How Does 3D Printing Work?
3D printing is a relatively new branch of the product development and manufacturing market that offers ‘real’ material 3D outcomes on rapid turnaround. These modeling services have become normalized in recent years, but the paradigm shift they represent cannot be overestimated.
3D printing is a very broad term that offers you a startling range of divergent technologies and material families.
The thing they all have in common is that their intended purpose is to apply a ‘raw’ or source material to build net shape 3D components with minimal extraneous material usage.
These components are generally built such that limited (and often no) post processing is required in manufacturing parts.
This net-shape approach works by additive building, applying material in points, strands, sheets or streams to construct a 3D outcome in a single automated production stage. The approach is also referred to as Additive Manufacturing, Rapid Prototyping and Digital Manufacturing.
The material can be laid down as tiny inkjet dots of liquid resins, cured in place; it can be plastic threads melt extruded to ‘draw’ model elements; it can be scanning dot or whole picture cured UV sensitive resin, catalyzed by the light; it can be metal wire or metal powders welded in place; it can be a sacrificial binder agents and build material particles sintered after building; it can be powdered build material and a liquid bonding agent; it can be a gelled slurry that is applied like toothpaste; in limited cases it can even be laser cut sheet slices bonded in stacks.
The end result is construction of a finished (or near finished) 3D object, generally with minimal extraneous material used.
While the methods used for metal 3D printing are very different from those used for plastics, for example, the principle holds true for all techniques and materials.
Material is quantized and used to construct a single slice of a 3D object on a work platform. That platform then moves by one slice and the next layer can be applied.
Each of these layers is a free-standing, essentially 2D object of finite thickness, and the key to construction of useful 3D outcomes is for these layers to be a) integrally bonded within the layer and b) securely bonded to the layers that surround it.
By this means of layer deposition, 3D objects of near infinite complexity can be constructed.
A key differentiator of the various 3D printing technologies (other than material) is the 3D dimensions of the quanta of raw material that can be delivered. As a guiding principle, a perfect 3D component has a resolution in which the quanta (point size, layer thickness) tend towards the atomic level, making its resolution effectively infinite.
In reality, 3D printing systems cannot operate at these infinite resolutions, so layer thicknesses and material quanta fall into broad categories.
The highest resolution constructions result from processes with around 16µm resolution in 3 axes. Many engineering systems offer lower resolution, often 1/4mm and sometimes as low as 1mm.
As a boundary example, building construction systems often use resolutions measured in tens of millimeters and more.
3D Printing Industry
There is a complex market for 3D printing equipment, services and materials. This market continues to evolve, both through technology/material developments and through shifting business models, often driven by the trickle-down of increasingly high end and capable equipment whose price points suffer extreme downward pressure.
At the upper end of the capability spectrum, the equipment and operating costs are high. It is common for materials to cost $1,000 per kg and more and for machines to cost many hundreds of thousands.
At this level, climate controlled space and highly skilled operators are a must, so the overheads are large.
For this type of setup to be productive, a steady flow of high value tasks is a prerequisite. It’s for this reason that the ownership of the best equipment is limited to very large organizations and very busy sub-contract prototyping/manufacturing agents.
This type of agency varies from smaller local suppliers with some limited, high end equipment, up to large integrated consultancies that offer a full range of 3D printing technologies and often include ‘extractive’ CNC services and other manufacturing methods, for a comprehensive portfolio.
A relatively new aspect to the market is agencies offering printer ‘farms’ for manufacture of larger volumes of parts. In plastics, these services tend to concentrate on the aesthetic rather than functional end of the demand curve.
However, it’s increasingly evident that some equipment in both plastics and metal 3D printing is capable of delivering production-standard. These are increasingly used in medical and aerospace manufacture, where cost sensitivity is outweighed by the need for speed.
At the more accessible end, simple home use machines are available that suffer none of these cost challenges. Machines costing a few hundred dollars, using materials that can cost as little as $20/kg are quite common.
However, their capabilities are considerably less and the resulting 3D printed parts are therefore of lower quality.
