Induction Furnace

What is an Induction Furnace

An induction furnace is an electrical melting furnace in which the metal charge is heated by induced eddy currents rather than by a flame, an arc, or a resistance element. An alternating current flows through a water-cooled copper coil, the coil sets up an alternating magnetic field, and that field induces circulating eddy currents inside the conductive metal charge. The charge's own electrical resistance dissipates these currents as heat, so the metal effectively heats itself from within. Because no electrode touches the metal and no combustion gas contacts the bath, the melt stays clean and the operator gains precise control over temperature and chemistry loss.

Structurally an induction melting furnace has four core subsystems: (1) the induction coil, a hollow water-cooled copper conductor wound around or beside the melting cavity; (2) the refractory crucible or channel that contains the molten metal and isolates it from the coil; (3) the power supply, which converts incoming line power into the medium-frequency current the coil requires and resonates with the coil through a capacitor bank; and (4) the auxiliary systems, including the closed-loop cooling water circuit, the hydraulic tilting frame for pouring, the lining sinter and push-out tooling, and the control and interlock cabinet. The same four subsystems appear in both coreless and channel designs, only their geometry differs.

Two physical effects beyond simple heating make induction melting distinctive. The first is electromagnetic stirring: the same field that heats the bath also exerts a body force on the liquid metal, creating a recirculating flow that homogenizes temperature and alloy additions without any mechanical agitator. Stirring intensity rises as frequency falls, so a line-frequency furnace stirs hard while a high-frequency furnace stirs gently. The second is the skin effect, where induced current concentrates in a surface layer whose depth shrinks as frequency rises, which is why furnace builders match frequency to charge size so the field couples efficiently into the metal.

It is important to understand what an induction furnace does not do. It melts and homogenizes, but it does not refine. With no arc, no oxygen lance, and only a thin, relatively cold and unstirred slag layer, it cannot run the oxidizing and reducing slag practice that an electric arc furnace or ladle-refining furnace uses to strip phosphorus, sulphur, or carbon. The charge must therefore be clean, oxide-free, and of known composition, and reactive alloying elements can still suffer some oxidation loss. Chemistry is controlled mainly through the charge mix and any downstream ladle treatment, not inside the furnace itself.

In application scale the induction furnace spans from precious-metal and laboratory crucibles under one kilogram to large foundry and steelmaking furnaces of 65 tonnes or more, with corresponding power supplies from about 10 kW to 42 MW. Across that range it serves grey and ductile iron foundries, steel and stainless steel melt shops, aluminium and copper casting plants, and investment-casting and precious-metal operations. Its growth in cast iron melting is driven largely by environmental advantage, since it produces far less dust and flue gas than a coke-fired cupola while giving tighter temperature control.

Furnace Types: Coreless and Channel

Industrial induction furnaces divide into two families defined by how the magnetic circuit couples to the metal: the coreless furnace and the channel furnace. Choosing the wrong family is the most consequential early mistake, because each is optimized for a different job. The coreless furnace is built to melt cold scrap from empty, while the channel furnace is built to hold and pour molten metal efficiently. The table below summarizes the core differences before the discussion that follows.

Coreless induction furnace. Here the entire metal charge sits in a refractory crucible that is surrounded by the water-cooled copper coil, with no iron core in the magnetic path. The coil's alternating field induces current directly in the whole charge, so the furnace can start from cold solid scrap, melt it completely, and pour out the full heat, leaving the crucible empty for the next batch. This flexibility, plus strong electromagnetic stirring that mixes alloy additions thoroughly, makes the coreless furnace the workhorse of scrap-melting foundries and steel melt shops. Without a flux-concentrating core its electrical efficiency is somewhat lower than a channel furnace, on the order of 75 percent at the coil.

