Why Moisture in the Charge Matters
Charge preparation is one of the few variables in a foundry or steelmaking operation that is fully under the buyer's control before the heat is tapped. Moisture is the part of charge preparation that is most often left to chance. Water carried into the furnace on the surface of scrap, in the porosity of sponge or briquetted material, or absorbed inside flux and alloy additions dissociates at melt temperature. The oxygen reports to the slag or to the deoxidation reaction; the hydrogen dissolves readily in liquid steel and liquid aluminium and is only partly removed before solidification.
The result is gas porosity. In aluminium alloys the solubility of hydrogen in the liquid is many times its solubility in the solid, so hydrogen rejected at the solidification front nucleates pores directly. In steel castings the same mechanism produces blowholes and subsurface porosity, and it interacts with any hydrogen already present from damp refractories, damp ladle linings or wet alloy additions.
The Mechanism in Practice
The hydrogen pick-up from a wet addition is not proportional to the total water in the charge; it depends on how much water is exposed to liquid metal and on the deoxidation state of the bath. A small amount of moisture carried in a flux that is added to a clean, well-deoxidised melt can produce a disproportionate gas content, because the liquid metal at that point has a high capacity to dissolve hydrogen and few sinks for it. Lime used for slagging is a frequent offender, as are damp ferroalloys charged into an already liquid bath, where dissolution is fast and the released hydrogen has no time to escape.
This is why the control point is not only the furnace but the store, the day bin and the charging sequence. Alloy additions that are added early, into a cold charge, give any hydrogen a chance to leave with the furnace gases. The same additions dropped into a liquid bath give it nowhere to go but into the metal.
Measurement and Practical Limits
| Material | What is controlled | Common practice |
|---|---|---|
| Iron ore and ore-based charge | Total moisture by drying to constant mass | ISO 3087 procedure, result reported with the lot |
| Lime and slagging fluxes | Free moisture after calcination | Delivery in sealed or lined containers, use within a short window |
| Ferroalloys and deoxidisers | Surface moisture, absorbed humidity | Covered store, dry floor, pre-heat or dry before charging |
| Cored wire and powder additions | Sealed fill, moisture barrier packaging | Keep sealed until the moment of use |
| Scrap | Rain and standing water, oil and coatings | Covered storage, drainage, pre-heat where the grade allows |
ISO 3087 is the reference procedure for determining moisture in iron ore consignments, based on drying a sample to constant mass and reporting the loss. For foundry additions, an in-house drying test at a defined temperature and time is usually more useful than an absolute figure, because it can be run per delivery and trends can be tracked. Whatever method is used, the sample has to be taken and sealed properly; a moisture result from a sample that sat open in the laboratory is meaningless.
Absolute limits depend on the process, and they should be set from plant data rather than copied from a generic table. The workable approach is to establish the level at which a specific foundry sees gas defects, then specify comfortably below it and monitor the delivery-to-delivery variation.
Storage, Drying and Purchasing Controls
The cheapest moisture control is physical: covered storage with a dry floor, drainage so that water cannot stand around bags, first-in-first-out rotation, and no outdoor storage of lime, cored wire or powder additions. Where a material is unavoidably damp, pre-heating or drying before charging is more effective than trying to dry it inside the furnace, and it should be written into the work instruction with a temperature and a holding time.
On the purchasing side, three clauses do most of the work. Require a declared moisture figure with the certificate for any hygroscopic addition. Require moisture-barrier packaging for lime, powders and cored wire, and state that torn or opened packaging is grounds for rejection. Require the supplier to state a shelf life and a storage condition, and check arrival weight against declared net weight, since an unexplained gain in weight on a sealed bag is a moisture signal.
Where a foundry is chasing intermittent porosity, the sequence that resolves it fastest is to audit the charging sequence first, then the store, then the certificates. Reworking the alloy specification is usually the least productive place to start.
Frequently Asked Questions
Q: How does moisture in the charge cause porosity in castings?
A: Water dissociates at melt temperature and releases hydrogen, which dissolves in the liquid metal and is rejected during solidification as gas pores.
Q: Which charge materials carry the most risk?
A: Lime and slagging fluxes, damp ferroalloys and deoxidisers, hygroscopic powders, and scrap stored in the open. The risk is highest when they are added to an already liquid bath.
Q: How is moisture in a charge measured?
A: By drying a properly taken, sealed sample to constant mass and reporting the loss, following a defined procedure such as ISO 3087 for ore consignments or an in-house test for foundry additions.
Q: Is there a universal maximum moisture content for charge materials?
A: No. Acceptable limits depend on the alloy, the melting route and the defect sensitivity of the product, so limits should be derived from plant data and specified per material.
Q: Does pre-heating scrap or alloys remove the risk?
A: Pre-heating removes surface moisture effectively and is standard practice where it is permitted, but it does not remove moisture trapped inside porous material.
Q: What should a purchasing specification say about moisture?
A: A declared moisture figure with each lot certificate, moisture-barrier packaging for hygroscopic products, a stated shelf life and storage condition, and clear rejection criteria for damaged packaging.