In the mid range of capability, it is common for SMEs to have in-house equipment. This can range from the low resolution/low cost end of the technology spectrum.
It is also common for SMEs to operate older, legacy equipment as purchases made even 5 years prior can be quite obsolete, but still represent value to the users.
The 3D printing sector as a whole is expected to meet or exceed $41 billion by 2026.
Examples of 3D Printing
3D Printing Technologies & Processes
The market was initiated by a series of early developments in polymer 3D printing – successes were based on filament/powder extrusion, powder bed laser or printed adhesive fusion/bonding and UV laser cured liquid bath processes.
The only major failure was modeling by cut and bonded laminates, which came to market several times but never gathered full momentum.
As these early commercializations began to succeed in the market, massive R&D effort led to rapidly increasing diversity of technologies and a multitude of companies developed unique approaches that achieved varied degrees of success.
Stereolithography (SLA) was among the first 3D printing processes to be commercially successful – various experimental processes developed at the same time, but the race to scale up was resoundingly won by SLA.
SLA uses a scanning laser to cure UV sensitive resin at the surface of a bath, rendering solid one slice of a model before lowering the build table and wiping a new layer of uncured resin into place, ready for the next layer.
Models are built up in this way from the build table, and then removed from the resin and washed to remove uncured material. The built model is then given an additional UV exposure, to complete the cure and then any support structures are removed as a final stage.
SLA printing makes parts of good detail, smooth surface finish, and good tolerances. Resolutions for most SLA equipment are of the order of ~50 µm, which has remained the typical build height/feature resolution of much of the industry.
SLA parts are accurate enough for testing the fit of an assembly, but materials remain somewhat brittle for stressed functional testing and they degrade in sunlight.
The system remains widely employed in the medical industry for anatomical models and microfluidics, although there is an increasing tendency to models colored in the printing process in this sector.
Many companies and agencies use SLA legacy and up to date equipment, but the process’ market dominance has passed, as options widen.
Selective Laser Sintering (SLS)
Selective laser sintering (SLS) was another early breakout success in the sector. The process never developed the traction of SLA, partly because the resulting models have a difficult to correct surface roughness.
The process uses a blade to sweep a precise ‘film’ of nylon powder onto the work area and then fuses the particles selectively, by scanning an image of the slice using a UV laser to melt the powder.
The table then drops (generally by ~50 µm) before the loose powder for the next slice is swept over the top. In this way, full density models of good resolution and powdery surface finish are built up. Finished models are taken from the bed of model/powder and simply shaken clean.
SLS produces models of similar resolution/accuracy to SLA. These models are among the strongest plastic 3D prints and find application in high stress testing and occasionally in finished part production.
Sanding and painting can produce moderate cosmetic quality and these models are durable, suitable for stress and functional testing and allow live hinges and snap-fits that are both durable and potentially dimensionally similar to molded parts.
A major advantage of SLS is that the unfused powder acts as support, so parts can be nested and layered to closely fill a build table, offering significant build cost/speed advantages.
Polyjet was the first process to deliver true high resolution, using a process related to bubble jet technology to build plastic models in resins similar to those used in SLA.
PolyJet was a second generation technology that was launched in the ‘00s and captured a strong market share, while it was the first tech to a) print at 16 µm resolution and b) offer practical, usable rubber like prototypes.
Polyjet uses a long bar or bubble jet ports in a print head to deposit microdrops of resin and literally print a 2D ‘picture’ of the slice. This slice is then fully cured in the same pass by a strong UV lamp. Models can be solely build material when simple, but a second fluid system/print head set allows the printing of friable, water soluble support material.
Diversity of print heads and offsetting repeated print stripes makes the equipment tolerant of blocked nozzles. Completed models are removed from the table and generally water blasted to remove support material. Dried models require no further work, making one of the resin brand-names ‘FullCure’.
The current PolyJet machines are capable of printing two and sometimes three build materials plus support. This allows material blending for custom properties and custom colors and layer based color transitions.
This does not allow full color printing, but it can be a useful feature. More useful, perhaps, in dual build material machines is the capability to print rigid and ‘overmolded’ rubber parts in a single process, simulating perfect overmolds on plastic parts.
