| 3D Printing Printers > 3D |  | Navigation Icons & Guide
|
3D printing is rather different. 3D print is also known as "RP" (Rapid Prototyping) because that tends to be the main aim at the moment. The output from a 3D printer is objects, not pictures.
The term 3D printing might be reserved for building objects using techniques like an inkjet. However there are a quite a range of interrelated techniques. For instance the UV setting polymers used in stereolithography can also be used in 3D printing and in ordinary 2D printing of outdoor signage.
3D printing methods are not all entirely new. Geographers and geologists used to make landscape models using layers of polystyrene tiles. Wooden objects can be made from gluing together successions of laminated wood shapes and then smoothing the result. Some printer parts are made using laminate techniques, for instance, for some time piezoelectric printheads have been made from stacks of thin stainless steel plates welded together. The aim with 3D print is generally to form successive layers already bonded together as they are made. The object is created in one piece and doesn't need much further processing.
Building something in layers is often called an additive process - contrasted with machining a part from metal or wood which is subtractive. Forming items by adding fractional amounts of material has been made much easier by the progress in microelectronics which allows highly controlled printheads to work for hours unattended.
At present 3D print is used in exceptional circumstances - for prototyping where the costs of making a first version of things are often high. Extreme situations such as military logistics is another use. Sports clothing design where items can be tailored exactly to an individuals need is another. In the long term digital print might become widespread taking over from many conventional production processes just as digital print is taking over from conventional presses.
3D print is potentially a revolutionary production technique with several advantages.
| Objects can be made on demand using just a few base materials and a digital description. In many industries the logistics costs of spares are very high. |
| Things can be customised - built for the specific use. Mass production is always a compromise. |
| Parts that could not easily be made in one piece by any other technique are possible. For instance engine and wing components for aircraft can have material precisely where the stresses occur and voids where material would be wasted. |
| Things can be optimised - made with the exact shape and amount of material needed to deliver the required performance. |
| Manufacture of the object (and possibly the parts for the 3D printer) can be close to the point of use - potentially diminishing the need for global scale factories. |
Several of the 2D printing techniques have some potential for 3D
Extending the thermal inkjet or laser printer process could make more complicated objects but both techniques run into limits. Inkjet ink needs to be runny to flow through the head. Laser toner needs an oppositely charged surface on which material from the OPC can be deposited.
Piezoelectric inkjets can print anything that doesn't block the nozzles or overheat the actuators. Samsung's ball-bearing mechanism would have similar potential.
Thermal transfer would have some potential although the costs of foil might get a bit out of hand and objects made from soft and easily melted wax or resin probably wouldn't last long.
3D print techniques tend to be rather different - notably laser power ises from the milliwatts into the kilowatts.
Handling 3D is obviously like 2D but sometimes a bit more complicated.
Engineering information is often held in computers in 3D - things are drawn as vectors using software such as SolidWorks or AutoCad and held as an STL file. In normal printing a projection is placed on a 2 dimensional printout. This kind of data has to be translated into something a 3D printer can act on - a slicing algorithm might make a bitmap of voxels rather than pixels for instance.Builders, surveyors, engineers, tailors and shoe-makers use various kinds of measure to get accurate dimensions. These measurements can be typed into a machine, (they often are) but its a slow process. Some kind of computer steered data capture mechanism using an encoder of some kind.
Acquiring information from objects usually needs a different technique because practical instruments don't tend to return vectors.
3D acquisition includes - stereoscopic cameras, 3D scanners and LIDAR. These produce a point-cloud, a collection of measurements that describe the space
Cameras help people judge the shape of things but the human mind uses all sorts of experience to make its mental picture of the world. Computers need a bit more help, to get an accurate measurement with a stereoscopic camera arrangement needs some clear target points that software can easily pick up - which nature doesn't always provide.
3D scanners can be used to get data from nature. Scanners ideally use optical time domain reflectometry, usually known as "LIDAR" - Light Detection And Ranging. Light travels at fractionally under 300,000,000 metres per second. Shine a bright light at something a metre away and in about 6 nanoseconds it reflects back (3 nanoseconds each way). If the object moves away from the light source the delay lengthens. Arrange a laser to pulse as it scans across a scene using a polygon mirror like those in a laser printer and the succession of measurements gives the coordinates of the surfaces nearby. A problem with LIDAR capturing small scale data is that transparent objects can give false readings.
Coordinate Measuring Machines (CMMs)
In human terms measuring intervals of a few picoseconds to get measurements to a millimetre or so appears difficult but communications and data recording lasers and sensors have to pulse at these sort of frequencies. Variations on the theme can be used. The measurement might use diffraction to create interference pulses.
Types of Printer include stereolithography, laser sintering (SLS) fused deposition modelling (FDM), laser engineered net shaping (LENS) and 3D printing.
Most but not all 3D printers are laser based, layer - additive devices. A laser scans successive cross-sections of the 3D object building it up from a liquid or powdered material.
Stereolithography.
