User:Steve Race

INMOLD PROCESS

It seemed, and was indeed, a novel idea to place granular MgFeSi into a mould cavity and pour molten metal over it. Even in 1968 the staff of Parkfield Foundry were well acquainted with experiments and even more familiar with the pyrotechnics generated by the porous plug SG method. It was then with some trepidation they awaited the trial, in the event the experiment appeared to be a failure as no flame or fume was observed, an immediate hypothesis was that the alloy had failed to dissolve and no further thought was given to the project.

Some days later when a sample was examined under the microscope, it was suddenly clear that the alloy had dissolved and produced SG of a quality never before seen. When Cliff Dunks, the instigator of the experiment reported back to his management it was already clear that something big had been discovered, how big, was yet to be determined. Work continued at Parkfield to verify the results and attempt to establish process parameters. What did emerge very quickly was the recovery of magnesium with the new method was double that of the two most commonly used processes, porous plug and sandwich.

Both porous plug and sandwich processes used around 2.0-2.2% of a 5% MgFeSi alloy, giving nodular structures at around 40% recovery, the new method replicated the magnesium levels at 85-90% recovery and thereby reduced usage of alloy to around 0.7-0.8%. It was also quickly realised that no traditional post inoculation was required, a further substantial saving in costs.

Within some weeks then, the new method and going no further than alloy usage, offered the prospect of cost reductions in the order of 45-55% against the two most common production methods. During a conversation on further patterns for trials, a young apprentice name unknown, enquired why they were putting alloy in-the mould. The name stuck and the INMOLD Process was born.

It would be fair to say that in Autumn 1968 while people were aware of the likely cost savings with INMOLD, attention focused on metallurgical aspects. Nodule counts, along with electronic Calculators, a phrase not yet invented, were perhaps double the normal levels achieved, the significance of this was unclear, certainly there appeared to be more ferrite in the matrix, but what about carbides,' in certain thin section castings at the end of a ladle, there was a likelihood, but with INMOLD nothing.

Prior to INMOLD, of course, what nodules there were, was what you got, but now it was clear that increasing numbers of nodules increased the levels of ferrite and guaranteed virtually no carbides, heat treatment could be eliminated. So within the first seven months of 1968 a process had emerged that offered an enormous boost to SG potential that to some extent had reached a plateau after Millis's pioneering work and invention.

Still there was work to do, an alloy had to be designed for this new process, many sizes had been tried and indeed 25mms appeared to give the best results but a l-4mms alloy was eventually selected as it was obviously easier to volume measure accurately than the lumps. In the event this was a correct decision, the l-4mms also presented the right density of packing to the metal flow, the metal flowed over the surface barely penetrating the chamber.

High Calcium?, low calcium? all of these parameters had to be determined, the final selection in 1968 was made and has stood the test of time to current periods. Up till November 1968 many castings had been produced using the INMOLD Process. No precise parameters had emerged at this stage however, allowing calculation of the now named Reaction Chamber, the Chambers were calculated on the basis of the Magnesium recovery known to be achieved, and the Magnesium required for the casting.

Christmas 1968 came and went, but when the team assembled early in the New Year a significant, and probably critical new component had been added to the puzzle. It was discovered that if the surface area of the chamber was divided into pouring rate of the metal, the answer inevitably came out 0.9-1.0. This was the Solution Factor, the missing link which allowed precise calculations to be made to establish the Reaction Chamber.

Pour Rate Ibs/sec = Solution Factor Chamber area/sq.ins Transposing the equation thus Pouring Rate (known) = Cross Section of the Chamber sq.ins. Solution Factor 0.9-1 gave the possibility to accommodate virtually all pouring addition. The levels of alloy had of course been established at 0.7-0.8, so converting the actual weight required into a volume, with now knowledge of the area required, fixed the length, breadth and depth of the chamber. Both the INMOLD concept and the Solution Factor were granted patents, the official installation of the worlds first INMOLD foundry was completed in March 1969.

It was clear that INMOLD, while offering many advantages to the mechanised foundry, was not going to be suitable for the jobbing foundry. This realisation led to two quite diverse paths, one was the invention of the FLOTRET Process and the other, discussion with FIAT to establish the INMOLD Process as a major production method.