Digital Light Processing (DLP)
Digital light processing is a recent development from SLA technology, using the same basic light cured resins. This system has enabled small format SLA like printers to reach down into the home/hobby market, creating a whole new subsector.
The technology works by exposing the light cured resin slice on the surface of an LCD that exposes the whole slice as a single picture, sandwiched between the upward facing, resin immersed LCD and the upward moving build ‘table’ This whole-slice exposure makes the print speed high in technically simple equipment – and the upwards withdrawal of the built model from the resin tank is a counter intuitive step that brings the advantage of draining uncured resin back into the reservoir as the build progresses.
The machines that deliver on this technology make prototypes that are of good resolution (generally 50 µm layers and element size in X-Y) and have very similar properties to SLA.
It is likely that this approach has advantages that will see it develop rapidly in resolution, platform size and material range.
Fused Deposition Modeling (FDM)
Fused Deposition Modeling (FDM) is the last of the initial 3D printing methods to come to market and it was a very significant component of the sector until the early patents from Stratasys lapsed. This led to an explosion of industrial and home machine suppliers coming to market, generally with simplified approaches that avoided the remaining protected IP issues.
The term FDM is still copyright of Stratasys, but it has become the catch-all term for the related group of methods which are more correctly deferred to as Fused Filament Fabrication FFF. FDM is performed in a controlled, closed environment, FFF is (generally) performed under less environmental control and can use open frame equipment.
Fused Deposition Modeling is a widely used industrial and desktop 3D printing technology, making plastic parts of modest quality/resolution (0.2 mm to 1mm range).
These printers function by extruding a plastic filament through a heated nozzle, ‘drawing’ a model slice onto the build platform to form cohesive sheets of plastic, which fuse internally and fuse to the next later as it is applied.
These models are anisotropic in properties and generally of moderate strength and surface finish. A very wide range of polymers can be printed in this way, and any machine is generally capable of printing with various filament materials.
Some machines offer multiple print heads, to allow two color printing (but not color blending) and to apply water soluble support structures.
It’s a low cost and rapid method for producing 3D prototypes. These can occasionally serve for functional testing in coarser and larger models, but poor resolution, stepped surfaces and relatively poor dimensional controls limit the applicability of the parts in demanding roles.
Multi Jet Fusion (MJF)
Multi Jet Fusion is a relative newcomer in the sector, but with HP backing it up, it is rapidly capturing market share. It is a blend of approaches that is similar to Binder Jetting and SLS but differs fundamentally from all existing technologies, on closer examination.
Similarly to SLS, a controlled depth later of nylon powder is swept onto the build table, but then a liquid, inkjet like print stage follows. Two materials are jetted onto the surface – one being a binder agent that fuses the layer into a lightly integrated picture of the slice.
The other is a ‘detailing’ agent that improves the edge quality. Once these materials are applied, an infrared lamp passes over the surface and some magic happens;
- The ‘binder’ component encourages melting, as the lamp passes over. This fuses the particles that are ‘binder’ coated, but leaves the rest unfused and undisturbed.
- The ‘detailing’ agent does the opposite, preventing stray melting/fusing at the slice extremities, making the outer surface crisper and more precise.
Finished models are extracted from the powder bed and the unfused material just shakes off, leaving finished parts.
This results in better mechanical properties and smoother surface finish than SLS, and because the fusing is wide area (rather than scanning laser point to point) the build rate is increased.
The equipment is relatively expensive, but the resulting models are ideal both for mechanical design/fit evaluation and for high stress use.
Binder Jetting is an early second generation technology that developed after the first SLS, FDM and SLA methods. Not a polymer based process, it exists in a category of its own. It is a side branch of the SLA powder bed processes.
A powder bed of plaster dust is layer bonded using bubble jet printed cyanoacrylate adhesive. These layers can then be full color inkjet printed to build up full color models, which was for a long time a unique capability.
The models are weak and of no use for functional purposes, but offer a low cost way to make full color models.
