stability
Mo 0.70–0.90 Solid solution strengthening, creep resistance High levels may increase segregation in heavy sections
Ni 0.45–0.65 Toughness, matrix stability Balance with Cr and C must avoid retained austenite
V 0.13–0.18 Precipitation strengthening, grain refinement Strong nitride former; helps reduce free nitrogen but high levels may form coarse nitrides
Nb 0.040–0.060 Precipitation strengthening, grain refinement Strong carbonitride former; controls free nitrogen and carbon
N 0.040–0.060 Solid solution and precipitation strengthening Excess free nitrogen can form gas porosity casting defect

I use precipitation and solubility relations to judge whether nitrogen is likely to remain in solution or form gas. For a carbonitride of the type (Nb,V)(C,N), the simplified reaction is

$$[Nb] + [C] \rightleftharpoons NbC(s)$$

$$[V] + [N] \rightleftharpoons VN(s)$$

$$[Nb] + [N] \rightleftharpoons NbN(s)$$

The solubility product can be expressed in a general form as

$$\log\left([\%Nb][\%C]\right) = A – \frac{B}{T}$$

where \(A\) and \(B\) are temperature-dependent constants for the specific precipitate. Similar expressions apply to VN and NbN. The practical consequence is that vanadium and niobium must be present in sufficient amounts to combine with nitrogen before solidification is complete. If the nitrogen content is too high or the nitride formers are too low, the residual free nitrogen can contribute to a gas porosity casting defect. I therefore include nitrogen in the internal composition specification and verify it immediately before tapping.

I also track the carbon-oxygen balance because carbon monoxide can form internally:

$$[C] + [O] \rightleftharpoons CO(g)$$

The equilibrium constant is

$$K_{CO} = \frac{P_{CO}}{a_C a_O}$$

where \(a_C\) and \(a_O\) are the activities of carbon and oxygen in the liquid steel. If the product of carbon and oxygen activities is high, CO gas can form. In AOD refining, I use argon dilution to reduce the partial pressure of CO, which shifts the equilibrium and allows deep decarburization without excessive oxidation of chromium. This is one of the main reasons I prefer AOD refining for this alloy: it helps suppress a CO-related casting defect while retaining alloy control.

3. Melting Route: Induction Furnace Plus AOD Refining

In my experience, an induction furnace alone can melt this alloy, but it is difficult to guarantee consistently low oxygen, hydrogen, and nitrogen when the melt is exposed to air. The strong electromagnetic stirring in an induction furnace produces a mound or “hump” on the melt surface. This exposes a large area of liquid steel to the atmosphere, increasing gas absorption. The exposed surface also makes it difficult to maintain a protective slag cover. If the melt absorbs too much gas, the result is often a gas porosity casting defect after pouring.

I therefore use a two-stage route: primary melting in an induction furnace followed by AOD refining. The induction furnace establishes the base chemistry and supplies superheated liquid steel. The AOD vessel provides decarburization, desulfurization, deoxidation, gas control, and final chemistry trimming. This combination gives me much better control over the casting defect risk than induction melting alone.

Process Stage Main Objective Critical Control Influence on Casting Defect
Raw material preparation Clean, dry, low-gas charge No oil, moisture, rust, or organic contamination Reduces hydrogen and nitrogen pickup, lowers gas porosity risk
Induction furnace charging Establish base melt Add cover slag, control power, minimize hump Limits atmospheric exposure and gas absorption
Induction furnace melting Melt and homogenize Short melt time, stable stirring, protective cover Reduces oxygen and nitrogen pickup that later forms casting defect
AOD decarburization Lower carbon and oxygen Ar/O2 ratio, oxygen volume, bath temperature Suppresses CO gas formation and oxide inclusion load
AOD reduction Remove oxygen and sulfur Al addition, lime addition, argon stirring Lowers gas and inclusion-related porosity
AOD nitrogen adjustment Set final nitrogen Nitrogen gas injection or nitrogen-bearing additions Controls free nitrogen and gas porosity casting defect risk
Final chemistry check Confirm internal specification Full chemical analysis, especially C, N, V, Nb Ensures precipitation control and gas balance
Tapping and pouring Deliver clean steel Temperature, ladle condition, pouring speed Reduces reoxidation and gas entrapment

