Prospective directions for the development of 3D printing in the framework of home technologies and an open community. Guide to action
Difficulties and problems of development of 3D printers available in the house. Possible ways of improvement are considered and topics for development in the field of 3D printing for printers printing with molten plastic, as well as for printers printing with a sinterable laser powder, for printers printing with metal wire and for composite and "baked products" are proposed.
The article expresses my opinions, which may not coincide with the Truth or your opinion, the proposed improvements may have already been implemented somewhere or even patented - I did not specifically look for it. I (we) reflected on this topic, based on my erudition, experience and information gathered from the Internet.
There are predictions and hopes that over time, the growth and development of the industry will reach such an extent that everything necessary for life will become available to everyone, regardless of what he does. (well described by Sergey Lukyanenko. The road to Vellesberg) That is, people will not have to work for food. An important part of this is the production of 3D printers, including Replicators - RepRaps, which, as it were, can produce themselves.
')
Picture 1. Horn of plenty - perfect 3d printer?
So, let's see within the framework of the heading of the article, what ideas for the development of technology, me and my young friend - let's designate it - Physicist, and he is cool to the extreme, came to mind during a series of brain storms, meditations and records a la Mendeleev meaning night jumping with recording). The obvious idea in the American spirit is that you can create services armed with powerful industrial printers and relevant consultants who use working as clock, mail and delivery service I did not consider. From my own experience I will say - when my machine is, I can do with it even what I did not hope for. A remote order requires honest experience and rarely happens without jambs. I often had to deal with the categorical: -This we can not. The problem is often that in this case there is a separation of the process between two specialists - one knows how to design the product and what is required of him and does not have exact knowledge of what technology he can use, and the second one knows only how to deal with the appropriate machine and does not want to think about how to change the manufactured product for greater manufacturability and suitability for manufacturing by this method. The result is an incomplete use of the possibilities by both.
What possibilities of improving existing printing technologies do I see (not without the help of Physics) Consider them:
1. Printers printing molten plastic bar.
photo 2. My printer is a rabbit for experiences
The main problems that we do not like about them:
1.1 - low resolution and printing accuracy, and the more accurate the print (the smaller the nozzle diameter), the slower the print speed;
1.2 - expensive consumables;
1.3 - strength and other properties of the materials used are not good enough for many applications;
1.4 - the complexity of using multiple colors / materials. This primarily increases the price of the printer;
1.5 - preparation for printing is sometimes not too simple, as it is not easy and to separate the part from the base.
In paragraph 1.1, resolution and accuracy are made up of several factors. Here and the roughness of the walls and the shrinkage of the plastic after printing and deformation during printing, which makes it difficult to withstand accurate dimensions, here and warping of large parts, and most importantly - the diameter of the nozzle determines the size of the minimum printed parts.
What can we do here in terms of improving printing and addressing identified shortcomings?
1.1 Print resolution and speed properties are opposite and mutually exclusive. What can you do ... But you can try. Imagine a head with a nozzle of the order of 0.1-0.15 mm, light, small, accurate and fast. Well, yes, because it is light and small - it means that the extruder is not large and not heavy (the flow rate is small). It prints us only a few layers of the base, perhaps with a raft, for better separation from the substrate, and then makes the side walls - as always happens when printing. We select the wall thickness - probably 3-5 layers will be required. This will correspond to a thickness of 0.3-0.5 mm. Moreover, we print walls to a height of, for example, 0.5–1 mm, after which the light and precise head moves back and the filling head with a diameter of 0.8–1.2 mm comes into effect. It fills the entire interior with the temperature calculated so that the filling sticks to the walls, but does not melt them. Then everything is repeated at the next level. The fabrication takes place in such rather thick layers, but with high detailing of the outer wall. I expect, based on the experience of my recent work, that such a small and lightweight head can be made. A lightweight extruder - or rather, its part with a motor, is not such a waste topic, but the question here is in the degree of lightness and in the power spent on development. High positioning accuracy on two axes will require the greatest effort, although the heads in DVD drives are positioned much more accurately. So you can do. Borrowing any of these drives. I am in terms of ideas.
It is clear that the Draft-head does not present any particular difficulties in the design, as well as its supply unit.
1.2 Consumables. Their price. There are hopes that the device representing the plastic crusher with its subsequent transformation into a filament will solve absolutely all problems.
Photo 3. Filament Wee makes a filament of granules
Link to this topic: http://3dtoday.ru/industry/domashnee-proizvodstvo-prutka-ili-ekonomika-dolzhna-byt-ekonomnoy.html
Here I have doubts and questions. The fact is that on the plastic in the crusher will fall dust particles, dirt, pieces of paint, chips. Not the fact that you can correctly determine the type of plastic. And if you have a head, at least 0.2 mm in diameter, it will be easy to score, and it is difficult to clean it. So my opinion: ABS granulated plastic costs many times less than filament. It will be quite enough if you save 90% of your expenses, without unnecessary risks, using industrial plastic granules to turn them into a filament in a home extruder. For a thin head in general, you can buy ready-made filament - to avoid and to improve, and homemade / conversion - to fill. Such a rough head is easy to clean.
