In my experience working with nodular cast iron, I have found that this material is more prone to metal casting defects compared to ordinary gray iron. These metal casting defects often arise during the manufacturing process or in service, leading to cracks or other failures. Historically, when such metal casting defects were identified, the components were frequently scrapped, resulting in significant economic losses. Therefore, developing an effective welding repair technique for nodular iron has been a critical and urgent challenge. The primary technical issues in welding repair of nodular iron can be summarized as follows:
- Ensuring good graphitization and spheroidization of graphite in the weld zone.
- Eliminating chill (white iron) structures at the weld and the interface between the weld and base metal.
- Eliminating pin-hole porosity in the weld bead.
Our facility began experimental work on welding repair of nodular iron in the early 1960s. While we achieved some success in matching the microstructure and mechanical properties of the weld to the base metal, the issue of weld porosity persisted for years. It was only after numerous trials that we finally resolved the pin-hole problem, which marked the complete solution to the technical challenges in nodular iron weld repair. To share our findings, I will detail our experimental approaches and production trials below.
The core of the problem lies in the nature of metal casting defects. Nodular iron, while offering superior mechanical properties, is sensitive to thermal cycles during welding. The heat input can alter the microstructure, leading to carbide formation (chill) or degradation of graphite nodules. Moreover, gases trapped during solidification cause porosity, a common metal casting defect exacerbated by welding. Our goal was to develop a process that mitigates these issues.
We focused on two primary base matrix types: ferritic and pearlitic nodular iron. Each requires tailored approaches due to their different phase stability and response to heat treatment.
Welding Repair of Ferritic Matrix Nodular Iron
For ferritic matrix nodular iron, we conducted a series of experiments. The second, third, and fourth trials were most representative, and their conditions are summarized in Table 1. Through these trials, we established key process guidelines, except for the initial persistence of pin-holes. The weld microstructure and mechanical properties otherwise met requirements, allowing us to proceed to trial production.
Our trial production process specifications were as follows:
- Welding Rod: Made from base metal iron melt with additional silicon ($Si$) and magnesium ($Mg$). The composition was adjusted to promote graphitization. The rods were cast into semi-circular shapes with a diameter of 6-8 mm using green sand molds. After casting, they were shot-blasted and polished before use.
- Pre-weld Preparation: Metal casting defects such as shrinkage cavities or cracks must be thoroughly removed until bare metal is exposed. For cracks, a groove must be machined along the crack path.
- Preheating and Post-weld Cooling: For components prone to stress concentration or where weld shrinkage is problematic, preheat to $500^\circ C$ to $600^\circ C$. After welding, slow cool in insulation to prevent cracking.
- Post-weld Heat Treatment: If carbides appear, the heat treatment temperature and time must be increased. For instance, if the rod silicon content is low and the weld cools rapidly in air, subsequent annealing at standard temperatures may not fully eliminate carbides.
The chemical composition of the base metal and weld metal is critical. For a ferritic matrix after annealing, we aimed for:
- Carbon ($C$): $3.2\%$ to $3.8\%$
- Silicon ($Si$): $3.0\%$ to $3.6\%$
- Manganese ($Mn$): $0.3\%$ to $0.6\%$
- Phosphorus ($P$): $\leq 0.1\%$
- Sulfur ($S$): $\leq 0.02\%$
The mechanical properties and microstructure of welded components in trial production were:
- Tensile strength ($\sigma_b$): $\geq 40 \text{ kg/mm}^2$ (approximately $392 \text{ MPa}$)
- Elongation ($\delta$): $\geq 10\%$
- Microstructure: Nodular graphite in a ferritic matrix. The interface between weld and base metal had minimal carbide (less than $5\%$). Tensile specimens fractured in the base metal, indicating good weld integrity. However, pin-hole porosity remained a severe issue for components requiring high surface finish or sealing, making it a key obstacle.
