Challenging horizons: making it happen.
Over the years, I have written many papers and made many presentations, and never had much difficulty in selecting a topic or deciding what to say. But this time it was different. The AFS Charles Edgar Hoyt Lecture is a unique opportunity, so I found myself wanting to make this a unique presentation.
I wanted to avoid those topics with which I have been identified in the past. What I finally decided on was a compromise of sorts: I am tying one thought that is a pet current topic, but which I have never formally presented, with another that has long been a favorite subject of mine.
I am concerned that some very important needs and opportunities in the foundry industry are, in fact, lingering needs--needs that we have long recognized but have not resolved. I will address one of them today.
I am reminded, too, that a development with which I am still closely identified was also at its inception seen as a very important need. It too lingered for a while, but that need was resolved. The successful development of the 390 alloy, in spite of what were often seemingly overwhelming problems, taught me a lasting lesson. Even the most "challenging horizon" can be "made to happen."
I don't want to expound on the array of major issues facing American industry or the American Foundrymen's Society and its membership. These issues--the economy, environmental regulations, foreign policies and the like--have enormous impact, but they belong to the economists, statisticians, politicians and philosophers. They are not my topics; they are not me. What I will talk about, however, also has enormous potential impact on two vital issues confronting metalcasters--markets and competitiveness.
Larry Krueger, in presenting his Hoyt Lecture last year, provided poignant insights regarding marketing challenges. He said we should eliminate any acceptance of a commodity approach to marketing our product in favor of an approach that defines castings as engineered components. He said, too, that we must make our markets grow to utilize all of our qualified capacity. We must cease commodity selling and accept constant innovation as the route to market growth.
If accomplished, we will be an industry poised to move forward, poised to meet the challenges of the future, and ready to make whatever changes are needed to maintain a healthy U.S. metalcasting industry. Those were sage remarks.
You are, of course, aware that my background is aluminum castings. Most of my career was in a research and development-type environment, so my bent is naturally technical. My remarks today will be somewhat aluminum oriented. I am sure, though, that those of you whose interests lie with other metals will find my remarks equally relevant in your case, too.
I am going to suggest an idea--a goal--that I believe embodies all of Larry Krueger's advice. It is a key to our future competitiveness and expanding markets. That goal is zero casting defects, zero casting scrap.
The Challenge: Zero Scrap
I know of a number of U.S. foundries that pride themselves on delivering to their customers scrap or returns that are measured in parts per million; in fact, in only a very few parts per million. This is a great accomplishment. it has resulted in a scenario wherein some U.S. foundries are now the preferred suppliers to foreign manufacturers who, just a very few years ago, would not buy our products because they felt our quality was unacceptable. This is a real turnaround.
I'm not talking about delivering zero scrap. I'm talking about producing no casting scrap, no casting defects--none! No shrinks. No porosity. No cracks. No tears. No misruns. No cold shuts.
Is zero scrap in this context feasible? Is it within the realm of possibility? Or am I pipe-dreaming? Well, I believe zero casting scrap, or very nearly so, can be accomplished. Is it a challenging horizon? I think you could call it that.
I know that some of you are thinking right now, "This guy is nuts. Zero scrap is impossible." Well, I say if we don't begin today doing everything in our power to achieve zero scrap, we are very likely to look back at some point during the next decade and wonder what our competitors did in the 1990s that placed them so far ahead us.
If zero scrap is feasible, if it is within the realm of possibility, then what are the steps needed to get us there?
* First, cast parts cannot, by design, cause scrap. Foundries and their customers must work together to design parts for manufacturability as well as for function.
* Second, we can never abandon basic metalcasting principles. We must incorporate all of our best knowledge--our best techniques regarding metal cleanliness, microstructure controls, impurities minimization, elimination of turbulent metal flow, directional solidification--during every pour.
* Third, and this is where what I want to emphasize today really begins, we must very precisely and accurately model the casting process, every important detail of the process. Then we must automate and monitor the process--measure as many important dependent and independent variables as we can in real time during every casting cycle. Finally, we must let a computer massage those measurements to manage and control the process, in real time, to make a defect-free casting every cycle. Humans, with the very best of intentions, simply cannot provide the needed control. To reach perfection, we must rely on computers and machines to avoid the need for human judgments and the possibility of human errors.
What will this require? Three all-important elements. One, it will require a detailed understanding of what transpires in molten metal and during the molten-to-solid transition. The key word here is detailed. I will say more about this later on.
Second, it will require that we develop the accurate thermophysical data that is needed to precisely model the process. The key word here is accurate. And I'll say more about that, too, later.
Third, and, perhaps the most important element, are the gurus, the champions, who will make it happen. The key thought here is making it happen. I'll say more about that right now.
