ANALYTICAL DEVELOPMENT OF CAST IRON ENAMELLING
Kevin Kibble - PhD, University of Wolverhampton, Great Britain
Martin Murphy, MSc, Waterford Stanley, Ireland
Jon Allen - BSc, University of Wolverhampton, Great Britain
David Hubble - BSc, University of Wolverhampton, Great Britain
Sheila Donegan - PhD, Waterford Institute of Technology, Ireland
Judith Coughlan - BSc, Waterford Institute of Technology, Ireland
Joe Doyle - BSc, MIRC at the University of Limerick, Ireland
Lorraine Cunning - BSc, MIRC at the University of Limerick, Ireland

Abstract
This paper describes selected work on an EC CRAFT project (ADCIE) on enamelling of cast iron. An expert system has been developed for use as a diagnostic tool in identifying enamelling defects and providing remedial solutions.
The routine for boil defects is described and the results of investigative work on the causes of boil defects given.

Introduction
The project was carried out under the EC CRAFT [1] programme. The project objectives are:

The SME partners in the project are Waterford Stanley Ltd, Eire, Escol ProductsLtd, UK, Quality Porcelain Enamelling Ltd, UK, TRICO V.E. Ltd, UK and Ouzledale Foundry Ltd, UK. The RTD partners are the University of Wolverhampton, (UNOWV), UK, Materials Ireland Research Centre (MIRC) at the University of Limerick, Eire, and Waterford Institute of Technology (WIT), Eire.
UNOWV is responsible for developing the expert system and for carrying out compositional and structural analysis of enamels; MIRC is responsible for carryingout physical property measurements and WIT for carrying out enamel failure investigations.
In this paper only work on the expert system and investigations into the "boil" enamel defect are reported. Further project work will be reported elsewhere.

Factors influencing "boil" in the enamel cover coat
A major defect in enamelled cast iron is boil (surface porosity in the enamel) and this is linked to the cast iron quality, effectiveness of the shot blast operation and the groundcoat layer.
To produce enamelled cast iron with deep saturated colours a groundcoat is essential. Boil defects occur when the groundcoat layer is either missing or too thin. An example of the large bubble structure found in an enamel boil is shown in figure1.
A groundcoat must have sufficient porosity to absorb or dissipate casting gases but if the porosity is excessive than enamel will fail through loss of adhesion to cast substrate, compare figures 2 and 3.
Strict control is necessary to achieve the optimum groundcoat and a matching cast iron substrate.

Castings without a groundcoat layer are prone to boil defects unless a directon cover coat is utilised, figure 1. In single coat systems dissipation of gases relies on the fired coat working range being of sufficient time for healing of large bubbles to occur. Two coat systems rely on the groundcoat having sufficient porosity to dissipate the gases.
A relatively thin groundcoat layer (around 75-100 μm) allows "new" castings, i.e.
castings processed direct from the production cycle with minimum storage after manufacture and newly shot-blasted, to be processed without boil occurring, see figure 2. This observation suggests that casting gas evolution, during enamel firing, is probably linked to corrosion of the substrate during storage or even while waiting for processing.
Will be easier along microstructural interfaces, i.e. flake graphite-metal interfaces. Industrial trials at one of the partner SMEs have shown that boil is linked to casting age (i.e. storage).
An increase in groundcoat thickness enables "older" castings to be utilised without boil occurring, figure 3. The increase in the depth of the groundcoat porous layer (increase in porosity volume) will allow older casting stock to be successfully enamelled. It is probable that the porous groundcoat layer acts as a membrane to allow diffusion of gas, emanating from the casting, through to the cover coat, or to provide a reservoir space for casting gases.
The result is that large bubbles evolved at the casting surface are dissipated.
However, thick groundcoats are more likely to result in poor enamel adhesion.
Coarser graphite flakes in the cast iron substrate cause boil. Dawson [2] has suggested that the reason for boil occurring is due to the formation of oxidation pits in the iron, during firing. However, when coarse graphite flakes are present it is also probable that corrosion, along graphite-metal interfaces, will extend to greater depths within the casting with a coarser microstructure.
Under these circumstances shot blasting may not remove a sufficient depth of cast surface (i.e. "corrosion layer"). Confirmation of this observation was found in an investigation of oven door castings, enamelled at one of the partner SMEs. This investigation showed that the castings, which have both relatively thick and thin cross-sections, had boil defects coincident with the thicker cast sections but absent from regions of the castings with thin cross-sections. Representative microstructures of both cross-sections can be seen in figures 4 and 5. In figure 4 the graphite is too coarse but figure 5 shows an acceptable microstructure for enamelling cast iron. With coarse graphite structures annealing the cast iron, prior to the application of ground coat, makes the situation worse in promoting the tendency to boil. It is probable that oxidation of the graphite flakes to FeO occurs to some depth within the cast iron.
Corrosion is more rapid with oxidised iron than bare metal. Shot blasting operations will not remove sufficient corroded metal.