1.3 The issue with the physical properties of materials used in FDM printers - these are the ones we are talking about here, limited to various types of thermoplastics. ABS and PLA are widely used with their advantages and disadvantages. They are easier to print. But there are still plastics that are quite applicable and have significantly different properties.
- For example - nylon, now it has become known as polyamide. Plastic is very durable, quite flexible, that is, it is difficult to break it. Very hard. Does not smell in the printing process. And if your printer does not burn with a blue flame, it doesn’t emit anything particularly harmful in dangerous concentrations, this was checked by rather tough experiments. Problems with printing polyamide, that large enough details are beginning to warp with great force. So it is necessary to apply, best of all, programmatically, the system of cuts, allowing to remove deformations. Need to experiment. I really did not use polyamide for printing, but simply a trimmer line - cut the grass. It is painfully cheap and always at hand. Here it jarred at large sizes. Anyway, it would be nice to understand how to make parts without warping and maybe, to introduce this mathematics into Slicer.
- I also like polypropylene. He has both positive and negative properties. It is chemically very resistant, not harmful. To a large extent. Its strength is less than polyamide and hardness is also not too. But he does not break. It is very cheap and in fact - always at hand. I had and have an idea to make a machine for turning a polypropylene rope into a fishing line - a filament for a printer. Get a fishing line with a large number of small air bubbles, quite flexible. And the price of such a rope is remarkably low! So here is the reference to paragraph 1.2. If you make the walls of the product from polyamide, then you can quite fill the polypropylene, I think. True, not in all cases. Well, it's according to needs. Molten polypropylene has one uncomfortable property - it slowly solidifies. But for layer-by-layer casting, this will be quite good, since the melting point of polypropylene is much lower than that of polyamide and ABS.
Photo 4. Nylon (polyamide), polypropylene rope, ABS
1.4 Regarding the change of colors or materials from ideas, only make an automated or semi-automated replacement of the working material in the head. This will allow you to do not 2-3-4 colors, but more. It seems very difficult mechanically and programmatically, but the idea is worth it. Remember the automatic machines that change the instrument themselves - beauty!
- This can be realized as a change of material with taking it out of the head and inserting a new one (as earlier in the automata the records were changed), or
- a somewhat lower quality, but incomparably simpler method with changing the feed of filament streams already inserted into the head. Those. they enter the melting zone of the head, let's say both, but one is served and the second is waiting for its turn. Then the first one stops, the second one goes. There will be some confusion, it is necessary to take into account the volume of the molten plastic already and to spend it, but this is not important, it can be solved by software.
Picture 5. Scheme of alternate feed with one head. Red - heater, green and yellow - two types of filament.
1.5 Of the simplifications, I consider it necessary:
- Make a simple and reliable system of positioning the head at a height above the table, and of course ensure its horizontal position. For this there can be many ways, I will not detail here.
“This can also include a control and adjustment unit for filament feed force. At the same time, the head remains fixed on the guide axes X and Y, and the filament feeding unit is fastened through a spring element of known stiffness. According to the degree of lifting of the extruder assembly, it is possible to determine the filing feed force. This will improve the reproducibility of the results, as well as reproducibly evaluate the properties of plastics (polymers ... may differ in the degree of polymerization, and therefore in properties).
- It seems to me more interesting, instead of a belt drive, to move the head in the printer, try using an arrow switch - similar to that used in hard drives for positioning the head. Structurally, it will look like a bracket fixed on the axis of the stepper motor. It is clear that when rotating it will describe an arc. Alas, this is only for one maxi - two. Do not worry - you can print on a concave surface. Mathematically it is clear that with layer-by-layer filling on the inner surface of a cylinder or sphere it will not be difficult to print a product of any shape. Only a few more complicated calculations. The question here is whether you need such a high speed of moving the head, which this positioning method will allow, or the extrusion speed for light heads of small diameter will not be able to be raised too much.
Picture 6. Attempt to show how to print a plane on a concave cylinder. Different height of the printing unit is highlighted in color.
- In a closer perspective, we see the manufacture of light and small heads with new heating elements and unconventional for 3D, temperature measurement methods, and to facilitate the replacement of standard ones, you can make an emulation board for the inputs / outputs of standard heads, with a number of necessary protections and adjustments. The scarf will be small. I can't get more details yet - here the topic overlaps with the topic of my work on developing a small LixPen handheld printer. I hope that in the near future I will receive permission to publish unused finds made during these works.
2. Stereolithographic printers.
These are printers where the photosensitive resin is cured by a laser or projection of light through a computer LCD projector. While I do not feel much enthusiasm from them. Rather, I do not like them. High price of resin for them. The polymer itself is not too strong. He is also, most likely, quite poisonous and not very stable. Nothing has been thought out yet, and I do not want to.