Making It Happen: 390 Alloy
Anyone who spends much time in a research environment realizes that only a small percentage of the "great ideas" that flow so freely from the "great minds" and "natural innovators" who staff research organizations, will ever become realities. In time, it becomes easy to accept that not every pursuit is going to be successful. Some good ideas, perhaps even great ideas--great new products, great new processes, businesses or technologies--fail because they were merely given the opportunity to succeed. They were, in fact, not made to succeed.
On the other hand, some seemingly impossible tasks succeed solely because an individual, some champion, some guru, simply would not give up.
The story of the 390 alloy fits the later case, except that it had several determined champions. Many of you know that the target of that development was a bare-bore aluminum engine. It was an alloy that would perform like iron in a cylinder bore. Actually, we started with an alloy that was already under development at General Motors in the mid-1950s. We were told that their alloy was not quite fully developed, that it performed well under some circumstances, but failed in too many other important tests.
We were asked to continue and refine the development and to arrive at a composition that would survive in all engine environments (and the engine guys had devised some seemingly very unrealistic tests). It seemed to many people an impossible task.
The metallurgical approach was to tailor an alloy for wear resistance, hardness, elevated temperature properties and the like. That part of the development proceeded smoothly and in timely fashion. Then it was discovered that something other than alloy composition was leading to bore failures under certain engine conditions.
Failed bores were analyzed to determine why they failed and solutions were developed. In these cases, failures were due to the bore surface. Three special surface finishing techniques were developed--one mechanical, one chemical and one electrochemical.
Finding out why bore failures occurred led to a bore system, an alloy, surface finish and piston treatment that overcame every performance shortcoming. We had a robust bore system.
But that alone was not enough. The casting experts told us that we could never mass produce components from a high-silicon alloy like 390. Required casting temperatures were entirely too high. Microstructures were almost impossible to control. The alloy, they said, would wear out molds, dies, crucibles and furnace tools much too fast.
Well, we learned how to cast 390. One by one, we addressed concerns of the foundry experts, learning along the way not necessarily perfect solutions but workable solutions. We learned how to handle the alloy's high heat of fusion; how to do cyclic die cooling; how to fill molds in a way that did not require excessive temperatures; and how to control microstructures with inoculants and by controlling casting conditions. Just as importantly, we taught diecasters to cast the alloy by comparing its requirements to those of their more familiar 380 or 383 alloys.
Still, that too was not enough. The machining experts complained that the 390 alloy wore cutting tools much too fast in mass production machining operations. Again, solutions were found by first developing an understanding of why 390 was difficult to machine, and then by engaging cutting tool experts to combat those difficulties.
Throughout its entire development, covering almost an entire decade beginning in the late 1950s, successes came because of the dogged persistence of champions of the cause who would not let the project die. Senior management continued to support the effort in spite of frequent setbacks. Key industry experts and companies, too, recognized the alloy's potential and brought their special resources to bear.
The 390 technology would not be with us today--all of those high-performance Mercedes, Porsche, BMW, Audi and other engines, and all of those pistons, transmission components and compressor parts made in North America--all would be dreams and not realities if 390 had merely been "left" to happen. It is a commercial success today, measurable in hundreds of millions of pounds of castings annually, only because it was made to happen.
Making Zero Scrap Happen
Like the 390 alloy development, zero scrap will never be a reality if we merely let it happen. We must make it happen.
Being competitive in today's markets no longer means merely being lowest cost. Customers throughout the world are demanding higher quality, greater reliability, greater durability, thinner cast sections, lighter weight and closer dimensional control in near-net-shape components. Foundries now play key roles in bringing new products from inception to market in tightly compressed time frames.
Casting suppliers must be able to assist customers from the onset of new programs with product design, both for function and manufacturability. They must be able to assist with the development, testing and validation of new cast products.
All of this must not only introduce the customers' product at the lowest possible cost, but continuous improvements are expected to further reduce costs at regular intervals over the life of the product.
Competing in that kind of environment is only possible through the extensive use of computers. The official AFS/USA exchange paper by Naysmith, Aucha and Ruff at the 1991 World Foundry Congress in Krakow, Poland, explained how computer-aided engineering technologies are used to optimize casting design and weight. It described how computers are used to improve casting quality and performance, to minimize implementation and validation time, and to arrive at the lowest possible casting cost.
The authors described in considerable detail a fully integrated system of computers and software to perform product and tooling designs; mechanical and thermal stress analysis; fatigue analysis; fluid flow; heat transfer and solidification analysis; gating and risering design; parametric analysis; design for manufacturability and assembly; and statistical variation simulation. That level of capability must become state of the art in truly modern foundries. It is state of the art in my new home, CMI International, Inc.
