For 3D prototypes and visual models
What is rapid prototyping?
In modern CAD-technology, it is easy to write (in code) a 3D model of a given design. To view the model, you use a 3D printing system. In the olden days, you printed or plotted a 2D-drawing and had to build the 3D image in your head. Nowadays you hold the actual 3D representation in your hands. Real "see and feel" models can be helpful, even essential, in visualising things.
In former days, visual models (non-functional) and prototypes (at least partly funcional) were made by laboriously removing material from a block, usually by CNC milling. Over the last few decades, new methods have been developed in which a prototype can be built up directly from the CAD-code instead of milled down . Most of these technologies are quick and easy. They are grouped under the general header "rapid prototyping".
Various CAD languages
Starting point for a rapid prototyping method is "a CAD description". Worldwide, quite a number of different types of CAD software are used. A good rapid prototyping machine ought to be able to communicate with all major CAD systems. To this end, most machines use their own language, adding software converters to translate conventional CAD-files as required. These conversions often are not foolproof.
Maintaining (sufficient) accuracy and (preventing) deformations
During rapid prototyping, various deforming effects - such as shrinkage - (can) occur. Modern rapid prototyping machine software automatically modifies manufacturing parameters to compensate for these distortions. There is a risk in this. The printer does not tell you what and how it modifies the original desig. The printed model is "magically created" and that is not a comfortable thought for the designer.
Laser scanning stereo lithography
In laser scanning stereo lithography, a liquid photopolymer resin is selectively cured using an ultraviolet beam from a laser. The laser beam is focused onto the surface of a vat of this resin. It scans the surface in parallel lines. In the moving focal point of the beam, the resin is hardened almost instantaneously. In this way, a thin (virtually 2D) layer is built of a shape corresponding to a horizontal slice of the final model. The machine's software slices the CAD model of the 3D shape to be produced into the required layers. The computer directs the laser beam by means of a set of galvanometer-controlled mirrors.
Just below the surface of the liquid photopolymer, a horizontal elevator tray supports the layer of cured resin. As soon as a layer is finished, the elevator tray is lowered one step (one layer thickness) to submerge the finished layer and fresh (liquid) resin floods onto the surface, covering it ready for the next layer. When the tray elevation step is set correctly, each layer neatly bonds to the previous layer.
In this way, the process is repeated, building layer after layer, until the complete model is finished. It is fully submerged in liquid resin and rests on the elevator tray. After the final layer is scanned, the elevator tray is slowly raised. The model emerges from the bath and can be removed for final curing and cleaning.
Laser scanning stereo lithography machines can be operated with various ultra-violet emitting lasers depending on power requirements. This gives the possibility of exchanging lasers or upgrading the system as required. The laser output is controlled by computer.
The laser beam is focused using a flat field lens, which allows very small spot-sizes across the entire vat. The two-dimensional galvanometer scanning system offers a high vector speed (hardly any mass to be accelerated / decelarated. The positioning accuracy ± 0.05 mm.
Mask exposure stereo lithography
In mask exposure stereo lithography, models are built from horizontal layers of ultraviolet-cured photopolymer. The CAD representation of the 3D shape is used to calculate exact raster images of a sequence of horizontal sections. The vertical stepping distance between these sections equals the layer thickness that will be deposited in the process. The created raster images are used as masks during exposure of the photopolymer. One by one, the images are sent to the mask plotting unit of the machine, where a high resolution optical image is projected on a flat glass platen. The image is then treated with black toner powder to form the actual mask.
At the same time the mask is being generated, a thin layer of liquid photopolymer is spread on the working platform of the machine. The platform and the toner mask are then positioned under the exposure station and carefully brought in register. A shutter opens, to expose the workpiece for about 3 seconds through the mask, to a high power (2,000 W) ultraviolet lamp. The UV light passing through the clear sections of the mask fully cures a horizontal section of the pattern of that slice. Unexposed areas of the layer surface remain in a liquid state.