The central decarburization reaction in AOD is

$$[C] + [O] \rightarrow CO(g)$$

When argon is injected with oxygen, the CO partial pressure is reduced according to the gas ratio:

$$P_{CO} = \frac{n_{CO}}{n_{CO} + n_{Ar}} P_{total}$$

Lower \(P_{CO}\) drives the reaction further toward CO formation, allowing carbon to be removed to very low levels. I typically aim for a carbon level below 0.08% before entering the reduction period, and then I adjust to the final internal range. The argon stirring also acts like a dispersed vacuum system. Each argon bubble provides a low-pressure cavity into which dissolved gases such as hydrogen, nitrogen, and CO can diffuse. As the bubbles rise, they carry gas and inclusions into the slag.

I describe the gas removal tendency qualitatively by

$$J_g = k_g A (C_g – C_g^*)$$

where \(J_g\) is the mass transfer flux of dissolved gas, \(k_g\) is the mass transfer coefficient, \(A\) is the interfacial area, \(C_g\) is the dissolved gas concentration, and \(C_g^*\) is the equilibrium concentration at the bubble surface. AOD stirring increases \(A\) and reduces \(C_g^*\), so \(J_g\) increases. This is a major reason the AOD route lowers the probability of a gas porosity casting defect.

AOD Stage Typical Gas Ratio Purpose Key Reaction or Control
First oxidation stage Ar/O2 around 1/4 Heat, melt slag, begin decarburization Carbon oxidation, slag formation, temperature rise
Main decarburization Progressive increase of Ar/O2 ratio Lower carbon without excessive Cr oxidation CO partial pressure reduction, Cr retention
Deep decarburization High argon fraction Reach low carbon CO equilibrium shift, gas removal
Reduction period Pure argon stirring Deoxidize and desulfurize Al and lime additions, inclusion flotation
Nitrogen adjustment Nitrogen injection as needed Set final N Nitrogen gas dissolution, nitride balance
Final holding Argon stirring Homogenize and float inclusions Cleanliness improvement before tapping

Before the liquid steel enters the AOD vessel, I make sure the vessel has been dried thoroughly. Moisture is a direct source of hydrogen, and hydrogen can produce a pinhole or gas porosity casting defect. I require a minimum drying time and a preheat temperature high enough to eliminate surface moisture. I also add lime that has been stored dry and is added in a controlled amount. The slag must be basic enough to absorb oxides and sulfides but not so thick that it interferes with argon stirring.

After AOD refining, I check the oxygen, nitrogen, and hydrogen potential through process indicators and final chemical analysis. The final nitrogen content must be inside the internal range, usually 0.040–0.060 wt.%. If nitrogen is too low, the alloy may not achieve the required precipitation strengthening. If nitrogen is too high, the risk of a gas porosity casting defect increases, especially in heavy sections. I therefore treat nitrogen as both an alloying element and a gas-forming species.

4. Casting Process Design and Mold Gas Control

The casting process is the second major front in controlling a gas porosity casting defect. In resin-bonded sand molds and cores, the binder decomposes when it contacts liquid metal. The decomposition products include hydrogen, carbon monoxide, methane, and other hydrocarbons. If the gas generation rate exceeds the venting capacity, the gas pressure at the mold-metal interface rises. The gas can then be pushed into the liquid steel, creating blowholes, pinholes, or a larger gas porosity casting defect.