Foundries, in fact, already display a great deal of sophistication utilizing computer technologies to manage business and to engineer products and processes. Now, we must move to the next level. We must allow computers to monitor and control the casting process.
The lead article in the October 1993 issue of modern casting provided a vision of computer-aided casting in the year 2000. Orogo, Callihan, Sigworth and Kuhn described ascenario in which almost all foundries will utilize computer capabilities at least as sophisticated as was described at the 1991 World Foundry Congress.
Those authors also advanced the concept that process sensors will automatically record relevant process information to provide an ongoing operational database for automatic correlation with casting performance for statistical process control.
That kind of process monitoring and process automation begins to get at what I believe the industry must do in order to be and to remain competitive.
Competitors Are Smart, Too
We must take care that we do not embark on any zero scrap campaign believing that "we alone have this brilliant idea." It is simply not so.
Think for a moment about the impact that zero scrap would have. We would avoid the cost of initially making more parts than needed (or remaking parts) to accommodate some level of scrap. I am constantly surprised and appalled at the high levels of scrap that some foundries are willing to make as long as they can rework scrap and meet delivery schedules.
Just as important, we could avoid the costs of 100% inspection, rework and re-testing, etc. We would conserve energy and materials by producing and heat treating no more than we needed to ship. Perhaps most importantly, it would improve product reliability and our customers' comfort level and satisfaction.
Remember, our competitors understand all of these facts, too.
Before I go on, I want to make clear the fact that I claim no expertise in either zero scrap or process automation. I can only call on several years of experience in dealing with casting defects and in seeking ways to reduce scrap to support my sincere belief that automation is the only way to reach near perfection.
More than 30 years ago in 1963, Walter Sicha addressed this body as the Hoyt lecturer and described the importance of testing--of measuring and understanding--fluidity, hot tearing and cracking, shrinkage feeding, solidification cooling curves, secondary dendrite arm spacing, and hydrogen and non-metallic inclusion control. The understanding of all of these is still accepted as being essential to the making of sound and reliable aluminum castings.
In the years since his address, there have been many studies that have enhanced our understanding in some of these areas. And, there have been important breakthroughs in measuring, removing and controlling hydrogen; filtering out harmful nonmetallic inclusions; providing relatively long-lasting eutectic silicon modification; and in thermal analysis to assess the effectiveness of microstructure control additives before the melt is poured. These have been important developments. They have enhanced immensely our ability to make good aluminum castings.
Still, they are not enough. If, today, you revisit Sicha's Hoyt Lecture, you quickly realize that much of what he described in 1963 remains state of the art today in a great many aluminum foundries. In one sense, it is testimony to the sound technology and understanding that was developed by Sicha, his associates and others of that era. In another sense, however, it says that many aluminum foundries simply have not felt the need to really understand differences between the fluidity, hot shortness or shrinkage feeding characteristics of alloys. They can find comparisons, usually a rating of 1-5 or excellent to poor, in tables of characteristics in the data books.
Perhaps experience, their own or someone else's, has taught them to avoid, if possible, certain broad freezing range alloys, but they don't understand why. Experience may have taught them to favor certain gating and risering techniques, but they do not differentiate between alloys. They cannot relate fluidity or hot shortness or feeding characteristics to an alloy's specific chemistry or solidification pattern.
All of us realize that lack of technical understanding simply cannot support our future competitiveness in world markets. That's why we are here. That's why we have these Casting Congresses.
We have, indeed, heard some fine presentations these past few days: papers that address exactly the technical issues that are so important; papers that help us understand better how to control oxides, microstructure and porosity, and the effects of these on mechanical properties and scrap.
Yet, even as I listened and realized how well they address needs that can get us eventually to zero scrap (especially some of the modeling needs), I also could not help but feel that we are still thinking too genetically.
We study 356 alloy or 319 or 201, and this is fine. But we already know that 356 at its low silicon level is quite different in terms of fluidity and solidification and microstructural features than 356 at the high end of silicon. If we are to accurately model the 356 alloy, we must better understand and consider the small and seemingly insignificant differences, too.
Understanding Is Available
A few years ago, AFS and the aluminum foundry industry supported studies by Drs. Lennart Backerud and Lars Arnberg and their associates in Scandinavia that resulted in publications on the "Solidification Characteristics of Aluminum Alloys: Volumes 1 (Wrought Alloys) and 2 (Casting Alloys)." More recently, we supported their continuing studies measuring dendrite coherency during solidification of the same alloys. That work will be published as Volume 3 in the series early next year.
Backerud and Arnberg and their associates have now proposed additional studies to include measuring dendrite permeability, alloy fluidity, etc. and to develop from all of the studies computer models for casting.