3. Printers that print by selectively sintering the powder.
Photo 7. An example of the operation of the printer project OpenSLS
These printers work mainly in this way: - A thin layer of polymer powder (or even metal) is applied, then a laser beam (there are options where this is done with an electron beam or micro burner) powder is selectively sintered where it should be a solid material. Then the next layer is applied on top, and again is sintered. The green powder is separated and usually returns to production.
An obvious advantage is that the laser beam (as well as the electron beam) can be deflected at almost any speed. The speed determining factor here is the laser power. So far, the main reason for the small spread of such devices, in my opinion, is that high-power lasers are expensive, gas CO2, infrared are usually used. This is a hefty product, fragile, with high voltage applied. They do not twist, therefore, apply the deflected mirrors.
In the transition to solid-state / semiconductor lasers, there are certain problems with the wavelength emitted by gas lasers - it is different and used to it even at the level of materials. Powder poured into the working chamber is pre-heated to a temperature close to the beginning of sintering. This technique is able to reduce the laser beam energy required for full sintering in the working spot. At the same time, I suspect that the fact that at elevated temperature a significant part of the gases sorbed by it is desorbed has some significance. This may seem insignificant, but the greater the specific surface area of the powder (i.e., the smaller the particles), the more it sits on the atmosphere and the harder it is to remove. So, in the case of sintering metal powders, the use of an inert atmosphere (argon) or vacuuming is likely.
Some of you will ask a question - what if the laser is replaced with a powerful light source? I thought it through, and even picked up suitable light sources - sodium lamps. They have a very high efficiency. At the same time, they emit light in a fairly narrow range. Alas, after the explanation of Physics, I found it not promising. That is, it will be more expensive than with a laser. The fact is that the lamp emits light in all directions and collect it at one point, which is what we need, it turns out with great difficulty and loss - unlike a laser. So say the harsh laws of optics. If anyone needs more details - I nag from my correspondence with the Physicist, his explanations - there it is considered that explained at a very simple level.
The accuracy and resolution of such printers is close to the desired perfection.
The materials used for sintering are also remarkably good. This polyamide (nylon), titanium, stainless steel. It is possible to imagine the sintering of copper and aluminum powder. The disadvantages of these printers are the high price, rather expensive powders.
So what measures to improve these devices can be applied to bring them closer to our homes and our practice?
The first and obvious: for sintering use semiconductor lasers. This will seem to be much cheaper and will significantly simplify the beam deflection system. However, when studying the issue in more detail, I began to get the impression that the price of semiconductor lasers above a certain power grows too fast and that 40W CO2 lasers are already much cheaper than semiconductor lasers. We have several possible ways to control the development of lasers. Two are more interesting: The first is to move the entire head along two axes at a strictly fixed distance, just like in the usual FDM printers. The second is to place the laser well above the sintering level and control its movement by turning it at a given angle. It certainly must be a very fast way, only there are a number of subtleties. So, the laser beam can be collimated or focused. What's the Difference? The collimated laser beam is almost not expanding. This is probably the secret of love for IR CO2 lasers — a collimated beam is probably used there and can be controlled by deflecting it with mirrors. The density of the luminous flux is sufficient to quickly cut different things. Even metal. In our availability - powerful laser pointers. There, the beam is also collimated.Only the pointer is not the hyperboloid of engineer Garin, even with the most powerful, the energy density is such that it sluggishly melts plastic, sets fire to matches, cuts black paper. The speed of movement of the cochlea when the printer is working and the sintering spot diameter of 2 mm will not suit us. Therefore, we have to focus the beam. It is easy and power density will immediately suffice for many rather serious cases. Only here the focus will strictly depend on the distance. You can certainly imagine a system in which the focus changes depending on the angle of rotation of the head in order to compensate for the different distances to the working layer. But oh, this task will be difficult! The conclusion is that with laser power available to us (from DVDs to 1 watt), the real solution will be the direct movement of the laser head along two axes as in an FDM printer. Alas!We are waiting for cheap and powerful lasers.
So, what can we imagine on the road to improving this technology?
-3.1The simplest and most obvious is to use part of the DVD-RW drive mechanics along with the laser. - It can reach a power of 1 W, there is a controlled focusing lens, while its positioning along the same axis will be with micron precision. There may be objections that such accuracy relies on reading the disc track markup. Fine, we take two drives - in one disk with a marking, the second stupidly copies all movements. It will remain to achieve the same positioning accuracy along the second axis using DVD drive elements and this is already good. Well, the Z axis, of course. Of course, the working surface area will be about 5x5 cm, but even with a resolution and accuracy of 0.05-0.02 mm, this will be quite a significant step forward. Power should be enough for sintering plastics.
-3.2If we get low-cost lasers with enough power for sintering metals, I think we will have to take care of the protection from oxidation of the powder with oxygen and not only air. As an option, the use of the old amateur technology for producing small amounts of hydrogen / explosive gas by electrolysis of an aqueous solution of something like sodium sulfate. With the subsequent drying of the gas with a mixture of means for cleaning sewage and burnt lime, which is known under the name “soda lime”
4. Printers printing molten metal.
Here I am not talking about direct sintering of a powder or an electron gun by a laser. If you used a welding machine, you know that in this way, with due diligence, you can weld a protrusion of the desired shape using an electrode like a filament. It is rather difficult to adapt this process for 3D printing, although there are reports of printers and printed products built on this principle. Products turn out clumsy. The problem, in my opinion, is that the arc is a high-temperature thing, splashing occurs, bubbles form, the metal flows, its viscosity is small, the temperature gradient is high.