The workpiece passes under a wiper, that removes all residual liquid photopolymer, leaving behind only the cured layer. It then travels to the next station, where a thin layer of melted wax is spread over the surface, filling all the voids and cavities left after removal of the untreated photopolymer. A cold plate is then lowered onto the wax surface, cooling and solidifying it. The results is a fully solid layer that is composed partly of cured polymer and partly of wax.
Next station. The workpiece is brought under a milling cutter that trims the layer's thickness down to the predetermined thickness, creating a flat smooth surface ready for the next layer. Also, the glass mask plate is returned to the mask plotter, where it is discharged. The previous mask is cleaned off. The platen is then ready to receive the next mask.
The process can now be repeated, building the next layer, until finally all layers are finished. When the last layer has been completed, we are left with a block of wax, within which the model (or models!) are embedded. The wax can be melted away in a microwave oven, by the use of a hot air gun or even by using warm water (60° C), after which the finished models are ready for use.
A special advantage of this method is the layer thickness milling operation, which ensures high accuracy of the finished model(s). No final curing of the finished model is required, as with laser scanning stereo lithography. These machines can generate layers from 0.05 mm to 0.15 mm thickness. The finest resolution is only used where sharply radiused curves in the vertical direction occur. Elsewhere a larger thickness is chosen, to speed up the building process.
Shrinkage and distortion
Shrinkage and distortion are relatively small with mask exposure stereo lithography. Accuracy is within 0.1%. Undercuts (overhanging shapes) and concave surfaces cause few problems. A special feature of the mask process is, that multiple models can be built at the same time in the same process. Very good for improving productivity. Also, the process can easily be modified to produce wax models for use in the lost wax casting (investment casting) process. The latter method is particularly useful when casting small products in difficult metals and/or complex shapes.
Selective laser sintering
The selective laser sintering process is a free-form fabrication method to create components by precise thermal fusing (sintering) of powdered materials. Parts of complex geometries are built in successive layers, that define subsequent cross sections of the component.
The sinter powder is power-fed to the process chamber from two cartridges flanking the partly built product. This allows for bi-directional powder feeding to the roller that lays powder across the top of the product, thus improving building speed. Unsintered powder is returned to the powder feeding cartridges, to be recycled. The process chamber is filled with nitrogen to obtain safe material sintering conditions.
Materials most often used in selective laser sintering are polycarbonates and nylon. For polycarbonates, densities of 75-92% of that of the standard injection-moulded material can be obtained; for nylon even better values, 87-93%, have been realised. Another material that is well suited for the process is investment casting wax. This allows the direct construction of wax patterns for foundry use.
Considering the density results just mentioned, selective laser sintering is particularly suited to build fully functional prototypes in polycarbonates and/or nylon, that closely approximate the mechanical properties found in the final (i.e., injection-moulded) products. Such models can be tested before any capital outlay for the construction of injection moulds becomes necessary.
Laminated objects manufacturing
The laminated objects manufacturing method creates models from layers of paper or polyester laminate, bonded together. The paper, coated with a thin layer of heat-sensitive adhesive, is supplied on a roll. It passes up over a vertically�moving table to a take-up reel. At the start of each layer, a heated roll passes over the newly exposed sheet of paper causing it to adhere to the previous layer. Papers of varying thickness can be used.
First, a laser beam directed by an X-Y device and mirrors, cuts a rectangle free from the bulk of the paper. Next, the shape of the layer is cut into this rectangle with the laser. The waste paper, surrounding and within the model, is cut into small squares, to make it easier to (later) remove them from the model.
When a layer is finished, the table on which the workpiece is supported, lowers one step (one paper thickness) and the paper is advanced to supply fresh material for the next layer. The cycle then starts again.
The laser motion is controlled directly by the computer on the basis of a 3D CAD model of the product to be manufactured. The system software has a scaling facility and thus can compensate for shrinkage. In other words, it can produce foundry patterns suitable for e.g. sand casting directly from the original shape. There is no need to first correct this for shrinkage. The software allows production of several models at the same time, from the same roll of paper.