I estimate the gas evolution behavior using a rate expression:

$$\dot{V}_g = f(m_b, T, t)$$

where \(\dot{V}_g\) is the gas generation rate, \(m_b\) is the binder mass, \(T\) is the local temperature, and \(t\) is time. The exact function depends on binder chemistry and sand composition, but the practical conclusion is clear: reducing resin content and improving core dryness lowers \(\dot{V}_g\). I also consider the venting capacity:

$$Q_v = C_v A_v \sqrt{\frac{2\Delta P}{\rho_g}}$$

where \(Q_v\) is the volumetric venting rate, \(C_v\) is a discharge coefficient, \(A_v\) is the vent area, \(\Delta P\) is the pressure difference, and \(\rho_g\) is the gas density. A larger vent area or a higher permeability reduces the pressure difference and lowers the chance that mold gas enters the steel. This is why I combine reduced gas generation with improved venting.

Control Measure Process Change Mechanism Expected Effect on Casting Defect
Resin content reduction Lower binder addition within strength limits Less binder decomposition gas Reduces gas porosity and blowhole formation
Core drying improvement Longer drying, higher temperature, better air circulation Removes moisture and residual volatiles Lowers hydrogen and water vapor gas defects
Stepped gating Multiple ingates at different heights More progressive filling, less turbulence Reduces air entrapment and gas-shrinkage porosity
Slag trap Ceramic or sand trap before ingates Traps slag and inclusions Reduces inclusion-related gas nucleation
Top vents Multiple vents in the cope Allows mold gas and steel gas to escape Reduces internal gas pressure and blowholes
Chromium ore sand Local high-chill facing sand Increases local heat extraction Shortens solidification time, limits gas precipitation
Chills External or internal chills at hot spots Steepens thermal gradient Suppresses gas-shrinkage porosity
Pouring temperature control Lower superheat while maintaining fill Reduces gas solubility and reoxidation Lowers gas porosity risk

For the specific valve housing geometry I studied, the part is thick and has several isolated heavy sections. These heavy sections solidify slowly. The solidification time can be approximated by a Chvorinov-type relation:

$$t_s = B\left(\frac{V}{A}\right)^n$$

where \(t_s\) is solidification time, \(V\) is volume, \(A\) is surface area, \(B\) is a mold constant, and \(n\) is typically between 1.5 and 2.0 for many castings. A large \(V/A\) ratio means a long solidification time. During that time, the solubility of hydrogen, nitrogen, and oxygen decreases. The gas that cannot remain in solution precipitates. If the local pressure is low and the feeding is insufficient, the precipitated gas becomes a casting defect. I therefore use chromium ore sand and chills to reduce the local solidification time and increase the thermal gradient. In my trials, this increased the local solidification rate by about 30%, which reduced the amount of gas that could precipitate before the section became solid.

I also pay great attention to gating because turbulent filling is a direct cause of air entrapment. The Reynolds number is a useful indicator:

$$Re = \frac{\rho v D}{\mu}$$

where \(\rho\) is density, \(v\) is velocity, \(D\) is a characteristic dimension, and \(\mu\) is dynamic viscosity. High \(Re\) indicates a greater tendency toward turbulence. In gating design, I aim to keep the flow as smooth as possible and to avoid free falls and abrupt expansions. A stepped gating system with a slag trap reduces velocity at the ingate and helps separate slag and inclusions. This lowers the chance that a gas porosity casting defect will form from entrained air or from inclusion-assisted bubble nucleation.

Venting is equally important. I place multiple vents in the upper mold, especially above thick sections and core prints. The vents must be clean and connected to the atmosphere. If a vent is blocked by sand or coating, the local pressure rises and gas is forced into the steel. In one production trial, simply increasing the number and diameter of top vents significantly reduced the radiographic indication of gas porosity. This confirmed that a casting defect in this alloy is not only a steel cleanliness problem; it is also a mold atmosphere problem.

5. Solidification, Feeding, and Gas-Solubility Interaction

The interaction between gas precipitation and shrinkage is central to my understanding of casting defect formation in heavy ZG1Cr10MoVNbN parts. As the alloy cools, the liquid contracts. If feed metal is available, the contraction is compensated. If feed metal is not available, shrinkage porosity forms. At the same time, gas solubility decreases. If gas is present above the local solubility limit, gas bubbles form. The final defect may be a rounded gas pore, an irregular shrinkage cavity, or a mixed gas-shrinkage cavity. In thick sections, the last liquid is often isolated, so both problems occur at the same location.