Channel induction furnace. This design behaves like a transformer. A primary coil is wound around a laminated iron core, and a closed loop of molten metal in a refractory channel acts as a single-turn short-circuited secondary. Current induced in that loop heats the metal, and natural convection carries the heat up into the main bath above. Because the secondary must remain liquid, the channel always has to keep a molten heel and cannot start from cold. In return, concentrating the field through the iron core gives the channel furnace higher electrical efficiency and very low energy loss when simply holding metal at temperature, which is why it excels as a holding and pouring furnace.

The practical division of labour follows directly from these mechanics. Coreless furnaces melt: they accept cold scrap, run across a wide frequency band tuned to capacity, and empty fully between heats. Channel furnaces hold and pour: vertical channel furnaces store large volumes of molten iron, hold it precisely at pouring temperature, and decouple the melting operation from the casting line so that a continuous or high-volume casting machine never starves. Channel units are also well suited to aluminium and copper, where their efficient holding offsets their inability to cold-start. Many large plants run both: coreless to melt, channel to hold and pour, in a duplex arrangement.

One operational caution specific to channel furnaces is build-up, the gradual accretion of oxides and intermetallics that narrows the channel cross-section over time, raising electrical losses and eventually restricting flow. Build-up is managed through metal cleanliness, channel design, and periodic thermal cleaning cycles, and its rate is a real selection factor when a channel furnace will run on a given alloy for long campaigns.

Power-Supply Technology and Frequency

The power supply is the part of the furnace that converts incoming three-phase line power into the medium-frequency alternating current the coil needs, and it resonates with the coil through a capacitor bank to deliver real power into the metal. Two switching technologies dominate, the older SCR (thyristor) converter and the now-mainstream IGBT converter, and they differ in how they regulate power and in the power factor they present to the grid. The table below compares them on the metrics that drive purchasing.

SCR power supplies were the early standard. A parallel-resonant SCR supply uses full-bridge rectification with the capacitor bank in parallel with the coil, while the later series-resonant SCR supply uses half-bridge rectification with the capacitors in series. Both regulate output power by varying the thyristor conduction angle, which is effectively a voltage adjustment. The drawback is that throttling the conduction angle pulls down the grid power factor at part load, so plants running SCR furnaces usually have to add a power-factor compensation capacitor cabinet to avoid utility penalties. SCR designs are robust and still in service, but they are no longer the default for new installations.

IGBT power supplies are series-resonant inverters built around insulated-gate bipolar transistors, and they regulate power by shifting frequency rather than chopping voltage. The practical consequences are significant: the grid power factor stays high and is not pulled down by part-load operation, output power remains essentially constant even as the charge volume changes during a melt or as the lining thins over a campaign, and melt times are faster with lower element burn-off. IGBT supplies have therefore largely displaced the older KGPS SCR designs for new medium-frequency furnaces, and they are the appropriate default unless a specific legacy or very-high-power constraint dictates otherwise.

Frequency selection is tied to furnace capacity and charge size through the skin effect. Large furnaces holding several tonnes run at line frequency or a few hundred hertz, because at low frequency the induced current penetrates deep into the charge and couples efficiently into a large mass while stirring it vigorously. Small furnaces of tens to a few hundred kilograms need a higher frequency, typically 1 to 10 kHz, so that the shallower current layer still transfers power into a small charge without violently ejecting light scrap. Most foundry coreless melting furnaces therefore operate in the medium-frequency band of roughly 150 Hz to 10 kHz, with capacity setting the exact value.

Frequency also trades off against metallurgy. Lower frequency means stronger electromagnetic stirring, which speeds alloy homogenization and temperature uniformity but increases lining wear and the risk of picking up oxides from a turbulent surface. Higher frequency means gentler stirring, which protects the lining and limits alloy and carbon pickup but mixes more slowly. The furnace builder selects a frequency that balances these against the chosen capacity, and on variable-power IGBT systems the effective operating point can shift somewhat across the melt as the charge consolidates.