The speed of cutting only depends on the peripheral length of each cross section rather than on the total surface area as is the case with alternative technologies. Therefore, parts with thick walls are produced in LOM just as fast as those with thin walls. To prevent problems due to variations in paper/adhesive thickness, the height of the workpiece is measured and fed back to the computer. The profile to be cut is calculated to be exactly appropriate to the actual height of the model at that stage of build.
The waste material is left in place until the model is finished. The cubes could be removed during the process, but as long as they are left, they provide stability to the model. The bonding roller every time passes over a complete sheet and thin sections remain adequately supported. The size of the cubes can be varied to suit the detail required. They are removed (with hammer and chisel) after the model is taken from the machine.
Accuracy and finish
Overall accuracy is ± 0.1 mm, so a high degree of detail can be incorporated into the models. If necessary, they can be hand finished. The models exhibit characteristics similar to those of plywood, including strength and machinability. As the models are constructed at room temperature from laminates of solid material, there is no shrinkage. This ensures that the model has a predictable accuracy level.
Ease of use
These machines are free-standing, self-contained units, operating from a normal 240 Volts supply and requiring no special environmental considerations. The unit can be installed either on the shop floor or in an office.
Paper layer technology
In paper layer technology, the product is built up from layers of thin paper. The computer slices the 3D CAD representation of the product into separate layers. Each is plotted on the paper with a special toner. The toner forms a cross section at the level of the present layer. Each successive layer is placed on top of the previous one and lightly heated and pressed on. Through this, the toner bonds with the previous layer, after which a plotter knife is used to cut the surplus paper away. As the paper surrounding the actual cross section is not bonded (no toner being present), the waste material can easily be removed.
The paper model offers fine detail, due to the small paper thickness used (0.08 mm). Its surface is treated with an epoxy resin to add strength, smoothness and colour to the model. A wide variety of finishes is within easy reach. In comparison with alternative rapid prototyping methods, paper layer technology is cheap.
Paper layer technology is very often used to produce "visuals" or "touchables" for marketing purposes. The models are however also suitable as sand casting patterns; as well as patterns for making silicone rubber moulds, to cast plastics or low melting point non-ferrous alloys in.
Fused deposition modelling
Fused deposition modelling builds a product up in thin layers of thermoplastic wire-like filaments. According to the proponents: ".... there's no need for messy liquid photopolymers, no powders, no lasers. An FDM machine enables full desktop operation and is fully suitable for use in an office environment."
The process starts from a 3D geometric CAD model, which is computer-sliced into horizontal layers. In the machine, molten thermoplastic material (PLA, ABS) or wax is extruded through a fine nozzle. The extrusion head with its nozzle is governed in x- and y-direction by the computer in accordance with the 3D model's slices, to produce the required cross section.
A second nozzle at the same time creates a supporting structure as required, e.g., to allow the formation of overhanging edges. When the model is finished, the supporting structure is easily removed. The layers are built on a horizontal table, which is lowered step by step as each layer is added.
Layer thickness depends on the nozzle orifice, the feed rate of the material and the speed of the extrusion head. It ranges typically from 0.18 mm to 0.26 mm/s. Large(r)-sized models are assembled from separate parts. It is easy to use different colours for these parts.
The mechanical properties of models made by FDM are comparable with those of injection moulded parts. FDM parts can be functionally tested, before any capital outlay for the construction of injection moulds becomes necessary.
Maintaining the molten material at precisely the right temperature (within 0.5° C above the solidification point is fundamental to the FDM process. Only then, the material solidifies quickly enough upon extrusion to generate usable wires, while still bonding to the previous layer through thermal fusion.
Multi jet modelling
In the multi jet modelling process, a product is built up of thin layers of wax. The mechanical properties of the model are less important, the easy formation of complex shapes is more important. Models generated by MJM are used mainly to help designers in visualising complex shapes.
Layer by layer is plotted on a wax plotter, under computer control, on the basis of the usual 3D CAD description. The plotting head has a large number (up to 100) individual wax jets. It is moved from left to right and back to "print" narrow strips. The wax is deposited in a low-strength raster-like structure. Where necessary, a supporting structure can be added - in the same material - which is removed easily once the model is finished.