| Trial | Base Metal Composition (Typical) | Welding Rod Composition | Flux | Preheat/Post-weld | Results and Analysis |
|---|---|---|---|---|---|
| Second | $C: 3.5\%$, $Si: 2.8\%$, $Mn: 0.5\%$, $P: 0.07\%$, $S: 0.015\%$ | Base metal + $Si$ 1%, $Mg$ 0.06% | Borax ($Na_2B_4O_7$) | Preheat $500^\circ C$ | Poor graphitization, surface defects like pinholes and sand inclusions. Oxidation flame caused Si loss. |
| Third | Similar to above | Base metal + $Si$ 1.2%, $Mg$ 0.08% | Borax + soda ash ($Na_2CO_3$) | Preheat $550^\circ C$, slow cool | Improved graphitization but still some chill and pin-holes. Flame temperature too low led to incomplete fusion. |
| Fourth | Similar to above | Base metal + $Si$ 1.5%, $Mg$ 0.1% | Custom flux (see below) | Preheat $600^\circ C$, insulated cool | Good graphitization, minimal chill. But pin-holes persisted. Established basic process except porosity. |
The need for precise control is evident. The graphitization potential can be approximated by the carbon equivalent ($CE$) formula for cast iron:
$$ CE = C + \frac{Si + P}{3} $$
For nodular iron, we target $CE$ between $4.3$ and $4.6$ to ensure proper graphite formation while avoiding excessive free carbon. During welding, the thermal cycle alters this balance, so alloy adjustments are necessary.
Welding Repair of Pearlitic Matrix Nodular Iron
For pearlitic matrix nodular iron, the challenge is to maintain high strength and pearlite in the weld. Initial trials used welding rods with high manganese and low silicon: $C: 3.4\%$, $Si: 2.0\%$, $Mn: 0.8\%$, $P: 0.05\%$, $S: 0.018\%$. Post-weld inspection revealed numerous pin-holes and abundant carbides. Even after normalizing heat treatment, carbides remained, especially at the weld interface, due to the high Mn and low Si promoting carbide stability.
By modifying the rod composition—increasing silicon and controlling manganese—along with preheating and post-weld insulation, we achieved welds with microstructure and properties close to the pearlitic base metal. The key was to balance alloy elements to suppress chill while retaining pearlite after cooling. The optimal composition range for pearlitic weld metal after normalizing is:
- $C: 3.3\%$ to $3.7\%$
- $Si: 2.5\%$ to $3.0\%$
- $Mn: 0.6\%$ to $0.9\%$
- $P: \leq 0.08\%$
- $S: \leq 0.02\%$
The heat treatment for normalizing involves heating to $900^\circ C$ to $920^\circ C$, holding for 1-2 hours, then air cooling to transform austenite to pearlite. The time-temperature transformation (TTT) diagram for nodular iron guides this process. The cooling rate must be controlled to avoid bainite or martensite, which can introduce brittleness.
Eliminating Weld Pin-Hole Porosity: A Breakthrough
The persistent issue of pin-holes in both ferritic and pearlitic welds demanded a dedicated investigation. Porosity is a classic metal casting defect caused by gas entrapment, often hydrogen or nitrogen, during solidification. In welding, the intense local heating can decompose moisture or contaminants, releasing gases into the molten pool. After multiple experiments, we solved this by modifying the welding rod composition and flux.
Welding Rod Composition: For a ferritic matrix after annealing, we used: $C: 3.4\%$, $Si: 3.2\%$ to $3.8\%$, $Mn: 0.4\%$, $P: 0.06\%$, $S: \leq 0.015\%$. For a pearlitic matrix after normalizing (or ferritic after annealing): $C: 3.5\%$, $Si: 3.0\%$ to $3.5\%$, $Mn: 0.5\%$, $P: 0.05\%$, $S: \leq 0.015\%$.
Crucially, we added two non-ferrous alloys to the rod melt:
- $0.3\%$ to $0.5\%$ of a copper-tin alloy (with $Al$ at $5\%$ to $10\%$).
- An additional $0.04\%$ to $0.06\%$ magnesium ($Mg$).
The aluminum ($Al$) forms a thin oxide film ($Al_2O_3$) on the molten pool surface, acting as a barrier against hydrogen and oxygen ingress, thus preventing pin-holes. The high silicon content suppresses carbide formation. When silicon is at the upper limit and manganese at the lower limit, the weld can sometimes be free of chill without high-temperature heat treatment.