Refractory Linings and Materials

The refractory lining is the consumable heart of a coreless furnace. It forms the crucible wall that holds the molten metal, isolates it from the copper coil, and must survive thermal shock on every heat, chemical attack from metal and slag, and mechanical erosion from charging and stirring. Lining material is chosen primarily against slag chemistry, and the three mainstream families are acidic silica, neutral alumina-spinel, and basic magnesia. The table below gives the typical composition, sinter or service temperature, and best-fit duty of each.

Acidic silica ramming mass is the most common lining for cast iron. With silica content of 99 percent or more and very low alumina and iron oxide, it sinters into a hard working face at around 1,280 degrees Celsius and offers excellent thermal-shock resistance, which suits the rapid heat-to-heat cycling of a batch melter. Its limitation is chemical: a basic slag attacks silica rapidly, so it is unsuitable for steel grades or operations that generate basic slag. Silica is also low in cost and easy to install and push out, which keeps relining time and consumable spend low for iron foundries.

Neutral alumina-spinel ramming mass is the general-purpose choice for steel and demanding alloys. It is mainly alumina with a magnesia addition that reacts in service to form magnesium-aluminate spinel (MgO and Al2O3 combine into MgAl2O4) directly in the matrix, giving a rigid, erosion-resistant bonded structure with a service temperature up to about 1,750 degrees Celsius. The in-situ spinel formation resists both metal penetration and slag attack across a broad chemistry window, which is why neutral linings dominate steel melting where silica would fail. They cost more than silica and need disciplined sintering, but they deliver longer, more predictable campaigns on aggressive heats.

Basic magnesia linings carry roughly 88 percent MgO and tolerate high-basicity steel slag up to about 1,750 degrees Celsius, with a very high pure-MgO melting point near 2,800 degrees. Their weakness is thermal cycling: magnesia spalls under repeated heat-up and cool-down, so it is favoured where the furnace runs hot and steady on basic slag rather than cycling cold between heats. Lining choice is therefore not only a chemistry decision but an operating-pattern decision, matching the material's failure mode to how the furnace will actually be run.

Lining life is reported in campaign heats, but it is not a single fixed number, because it depends on metal grade, slag chemistry, the severity of thermal cycling, charging practice, and how aggressively the bath is superheated. The reliable approach is to track heats per campaign and measure the residual wall profile, then schedule a reline before wear thins the working face past the coil-safety limit, rather than running to failure. Push-out and re-ramming tooling, sinter curves, and lining-wear monitoring are part of the furnace package and should be specified alongside the furnace itself, because lining management dominates the operating cost and the safety case of a coreless installation.

Key Specification Parameters

Reading an induction furnace datasheet means looking past the marketing capacity figure to the parameters that actually govern throughput, energy cost, and integration. The seven that drive most selection decisions are rated capacity, rated power, melt rate, operating frequency, specific energy consumption, electrical efficiency, and cooling-water duty. Each is explained below, with representative values to calibrate expectations.

Rated capacity is the mass of molten metal the crucible holds, quoted against a reference metal because density matters: a crucible rated for one tonne of iron holds far less aluminium by mass. Always confirm the reference metal and remember that capacity alone says nothing about how fast the furnace melts. Rated power is the electrical input the power supply delivers to the coil, and together with capacity it sets the melt rate. As reference points, small workshop coreless furnaces commonly run 15 to 160 kW for 1 to about 200 kg of ferrous charge, mid-size foundry furnaces of 0.3 to 6 tonnes draw up to roughly 4,800 kW, and the largest 8 to 65 tonne units can exceed 40 MW.

Melt rate is the tonnes or kilograms of metal the furnace can bring to pouring temperature per hour, and it is the number that must match downstream demand. A 10-tonne medium-frequency coreless furnace at around 3,000 kW, for example, reaches a melt rate near 5 tonnes per hour on iron, since real consumption sits close to 600 kWh per tonne. Sizing the furnace by melt rate against the casting line, rather than by crucible volume, is the single most important capacity decision. Operating frequency, covered in Chapter 3, is listed here because it appears on every datasheet and constrains both the minimum practical charge size and the stirring intensity.

