The basic idea behind it is simple, a plasma beam is directed at a surface, where the plasma cools and solidifies creating a thin nanometer film on that surface. Many such films will be use to build three dimensional objects. If we can precisely control the plasma ion beam, that is, be able
to scan the beam across the surface in the x y direction, while the surface lowers by the z axis, we should be able, in theory, to build any part, any shape in three dimensions.
THE PARTICLE FOUNTAIN
The first part of the device, is what I call the particle fountain. Basically it is the source of plasma for the machine. A massive electron beam vaporizer.
Mixed matter is dumped into a large hopper built into the ground. Large trucks pull onto a concrete pad, the hopper doors are automatically opened and the truck dumps its load of forty or so tons of toner. Depending on the size of the AD itself, the matter stock hopper maybe small, being able to fit into someones basement, or even be a table top device, or as large as a city block, being fed loads brought by trains. The hopper has massive air tight doors, so that high quality vacuum is maintained during the loading of the hopper and the operation of the AD is not interrupted. Arrays of hoppers may be daisy chained so that the supply of stock is continuous, and the vacuum is not destroyed when stock is being loaded into the hopper. Think of the hopper as a massive airlock. Some advanced units later on might use a plasma window to maintain the vacuum while allowing stock to pass through.
From there, huge steel rollers crush all incoming stock to the consistency of fine sand, the stock, may be dry or wet, this is irrelevant. Through a bunch of pipes the sand is brought to the particle fountain generator. The stock material goes into a small inductively heated crucible, where the stock is heated to a high temperature. Above the top of the crucible, a large electron gun, with a short linac, which accelerates the electrons to a quarter of light speed, focuses and fires the beam into the center of the materials stock in the high temperature crucible. The kinetic energy of the electrons is converted into heat flux and the stock is instantaneously flashed into a plasma. The electron gun is pulsed, synchronized with the inductive or capacitive heater in the crucible, so that their fields would not defocus or interfere with the tightly focused electron beam.
Due to the high em fields present in the crucible the plasma is shot out of the crucible at high speed, a magnetic mirror is used to block the plasma from flowing into the electron gun. A magnetic solenoid valve and a magnetic solenoid focusing coils, bent and focus and guide the plasma beam into a resonant plasma wave guide system. Basically a glorified neon tube. And again all of the magnetic solenoids are pulsed, so as not to interfere with the electron beam, and the em field of the crucible.
The operation of the particle fountain is as follows. Stock is brought to the em crucible, were it is heated instantly by the em oven. Than the inductive and capacitive oven is turned of, The electron beam is pulsed into the center of the oven, producing a cloud of plasma. As the electron beam is turned of the magnetic solenoid valve and solenoid focusing coils are turned on guiding the plasma into the wave guides. As soon as the stock plasma has cleared the particle fountain, the magnetic valve and mirror are turned of. The inductive and capacitive heating starts again. Ideally this cycle would be repeated, several million times per second. As the amount of plasma that is produced with each cycle is fairly limited, possibly a single pulse would produce a spherical plasma cloud that would measure several microns on a side. However in a mature, advanced technology, I see no reason why the particle fountain could not run in the megahertz range, or even in the gigahertz range. These tiny plasmids would quickly add up. Indeed to the human senses it would look like a solid beam.
THE PLASMA WAVE GUIDE.
From here on, guided by the magnetic valve and focusing lens, the plasma pulse, (it is useful and easier to just follow the footsteps of a single plasmid, rather than trying to keep track of the whole beam)
plasma pulse enters the plasma wave guide. The wave guide could possibly be nothing more that a glass tube with a spiral copper wire wrapped around the length of the glass tube. That way the plasma induces on electric field in the copper spiral that causes the plasma to self constrict. The em interactions are complex and the exact technology will need to be experimentally verified, trough trail and error.
From here on the plasma is dumped in to a linac again, and more focusing solenoids. Going through the linac the plasma is boosted to higher speeds and energies. Since the plasma is a mix of various elements at this stage it needs to be purified.
THE PLASMA PRISM
In comes the plasma prism. The plasma prism is nothing more then a very powerful, superconducting solenoid positioned on the z axis relative to earth and the plasma beam.
The plasma stock flows through the solenoid, and each element gets deflected by magnetic field in the prism according to the plasmas energies and weight. Heavy elements would be deflected less, while lighter stuff would be deflected more. So after passing through the prism the beam would be split into as many distinctive beams as there were elements present in the stock.
