Marble Machines, Delivery Systems and Gob Feeders
By Steve Sturtz © June 1, 2012
Judy Huxford, Gary Huxford, Scott McBride, guests, fellow marble collectors and friends. It is such an honor to be speaking here this evening so, I have chosen to share with you original research on Marble machines, Gob feeders and Delivery Systems.
My parents gave me my first sulphides and latticinos in the early 1960’s. I had never seen such wonders. These gifts started me on my adventures in collecting marbles.
In the last seven or eight years, the process of marble making has fascinated me. JABO has led me down many wonderful paths. Learning how delivery systems work is one of those paths. I realize that if I understand how the system works, I can ID marbles more accurately.
My curiosity about how marbles are made started in late 2007 when David McCullough and I had started talking about our goals for saving JABO. We wanted to keep the jobs on shore. We wanted to make the prettiest marbles ever made. Our daily conversations included something about the process, the tanks, the nozzles, and the placement of air flow. How did he try to make corkscrews? How did he try to make flames? How did he think flames had been made in the past? What happens if he made a tank this way? What are some things that he had always wanted to try? So, day by day, my knowledge of what he was doing grew.
I hope that after this evening, you will begin exploring some new thoughts about ID-ing marbles. Ideas based on the manufacturing process. Can we change how we ID marbles through understanding engineering, scientific method, chemistry, and physics that are applicable to marble making?
I have spoken with people from around the country. Many have the same interests as I. Gerald Witcher has joined me in furthering this vein of thought. He has written an article about Marble King which includes how their delivery system works. Don Miller has added his insights in his article about Anacortes which shows how that delivery system works. Both articles will appear in the near future in different venues. This speech and the articles form a troika that will get people exploring a new thought. Together they open a new door and chapter of how to ID marbles more accurately. The impressionistic observations all of us have made in the past, are now making us all dinosaurs. ID-ing will become more scientific.
I hope that those of us who are doing research are in a position to figure some of these things out. It is this research that puts collectors in a position to establish value and relative value.
Patents are not the point of this speech. I want you to “see” the general flow of glass through any system. The pieces of the puzzle are based in scientific method, fluid flow analysis, chemistry, physics and mechanical engineering. They can be quantified machine by machine, delivery system by delivery system, company by company. We can go back into the past and reengineer it. New knowledge of the present will allow us a better understanding of the past.
I thought each system could only make a very narrow range of specific items. Thanks to John McCormick, David McCullough, Raelyn Dolton, and Todd Burnworth, I have learned that these systems have a great deal of flexibility depending on how good the marble maker is. Small adjustments make huge changes in the pattern.
There are three main components in a machine-made marble delivery system. They are the tank, gob feeder, and machine. I will follow the glass flow from the back where it enters the tank all the way through to where it becomes a round marble.
The heart of any marble delivery system is the tank and how it is built. An easy way to understand a tank is by comparing it to a bathtub and drain. A marble tank is just a big tub that holds and heats glass as it melts and comes forward by gravity and exits through the drain which is called the exit orifice or “bushing”. That drain, like your kitchen sink, bathtub, or toilet, will be spinning in a counter-clockwise rotation, if you are north of the equator. Direction of spin becomes important when the marble patterns and number of colors become more complex.
Marble King tanks are about two tons. At JABO, they hold about one ton. These tanks are 2 inches deep at the back, the bottom is sloped forward and the glass is 5 inches deep at the front. JABO could not heat glass that was deeper, all the way through. Vitro Anacortes, on the other hand, used a tank that was a 16’ tub style. They ran their glass 10 inches deep. The tank bottom was level to the floor. Other companies used other set-ups. The length and width of the tank have an effect. It seems that a longer tank gives better consistent heat all the way through. They “seem” to flow better. The reason is a more thorough and consistent heat which keeps the molecules of the glass “at temperature”.
This drawing is one of the many I have received, signed and dated, by David McCullough. David gave me this after we had discussed various tank configurations and how these configurations would change the architecture of a marble. It is a drawing of a tank he has used. It is interesting to see the evolution of his tanks over time as shown in his drawings.