Specific energy consumption is the electricity used per tonne of metal melted and superheated, and it is the dominant operating cost. For a modern coreless furnace it typically falls between 500 and 800 kWh per tonne: cast iron around 550 to 600, ductile (SG) iron 550 to 650, mild or stainless steel 600 to 680, and aluminium roughly 650 to 700. For context, the theoretical minimum to heat one tonne of iron to about 1,500 degrees Celsius is near 396 kWh, so the gap between that floor and the real figure is the efficiency and operating-discipline opportunity.

Electrical efficiency, the share of purchased electricity that ends up as useful heat in the metal, generally sits between 60 and 75 percent for a coreless furnace, with channel furnaces higher when holding because the iron core concentrates the field. The remainder is lost in the coil, the power supply, and radiation from the open bath, which is why lid discipline and dense charging matter so much in practice. Cooling-water duty is the closed-loop flow and heat rejection the coil and power electronics require; it is easy to overlook at purchase yet it sizes the chiller or cooling tower and is a hard prerequisite for commissioning, so it belongs on the specification alongside the electrical numbers.

Selection Decision Factors

To turn the preceding five chapters into a specific furnace order, work through the decision sequence below. Most selection failures come not from a single wrong number but from deciding capacity before defining the metallurgical and integration requirements that should constrain it. These eight steps can serve as a fixed RFQ template.

1. Function and furnace type: First decide whether the furnace must melt cold scrap (coreless) or hold and pour an existing melt (channel), and whether a duplex melt-plus-hold arrangement is justified. This single decision constrains almost everything that follows.
2. Metal and grade: Define the alloys to be melted and their slag chemistry, because grade drives the refractory family (acidic silica for cast iron, neutral alumina-spinel for steel, basic magnesia for high-basicity slag) and influences frequency and stirring tolerance.
3. Capacity and melt rate: Size by hourly melt rate against the casting or pouring demand, not by crucible volume alone, and confirm the reference metal for the stated capacity. Match rated power to the melt rate you actually need.
4. Operating frequency: Confirm the frequency band suits the capacity and charge size, balancing stirring (faster homogenization at low frequency) against lining life and alloy pickup (gentler at high frequency).
5. Power-supply technology: Prefer an IGBT series-resonant supply for stable power factor and constant output as the charge and lining change; treat SCR (KGPS) as a legacy option only where a specific constraint requires it, and budget a PFC cabinet if so.
6. Refractory and lining system: Specify the lining material against slag chemistry and operating pattern, and include sinter curves, push-out and re-ramming tooling, and lining-wear monitoring, since lining management dominates operating cost and the safety case.
7. Utilities and infrastructure: Confirm electrical supply and transformer rating, the closed-loop cooling-water flow and heat-rejection duty, hydraulic tilting, fume extraction, and the physical foundation and coil-safety interlocks the installation demands.
8. Total cost of ownership (TCO): Sum purchase price, installation, lining consumables and relining labour, electricity at the realistic specific-energy figure for your grade, cooling and maintenance, and downtime risk. A cheaper furnace with a poor lining system or low efficiency often costs more within two to three years of production.

One last commonly overlooked dimension is manufacturer serviceability: local availability of refractory consumables, coil rewind and power-supply repair support, spare-part lead time, and field engineering for commissioning and lining qualification. These seem secondary at the purchasing stage but determine repair response and uptime once the furnace is on the casting line. Established suppliers such as Inductotherm (with Ajax Tocco), ABP Induction, and Otto Junker maintain service and parts networks across major foundry regions, while lower-cost IGBT medium-frequency furnaces can suit non-critical melting duty provided power-supply type, lining system, and local service are confirmed before comparing on price alone.