Photo 8. Here is a gear made on an arc open printer community in Michigan. But from the metal.
I also draw your attention to the fact that metal, unlike plastic, behaves much differently. Its viscosity is much lower, and its surface tension is much higher than that of molten plastic. Thus, the molten metal strives to gather in droplets. (for tin it is very characteristic) Another disadvantage is that the metal usually interacts with the atmosphere, which means a protective environment is required. To obtain it, you can attract the idea from the previous point - a hydrogen generator, or use the flame of a butane burner to create a zone free from oxygen. But how then did we manage to weld the protrusions on the metal with a welding machine? - It was a fusion. The arc melted the substrate and the material of the electrode and the metal gradually grew solidified by the cold of the substrate. This fact,by the way, it reduces interest in systems based on FDM printers with the idea of fusing molten tin tracks connecting the inserted components. Fusing on a small droplet will not only give an uneven and thick track, but is also doubtful from the point of view of a good electrical as well as mechanical contact. - Simply put, the stability and reliability of such a connection will not be very.
What improvements can we expect to see in this field?
-4.1 I was here looking at the store for welding wire for automatic welding machines. Cheap. Durable. It is also covered with copper. I think that it is worth trying to use for laying this wire, not arc heating, but resistive. However, as the Physicist suggests, simple softening is not enough to create a good contact. So you have to combine resistance heating and stacking with spot welding for the connection between the layers and in the places of the head turns. Cutting the wire can be a sharp increase in the resistive current to complete melting with the rise of the head, but without wire feeding.
The difference in my understanding between the resistive current and the welding current is that resistive is a very large current (hundreds of amperes, up to thousands) with low voltage, so that the current flowing through a steel wire with a diameter of 0.2-0.3 mm heats it until softened, but the voltage is so small that at a sufficiently small length, so that there is no electric arc. The arc current is a current with a voltage of 20-50 volts (select, of course), creating an arc that melts and welds different layers between them. The algorithm will be as follows:- the head is not high, 2 millimeters above the grounded steel substrate (it will be the base of the part-Alas). The tip of the welding wire extends with applied welding voltage. Down - up, before the arc. After a certain time, the wire drops sharply into the hole of the molten metal, the welding voltage is turned off. When this occurs, a stable electrical contact. We have 2 mm wire from the base to the head. We supply resistive current to the head, 2 mm of the wire softens, we move the head with the retractable wire and the resistive current, for example, 5 mm, again we weld to the substrate, and so on. If you need to cut the wire - we increase the resistive current by several times and sharply pull the head up. The head at the same time copper, cooled.
- 4.2Another option is a ceramic heat-resistant head operating in the arc mode and riveting molten metal in a droplet. This option will give a denser and more durable products, only there are quite great difficulties with the manufacture of such a heat-resistant ceramic head, and even tolerant to shocks in a heated state.
5. Printers "baked products."
Well, here are some crazy ideas:
- 5.1Sintering powder printer. We have two nozzles in the head feeding the powder. One of the powders is powder that is sintered into metal, the other is ballast, for example, alumina or magnesium oxide. Pouring is done in thin layers, the ballast powder is limited to spreading the worker. Then perhaps rolling a layer to seal or not - you need to check. The whole working volume is filled up in layers, with the figure of the future product enclosed inside. Then the working chamber is placed in the stove and is sintered. From aluminum oxide (it is more free-flowing and conducts heat better), we remove the finished, sintered product by cooling. Accuracy of printing will not be very. Yes, and the stove is needed.
-5.2And here's another: - sand castles method. In the working chamber there is a substrate made of porous ceramics, air is pumped out from below. The head is applied to the tracks of a fairly fluid mixture of water and metal powder, stabilized with something like modified starch, gelatin or guar gum. Like the usual FDM-printers. Due to pumping from the bottom of the substrate, the moisture from the mixture is sucked off and it becomes relatively solid. In childhood, sand castles were made like this. In this case, some adhesive component is needed, so that when the drying process does not crumble, a modified starch, gelatin or guar gum should suffice. Again in the oven and bake.
6. Composite printers.
We are talking about printers that use to build the parts are not homogeneous material, and the material with different inclusions, such as wire, electronic components and what else is invented. This is most likely related to FDM-type printers with an additional manipulator / head that can arrange electronic components, lay wires to connect them or to harden parts. Connections of electrical wires with parts can be made by melting spark / micro-arc solder wire or contact welding. To increase the longitudinal strength, you can provide holes for steel inserts, which will be enough to melt from two ends with a soldering iron, or grooves for hardening by winding.