I use a simplified pressure balance at the solidification front:

$$P_{local} = P_{metalostatic} + P_{gas} + P_{capillary}$$

If the local pressure is lower than the sum of the gas pressure and capillary resistance, a pore can grow. If the local pressure is higher, the pore may be suppressed or collapse. This is why chills and chromium ore sand are effective: they reduce the duration of the low-pressure, high-gas period by accelerating solidification. They also improve directional solidification, which helps feeding.

I also consider the segregation of nitrogen and hydrogen at the dendrite interface. The concentration in the residual liquid can be estimated by a Scheil-type relation:

$$C_L = C_0 (1 – f_s)^{k – 1}$$

where \(C_L\) is the concentration in the liquid, \(C_0\) is the initial concentration, \(f_s\) is the solid fraction, and \(k\) is the partition coefficient. When \(k\) is less than 1, the solute is rejected into the liquid. For nitrogen and hydrogen in steel, \(k\) is usually small, so the residual liquid becomes enriched. This enrichment promotes gas bubble formation. If the gas bubble forms at the same time as the last liquid is being consumed by shrinkage, the result is a mixed casting defect.

To reduce this risk, I combine three actions:

1. I reduce the initial gas content by AOD refining.

2. I reduce the local solidification time by chills and chromium ore sand.

3. I improve feeding by proper riser design and gating so that the last liquid remains under positive pressure.

These actions act together. A single action may reduce one casting defect type but leave another. For example, lower pouring temperature reduces gas solubility but can cause cold shuts if taken too far. Higher chill intensity reduces gas precipitation but can create cracks if the thermal gradient is too steep. I therefore treat the process window as a balance rather than as a single optimum.

6. Heat Treatment and Microstructural Control

After casting, the material receives a preliminary high-temperature anneal and then a final normalizing and tempering treatment. In my process, the annealing treatment is followed by furnace cooling to below 250 °C before removal. The final heat treatment is normalizing at 1090 °C ± 10 °C and tempering at 710 °C ± 10 °C. The purpose is to produce a tempered martensitic matrix with fine carbonitride precipitates. If the heat treatment is not controlled, the microstructure may contain retained austenite, coarse precipitates, or an unfavorable carbide distribution. These features do not directly produce a gas porosity casting defect, but they can amplify the effect of any porosity that is already present by lowering toughness and fatigue resistance.

Heat Treatment Step Temperature Atmosphere or Cooling Purpose Quality Link
Preliminary annealing High-temperature anneal Furnace cool to below 250 °C Relieve stress, soften structure, prepare for machining Reduces crack risk around existing casting defect
Normalizing 1090 °C ± 10 °C Controlled heating and cooling Homogenize and form martensitic structure Develops uniform matrix around sound sections
Tempering 710 °C ± 10 °C Controlled heating and cooling Precipitate carbonitrides, improve toughness Enhances tolerance to small defects

I also use heat treatment as a verification tool. If a casting contains a significant gas porosity casting defect, the defect may open slightly during thermal processing or be revealed by subsequent inspection. This is why I prefer to perform radiographic testing and penetrant testing after the final heat treatment. In the production validation described below, the finished valve housing castings were examined by radiographic testing and penetrant testing. No gas porosity, shrinkage porosity, or other unacceptable casting defect was found.

7. Industrial Production Validation

To verify the combined casting and melting measures, I evaluated several industrial heats of ZG1Cr10MoVNbN. All heats used induction furnace primary melting plus AOD refining. The internal chemical composition was controlled to the ranges shown earlier. The castings were poured using the optimized resin sand mold and core system, stepped gating with a slag trap, top venting, chromium ore sand, and chills at critical sections. After casting, the parts were heat treated with the normalizing and tempering schedule described above.