FIAT are now and were in 1970 a major European casting producer for their automobiles, their basic production method was the Sandwich Process which despite being well engineered still left much to be desired in terms of handling and environment, as the Sandwich was the state of the art at that time, the high levels of alloy usage and associated costs were assumed to be a universal constant. The idea of INMOLD therefore generated a good deal of interest and doubt. The work at FIAT took over one year, very early on, it was clear that INMOLD could produce SG iron, but what about consistency, physical properties? During the twelve months FIAT were able to establish many new parameters but also to the gratification of the original Parkfield team, confirm that was in the original manual produced in 1968 was correct. FIAT were able to establish a number of fundamental aspects of the Process and point out many more which were at the time of Parkfield, not considered significant.

It was gratifying to the original team that, the basics of the Process were as written in the first installation manual in 1968: What was established between 1968 and 1971 was as follows:- 1) At 0.010% S the INMOLD Process used between 0.70% and 0.85% of alloy. 2) The Process was fume free. 3) The inoculation provided by the INMOLD Process was at the optimum, examples of cementite derived from fade were eliminated. 4) With ladle treatments and manual pouring losses of metal could be as high as 3% from deslagging and pour down this metal with Sandwich would have been treated with MgFeSi and post inoculated. Any pour down with INMOLD was untreated. As the scale of production increased this became an increasingly important cost factor. 5) It was noted that with ladle pouring as practical at FIAT savings of 50°C of metal temperature was gained which in large foundries proved quite significant. 6) Physical properties with INMOLD were significantly better than Sandwich across the board. 7) There was a loss of yield, because of the chamber of approximately 4.5%. 8) Means of placing the alloy, confirming the addition would need to be developed. This in fact was achieved quickly and ingeniously, alloy was weighed automatically and dispensed into the chamber, confirmation of the addition done by light reflection from the bright alloy being picked up by a photoelectric cell. If no reflection was acquired the line would automatically stop.

The Critics

As news of INMOLD began to emerge in the press critics began to assemble, these were in three groups. 1) Alloy suppliers 2) Advocates of pure magnesium treatments 3) Universities The critics began with the yield question, there was of course no way the use of the chamber could be avoided and apparently a loss of yield of 4.5%, a further loss could be incurred if an impression had to be removed, it should have been a simple question of accounting, savings on treatment costs, inoculation, fume productivity etc., versus loss of yield. But INMOLD was not destined for an easy passage, despite the fact that at this time FIAT were producing circa 10,000 castings a day.

During the study period at FIAT it was revealed that in their situation they could obtain 1 ton of castings from 2.6 tons of Sandwich metal, but with INMOLD they used only 2.1 tons, despite the chamber. Furthermore when considerations were given to feeding and shrinkage potential it was concluded the feeding requirements for an iron containing 0.045-0.050% Mg, the as tapped Sandwich level fading to 0.035% Mg were appreciably higher than the consistent 0.035% Mg with INMOLD. This then offered further potential for improving the yield by reducing the diameter of the feeders, clearly pattern alterations for existing work could be phased in over a period when replacement was due.

Consistency and Inclusions Laboratory studies had clearly shown that the INMOLD Process could not work. The Process was inconsistent and produced inclusions. FIAT, working for their own edification rather than contradiction of the critics devised a most ingenious experiment, a quartz window placed in a mould, allowed a slow motion film to be produced, that in fact showed the solution of the alloy was entirely regular further confirmation of consistency was shown by the simple expedient of analysing for Magnesium and Silicon across the casting.

Early work at Parkfleld had shown an increase in Magnesium Silicate inclusions on certain castings, inclusion possibilities however, were and are a feature of SG iron production by any method. It should be noted that at this time of a40-DMrto~f4 aftd no colour television, filters were not generally available. FIAT achieved their targets using a small hole filter and by the mid-seventies 50,000 castings a day were being produced world-wide, some with, most without filters. The producers at least were not too discouraged.

The final throw on the inclusion debate came in 1977 when a paper presented to the World Congress noted heavy inclusions of Magnesium Sulphide and Silicate on casting copes, the base Sulphurs were 0.06% (5 times the established limit). Magnesium's of 0.12% (4 times the recommended level, using a 9% alloy as opposed to the 6% recommended.