From then on the individual beams of plasma will be picked up by a wave guide pick up head and conveyed to a plasma storage system. Each element in the periodic table would have its own wave guide and a pick up head past the plasma prism. There would also be as many storage rings as there are elements in the periodic table. Or at least the storage rings for the plasma stock that an AD commonly works with. The plasma prism device functions similar to the electron spectroscope.
THE STORAGE RINGS
As the plasma leaves the prism, after being picked up by the pick up solenoid, it is guided to its own storage ring, basically a large tokamak like device, were the plasma circulates until needed. I like to call it the plasma capacitor. From there, the some what clean plasma, can be sent, and is provided a recycler pathway, more wave guides, where the plasma is simple dumped back into the linac stage before the plasma prism, to go through the process again, if needed many thousands of times per second, until it is completely pure. Before being stored indefinitely or sent to the plasma buffer. Ideally a single plasma storage ring would be able to store several tons of pure elemental stock plasma.
THE PLASMA BUFFER
This is a small magnetic plasma can, where the stock plasma is kept in preparedness to be used by the AD work head.
THE PLASMA REGISTER
The stock plasma is called up by the plasma register from the buffer, in nano-sized bits, where it is held ready to be deposited by the AD work head. The plasma register is controlled by a computer which loads the appropriate plasma and the amount, from the plasma buffer, for any given build.
A CAD file is uploaded to the AD control computer, the control computer reads the CAD file and creates the materials and the mass inventory. The inventory data is sent to the register, which in turn queues the plasma buffer to prepare the plasma and to stand by.
Lets image that the AD control computer determines from the CAD files that one hundred pounds of aluminum will be needed for a given build. It then sends a command to the plasma register, which then in turn queues the plasma buffer to prepare one hundred pounds of plasma and hold it there until further notice. This will be done for as many elements as the work piece is made out of. By the work piece, I mean that which is going to be built. Each element has its own plasma buffer and register.
Depending on the precision and accuracy needed in the building of the work piece, the register might perform precision conditioning of the plasma. For highly accurate and precise work, the register will provide pulses of plasma as fine as a single nanometer across. For simple filler material the register can produce large plasma drops in bulk. All in all the register should be able to create a billion bits of plasma per second of nanometer size. Indeed the register treats each plasma bit as a standard digital bit or more precisely, a pixel.
For example. A three-dimensional CAD file is sliced into a billion, two- dimensional pictures. Each pictures is converted into a high resolution bit map, where each bit is precisely assigned a position in a x y coordinates system. Each two-dimensional picture will most likely contain hundreds of terabytes of data. The bit, that is, each pixel carries information, but in this case not the color information, but mainly the following data. Element type, that is, for example, hydrogen, iron, carbon or it could be any element of the periodic table or as required by the build. The size of that bit, could range from nanometer/s and all the way up to micrometers on a side. The weight of that bit, would most likely be measured in the, femto and pico litters, this is needed for precise control of the work piece density and weight. The temperature of the bit and so far and on.
When the AD work head reads the sliced CAD files, it reads all this information, which has been more or less generated automatically by the control computer, as per the engineers, strength, precision, accuracy and quality specifications. While the control computer generates the necessary data, the plasma register actually generates the actual hard data on the fly at the request of the AD work head. So the register provides the real hard data bits that are needed to fill in the bit map, by reading the information in the two dimensional picture.
THE WORK HEAD
The work head reads the data files, the position of the pixel or where the bit of plasma is going, its type, weight and temperature and requests the bit from the plasma register, once the bit arrives from the register, the work head targets the work surface. The work head focuses its magnetic lenses so that the plasma bit lands at the precise x y coordinates. While the work surface moves up and down on the z axis, for precision control the work surface is mounted on piezoelectric actuators. These can fine tune the rough movement of the work table to nanometer precision.
The work surface is charged opposite that of plasma bits, so that the plasma bits stick. The work surface is also cryogenically cooled to remove the massive heat flux from the plasma.
Once a single layer is completed the work platform lowers by a predetermined amount and a new layer is build. This process is repeated over and over again, in a straight forward serial manner. Until the work peace is done. The work head works like on electron microscope, but instead fires nanosized particles of plasma. Electron microscopes have been build with subnano resolution, so it is conceivably that the AD head will be capable of nanometer precision.