Heat Number Alloy C Mn Si P S Cr Mo Ni V Nb N
Y39058 ZG1Cr10MoVNbN 0.118 0.65 0.46 0.018 0.001 9.58 0.82 0.52 0.17 0.045 0.041
Y39134 ZG1Cr10MoVNbN 0.116 0.63 0.47 0.020 0.001 9.67 0.85 0.51 0.16 0.047 0.043
Y39386 ZG1Cr10MoVNbN 0.124 0.67 0.44 0.018 0.001 9.53 0.83 0.55 0.17 0.045 0.042
Y39593 ZG1Cr10MoVNbN 0.121 0.65 0.46 0.019 0.001 9.74 0.86 0.52 0.15 0.048 0.042

The measured chemical compositions were all within the internal control ranges. The sulfur and phosphorus levels were low, the chromium and molybdenum levels were stable, and the nitrogen content remained inside the target window. The vanadium and niobium contents were sufficient to combine with nitrogen and carbon, which reduced the amount of free nitrogen available to form a gas porosity casting defect. I consider this chemistry control to be a prerequisite for consistent casting quality.

Heat Number Alloy Test Temperature, °C Stress, N/mm² Standard Requirement, h Actual Time, h Result
Y39058 ZG1Cr10MoVNbN 630 147 ≥50 55 Pass
Y39058 ZG1Cr10MoVNbN 650 123 ≥80 90 Pass
Y39058 ZG1Cr10MoVNbN 660 98 ≥200 230 Pass
Y39134 ZG1Cr10MoVNbN 630 147 ≥50 53 Pass
Y39134 ZG1Cr10MoVNbN 650 123 ≥80 88 Pass
Y39134 ZG1Cr10MoVNbN 660 98 ≥200 235 Pass
Y39386 ZG1Cr10MoVNbN 630 147 ≥50 58 Pass
Y39386 ZG1Cr10MoVNbN 650 123 ≥80 91 Pass
Y39386 ZG1Cr10MoVNbN 660 98 ≥200 235 Pass
Y39593 ZG1Cr10MoVNbN 630 147 ≥50 60 Pass
Y39593 ZG1Cr10MoVNbN 650 123 ≥80 90 Pass
Y39593 ZG1Cr10MoVNbN 660 98 ≥200 240 Pass

The high-temperature creep rupture results confirmed that the material met the service requirements. At 630 °C and 147 N/mm², the actual rupture times ranged from 53 to 60 h, above the 50 h requirement. At 650 °C and 123 N/mm², the actual times ranged from 88 to 91 h, above the 80 h requirement. At 660 °C and 98 N/mm², the actual times ranged from 230 to 240 h, above the 200 h requirement. These results are important because they show that reducing a casting defect such as gas porosity does not come at the expense of mechanical performance. Instead, cleaner steel and better solidification improve both reliability and performance.

8. Discussion of Casting Defect Control Logic

My overall control logic for a gas porosity casting defect in ZG1Cr10MoVNbN can be summarized in five principles.

First, gas must be kept out of the steel before casting. This means clean raw materials, dry additives, protected melting, and AOD refining. The induction furnace must not be allowed to expose the melt to air unnecessarily. The AOD vessel must be dry and properly preheated. The lime and alloys must be dry. If moisture enters the melt, hydrogen will enter the steel, and a gas porosity casting defect will follow.

Second, gas that is already in the steel must be removed or fixed. AOD argon stirring removes hydrogen, nitrogen, and CO by providing a large bubble interface. Vanadium and niobium fix nitrogen as carbonitrides, provided the chemistry is balanced. Carbon and oxygen must be controlled to avoid CO gas formation.

Third, mold and core gas must be minimized. Resin content should be as low as practical. Cores must be thoroughly dried. Coatings must be dry and well bonded. Vents must be clean and open. A mold that generates a large volume of gas will create a gas porosity casting defect even if the steel is clean.

Fourth, filling must be smooth and progressive. Stepped gating and slag traps reduce turbulence and inclusion entrainment. A smooth fill avoids air entrapment. A slag trap prevents oxide inclusions from reaching the casting cavity and acting as nucleation sites for gas bubbles.