Pure Magnesium versus INMOLD

A most comprehensive study was made of pure Mg versus the INMOLD Process. The paper showed that with INMOLD a casting yield from several foundries with INMOLD was 39%. This was somewhat surprising as FIAT (as the world's biggest producer at this time) had reported a yield of 48% with INMOLD and 38% with Sandwich. The bad news continued until the bottom line revealed that the treatment costs of INMOLD castings were 380% higher than pure Magnesium. Could this have been the start of some creative accounting? This news however, did not appear to have diminished the interest in INMOLD generally, enquiries in Europe were booming and foundries such as Buderus Eisenwerke had trialled and accepted the Process. 10

THE USA

There were two developments in the early seventies which prompted a lot of interest in the US. The first of these was the increasing use of autopours and the discovery of the problems of holding treated SG the second problem concerned the O.S.H.A Regulations on the generation of fume from SG iron processing. It was estimated in 1980 that the cost to a medium sized foundry in cleaning up fume emissions was circa $250,000. The arrival of INMOLD solved both of these problems, plus as bonus over 50% reduction in alloy costs. The INMOLD was quickly accepted by FORD and General Motors, FORD noting through Mr. Ackerman in 1978, that they (FORD) had been able to achieve compliance with OSHA with minimal cost. Globe Metallurgical were selected as the initial producer of the new INMOLD alloy becoming the major supplier to US INMOLD foundries from 1973, a position they have maintained to this day. Other foundries followed the big two in adopting the INMOLD Process, to the extent it was estimated by 1980, 20% of all castings produced in the US (excluding pipe) were INMOLD. In the early eighties, Eastern Europe showed enormous interest in the INMOLD Process. This elegant Process had found its way into the text books and universities from Moscow to Prague to Sofia. The young students took this knowledge with them to the plants and INMOLD can now be found in almost all of the Eastern European countries.

FLOTRET, SIGMAT et al

Given in the early days the early recognition that INMOLD excelled at mass production but not at 'one oils' work had proceeded to use INMOLD principles in a new process, FLOTRET. The first production with FLOTRET started in England around 1975, little more than an INMOLD in a box core, the FLOTRET preceded, however, the Tundish system by a good ten years, contrary to claims otherwise. It was however, not until 1980 that a workable high production tool emerged, from development again this proved enormously popular particularly with the jobbing foundries who valued the flexibility offered by the Process and the big reduction in fume, fume by this time was becoming an issue in Europe. Other processes such as SIGMAT followed increasing efficiency and flexibility. Weight ranges by the late eighties ranged from 250 kg's to 15,000kgs, units were sited directly into autopours, in front of furnaces and some were even fitted with wheels to treat near the mould, avoiding post inoculation. INMOLD itself was still being developed for the inoculation of grey iron, for the production of CG as well as a post inoculation booster to orthodox treated irons. Between 1985 and 1992 ladle processes and flow systems had to some extent narrowed the gap between INMOLD, at least if only alloy usage was considered. The recession of the late eighties however prompted foundries to look more closely at their process costs overall rather than the invoice price of the alloy, the virtues of INMOLD were being rediscovered. Modern electronics and ancillaries now offer virtually absolute security on such overstated but contentious issues such as alloy addition, detection etc., as more and more green issues come to the fore the axiom 'the best way to limit fume and flare in the foundry is not to produce any' will certainly mean this great process will continue to develop for many more years. The most recent estimate for world wide production of INMOLD castings is put at over 3,000,000 tons annually.

AUTHORS:

B. Race Managing Director Materials & Methods Ltd. ENGLAND

K.B. Turner Former Managing Director Parkfield Foundry ENGLAND Director Horwich Castings Ltd.

R.D. Jenkins Vice President Globe Metallurgical Inc. USA

The authors would like to note the following people for their major contributions to the development of the INMOLD Process.

CLIFF DUNKS (Inventor) Deceased ALAN TACEY Deceased JIM McCAULEY Deceased GEOFF MANNION MARCEL JAMES Deceased SNR. REMONDINO FIAT STEVE WEISS Harvard Industries