At this, I stop the allowed speeches and I hope that, at least, it was interesting, and maybe inspiring to work in the outlined areas. If you have any thoughts, suggestions and ideas - please contact us. Thank.
The article expresses my opinions, which may not coincide with the Truth or your opinion, the proposed improvements may have already been implemented somewhere or even patented - I did not specifically look for it. I (we) reflected on this topic, based on my erudition, experience and information gathered from the Internet.
There are predictions and hopes that over time, the growth and development of the industry will reach such an extent that everything necessary for life will become available to everyone, regardless of what he does. (well described by Sergey Lukyanenko. The road to Vellesberg) That is, people will not have to work for food. An important part of this is the production of 3D printers, including Replicators - RepRaps, which, as it were, can produce themselves.
')
Picture 1. Horn of plenty - perfect 3d printer?
So, let's see within the framework of the heading of the article, what ideas for the development of technology, me and my young friend - let's designate it - Physicist, and he is cool to the extreme, came to mind during a series of brain storms, meditations and records a la Mendeleev meaning night jumping with recording). The obvious idea in the American spirit is that you can create services armed with powerful industrial printers and relevant consultants who use working as clock, mail and delivery service I did not consider. From my own experience I will say - when my machine is, I can do with it even what I did not hope for. A remote order requires honest experience and rarely happens without jambs. I often had to deal with the categorical: -This we can not. The problem is often that in this case there is a separation of the process between two specialists - one knows how to design the product and what is required of him and does not have exact knowledge of what technology he can use, and the second one knows only how to deal with the appropriate machine and does not want to think about how to change the manufactured product for greater manufacturability and suitability for manufacturing by this method. The result is an incomplete use of the possibilities by both.
What possibilities of improving existing printing technologies do I see (not without the help of Physics) Consider them:
1. Printers printing molten plastic bar.
photo 2. My printer is a rabbit for experiences
The main problems that we do not like about them:
1.1 - low resolution and printing accuracy, and the more accurate the print (the smaller the nozzle diameter), the slower the print speed;
1.2 - expensive consumables;
1.3 - strength and other properties of the materials used are not good enough for many applications;
1.4 - the complexity of using multiple colors / materials. This primarily increases the price of the printer;
1.5 - preparation for printing is sometimes not too simple, as it is not easy and to separate the part from the base.
In paragraph 1.1, resolution and accuracy are made up of several factors. Here and the roughness of the walls and the shrinkage of the plastic after printing and deformation during printing, which makes it difficult to withstand accurate dimensions, here and warping of large parts, and most importantly - the diameter of the nozzle determines the size of the minimum printed parts.
What can we do here in terms of improving printing and addressing identified shortcomings?
1.1 Print resolution and speed properties are opposite and mutually exclusive. What can you do ... But you can try. Imagine a head with a nozzle of the order of 0.1-0.15 mm, light, small, accurate and fast. Well, yes, because it is light and small - it means that the extruder is not large and not heavy (the flow rate is small). It prints us only a few layers of the base, perhaps with a raft, for better separation from the substrate, and then makes the side walls - as always happens when printing. We select the wall thickness - probably 3-5 layers will be required. This will correspond to a thickness of 0.3-0.5 mm. Moreover, we print walls to a height of, for example, 0.5–1 mm, after which the light and precise head moves back and the filling head with a diameter of 0.8–1.2 mm comes into effect. It fills the entire interior with the temperature calculated so that the filling sticks to the walls, but does not melt them. Then everything is repeated at the next level. The fabrication takes place in such rather thick layers, but with high detailing of the outer wall. I expect, based on the experience of my recent work, that such a small and lightweight head can be made. A lightweight extruder - or rather, its part with a motor, is not such a waste topic, but the question here is in the degree of lightness and in the power spent on development. High positioning accuracy on two axes will require the greatest effort, although the heads in DVD drives are positioned much more accurately. So you can do. Borrowing any of these drives. I am in terms of ideas.
It is clear that the Draft-head does not present any particular difficulties in the design, as well as its supply unit.
1.2 Consumables. Their price. There are hopes that the device representing the plastic crusher with its subsequent transformation into a filament will solve absolutely all problems.
Photo 3. Filament Wee makes a filament of granules
Link to this topic: http://3dtoday.ru/industry/domashnee-proizvodstvo-prutka-ili-ekonomika-dolzhna-byt-ekonomnoy.html
Here I have doubts and questions. The fact is that on the plastic in the crusher will fall dust particles, dirt, pieces of paint, chips. Not the fact that you can correctly determine the type of plastic. And if you have a head, at least 0.2 mm in diameter, it will be easy to score, and it is difficult to clean it. So my opinion: ABS granulated plastic costs many times less than filament. It will be quite enough if you save 90% of your expenses, without unnecessary risks, using industrial plastic granules to turn them into a filament in a home extruder. For a thin head in general, you can buy ready-made filament - to avoid and to improve, and homemade / conversion - to fill. Such a rough head is easy to clean.