I cite these as examples of studies recently supported by AFS research funds, and by industry and government funds. I don't mean to imply that they are the only studies of the kind being done certainly they are not.
My point is merely that it is bits of very specific information like dendrite coherency and permeability measurements that are needed to accurately describe and model with much greater accuracy than we are able to do today those solidification features of specific alloy chemistries in specific cast shapes cast by specific processes with specific tools that influence the formation and distribution of defects. This will ultimately allow computers to control the process and make defect-free castings.
Solidification and fluid flow modeling is, of course, already in widespread use. It, in fact, provides close-to-optimum placement of ingates, risers and mold cooling to ensure reasonably sound castings from very early pours utilizing a new tool. As successful as the application of computer-aided engineering is, however, it still relies on far too many human judgment factors and estimated thermophysical properties to accurately describe what occurs at the numerous heat transfer interfaces involved in metalcasting.
On occasions when I've expressed concern that pouring and solidification modeling is too generic--too insensitive to specific alloys, mold coatings, temperature regimes and the like--I've been told that developing the thermophysical data needed to more precisely model casting processes would be an immense and costly task.
I attended a Rand Critical Technologies Institute workshop in Cleveland a few weeks ago, together with quite a few other folks in this audience. One purpose of that workshop was for industry people to advise the federal government through CTI of the needs of the U.S. foundry industry. I proposed in one of the breakout sessions that the national laboratories be used to generate needed thermophysical data and materials properties.
Reduced military emphasis has certainly left many of those labs in search of useful projects. Some are already well equipped and staffed for the task. Whereas in the past, they developed data for space travel and models to describe chaotic events and nuclear explosions, they should find our modeling needs relatively mundane (by comparison) and quite simple.
Since that workshop, I have had several telephone calls from people in the industry and government who support the idea that such a use of national labs is one way that the government could really assist the foundry industry. I believe that developing the thermophysical data and materials properties would bring us much closer to the computer control capability that will eventually lead to near-zero casting scrap.
Synergistic Impact on Needs
The AFS Aerospace Structural Casting Council has proposed a strategic plan to increase the design and use of aerospace structural castings. They cite that no single industry contributes more to U.S. exports than the aircraft industry. Until the early 1980s, U.S. aircraft manufacturers accounted for 95% of the commercial jet airplanes manufactured in the world. That position has diminished to 65-75% today, primarily because of European government subsidized competition from Airbus.
Domestic airframes now use forgings; Airbus airframes are 30-40% castings. It is estimated that using cast structural components in commercial aircraft could save U.S. manufacturers over $2 million per plane. That would be a significant step in returning U.S. manufacturers to a competitive position.
AFS and the National Institute for Standards and Technology first identified the roadblocks to more extensive use of structural aerospace castings. Then the council was formed to devise a plan to overcome those roadblocks. Of the six tasks identified, four--resonable mechanical property design allowables, elimination of casting factors, consistency in casting processes, and measurement systems for quantity--would benefit from any industry thrust toward computerized controls and zero casting defects.
Any positive outcome from the aerospace-oriented initiative will benefit other users of castings, too. For example, the automobile industry is rapidly moving into lightweight structural castings. Improvements that applied to the aerospace castings would certainly improve automotive structural castings as well. So the total domestic impact of such an effort is enormous.
One more thought before I close. To support whatever improvements in technology, markets and competitiveness this industry chooses to pursue will require a cadre of well-trained, computer-literate foundry engineers to lead the charge. This industry is very fortunate to have an organization like the Foundry Educational Foundation that assures that metalcasting subjects and training remain an important part of the curriculum of some 30 major colleges and universities in North America.
FEF deserves the support of every person in this industry. It isn't just foundries that experience the effects of FEF's efforts. Suppliers and casting users, too, are benefiting by having sharp young graduates enter the work force with a hands-on knowledge and understanding of foundry processes and cast metals.
In closing, I want to re-emphasize that I believe our ultimate best approach to competitiveness and expansion of the markets for castings is to extend the way we now utilize computers, to allow them to monitor and control the casting process, and ultimately to achieve zero casting scrap.
To accomplish this end will require that we develop a better understanding and that we very accurately model solidification, feeding and hot tearing, etc. specific to the exact chemistry and configuration and process being used at the moment. It will also require that we develop better thermophysical property data and knowledge about the many heat transfer interfaces experienced in metalcasting.
These are huge tasks. The effort is indeed a challenging horizon--a very challenging horizon--but one that can and will be accomplished if we make it happen.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||1994 AFS Hoyt Memorial Lecture; zero scrap as goal of foundries|
|Date:||Jul 1, 1994|
|Previous Article:||A deeper look at casting solidification software.|
|Next Article:||Beyond borders: the challenge of Canadian foundrymen.|