This technology has had modest but steady market presence, allowing packaging samples, color figurines and cosmetic prototypes. Its surface finish is slightly powdery, but the technology found a niche in marketing.
Laminate Object Modeling (LOM)
LOM is an exception to the general trajectory of 3D printing techniques, in that it has successfully come to market at least twice using paper and plastic films to construct 3D models by cutting and laminating sheet components.
There are no successful LOM systems in the market at this point, but a number of companies are attempting to break into sales with solutions that are related to or derived from LOM.
These are generally composite material based and attempting to build high, strength and high value parts.
Metal 3D printing
Metal 3D technologies were experimental, in the very early market developments of 3D printing, but the higher energies for fusion and the more reactive and demanding nature of the metal parts market made the marketization of these incipient methods slower.
One of the earliest plastic prototype methods, SLS, also spawned a metal version. Like most metal processes, it was slower to commercialize, but as a class, metal 3D printing has come of age and moved into the mainstream, with a similar flowering of approaches that occurred earlier in plastics and is still very much ongoing.
Powder Bed Fusion Methods
This is given a variety of branding names and using a range of fundamental but small process differences. Powder bed fusion is perhaps the simplest and most common approach to metal 3D printing.
These machines use fine (and hazardous) metal powders to build a layer onto a build plate, then selectively melt dots to build a slice of the part, without disrupting the powder layer.
The two types of powder bed fusion in metal are Selective Laser Melting and Electron Beam Melting, differing only in the way that the high energy for local fusion is delivered.
Selective Laser Melting (SLM)
Also referred to as Direct Metal Laser Sintering (DMLS), Selective Laser Sintering (SLS), Direct Metal Printing (DMP), Laser Powder Bed Fusion (LPBF).
Most powder bed fusion machines use the Selective Laser Melting (SLM) approach. SLM uses a high powered laser to melt/fuse points/lines of metal powder to create a model slice, building up a solid metal part buried within an unfused powder bed. The finished part is detached from the build table and cleaned, ready for any post processing
SLM is often considered the standard for metal 3D printing, as it’s the most widely adopted. SLM printed parts are widely used in sectors from dental/healthcare to aerospace.
Machine build volumes range from 100 mm cube up to 800 mm X 500 mm X 400 mm, but as a point fused modeling process, print speed is relatively slow.
Operating SLM machines is hazardous, as the metal powders pose serious health risks. The machines are expensive and require very skilled setup and operation.
Parts are of moderate surface quality and generally require skilled post processing and precision machining, to deliver useful outcomes.
Electron Beam Melting (EBM)
EBM machines are functionally identical to SLM machines, apart from the use of an electron beam to fuse the powders. GE Additive is the only maker of EBM machines. EBM machines make less precise parts than SLM, but build faster for larger parts.
These machines pose all of the same difficulties in setup and operation as SLM machines, but are more widely used in aerospace and medical applications.
Direct Energy Deposition
Direct Energy Deposition machines use either powder or wire feedstock and a powerful laser to construct components. This dispense and melt process happens at a print head, so there is no powder bed and less (or no) handling of hazardous powdered metals
Also called Laser Material Deposition (LMD) and Blown Powder Deposition. DED machines blow precisely metered and directed powder from a nozzle onto a part, melting and fuzing the powder as it travels to the build point.
Parts printed with DED based methods are indistinguishable from those printed by SLM, without deep process knowledge. However, DED has a unique capability, which is to repair non printed parts, such as machined and cast components.
Their material options, post processing needs and powder management hazards are similar to SLM, though the degree of operator exposure to powder is much reduced compared with powder bed systems. The equipment is very high cost to purchase and operate.
This is often referred to as Electron Beam Additive Manufacturing (EBAM). The use of wire feedstock allows a technically simpler machine and this technology is more commonly used for large build volumes (up to 5 m x 1 m x 1 m) and it allows fast print times, producing lower precision outcomes.
These machines are very expensive and quite uncommon, other than in very specialist applications.