1.3 The issue with the physical properties of materials used in FDM printers - these are the ones we are talking about here, limited to various types of thermoplastics. ABS and PLA are widely used with their advantages and disadvantages. They are easier to print. But there are still plastics that are quite applicable and have significantly different properties.
- For example - nylon, now it has become known as polyamide. Plastic is very durable, quite flexible, that is, it is difficult to break it. Very hard. Does not smell in the printing process. And if your printer does not burn with a blue flame, it doesn’t emit anything particularly harmful in dangerous concentrations, this was checked by rather tough experiments. Problems with printing polyamide, that large enough details are beginning to warp with great force. So it is necessary to apply, best of all, programmatically, the system of cuts, allowing to remove deformations. Need to experiment. I really did not use polyamide for printing, but simply a trimmer line - cut the grass. It is painfully cheap and always at hand. Here it jarred at large sizes. Anyway, it would be nice to understand how to make parts without warping and maybe, to introduce this mathematics into Slicer.
- I also like polypropylene. He has both positive and negative properties. It is chemically very resistant, not harmful. To a large extent. Its strength is less than polyamide and hardness is also not too. But he does not break. It is very cheap and in fact - always at hand. I had and have an idea to make a machine for turning a polypropylene rope into a fishing line - a filament for a printer. Get a fishing line with a large number of small air bubbles, quite flexible. And the price of such a rope is remarkably low! So here is the reference to paragraph 1.2. If you make the walls of the product from polyamide, then you can quite fill the polypropylene, I think. True, not in all cases. Well, it's according to needs. Molten polypropylene has one uncomfortable property - it slowly solidifies. But for layer-by-layer casting, this will be quite good, since the melting point of polypropylene is much lower than that of polyamide and ABS.
Photo 4. Nylon (polyamide), polypropylene rope, ABS
1.4 Regarding the change of colors or materials from ideas, only make an automated or semi-automated replacement of the working material in the head. This will allow you to do not 2-3-4 colors, but more. It seems very difficult mechanically and programmatically, but the idea is worth it. Remember the automatic machines that change the instrument themselves - beauty!
- This can be realized as a change of material with taking it out of the head and inserting a new one (as earlier in the automata the records were changed), or
- a somewhat lower quality, but incomparably simpler method with changing the feed of filament streams already inserted into the head. Those. they enter the melting zone of the head, let's say both, but one is served and the second is waiting for its turn. Then the first one stops, the second one goes. There will be some confusion, it is necessary to take into account the volume of the molten plastic already and to spend it, but this is not important, it can be solved by software.
Picture 5. Scheme of alternate feed with one head. Red - heater, green and yellow - two types of filament.
1.5 Of the simplifications, I consider it necessary:
- Make a simple and reliable system of positioning the head at a height above the table, and of course ensure its horizontal position. For this there can be many ways, I will not detail here.
“This can also include a control and adjustment unit for filament feed force. At the same time, the head remains fixed on the guide axes X and Y, and the filament feeding unit is fastened through a spring element of known stiffness. According to the degree of lifting of the extruder assembly, it is possible to determine the filing feed force. This will improve the reproducibility of the results, as well as reproducibly evaluate the properties of plastics (polymers ... may differ in the degree of polymerization, and therefore in properties).
- It seems to me more interesting, instead of a belt drive, to move the head in the printer, try using an arrow switch - similar to that used in hard drives for positioning the head. Structurally, it will look like a bracket fixed on the axis of the stepper motor. It is clear that when rotating it will describe an arc. Alas, this is only for one maxi - two. Do not worry - you can print on a concave surface. Mathematically it is clear that with layer-by-layer filling on the inner surface of a cylinder or sphere it will not be difficult to print a product of any shape. Only a few more complicated calculations. The question here is whether you need such a high speed of moving the head, which this positioning method will allow, or the extrusion speed for light heads of small diameter will not be able to be raised too much.
Picture 6. Attempt to show how to print a plane on a concave cylinder. Different height of the printing unit is highlighted in color.
- In a closer perspective, we see the manufacture of light and small heads with new heating elements and unconventional for 3D, temperature measurement methods, and to facilitate the replacement of standard ones, you can make an emulation board for the inputs / outputs of standard heads, with a number of necessary protections and adjustments. The scarf will be small. I can't get more details yet - here the topic overlaps with the topic of my work on developing a small LixPen handheld printer. I hope that in the near future I will receive permission to publish unused finds made during these works.
2. Stereolithographic printers.
These are printers where the photosensitive resin is cured by a laser or projection of light through a computer LCD projector. While I do not feel much enthusiasm from them. Rather, I do not like them. High price of resin for them. The polymer itself is not too strong. He is also, most likely, quite poisonous and not very stable. Nothing has been thought out yet, and I do not want to.
3. Printers that print by selectively sintering the powder.
Photo 7. An example of the operation of the printer project OpenSLS
These printers work mainly in this way: - A thin layer of polymer powder (or even metal) is applied, then a laser beam (there are options where this is done with an electron beam or micro burner) powder is selectively sintered where it should be a solid material. Then the next layer is applied on top, and again is sintered. The green powder is separated and usually returns to production.