The base glass, whether it be cullet or batched, goes into the back, melts for 6 to 12 hours to a temperature of 2200° and then gravity moves it forward to the exit orifice. The common thread in all machine-made marble systems, whether internal delivery, external delivery, or multi- stream, is that the colors come together near the exit point at the bottom front of the tank. They might all do it differently, but the result is the same.
Tank preparation i.e. the formula of base glass and striking glass, how long the glass is heated before the exit orifice is opened, and the placement of the pots and crucibles are important keys in making pretty marbles. There is a huge distinction between making great glass and making pretty marbles.
Base glass and striking colors are interesting. Some companies did what is called batching, which means they custom made their glass using a formula. Akro would batch certain colors in big batches and others in small batches so they would have better control. They sometimes used cullet. The real key is to have glass with similar coefficients of expansion in both striking color and base color. Glass expands and contracts at different rates. If the coefficients do not match, the marbles will crack. There are a few exceptions to this.
The art of marble-making is determining which colors go well together, which colors run well together, and what is pretty. The complexity of choices varies. If you took 10 different groups of people, the same glassmaker and the same palette, you would get totally different results. Color selection will change the pattern somewhat. Darker colors seem to melt more slowly, and the light colors melt very quickly. Sometimes yellow runs through these tanks like thin, water-based paint, yet black, purple and other dark colors stay thick.
Until the last couple of years, JABO tanks had a dam. The dam was in the middle of the tank to the rear of the exit orifice, behind the pots or crucibles. The dam did two things. 1st - it kept debris from coming down the middle, and blocking the exit orifice or getting into the marbles. 2nd - it slowed the flow a little bit, and gave the pattern a little more time to develop.
To make a multi-colored marble, you have to get 3, 4 or more colors delivered to the base glass without it melting into and changing the color of that base. I think it can be compared to a hot fudge sundae. If you get a hot fudge sundae and eat it right away, you are eating ice cream and chocolate. If you let it sit in the sun, the whole thing mixes together and becomes a tan, muddy color. To keep your striking colors separate, you do not put them in the back of the tank where they will usually taint your base glass. Generally, you enter striking colors through crucibles or pots at the top of the tank.
For the sake of this speech, I will refer to crucibles as the tubes that are suspended from the top of the tank and come down towards the glass. It is here that the striking color is added, a coffee can at a time. Some crucibles do not touch the molten glass. As the striking color melts, it dribbles into and through the base glass to the bottom of the tank and moves to the exit orifice. Other crucibles go from the top of the tank down into the base glass where its contents are released.
The pots are set onto the floor of the tank. There is a small hole drilled in the side at the bottom of it to let the molten striking color come out. It proceeds to the exit orifice. Whether by pot or crucible, the striking colors go through the main glass or under it. They swirl counter-clockwise out the exit. The distance from the pot to the exit orifice will change the pattern. However, as the pot wears and the hole gets bigger so will the width of pattern.
Cat’s-eyes are created with a nozzle inside a nozzle configuration. If the distance between the two walls of the nozzles is close, it will have a thin pattern. With Anacortes’ caged cat’s-eyes, as the distance between the walls gets wider, the pattern gets fatter. The pattern also gets fatter as the cat’s-eye crucible wears out from friction. The holes that extrude the color get bigger. More striking color comes out and the pattern gets fuller. A new crucible has holes that are 5/8 of an inch in diameter. After about 6 weeks, these holes enlarge to ¾ of an inch. The crucible, at this point, has become ineffective and needs to be replaced.
When the base glass goes out the exit orifice, it goes out through a ceramic funnel called a bushing. When the glass is molten, it is ready to come out of the tank. At this point, the gob feeder, which is blocking the exit orifice, is opened and the glass begins to flow. The gob feeder can be thought of, quite simply, as your garden hose. When you put your thumb over the end of the hose, you get more pressure and a direct, precise stream. That is how a gob feeder works. There are many types of gob feeders, from simple to complex. But the common thread in all of them is that they open or restrict flow of glass in a marble delivery system, hence controlling the size of a marble.