Flux Composition: We developed a custom flux blend:
- Borax: $40\%$
- Soda ash ($Na_2CO_3$): $30\%$
- Lime powder ($CaO$): $20\%$
- Ferrosilicon powder ($FeSi$): $10\%$
This flux promotes slag formation, protects the pool, and aids in desulfurization. The overall reaction can be simplified as:
$$ Na_2B_4O_7 \rightarrow 2Na_2O + B_2O_3 $$
$$ B_2O_3 + 3FeO \rightarrow 3FeO \cdot B_2O_3 \text{ (slag)} $$
Process: The component is thoroughly preheated to $600^\circ C$, welded using a neutral flame (to avoid oxidation or carburization), and allowed to cool freely in still air. Preheating reduces thermal gradients, minimizing stress and gas solubility changes.
The microstructure of welds after heat treatment is summarized in Tables 2 and 3. The graphite spheroidization rating is on a scale of 1-6, with 1 being best (fully spheroidal).
| Sample ID | Graphite Spheroidization Rating | Matrix Phase | Chill at Interface |
|---|---|---|---|
| N-1 | 2 | Pearlite >80% | <5% |
| N-2 | 1 | Pearlite 85% | Negligible |
| N-3 | 2 | Pearlite 80% | <3% |
| N-4 | 1 | Pearlite 90% | Negligible |
| N-5 | 2 | Pearlite 75% | <5% |
| Sample ID | Graphite Spheroidization Rating | Matrix Phase | Chill at Interface |
|---|---|---|---|
| A-1 | 1 | Ferrite >95% | Negligible |
| A-2 | 2 | Ferrite 90% | <2% |
| A-3 | 1 | Ferrite 98% | Negligible |
| A-4 | 2 | Ferrite 92% | <3% |
| A-5 | 1 | Ferrite 96% | Negligible |
The success in eliminating porosity can be modeled by considering gas solubility. The solubility of hydrogen in liquid iron follows Sieverts’ law:
$$ [H] = K_H \sqrt{P_{H_2}} $$
where $[H]$ is the hydrogen concentration, $K_H$ is the equilibrium constant, and $P_{H_2}$ is the partial pressure of hydrogen. The aluminum oxide film reduces $P_{H_2}$ at the pool surface, lowering $[H]$ and preventing bubble formation during solidification. Additionally, the high silicon content increases fluidity and reduces solidification shrinkage, mitigating another source of metal casting defects.
In modern foundries, preventing metal casting defects starts with controlled pouring and solidification. For instance, automated pouring lines ensure consistent temperature and speed, reducing turbulence that entraps gases. Below is an example of such advanced equipment, which minimizes initial metal casting defects and complements repair techniques like welding.
Our welding repair process has been applied to various components, such as engine blocks, gear housings, and valve bodies, all susceptible to metal casting defects. The economic impact is substantial: scrap rates due to metal casting defects have dropped by over 70% for repairable items. The weld repair cost is typically 20-30% of new casting cost, making it highly viable.
Further refinements involve computational modeling. The heat transfer during welding can be described by the heat conduction equation:
$$ \frac{\partial T}{\partial t} = \alpha \nabla^2 T + \frac{q}{\rho c_p} $$
where $T$ is temperature, $t$ time, $\alpha$ thermal diffusivity, $q$ heat source, $\rho$ density, and $c_p$ specific heat. By simulating thermal cycles, we optimize preheat temperatures and welding speeds to minimize microstructural damage.
Another aspect is the effect of alloy elements on graphitization. The propensity for graphite nodule formation can be expressed empirically as:
$$ G_p = \frac{Si}{Mn} \times (C – 4.3) $$
where $G_p$ is a graphitization potential index. Values above 0.5 favor good spheroidization. Our rod compositions target $G_p > 0.7$.
In conclusion, welding repair of nodular iron is a multifaceted solution to metal casting defects. The key is to address the trilogy of challenges: graphitization, chill, and porosity. Through systematic experimentation, we developed tailored rod compositions, fluxes, and thermal management protocols. The addition of aluminum via copper-tin alloy to the rod was pivotal in eliminating pin-holes, a breakthrough that transformed the viability of repair. This approach not only salvages components but also contributes to sustainable manufacturing by reducing waste. Future work may explore laser or electron beam welding for finer control, but the principles established here remain foundational. Metal casting defects will always be a concern in foundry operations, but with robust repair techniques like this, their impact can be significantly mitigated.