Metal Binder Jetting
Metal Binder Jetting is a relative newcomer into the metal 3D print space. It’s worth noting that there is nothing new in this technology, but it represents a clever reimagining of the technologies that came before. Its market momentum is building
A Metal Binder Jetting machine is, at its core, a powder bed system. A layer of powder is put down, defining the vertical build resolution – though not the final model resolution, for reasons that we will explain.
An inkjet printhead then traverses the slice and prints a binder agent that glues the particles together in a very precise reproduction of the slice in X-Y. The table then drops by one slice thickness and the print stage repeats, bonding the slice and adhering it to the slice below.
So far, just like all of the powder bed bonding processes. The big difference is in the next stage. The ‘green’ part – printed, cured and made of polymer coupled metal particles – can be used directly, looks and feels like metal but lacks the strength of the ‘real’ material.
The next step is to sinter the ‘green’ part in a vacuum furnace, such that the binder is burned away and the particles gel and fuse (not really melting). This shrinks the part to fill the now empty binder cavities and makes a 100% density, precise and fully metal part.
The shrinkage is very predictable and the final parts demonstrate all metal properties and precise dimensions.
This process allows fast builds, offline sintering and works well with large multipart batches, as parts can stack in the powder bed. The materials remain hazardous and the equipment os very expensive.
Bound Powder Extrusion
This technology is also referred to as Atomic Diffusion Additive Manufacturing (ADAM) and Bound Powder Deposition (BPD)
Bound Powder Extrusion (BPE) is another newcomer to the 3D print market. It differs from the powder bed systems in that the powder is premixed with a wax like binder, so it is heated and extruded in an analog of powder based FDM.
The wax bonded material is safe to handle, unlike all of the loose powder processes. The printed part is considered ‘green’ in the same way as in Metal Binder Jetting, but it is very weak and not usable in the green form. The two-step post processing of a wash stage to remove most of the wax binder, then a sinter stage in a vacuum furnace to shrink and particle fuse the part. This is offline to the machine, not affecting printer productivity.
This process offers lower resolution than Metal Binder Jetting, closely mirroring the FDM/FFF process it is cousin to, and ‘stacked’ builds require support structures that risk part damage, making bulk builds less practical.
However, parts are still good resolution and the low hazard nature of the build material is a huge advantage. Importantly, these machines cost around 10% of the price of Metal Binder Jetting systems and can be operated safely by less skilled staff.
The materials range available to you falls into two divergent categories; ‘real’ materials and analogues. The increasing focus of the development of the sector is on real materials, and modeling processes that result in component properties that are close to or match the solid material properties.
The metal based processes vary in their material options, but in general the materials available across the technology range as powders, powder pastes and wires to suit each technology are;
- Stainless steel (various grades are available)
- Tool/alloy/high Carbon steel
- Bronze (particularly specialist grades such as NIckel Aluminium Bronze)
- Precious metals (Platinum, Gold and Silver)
- Aluminum (various alloys and an increasing range)
The plastic 3D print material family is a lot larger, as the range of properties and intrinsic material types are very diverse. This list indicates which processes are appropriate for each material class;
- Nylon (SLS, FDM, FFF, MJF)
- General engineering polymers ABS, PET, PETG, PEEK, PMMA, PP, PBT, PC (FDM, FFF, MJF)
- UV cured modified acrylics (Polyjet)
- UV cured polyester resins (SLA, DLP)
In addition, many processes are now capable of building with composite materials, enhancing properties with additives. These are most common as additives in FDM/FFF filament materials;
- Carbon fiber, both single wall Carbon nanotubes (SWCNT) and multi wall (MWCNT)
- Kevlar fiber
- Metal particles and fibers – to render prototypes conductive
- Glass fiber
- Wood/cellulose fiber
Rapid prototyping of products and components has been experiencing a rolling revolution for 3 decades. This has lead to an explosion of service providers, targeting a huge range of niches – as well as generalists who cover a wide range of capabilities and services;
These suppliers offer a spectrum of capabilities that is not limited to 3D printing, although this is the visible, headline part of their provision.
In reality, the spectrum of services includes fast evolving interpretations of the more traditional processes. Metal prototyping services can include;
- 3D printing in metal. Resolutions and surface finishes vary directly in proportion to equipment purchase and operating costs. At the leading edge, near perfect parts that are, for many purposes, ready to use.