The marble machine consists of the cutter head and rollers. It does two things. First, it cuts a gob of glass and second it rounds the marble. These are the purposes of a marble machine. Most cutter heads go back and forth to cut the gob. The gob then goes down the chute into the rollers. There are two exceptions. The cutter head on the House of Marbles machine goes back and forward and some Akro machines used a spinner cup after the gob is cut.
Once the base and striking color exit the orifice together, the spinning stops. It is gravity-driven into the cutter head. While falling toward this cutter head, the colored glass is in stripes perpendicular to the floor.
As you change the height of the machine, the pattern will change. In my 2009 article, “Hey JABO, your Butt Crack is showing”, I showed how this happens. If the machine (rollers and cutter head) is as high as possible and therefore as close to the tank and exit orifice as possible, it can make opaque game marbles without striation. As it is lowered, it moves farther away from the tank. It will extrude a short fat gob that folds over on itself and creates either a Vitro “V” or a Butt Crack. Lower still, you get West Virginia swirls. To see this more clearly, think about taffy being pulled or ice cream coming from an ice cream machine. The farther it falls, the more stringy, it becomes. At some point, as it is falling, you get a whipping effect where it goes back and forth. The longer the distance from the tank exit to the cutter head, the more whipping motion, the more stringy the glass, and hence the more complex pattern. The pattern is developed in space between the exit orifice and where it piles up on the cutter blade before it is cut and drops down the chute into the rollers.
David McCullough discovered two different ways to make flames. He knew from experience at Champion that the flames came when the tank was running out of glass and the color was changing. When he was making Chinese checkers and was changing from yellow to blue, there would be a transition where he would get yellow and blue flames until the yellow ran out and only the blue remained. He was sure that this occurred at Alley and Christiansen, also.
The longer a tank was in use, the more etched the bottom became and this, too, would make flames. He tried manually etching the bottom of the tank with a single groove running from one pot in the corner to the exit orifice. He later etched three grooves that went from the front right pot as you face the tank to the exit orifice. Oddly enough, on both of those tanks, even though he etched all four corners only the front right seem to work well, whether there were one or three grooves. The front left worked a little. Neither of the back sets of grooves worked. David is the first marble-maker to etch a tank in an attempt to get flames. The resultant marbles were fairly distinct, three-colored flames.
People have asked why flames are not made all of the time. The answer is simple. The bottom of tanks deteriorate rapidly becoming very rough and exceedingly sharp. A tank would make flames for a day or two, but then, JABO would have to switch to industrial marbles for 6 weeks. When David was able to get back to making Classics, he would only get a few flames. The sweet spot was gone.
David was encouraged to get the Vitro peewee machine out of storage in 2011 and he made a few runs. With it, he was creating the flames the way Christiansen did. The reason I say this is I talked to a friend who is based in scientific method. He told me how CAC made flames. And then when David explained how he had accomplished it, the explanations matched. I realized that both of them understood what CAC had done; one from the school of hard knocks and the other from academic training.
There are three ways to make a patch and ribbon. Popular wisdom is that you have to set your tank to nozzle and nozzle. That is one way to do it. Marble King has made patch and ribbon this way for years. The second way is to adjust the height of the machine until you have the proper distance from the exit orifice to the cutter head. The third way is to knock the machine out of alignment with the exit orifice so that the glass flow is off center. All three work.
I wondered if a system that makes West Virginia swirls can make patch and ribbons or flames. The answer is yes. Can all of the machines and delivery systems make any kind of marble? No. A cat's-eye tank will make cat's-eyes because of the nozzle set up. A patch and ribbon system might be able to make a swirl. But, I have not seen an experiment to show how broad that system can be.
Roger Howdyshell told Dennis Webb in an interview how he made cat's-eyes and patch and ribbons which are the opposite of each other. They use the same technique of one nozzle inside another nozzle. With the patch and ribbon, the inner nozzle delivers the base color, and the veneering is delivered on the outside through a multi-holed nozzle. For the cat's-eye, the inner nozzle has holes that put the color in the vanes of the cat's-eyes and the base glass on the outside.