Fifth, solidification must be directed and rapid at hot spots. Chills and chromium ore sand increase the local cooling rate, reduce the time available for gas precipitation, and improve feeding. In my experience, a 30% increase in local solidification rate can significantly reduce gas-shrinkage porosity in heavy sections.

These principles are not independent. For example, if the carbon and oxygen balance is poor, CO bubbles may form even if the mold is well vented. If the nitrogen content is too high, even excellent AOD practice may not prevent a gas porosity casting defect. If the gating is turbulent, air may be entrained even when the steel is clean. I therefore use a combined specification and a combined process audit for every heat and every mold.

9. Inspection and Acceptance Strategy

For this alloy, I use radiographic testing and penetrant testing as the main acceptance tools for casting defect detection. Radiographic testing is effective for internal gas porosity, blowholes, shrinkage porosity, and gas-shrinkage porosity. Penetrant testing is effective for surface pinholes and surface-connected porosity. In critical regions, I may add ultrasonic testing to detect internal discontinuities that are not oriented favorably for radiography. Metallographic examination is used during process qualification to confirm the morphology of any casting defect and to distinguish gas porosity from shrinkage porosity.

Inspection Method Target Defect Advantage Limitation My Use
Radiographic testing Gas porosity, blowholes, shrinkage, inclusions Volume inspection, image record Orientation sensitivity, thickness limits Primary internal casting defect inspection
Penetrant testing Surface pinholes, cracks, surface porosity Simple, sensitive to surface openings Only surface-connected defects Final surface casting defect inspection
Ultrasonic testing Internal porosity, inclusions, cracks Deep penetration, sensitive to planar defects Geometry dependent, interpretation skill Selected critical sections
Metallography Micro-porosity, inclusion-related porosity Morphology and mechanism identification Destructive, local sampling Process development and failure analysis
Chemical analysis Gas-related elements C, N, O, H Direct process control Requires sampling discipline Every heat before and after refining

After the process changes were implemented, the production valve housings were inspected and no unacceptable gas porosity or shrinkage porosity was found. This was a significant improvement because earlier production had shown a greater incidence of a gas porosity casting defect in the thick sections. The improvement came from the combination of lower resin content, better core drying, optimized stepped gating, more top vents, and local chilling with chromium ore sand. It also came from cleaner steel produced by the induction furnace plus AOD route.

10. Practical Process Window and Summary of Findings

The practical process window I use for ZG1Cr10MoVNbN can be summarized as follows. Carbon is held near 0.12 wt.% to balance strength and toughness. Chromium is held near the middle of the 9.20–9.80 wt.% range. Molybdenum is held near 0.80 wt.% for creep resistance. Vanadium and niobium are controlled to form fine carbonitrides and to reduce free nitrogen. Nitrogen is held between 0.040 and 0.060 wt.% and verified before tapping. Sulfur and phosphorus are kept as low as practical. The melt is refined in AOD with argon stirring to lower oxygen, hydrogen, and nitrogen to acceptable levels. The mold uses reduced resin content, dry cores, stepped gating, a slag trap, top vents, and local chills. The casting is normalized at 1090 °C ± 10 °C and tempered at 710 °C ± 10 °C.

Control Area Key Variable Target or Action Effect on Casting Defect
Chemistry C 0.10–0.13 wt.% Balances strength and CO gas risk
Chemistry N 0.040–0.060 wt.% Controls free nitrogen and gas porosity
Chemistry V and Nb 0.13–0.18 and 0.040–0.060 wt.% Fixes nitrogen as carbonitrides
Melting Induction furnace cover Cover slag, controlled power, short melt Reduces atmospheric gas pickup
Refining AOD argon stirring Decarburization, reduction, inclusion flotation Lowers gas and inclusion-related porosity
Molding Resin content Minimum practical binder Reduces mold gas casting defect
Molding Core dryness Thorough drying and venting Reduces hydrogen and water vapor defects
Gating Stepped gating and slag trap Smooth fill, slag separation Reduces air entrapment and inclusion nucleation
Solidification Chills and chromium ore sand Increase local cooling rate by about 30% Suppresses gas-shrinkage porosity
Heat treatment Normalizing and tempering 1090 °C ± 10 °C and 710 °C ± 10 °C Develops microstructure tolerant to minor defects