- Post processing of metal 3D prints. Many metal parts require dimensional precision and surface qualities (in some areas) that are beyond 3D printing, so it is quite common to post machine 3D printed metal parts, where they interface to other components.
- CNC machining. It is common for metal parts to require high surface finish and precision on all or most surfaces. Once post machining of the majority of the part becomes necessary, then the benefits of 3D printing are no longer relevant and precision CNC machining takes over.
- Rapid casting (often using 3D printed patterns)
- Rapid forging (using Electrical Discharge Machined (EDM), CNC machined or 3D printed alloy steel forge tools)
- Rapid plastic/die cast tooling (using Electrical Discharge Machined (EDM), CNC machined or 3D printed alloy steel cavity components in standard ‘bolster’ tools)
- Barrelling/polishing. Where cosmetic finish is important, it is common to barrel or hand polish metal parts, both 3D printed and machined components and this is a key service provided by prototyping shops.
Plastic prototyping services offer greater material diversity and a similar breadth of approaches;
- Various levels/qualities of 3D printing in both engineering polymers and simulation polymers
- Vacuum casting – using silicone rubber mold tools cast around a 3D printed part to reproduce the part in volume by filling the resulting cavity with urethane resin to make you multiple copies.
- Pattern making for your sand and investment casting in metals
- Component surface finish improvement by CNC machining and barrel/hand finishing.
- CNC machining from solid plastic. This results in the closest properties analogue to injection molded parts.
- Rapid tooling for injection molding.
Car design and manufacture is a constant driver of the development of all 3D printing processes. The demand for large plastic and metal parts, precision and real material properties is influential in the R&D process – and a big driver of cost-down as they tend to be cost conscious, as a sector.
The Aerospace sector is a huge consumer of 3D printed parts, particularly for engine components (experimental turbine blades, flow guides, structural members, rocket engine parts, airframe parts and more).
This sector, perhaps more than any other, has driven the metal prototyping industry. There is lower cost sensitivity in aerospace development, so part cost and equipment cost tolerance is much higher.
Organizations like NASA have take a positive position in driving demand, operating print farms to make parts fast and allow rapid iteration – without much focus on cost.
Much publicized International Space Station components have even been printed, although these are more about publicity than function and are generally non critical parts.
The construction sector is beginning the rapid build process, with a variety of startups developing methods for extrusion printing buildings and building components. These companies generally employ gelled concrete extrusion, analogous to FDM/FFF to directly print finished parts.
A simplicity advantage in construction is that surface finish and precision requirements tend to be more modest, encouraging relatively simple and single stage (i.e. no finishing) processes to be developed.
The engineering development of consumer products has accelerated dramatically in the last two decades. Much of this is a direct result of CAD processes allowing precision and simulation without having to make actual parts.
The remainder of the acceleration is directly attributable to the explosion of 3D printing (and related) prototyping services. Fast iteration and extremely high design confidence prior to mass production, combined with increasingly agile processes, have sped up development greatly.
The healthcare sector makes extensive use of 3D printing, in a variety of ways;
- In development of new products, the process has been accelerated by the provision of 3D printing in metal and plastics, in exactly the same way as automotive and consumer product sectors have.
- In development of custom components for patients, metal and plastic 3D printing capabilities allow surgeons and physicians to customize treatment in ways that were previously impossible. From hip joints to external braces, parts can be made fast and precisely tailored to the patient.
- In teaching and visualizing, the USE of VR and colored prototypes to visualize and plan surgery is becoming increasingly widespread. Planning routes into congested surgical sites can make the difference in quality of outcome.
A variety of prototyping tools have developed for the food niche, and there is pressure for this technology to broaden out from the decorative and colored sugar printing that is relatively widespread, into synthetic food such as printing meat substitutes.
It is increasingly common for educational establishments to have 3D printing equipment for use in their curriculum.
This is generally limited to plastics and low power laser cutting at the school level, but Universities and research institutions now always have prototyping departments, which are often state of the art and ahead of industry norms.