I have picked up the word Snowball from David McCullough within the last couple of months. I was talking to him about Dennis Webb's work and particularly his interview with Roger Howdyshell. David, in fact, was present during that interview. Dennis had introduced David to Roger. I wanted to know if everything that Webb wrote was correct and complete. David did not answer that question directly which he often does not. He said, “No, we threw him a bunch of Snowballs.” I asked him exactly what a Snowball is. He said, “It's a story that could fall apart pretty easily.”
I realized why they used the term Snowballs. They were both happy to tell Dennis Webb the process in layman's terms ... rough stroke. But, Dennis must have asked some good questions. He did not get complete answers. Roger and David gave him as good an answer and as complete an answer as they possibly could, without giving up a trade secret. They would tell the story so you could understand the flow in your mind's eye. You could explain it in print, but you would be missing enough detail, so you could not reproduce it. These guys were not telling lies to tell lies. They were not even telling lies. They were telling incomplete stories which they called Snowballs. All they were doing was protecting proprietary information so it did not leak. I like the term Snowball. The lore of marbles; of who did what to whom; of how the systems work; might be just that...lore. Slowly, as the scientists get involved, more of these Snowballs will melt, and we will find that as they melt away, science will put us on a very solid foundation.
Two occurrences that are external to the marble are cold roll marks and heat sinks. Both are identifiable on marbles and both have to do with heat. Cold roll marks are common. When the heat is dropping too quickly as the marbles go down the rollers, the rollers cannot round the marble all the way.
Heat sinks and things that cause the temperature to drop are a big issue. One of the places where it is particularly noticeable is with the gold Lutz, which is applied externally. During the 2009 Tribute Friendship run, we were putting Lutz rod on 5/8ths. We expected that since it went well on the ¾ marble, there would be a bigger stripe on the 5/8ths. Instead, the Lutz was thin and sporadic. The hot marble sucked the gold Lutz into its center.
Blowing frit to make a confetti marble is also a heat sink. David McCullough uses a system that he has employed for thirty years. He uses a pipe to blow frit right onto the glass as it exits the tank and before the hot glass hits the cutter head. It is still at 2200° and melts the frit smoothly on the outside of the marble. Other methods such as adding the frit after the cutter head when the marbles are already cooling creates a heat sink and the frit goes to the center of the marble instead of sticking to the surface.
Anomalies, vagaries and all those sorts of things should be addressed. I would look at a JABO from a particular year, and I would have 50 or hundred of them. As I grouped them together into families, there would be a very attractive three color marble. All of a sudden I would get one that had four or five colors. How did that happen? That came in one of two ways. One: the bottom of the tank is very rough because of erosion. If a little color gets stuck in one of the cracks and crevices, eventually, it pops out and goes into a few marbles. Bang! Your three color family of marbles now has a few marbles that have 4 or more colors. The second anomaly occurs when the colors in the pots are changing. As one color starts to melt and go down into the tank heading for the exit orifice, another color is introduced. This one has not melted yet and is therefore heavier. Visualize breaking a Popsicle in half and you put one half in a hot tube. As it is melting, it becomes a puddle and the liquid goes out the bottom. Then you put in the second piece, which is now heavier than the first piece. It pushes the first half out of the funnel. There is moment where you might have a transition of the two colors. This is also true of marbles. You get a few marbles that will have two or three or four colors or the blending of colors that are in there. And then, eventually, they will switch and the second color will come in and you will go back to three color marble. Understanding that the color in those pots is transitioning on a regular basis is very important. As we learn more about this transitioning, our hobby will grow in the understanding and recognition of hybrids.
JABO experimentals have proven the cost of goods sold need not be the issue. The issue is the quality of product. If you make a quality product that is pretty, it will sell. Making a 21st century marble delivery system is really a simple matter of using available technology. I do not think it would have to be more complex than using a PC. My first hint of this was looking at the Anacortes tank. Dick Ryan, an engineering graduate of Clarkson University in Potsdam, New York took a different approach to marble making. He started while Vitro was still in West Virginia. He put dials and meters on the tanks. These instruments gave him glass temperature, gas flow rate/volume and air rate/volume. Barometric pressure can be controlled by changing your balance of gas and air so you have a neutral flame that neither oxidizes nor reduces. The colors will be more consistent. It is very controllable by putting in thermal couplers and gauges on your gas and air both by volume and by rate of flow. This set-up allows you to run at a constant, predetermined gas/air ratio. You then can change your flame construction so it provides a more even and accurate glass temperature. With the sensors, the computer can then change the gas air mixture so that it is always neutral regardless of the temperature or what it is doing outside. You are going to have better control.