Stereolithography builds solid plastic objects from a liquid photopolymer which solidifies on exposure to laser light. The stereolithography technique typically builds objects on a table which starts off just submerged in the liquid. A laser scans the polymer and the injection of energy triggers solidification of the polymer. The process successively scans, lowers the table and rescans until the object is finished. Layer thickness is typically from a thousandth to a hundredth of an inch and the esolution of the laser needs to be about the same. A blade might be use to wipe new liquid resin across the part between laser exposures, overcoming surface tension. Newer machines use a perforated table and a pump to do this. The resin becomes harder but not fully cured as the process works so large voids in a structure may need temporary support.
The laser scanner is working on the 2D surface of the liquid resin and can be a scaled up version of that in a laser printer with a rotating polygon mirror to create the raster. This can be quick but with elatively small work area - a build envelope roughly the same as an A3 page. An X-Y plotter action like those used in laser cutters will give a larger build envelope but then each layer make take a minute or more. Most photopolymers need an ultraviolet wavelength to activate so the laser is completely different to those used in an ordinary desktop printer - a large Neon-Helium device rather than the little semiconductor infrared device. Semiconductor lasers are a long way from delivering the power needed (?). Other powerful light sources can be used.
The process is not particularly fast as a large quanta of light is needed to activate the esin. There is a tradeoff between speed and accuracy so producing thicker layers will render the object more quickly but less accuracy. Overall accuracy can be better than 100th of an inch (some sources say +/-0.1% but that presumably depends on the resin).
On completion the part needs cleaning of any surface resin by immersion in a bath of solvent. It may then need further curing in a UV oven to completely cure the plastic.
Parts made by stereolithography usually have high accuracy and a good surface finish. There are a variety of photopolymer materials, some equire further curing in an oven.
The standard photopolymer is SOMOS 11120 which sets to a material like ABS plastic. Photopolymers are often the same as those primarily used in flexographic platemaking; they include DuPont Cyrel and BASF Nyloflex , Novacryl, Elaslon, McDermid and Soleflex Exaprint. Resins can be flexible or hard and they can be good for making evaluation items or for producing durable finished parts. The resins aren't cheap, typically in the hundreds to thousands of pounds per gallon class. Resins aren't all non-toxic and the solvent is typically alcohol so the process needs fume handling apparatus which drives up the equipment cost.
Stereolithography machines themselves tend to be in the £100,000 to £500,000 class.
Parts made by stereolithography can be rugged enough to use directly but it is obviously an expensive production process so it is more normally used for rapid prototyping or as the pattern for a sand casting, or to make the mould for room temperature vulcanization (RTV). In RTV the part is surrounded by silicone rubber, which sets and is then cut to etrieve the stereolithographed piece. The resulting 2-part mold can then be used to make about 25 urethane plastic or epoxy resin parts.
The Stereolithography process was invented by Charles Hull, who founded 3D Systems, Valencia, California. The first machines were shipped in 1988. The process became much more practical in the early 1990s as chemical manufacturers developed a range of photopolymers. 3D systems stereolithograph apparatus (SLAs) are industry standards and they are also leaders in producing selective laser sintering machines (SLS).
Laser Sintering.
Laser Sintering uses a laser to fuse a bed of powdered material. Quite a range of materials can be used - polystyrene, nylon, metal, ceramic or foundry sand. Anything a laser can melt can potentially be handled.
Physically the powder bed might melt, partially melt or be sintered. Which technique is used gives the density of the material - sintering leaves voids which may be useful in producing foundry moulds whereas full melting giving 100% density may make parts that are superior to those produced by conventional methods.
Laser power has to be sufficient for localised melting whilst raster or vector scanning the powder bed. Plastic materials can be handled by a mid-power carbon-dioxide laser - although the low powers used for laser cutting and engraving will produce objects slowly.
Because the machine is always working on a bed of powder the object doesn't usually require any support as it is constructed.
The SLS process has a much wider range of plastic and other materials available to it, where SLA is limited to photoplastics. However the localised melting inherent in SLS limits it's accuracy to about 0.010 of an inch whereas the SLA process can be held to 0.005 and parts come out with a better surface finish.
The process is generally known as "SLS" Selective Laser Sintering - which is a trademark of 3D Systems, Inc.
R.F. Housholder developed a process similar to SLS and patented it in 1979 but did not develop its commercial potential. Dr Carl Dckard at University of Texas at Austin developed and patented SLS in the mid 1980s and licensed the technology to DTM of Austin, Texas. DTM was aquired by 3D Systems in 2001 and they now manufacture the "Sinterstation" and supply powder materials.
SLS using metals can be known as DMLS "Direct Metal Laser Sintering"
Fused Deposition Modelling.
FDM (Fused Deposition Modelling) extrudes material from a nozzle moved by a 3D arm or plotter mechanism to create the required object. The material can be wire fed from a reel making a metal item, or a plastic such as ABS, polycarbonate, polyphenylsulfones and wax. The heat source can be electric conductance -so a metal object is built by continual welding. Plastics and waxes need to be melted by a heated nozzle.Soluble materials can be added as temporary supports as an object is built.
