By John Campbell, Emeritus Professor of Casting Technology, Department of Metallurgy and Materials, University of Birmingham, UK
When liquid metals are poured, this innocent action separates the liquid into splashes and droplets, and each new surface becomes instantly coated with an oxide film, making the surface of the splashes and droplets effectively ‘dry’. In coming together again the dry surface films meet up with other dry surface films, forming double films. The double films are submerged in the liquid metal and float around, remaining in suspension. Depending on the metal or its impurities, the films are often Al2O3 or similar highly stable ceramic. The double ceramic films cannot bond across their dry central interface, so they act as cracks in the liquid. Turbulent pouring fills the liquid metal with cracks. The cracks are frozen into the cast metal, filling many of our engineering metals with cracks.
For the past 6000 years the presence of these bifilm cracks in metals has been overlooked. They have remained undetected and unsuspected because they are often only a few molecules thick. However, they can form large and serious cracks, some the size of newspapers. More often they are one or more millimetres across. However, they are often present in huge numbers.
The presence of dense populations of bifilm cracks in our liquid metals explains practically all of our problems with cast materials: the formation of gas porosity becomes easy as a result of the gas in solution precipitating into and expanding the bifilm into a bubble. Shrinkage porosity is similarly explained by the reduced pressure in the liquid sucking the bifilms apart to grow a pore. Hot tears and cracks are simply explained: they are the bifilms; I always cure hot tearing by improving the filling system design for the casting – the hot tear disappears like magic!
The presence of the cracks in castings and ingots explains their variable and sometimes unreliable properties. In addition, they survive significant plastic working during forging, rolling and extrusion, leading to cracking effects of many kinds, such as forging cracks, edge cracking during rolling, and crocodile cracks during extrusion.
For the metallurgists and those interested in the technical details among us, it seems probable that every crack in every metal is started by a bifilm (the only exception to this may be the formation of ratcheting cracks from slip planes in a fatigue condition). This radical view is the outcome of molecular dynamics studies which appear to find that no viable lattice mechanisms to initiate cracks exist. This is contrary to traditional thinking which has assumed such mechanisms, for instance, as wedge cracks from dislocation pile-ups at inclusions. In other words, cracks should not be able to initiate in metals. It seems probable that the bifilm is the only crack initiating mechanism. Possibly therefore, as dislocations are the explanation for plasticity, bifilms are the explanation for the initiation of cracks.
Bifilms also seem to be involved in the initiation of corrosion pits and intergranular corrosion, because, if the bifilm contacts the surface of the metal, liquid corrodants can enter the metal by capillary attraction. The attack is further concentrated because of the inter-metallics and other phases which preferentially precipitate on the outer wetted interfaces of the bifilm, creating efficient corrosion couples. Stress corrosion cracking and hydrogen embrittlement might also eventually be found to involve bifilms, explaining the previous baffling and intractable behaviour of these phenomena.
It is welcome news that some casting operations in the world are now tackling the problem, reducing bifilm populations in their cast material with encouraging and sometimes dramatic benefits to properties.
Elongations to failure in Al alloys are improved by at least 10 or 20 times, inclusions and cracking in steel castings are reduced by a factor between 10 and 1000 times, and there is evidence that creep life of Ni superalloys is extended by at least 10 times. There are reasons to suspect these benefits are only the start. Furthermore, these effects are expected to be universal in metals which form stable oxides, or which contain alloys or impurities which form stable oxides. This includes most of our engineering metals; whereas the oxidation-resistant metals such as gold are in principle free of bifilm problems, the addition of only minute traces of impurity elements ensure the presence of these defects.
The engineering techniques required to make a start on this problem, to achieve a mere order of magnitude benefit, are easily implemented in current foundries at negligible cost, merely by following techniques already well developed and documented. These techniques involve the pouring of metals without the introduction and entrainment of air into the melt. (Interestingly, vacuum melting and casting is of little use at this time because industrial vacuums used for melting and casting are not sufficiently good: so far as the liquid metal and its reactivity are concerned, current vacuums are equivalent merely to dilute air.) The channels which convey the liquid metal into the mold have to be designed with great care to avoid the ingress of air at any point. This is relatively easily attained by following simple rules. An Al alloy foundry making parts up to 2000 kg and a steel foundry producing castings up to 3000 kg are using these techniques to produce extraordinarily perfect castings. I have used the same techniques to demonstrate a uniquely defect-free 50,000 kg steel casting.
For the future, there is every reason to believe that there is much more to be gained from total elimination of bifilms, which, in principle, appears to be attainable with current technology. Foundries are now being conceived in which the pouring of liquid metals is completely avoided at every stage: the metal either flows horizontally or travels uphill, against gravity, never downwards. Naturally, such new ventures require substantial funds, but are no more costly than traditional casting processes. Also, of course, they tend to be highly environmentally beneficial, especially since they tend only to make castings, not scrap. Foundries which are on the drawing board, being built, or currently working, eliminating pouring and employing counter-gravity filling of molds include an Al alloy foundry making precision castings weighing only 10g, an Al alloy foundry making automotive wheels, and a bronze foundry making 4000 kg propeller blades for ships. A steel foundry to make castings up to 500 kg is at the concept stage.
These are exciting times. The discovery of the bifilm, and the development of techniques to reduce or eliminate it promise a new era of metallurgy and engineering. It will be a revolution!
Further reading by the author
“Complete Casting Handbook” Elsevier 2011
“Stop Pouring; Start Casting” The Hoyte Lecture; American Foundry Society 2012.