The 21st-century glass tank would be much like a 21st-century automobile. I say that because the internal combustion engine really has not changed much over the years. It is still a closed air pump just like the human heart. But, as auto technology has improved, more and more sensors have been put onto the engines. Most measure performance. Some change performance. But, the basic principles have not changed. With these computers, for example, we can now drive a fuel injected car from sea level to 12,000 feet and never have to make any air adjustments because the computer sensors know that as you get higher and higher, the air gets thinner. The engine is being air starved and it needs more air. The computer adjusts and keeps the proper balance of gas and air. The same can be done with marble delivery systems. Once done, it becomes easier to be a glassmaker.
But the mystery and mystique of being a great marble maker will not change. The marbles will only be as pretty as the palette, the imagination, and tank design of the marble maker. I think marble makers should use current technology. I think the beauty and quality of American machine made marbles would jump to a new level.
Identifying the marble-making process and looking at the resultant marbles, helps ID-ing marbles that have been made in the past. In March 2008 David learned how to control aventurine both by width and density of product in a marble called Sammy Sparkle. The following weeks found him experimenting again in preparation for the first of many Experimental runs that has kept JABO alive for the past five years. He wanted to put aventurine in a three quarter inch marble. The marble Sammy Sparkle’s Big Brother has long burnt out stripes with a few flecks of aventurine. Howard Powell saw this marble and knew he had seen it before in an Alley he owned. This discovery has helped us ID Alley gold Lutz marbles from all three of the Alley factories in different color combinations. Then, I found a gold Lutz in a Marble King that I had purchased from Bruce ‘Triker’ Burkhart. To my knowledge, that was the first one we found in the wild. I gave it to a friend who has a very significant Marble King collection. Later on, when I was having dinner with Beri Fox, I invited this friend to come along, too. At that dinner, I asked Beri whether or not her father had ever experimented with gold Lutz. She told us how her father did in fact make some marbles with Lutz and they were her favorite marbles. She was surprised to find out that some of those same marbles had made it out the factory door and into the general public. We have seen something in the present and we connected it to the past. We have now found Gold Lutz in Peltier marbles.
One of the things I need to make very clear…. My understanding of delivery systems has changed so dramatically that if this speech could have been heads or tails, I started down the tails path. My learning curve has become steep. I am on the heads side of the coin, now. I am learning the flexibility and the broad range of each delivery system is larger than I had originally understood. I thought they were each very specific. Some really are. By and large, there is much more flexibility in what they can make than we recognize….. than I had recognized.
My thoughts down this path came, at first, from whimsy when RaeLyn gave me an Anacortes confetti. I met John McCormick. He shared his research and guided me. Because he is an engineer, he is based in scientific method. It is provable. It is not what I think. It is not an opinion. It is what can be proved. The laws of physics are not going to change for anybody. All these things have to meet the laws of physics, the laws of chemistry and fluid dynamics. And so the help that people gave me in general and specific terms made this possible. Much more needs to be done on how marble delivery systems work, ..... how the atmosphere in the tank works... how you control the color.... all those things. I believe this is the first original work done concerning these wonderful old marble delivery systems. Hopefully, as my research continues to evolve in depth and breadth, more information will come to light. Our hobby will, then, be pushed forward quickly.
I want to suggest to you that the first thing you want to do as a marble collector is to buy pretty marbles. If you buy marbles that you think are pretty, you will enjoy them. But, if you are going to take the next step forward into this process, learn how the marbles were made. By identifying the process, it will help you identify marbles.
JABO: A Classic. Sturtz, Steve; Johnson, Michael.
David’s JABO Renaissance. Sturtz, Steve; Johnson, Michael.
2008 JABO Classics: The Experimentals. Sturtz, Steve.