For successful enamelling to cast iron a graphitic microstructure is essential as the iron will deform easily during shot blasting.
The irregular surface created plus the graphitic structure allows keying-in of the enamel and therefore better adhesion. Where large carbides, formed during solidification, occur the cast surface is too hard to deform, cleaning of the surface through shot blasting is more difficult and there is the likelihood of an enamel boil occurring. The presence of these carbides is termed "chill" an example of "chill" at the casting edge is shown in figure 6. The micro cross-section, figure 6, was taken from an enamelled casting and shows that carbides still remain after enamel firing. In figure 6 surface inclusions can be seen, these have not been removed in the shotblasting operation.
Wherever surface inclusions are present in the cast iron there is always the possibility or either boil or pinhole defects being found.

figure 1 - "Boil" from large bubbles in a cover coat. The largest bubble is 100mm in diameter. The lack of groundcoat is due to faulty spraying

figure 2 - Groundcoat (GC) and black cover coat showing the typical bubble structure. The groundcoat layer is around 75 - 100 mm thick

figure 3 - A thick groundcoat (GC) that is more tolerant to quality of cast substrate but is more prone to failure through poor adhesion

figure 4 - Microstructure from a thick section of an oven door casting, showing coarse graphite flakes ("kish") in a ferrite-pearlite matrix. The graphite structure is much coarser than that observed in figure 5. The cross-section was taken from a region with boil defects in the enamel

Underfiring in the enamelling furnace is influential in boil defect occurrence. Underfiring is caused either by too low a firing temperature or insufficient time at temperature. Boil in the form of large blisters can heal with sufficient time at temperature. Thick cast sections present a problem, as they take longer to heat in the furnace with the attendant danger of underfiring, and they promote coarser graphite microstructures unless careful control of iron composition and casting practice is exercised.
In summary, sources of gas evolution that cause boil from the cast iron are: corrosion, graphite, carbides and inclusions.

figure 5 - Microstructure from a thin section of the oven door casting showing a finer graphite structure (in a ferrite matrix). Boil was absent from the enamel in the region of the casting where this cross-section was taken

figure 6 - Cross-section showing the cast surface consisting of eutectic carbides ("chilled" iron) in a ferrite matrix, at the casting-groundcoat interface. Some graphite is present in the form of fine "rosette" clusters, almost certainly formed through pearlite decomposition during the enamel firing operation. Note the two casting surface defects which are either porosity or inclusions. The surface has been too hard to remove in the shotblasting operation

An example from the ADCIE Expert System
The ADCIE expert system begins from the screen shown in figure 7. The user selects his company system and is introduced to a section giving the options of a brief description of enamelling, a glossary of enamel terms and examination of enamel defects. The menu for examination of enamelling defects is shown in figure 8. If boil is selected then a description of the boil defect is given, figure 9, with accompanying macro and micrographs.
The user is then led through a series of option questions concerning the cause of the defect. If the composition of the iron is selected as causing an unsatisfactory microstructure for enamelling then the casting composition menu is selected, figure 10. If the composition is known, it can be entered in the composition range screen, figure 11, and then checked for Carbon Equivalent and manganese : sulphur ratios. For the example composition shown in figure 11 the Carbon Equivalent Value (CEV) is determined to be 5.1 wt.%.
This is too high and will result in a coarse graphite microstructure prone to causing boil defects in the enamel. Recommended compositions are given in the expert system. Similar routines are given for the other enamel defects.
The ADCIE expert system has also been developed as a training tool for experienced and non-experienced enamellers. Information on good practice is also included in the expert system, e.g. quality control methods specific to the processing of enamelled cast iron. The ADCIE expert system will be made available for enamellers of cast iron.

figure 7 - Opening page

figure 8 - Enamel defect menu

figure 9 - Boil title page

figure 10 - Casting composition menu

figure 11 - Composition range

Conclusions
An expert system (ADCIE) has been developed as a diagnostic tool for identifying enamel defects and providing remedial solutions. The expert system also includes information on good practice and can be used as a training tool for experienced and non-experienced enamellers.
Experience gained from investigative work in the partner SMEs of the ADCIE project has been utilised in developing the expert system.
A description of the some of the investigative work has been presented in this paper specific to the boil defect. This work has shown that sources of gas evolution that cause boil from the cast iron are: corrosion, graphite, carbides and inclusions.

References
1. EC CRAFT project BRST-CT98-5322/BE-S2-5571.
2. Dawson J.V. "Some factors influencing the occurrence of boil defects in wet process vitreous enamel on cast iron", BCIRA Journal, V9(6), November 1961, pp.820-847.