The AD head communicates with all of the registers, all registers cycle one step ahead of the AD, so that the AD can work smoothly and continually. The registers keep track of the work flow of the AD head and at the same time communicate with the control computer, to stay one step ahead of the AD.
All registers are also linked by the recycler waveguides to the plasma prism, so the register can purge itself, if it messes up. Registers can also request a pause if they fall behind due to error or malfunction, or for high priority builds, registers can automatically delegate their shares to other elements register if those registers are idle.
Also many dedicated registers would be build for creating plasma alloys, of varying compositions and ratios. So that during design engineers can specify virtually on unlimited amount of materials from witch the AD can build the part.
There are two ways for creating alloyed builds. A dedicated register that homogeneously premixes the plasma. And the less homogeneous way of mixing materials during the deposition process. The exact working of the registers, Ill keep close to my heart, for now!
Indeed the material possibilities are endless.
WHAT IT MEANS
A square micrometer contains one million hard nano bits. If the deposition rate is one gigahertz, then one square micrometer will be filled in by the AD in a thousandth of a second. A square millimeter per second, at one nanometer resolution! Considering the fact that hardly anything requires such high precision in everyday life, this resolution would only be used for specialty items.
More closer to home, a square millimeter, has one million micrometers, that would be filled in by the AD in one thousandth of a second. Or the AD would fill in a thousand square millimeters per second! That is one hundred centimeters square. Large for almost all house hold goods. But wait, this is just a single layer. There are one million layers in one hundred centimeters. Therefore it would take the AD one million seconds to build a solid cube one hundred centimeters in all dimensions. That would take roughly two hundred and seventy seven hours! Or eleven days!
However this is just a single AD head working! The AD heads are microscopic. A single AD head plate will contain millions of these work heads all working in parallel! Divide a million by a million and you get one, one second that is! The AD device will be able to manufacture virtually everything instantly. Its actual speed will be limited by how much heat can the work surface dissipate. And with cryogenic coolers, Id say a lot.
NOTHING IS SOLID
Consider for a moment the fact that hardly anything is made out of a solid block of building materials. But rather everything is made out of engineered thin walled metals, such as sheet metal. As in a body of a car. Or household items are made out of, mostly, thin plastics. Such objects would take even less amount of time to build!
Even faster build times can be attained by using less dense but structurally stronger filler material. Such as three dimensional isogrid structures. Three dimensional honeycomb structures, even all kinds of three dimensional geometric fractal shapes can be obtained and build right into the part. While using less material, being faster to build and lighter, the part would still offer strength advantages.
For example, a car engine block would easily lose a significant amount of its weight, if the internal solid metal was replaced by a three dimensional nanometer sized fractal tetrahedron grid. Similarly almost all man made structure would benefit from this technology.
THE BULK DEPOSITION MODE
Also the internal structure that require no precision, could be filled by a different deposition mode. The bulk deposition mode. Where the stock plasma is deposited amorphously without accurate targeting, as long as the bulk deposition mode stays inside the precision boundaries of the part, the build time will be drastically improved. This will also will slash computer use. The bulk deposition mode is like a garden spray hose. It will be used to quickly rough out the structurally none critical parts of the build. Then switch to the high precision mode to finish the more demanding areas of the part. Different mode of precision and bulk will be mixed and matched to get the desired part. Indeed there well be many intermediate modes (thousands, most likely) between the bulk and the precise modes.
No doubt in a short amount of time, the deposition of the stock material itself will evolve into a complex science itself. It is easily foreseeable that thousands of deposition patters will be utilized. Spanning from nanoscopic to microscopic. Being either simple two dimensional or three dimensional patterns. Simple repeating ones, or highly complex fractal ones, all of these will be possibly. Given that the patterns and three dimensional structures can be changed on the fly, and at the same time changing the stock element, it will be possibly to obtain of unheard of performance from the structural parts that are build this way. All parts most like will end up being a mixture of thousands of different patterns, hundreds of different structural filler, and hideously complex hyper alloys.
These hyper alloys will be a mixture of many different alloys, trough the entire part. And the whole different areas of the part will be made from different alloy. The parts need not be homogeneous. No doubt, this AD technology would revolutionize even our basic assumptions about design thought.
STRUCTURES AND PATTERS DATABASE
As time goes by, a massive database of structural fill modes and patters will be obtained. Through experimentation and trial and error, the best ones will be found, cataloged, numbered and their structural load properties, thermal, electrical and all of the materials properties, will be described in the database. It is even possibly, even desirably, that a materials testing consortium be set up, for the evaluation and verification of the claimed properties of stock.