In my assessment, the most important finding is that a gas porosity casting defect in ZG1Cr10MoVNbN cannot be solved by a single remedy. It is a system problem. The alloy contains nitrogen, vanadium, and niobium, so gas chemistry and precipitation chemistry are coupled. The castings are thick, so solidification time is long and gas solubility loss is significant. The molds and cores are resin bonded, so mold gas is always a potential source. The gating and feeding system must therefore be designed to minimize turbulence, avoid air entrapment, and maintain positive pressure during solidification. The AOD process must be used to produce clean steel with controlled gas content. The heat treatment must produce a microstructure that can tolerate the small discontinuities that inevitably remain in large castings.

11. Conclusions

I have analyzed the causes of gas porosity in ZG1Cr10MoVNbN supercritical martensitic stainless steel castings and applied a combined casting and melting control strategy. The main conclusions are as follows.

First, the gas porosity casting defect in this alloy is driven by a combination of high dissolved gas content, mold and core gas evolution, turbulent filling, long solidification time in thick sections, and insufficient feeding. Nitrogen is a deliberate alloying element, but free nitrogen can become a gas-forming species if it is not combined with vanadium and niobium. Hydrogen from moisture and binder decomposition can also contribute to pinholes and subsurface porosity. Carbon monoxide can form if carbon and oxygen are not balanced.

Second, AOD refining after induction furnace melting greatly improves steel cleanliness. Argon stirring lowers the partial pressure of CO, promotes deep decarburization, removes hydrogen and nitrogen, and floats inclusions into the slag. The AOD route also allows nitrogen to be adjusted directly by gas injection, which is more flexible and more accurate than using nitrogen-bearing alloy additions alone.

Third, mold and core gas must be controlled by reducing resin content, drying cores thoroughly, and providing adequate venting. A casting defect of the gas porosity type is more likely when the mold generates more gas than the vents can remove. Top vents, stepped gating, and slag traps reduce the chance that mold gas, air, and inclusions enter the casting.

Fourth, local solidification control is essential in thick sections. Chromium ore sand and chills increase the cooling rate, reduce the time available for gas precipitation, and improve directional solidification. In the trials I reviewed, increasing the local solidification rate by about 30% was effective in suppressing gas-shrinkage porosity.

Fifth, the internal chemical composition must be tightly controlled. Carbon near 0.12 wt.%, chromium in the 9.20–9.80 wt.% range, molybdenum in the 0.70–0.90 wt.% range, vanadium in the 0.13–0.18 wt.% range, niobium in the 0.040–0.060 wt.% range, and nitrogen in the 0.040–0.060 wt.% range provide a good balance of strength, creep resistance, and gas resistance. Sulfur and phosphorus should be kept as low as practical.

Sixth, the final heat treatment of normalizing at 1090 °C ± 10 °C and tempering at 710 °C ± 10 °C produces a tempered martensitic structure with fine carbonitride precipitates. This structure supports the required high-temperature creep rupture performance. In the production validation, the actual rupture times exceeded the standard requirements at 630 °C, 650 °C, and 660 °C.

Seventh, the industrial castings produced with the optimized route showed no unacceptable gas porosity, shrinkage porosity, or other casting defect after radiographic testing and penetrant testing. The combination of clean steel, controlled mold gas, smooth gating, local chilling, and proper heat treatment produced reliable ZG1Cr10MoVNbN valve housings for supercritical service. I therefore consider the integrated process route to be effective for controlling casting defect formation and for meeting the demanding reliability requirements of supercritical and ultra-supercritical steam turbine components.

Scroll to Top