An obvious advantage is that the laser beam (as well as the electron beam) can be deflected at almost any speed. The speed determining factor here is the laser power. So far, the main reason for the small spread of such devices, in my opinion, is that high-power lasers are expensive, gas CO2, infrared are usually used. This is a hefty product, fragile, with high voltage applied. They do not twist, therefore, apply the deflected mirrors.
In the transition to solid-state / semiconductor lasers, there are certain problems with the wavelength emitted by gas lasers - it is different and used to it even at the level of materials. Powder poured into the working chamber is pre-heated to a temperature close to the beginning of sintering. This technique is able to reduce the laser beam energy required for full sintering in the working spot. At the same time, I suspect that the fact that at elevated temperature a significant part of the gases sorbed by it is desorbed has some significance. This may seem insignificant, but the greater the specific surface area of the powder (i.e., the smaller the particles), the more it sits on the atmosphere and the harder it is to remove. So, in the case of sintering metal powders, the use of an inert atmosphere (argon) or vacuuming is likely.
Some of you will ask a question - what if the laser is replaced with a powerful light source? I thought it through, and even picked up suitable light sources - sodium lamps. They have a very high efficiency. At the same time, they emit light in a fairly narrow range. Alas, after the explanation of Physics, I found it not promising. That is, it will be more expensive than with a laser. The fact is that the lamp emits light in all directions and collect it at one point, which is what we need, it turns out with great difficulty and loss - unlike a laser. So say the harsh laws of optics. If anyone needs more details - I nag from my correspondence with the Physicist, his explanations - there it is considered that explained at a very simple level.
The accuracy and resolution of such printers is close to the desired perfection.
The materials used for sintering are also remarkably good. This polyamide (nylon), titanium, stainless steel. It is possible to imagine the sintering of copper and aluminum powder. The disadvantages of these printers are the high price, rather expensive powders.
So what measures to improve these devices can be applied to bring them closer to our homes and our practice?
The first and obvious: for sintering use semiconductor lasers. This will seem to be much cheaper and will significantly simplify the beam deflection system. However, when studying the issue in more detail, I began to get the impression that the price of semiconductor lasers above a certain power grows too fast and that 40W CO2 lasers are already much cheaper than semiconductor lasers. We have several possible ways to control the development of lasers. Two are more interesting: The first is to move the entire head along two axes at a strictly fixed distance, just like in the usual FDM printers. The second is to place the laser well above the sintering level and control its movement by turning it at a given angle. It certainly must be a very fast way, only there are a number of subtleties. So, the laser beam can be collimated or focused. What's the Difference? The collimated laser beam is almost not expanding. This is probably the secret of love for IR CO2 lasers — a collimated beam is probably used there and can be controlled by deflecting it with mirrors. The density of the luminous flux is sufficient to quickly cut different things. Even metal. In our availability - powerful laser pointers. There, the beam is also collimated.Only the pointer is not the hyperboloid of engineer Garin, even with the most powerful, the energy density is such that it sluggishly melts plastic, sets fire to matches, cuts black paper. The speed of movement of the cochlea when the printer is working and the sintering spot diameter of 2 mm will not suit us. Therefore, we have to focus the beam. It is easy and power density will immediately suffice for many rather serious cases. Only here the focus will strictly depend on the distance. You can certainly imagine a system in which the focus changes depending on the angle of rotation of the head in order to compensate for the different distances to the working layer. But oh, this task will be difficult! The conclusion is that with laser power available to us (from DVDs to 1 watt), the real solution will be the direct movement of the laser head along two axes as in an FDM printer. Alas!We are waiting for cheap and powerful lasers.
So, what can we imagine on the road to improving this technology?
-3.1The simplest and most obvious is to use part of the DVD-RW drive mechanics along with the laser. - It can reach a power of 1 W, there is a controlled focusing lens, while its positioning along the same axis will be with micron precision. There may be objections that such accuracy relies on reading the disc track markup. Fine, we take two drives - in one disk with a marking, the second stupidly copies all movements. It will remain to achieve the same positioning accuracy along the second axis using DVD drive elements and this is already good. Well, the Z axis, of course. Of course, the working surface area will be about 5x5 cm, but even with a resolution and accuracy of 0.05-0.02 mm, this will be quite a significant step forward. Power should be enough for sintering plastics.
-3.2If we get low-cost lasers with enough power for sintering metals, I think we will have to take care of the protection from oxidation of the powder with oxygen and not only air. As an option, the use of the old amateur technology for producing small amounts of hydrogen / explosive gas by electrolysis of an aqueous solution of something like sodium sulfate. With the subsequent drying of the gas with a mixture of means for cleaning sewage and burnt lime, which is known under the name “soda lime”
4. Printers printing molten metal.
Here I am not talking about direct sintering of a powder or an electron gun by a laser. If you used a welding machine, you know that in this way, with due diligence, you can weld a protrusion of the desired shape using an electrode like a filament. It is rather difficult to adapt this process for 3D printing, although there are reports of printers and printed products built on this principle. Products turn out clumsy. The problem, in my opinion, is that the arc is a high-temperature thing, splashing occurs, bubbles form, the metal flows, its viscosity is small, the temperature gradient is high.