The FDM process can use materials such as ABS (Acrylonitrile Butadiene Styrene), E20 (Polyester based elastomer) and investment casting wax. Typically the materials can be interchanged on the same machine.
FDM was developed by S. Scott Crump in the late 1980s and commercialised in 1990 by Stratasys, Inc. which was still the industry leader in 2004. It has become the best selling RP technology with about 48% of the market reported. Simple machines have been sold under the Prodigy and Dimension trademarks.
Metallic FDM parts might have small voids but can be made stronger by wicking metal into the part after production.
RepRap is a project for an open source FDM machine capable of making most of its own parts. A RepRap machine should only need cheap commodity devices like electric motors to build a functioning copy of itself.
Laser Engineered Net Shaping.
Laser Engineered Net Shaping (or LENS ®) was developed by Sandia National Laboratories as a way to make metal parts. It might be interpreted as a cross between laser sintering and fused deposition. Metal powder is injected at the point where a laser beam is focused, melting and fusing into the object.The high power laser and powder source both lead into a deposition head. The laser might travel through the centre of the head with powder delivered by gravity or a gas jet coaxially around it. Since the laser spot is small surface tension will absorb the powder and maintain a smooth surface.
An inert gas shroud of the kind used in welding can shield the pool of molten material from atmospheric oxygen so that the metal doesn't degrade.
Microscopy studies show LENS parts to have no compositional degredation and mechanical tests show excellent mechanical properties.
The LENS process can work directly on metal materials - aluminium, titanium, stainless steel and many alloys. The techniques can be used to repair of pre-existing parts as well as rapid prototyping and limited run manufacturing.
Optomec specialise in LENS.
MultiJet modelling (MJM) uses a printhead with a large number of nozzles to spray hot thermoplastic and build up an object in slices. The technique is similar to FDM but with multiple nozzles using a hot material. The printhead used in the Xerox Phaser 2D printer which essentially handles hot wax is very similar to what is needed.
3D Printing.
3D Printing uses inkjet heads to form solid parts, typically by spraying a binder onto a thin layer of powder. Objects are built in layers from the bottom up by spreading a layer of powder then scanning with the inkjet head until the complete object has been formed. Surplus powder is then pored out of the voids.3D printing technology was first developed at Massachusetts Institute of Technology (MIT) and has been developed by Zcorp.
The powder can be more or less anything with plaster, corn starch and plastics all being used. The liquid component can be more or less anything that will pass through an inkjet head. Thermal printheads can be used if the binder is water or alcohol based because both will vaporise and act as a propellant.Piezoelectric heads will tend to have advantages because they can squirt out any material that doesn't actually block them and they run hot as with MJM.
An advantage of using inkjet mechanisms is that the multi-nozzle heads act quite quickly. The "ink" can also carry a dye or pigment just as it ordinarily would so the object is made in full colour.
Uses.
At the moment 3D manufacturing tends to be relatively expensive which tends to limit it to prototyping. Making prototypes, moulds and tools has always involved a lot of skill and specialist machines so 3D printers and their relatives greatly reduce costs. Parts for engineering visualisation and investment moulds are typical uses, dental and hip replacement prostheses are others.It is possible to capture a cathedral with a 3D scanner but currently totally uneconomical to replicate it full scale. Using 3D microfabrication techniques like photopolymerization it would be possible to make a rather accurate miniature model.
Smaller items like carvings can be captured using a 3D scanner and replicated in resin with a 3D printer.
Since 3D printers can print any fluid they can inject a substrate with organic cells allowing computer aided tissue engineering. Combined with suitable nutrients and support structures the ability to print eplacement organs seems to be within reach.
Voxels.
Digitising a two dimensional source gives picture elements or "pixels" - which are normally considered to be square. Digitising a three dimensional source gives volume elements or "voxels" which are normally considered as cubes.The sheer number of pixels used to be a problem for laser printers - the several megabytes of memory needed for a page used to pose a problem and still do where pages are large and in colour. Doubling printer page esolution quadruples the memory needed.
Three dimensions potentially poses real memory problems - even with today's very large memories. A small production space 100 millimetres on a side resolved at a crude 0.1mm is 1,000 x 1,000 x 1,000 voxels - a decimal gigabit. Adding properties like colour and material selection easily increases this to a gigabyte. Doubling the resolution to 0.05mm multiplies the memory equirement to 8 gigabytes.
Some 3D printers are like laser printers in that stopping would create a problem - heat would dissipate and the machine would have to slow or make preparation passes over existing work. It is generally helpful if what will be printed is ready to go as a memory bitmap as the production head needs it.
Memory use may not need to be quite so dramatic because physical structures are no usually quite so complicated as text and pictures. There are large volumes of open space, and smaller volumes of solid but undifferentiated material. Run length limited coding can substantially compress the memory needed.
Jobshop.com article on ProtoCAM
--
© Graham Huskinson 2010
This page is like all those in the "book" section in being under development.
If you think this page is wrong in some respect or have better information on how things are done let us know. Click here.