From then on a search engine will be made available to all engineers, where during the design phase, the mechanical engineer would input the desired properties for a stock, that is, strength, stiffness, resistance to heat or acid, or any parameter, into the search engine. The search engine would then find the materials that best match the desired criteria.Than all the engineer has to do is specify what is what in the CAD file, before uploading the files to the AD.
Naturally some combination's of plasmas would be forbidden from being utilized in the AD. Due to the self obvious reason, mainly that, that not all reactions are safe, or useful. These reactions might end up being destructive. This will be the job of a chemist to exclude various dangerous chemicals from being made. The exclusion list might have millions of items on it!
Obviously not all of the elements of the periodic table will be used, such as the radioactive ones, or the environment polluting ones and the generally dangerous ones. The chemist and or a materials scientists will utilize the AD itself to test and evaluate generated new material samples. The testing itself will be done on nanosized samples to prevent large scale destructive events if two elements that are explosive are brought together. Once the parameters of the sample are found they are uploaded to the database, ready for use.
THE INTEGRATED TESTING CIRCUIT
This we be an IC like device that the AD builds. The AD, during manufacture of the ITC, will have built in all and every conceivably test that the material scientist can conceive of. The sample will be included in the construction of the ITC. The ITC will be very small, measuring millimeters on the side, and many millions would be build to evaluate a large number of sample in one go. Basically it will be a lab on a chip, integrated with MEM device for structural testing.We would be able to TUNE the work material to get the desired property. By producing a million microscopic variations of a material and looking for the one that matches our desired properties we will be able to evolve super materials. Take solder for example, made of tin and lead, both have a higher melting point than solder, but, when mixed together, they melt at a much lower temperature! So these are only two different materials, if one were to take in to account all the elements in the periodic table and all the different mix proportions that are possibly, the amount of materials that one can make are in the trillions!
The AD would construct a integrated testing circuit, it would come complete with on array of tests and sensors that one could image! Inside the ITC will be the material that will undergo the destructive test , it was included there when the ITC was build. The ITC itself is tiny, comparable to a CPU. The ITC is made from a million cells. Each individual cell would test the material for a specific parameter. The information from each test would go to on outside supercomputer which controls the device. The supercomputer would find the two samples that more closely match the programmed parameters and mate them. The mating result would be another million samples inside a new ITC. And so the cycle would repeat until the parameters were matched. In a mature technology the cycle ideally would last seconds and be massively parallel, that is, there would be a million ITC's constructed in each cycle.
Now imagine finding out you can generate new smart designer materials whenever needed. This would be simple. The AD, using the ITC, builds a million of arbitrary samples. Than tests each and one of the samples, until the sample that the most closely matches you desired parameters is found. From that sample another million of samples is build in their respective ITC. And the cycle is repeated until a one hundred percent match is found. There may be millions of such cycle until a match is found. In this way any conceivable material with outlandish properties can be evolved!
SELF IMPROVING TECHNOLOGY.
So lets say we spent sixty billion dollars developing a primitive AD, sixty billion, I think is sufficient. Will it be worth it? Imagine now that the primitive AD can now make parts that will improve itself. It will be as simple as uploading new CAD files and presto we have a new part. How imagine that that part only marginally improves the AD, say less that one percent. Now imagine that a year later the AD has manufactured thousands parts to improve itself. All improvements were very tiny baby steps, but by the end of the year, what would the end result would be? Massive improvement! At the end of the year the device would be hundreds of times better than what it was at the start of that year. So much so that we would quickly ran out of ideas on how to improve the machine!
At this point, we ask, now what? Well, cheap and fast duplication of this machine, for everyone! Within several years the world would be completely transformed. Even if the original machine cost sixty billion dollars, the copy/s would be dirt cheap, figuratively and literally!
Now each and every Joe Shmo has the ability to build, test and develop every outlandish idea that there can be. What sort of technologies will we see?! One cant even imagine.
Imagine, sitting at your computer drafting, once done, several minutes later you have you AD make the actual part. If it does not work, simple throw it back into the hopper and it will shortly be recycled. And back to the drawing board.
If this process used to cost hundreds of millions of dollars, and took years, now it will cost zero, and will take only the time you spent behind the computer with CAD.