Photo 8. Here is a gear made on an arc open printer community in Michigan. But from the metal.
I also draw your attention to the fact that metal, unlike plastic, behaves much differently. Its viscosity is much lower, and its surface tension is much higher than that of molten plastic. Thus, the molten metal strives to gather in droplets. (for tin it is very characteristic) Another disadvantage is that the metal usually interacts with the atmosphere, which means a protective environment is required. To obtain it, you can attract the idea from the previous point - a hydrogen generator, or use the flame of a butane burner to create a zone free from oxygen. But how then did we manage to weld the protrusions on the metal with a welding machine? - It was a fusion. The arc melted the substrate and the material of the electrode and the metal gradually grew solidified by the cold of the substrate. This fact,by the way, it reduces interest in systems based on FDM printers with the idea of fusing molten tin tracks connecting the inserted components. Fusing on a small droplet will not only give an uneven and thick track, but is also doubtful from the point of view of a good electrical as well as mechanical contact. - Simply put, the stability and reliability of such a connection will not be very.
What improvements can we expect to see in this field?
-4.1 I was here looking at the store for welding wire for automatic welding machines. Cheap. Durable. It is also covered with copper. I think that it is worth trying to use for laying this wire, not arc heating, but resistive. However, as the Physicist suggests, simple softening is not enough to create a good contact. So you have to combine resistance heating and stacking with spot welding for the connection between the layers and in the places of the head turns. Cutting the wire can be a sharp increase in the resistive current to complete melting with the rise of the head, but without wire feeding.
The difference in my understanding between the resistive current and the welding current is that resistive is a very large current (hundreds of amperes, up to thousands) with low voltage, so that the current flowing through a steel wire with a diameter of 0.2-0.3 mm heats it until softened, but the voltage is so small that at a sufficiently small length, so that there is no electric arc. The arc current is a current with a voltage of 20-50 volts (select, of course), creating an arc that melts and welds different layers between them. The algorithm will be as follows:- the head is not high, 2 millimeters above the grounded steel substrate (it will be the base of the part-Alas). The tip of the welding wire extends with applied welding voltage. Down - up, before the arc. After a certain time, the wire drops sharply into the hole of the molten metal, the welding voltage is turned off. When this occurs, a stable electrical contact. We have 2 mm wire from the base to the head. We supply resistive current to the head, 2 mm of the wire softens, we move the head with the retractable wire and the resistive current, for example, 5 mm, again we weld to the substrate, and so on. If you need to cut the wire - we increase the resistive current by several times and sharply pull the head up. The head at the same time copper, cooled.
- 4.2Another option is a ceramic heat-resistant head operating in the arc mode and riveting molten metal in a droplet. This option will give a denser and more durable products, only there are quite great difficulties with the manufacture of such a heat-resistant ceramic head, and even tolerant to shocks in a heated state.
5. Printers "baked products."
Well, here are some crazy ideas:
- 5.1Sintering powder printer. We have two nozzles in the head feeding the powder. One of the powders is powder that is sintered into metal, the other is ballast, for example, alumina or magnesium oxide. Pouring is done in thin layers, the ballast powder is limited to spreading the worker. Then perhaps rolling a layer to seal or not - you need to check. The whole working volume is filled up in layers, with the figure of the future product enclosed inside. Then the working chamber is placed in the stove and is sintered. From aluminum oxide (it is more free-flowing and conducts heat better), we remove the finished, sintered product by cooling. Accuracy of printing will not be very. Yes, and the stove is needed.
-5.2And here's another: - sand castles method. In the working chamber there is a substrate made of porous ceramics, air is pumped out from below. The head is applied to the tracks of a fairly fluid mixture of water and metal powder, stabilized with something like modified starch, gelatin or guar gum. Like the usual FDM-printers. Due to pumping from the bottom of the substrate, the moisture from the mixture is sucked off and it becomes relatively solid. In childhood, sand castles were made like this. In this case, some adhesive component is needed, so that when the drying process does not crumble, a modified starch, gelatin or guar gum should suffice. Again in the oven and bake.
6. Composite printers.
We are talking about printers that use to build the parts are not homogeneous material, and the material with different inclusions, such as wire, electronic components and what else is invented. This is most likely related to FDM-type printers with an additional manipulator / head that can arrange electronic components, lay wires to connect them or to harden parts. Connections of electrical wires with parts can be made by melting spark / micro-arc solder wire or contact welding. To increase the longitudinal strength, you can provide holes for steel inserts, which will be enough to melt from two ends with a soldering iron, or grooves for hardening by winding.
At this, I stop the allowed speeches and I hope that, at least, it was interesting, and maybe inspiring to work in the outlined areas. If you have any thoughts, suggestions and ideas - please contact us. Thank.
Source: https://habr.com/ru/post/376195/
All Articles