Friday, 11 May 2012

Brain Storm session

Hello everyone,

This is the result of our first brainstorm session. This time, we investigate two possible options for 3D metal printing. The first option is Direct Metal Sintering, or should we say Localized Metal Melting, using a LASER. The second one is metal paste (also called metal clay) extrusion.

Localized Metal Melting
The idea is to design a machine that will be able to guide a LASER beam aimed at metal powder. The machine will also have to evenly spread metal powder over a bed.
  • The process is rather simple ( No complicated manipulations)
  • The process does not require post treatment other than cleaning, polishing and annealing if required.
  • The parts comes out to scale.
  • The porosity can be very low.
  • The time of processing can be quick with a powerful laser.
  • The support  is ensured by the powder and is easy to remove.
  • The machine can easily be transformed into a laser cutter.

  • High power laser is required. Simple calculations tells that a 10 to 20 watt laser would be required to process 1 mm³ of metal per second. 
  • High power laser cost a lot. Even the cheap laser diode option is not very cheap. One also need to get optics and mirrors.
  • Laser are dangerous. Especially the high power ones. And any security mechanism (filter glass, safety switches, googles...) are not strong enough against the curiosity of a DIY user. 
  • Melting metal can produce hazardous gases that need to be vented.
  • To reduce metal oxidation, the atmosphere in the machine should be controlled (vacuum, nitrogen, hydrogen, argon ...)
  • The machine need a lot of powder to fill the bed, even if the printed part is small. Also, not all the loose powder can be recuperated.
  • Due to the need of a bed and hermetic enclosure, this process cannot be integrated directly into an existing low cost open 3D printer.
Metal paste extrusion
This process is very similar to the fab@home system. A metal paste is placed in an extruder device, such as a motorized or pneumatic syringe. This paste is extruded on a building platform to the desired shape. The metal paste can be made from very fine metal powder mixed with water and other chemicals. Once the piece is completed, it is left to dry. Then the part is heated to around 500°C in an oxidizing atmosphere to burn the remaining chemicals. The part is then sintered in a reducing atmosphere to the temperature required for the used metal. These high temperature steps can be done very cheaply using a propane/butane torch or, more controlled using a kiln.

  • Existing open 3D printer can be easily used to print the part.
  • No significant waste of material. (Unless you break the part in the process)
  • Low setup cost possible. If using the 'propane/butane' torch method.
  • Porosity can be moderate. Useful to save on metal when the part does not require high strength.
  • Metal properties can be controlled using different temperature profiles in a kiln.
  • For good quality parts, a kiln is a must. This considerably raise the setup cost and time of the process.
  • The parts shrink from 10% to 30% according to many parameters (metal paste recipe, sintering time and temperature ...)
  • Many 'post printing' operations are required (drying, burning, sintering, finishing)
  • The sintering and burning process produce a lot of gases and should be vented.
  • A high temperature kiln is not a safe thing to have on a desk at home. 

As you can see, no solutions are perfect. Although the metal paste extrusion + propane torch sintering offer a very low cost alternative, anyone that look for anything better will be required to buy/build a 300 to 1200 USD kiln or a 2000-3000 USD laser machine. Both solution require a controlled atmosphere to prevent metal oxidation. This means having to find a nitrogen/argon gas bottle supplier or having to make an hydrogen generator.

Because of the danger and high cost of the laser concept, we preferred to turn to the metal extrusion method for now. The Laser option will probably be investigated further in the far future. Some tests are under way to check the ability to extrude metal paste from a syringe. We hope to be able to post some results soon. 

Thursday, 3 May 2012

What's in the world of metal 3D printing?

Hello everyone,

In this post, we would like to take a tour of the existing 3D metal printing methods. The idea is to identify as much working solution as possible. Knowing their pros and cons will enable us to identify the best candidate technology for our system.

Metalized plastic

Let start with "metal like" printing. This method consist of using a mix of traditional 3D printing polymer which serves as a binder for  metal particles. One example is the Alumide material offered by i.materialise. The idea is to have the look and sometimes the feel of a metallic object while keeping some plastic properties, like flexibility and the relatively easy and low cost process. Since the metals particles are not fused together the light will be reflected in many directions making the object will look like a dull metal piece. With enough metal, the object can get some properties like high density, thermal capacity and conductivity. This can gives the piece a slight cold and heavy metal feel. The printing process used can be Selective laser sintering, where the thermoplastic binder is melted by a laser but not the metal, Fused deposition modeling, where the melted binder and solid metal particles are extruded from a filament. Although no examples where found, Stereolithography might also be a potential process used, but if it is possible at all, the layers would probably have to be kept very thin due to the presence of metal particles reflecting and scattering the curing light. 

Fused metal deposition
Like it's plastic counter part, fused metal deposition is the process of extruding molten metal from a hot nozzle. The source of solid metal can be from a spool of wire or a syringe of microbeads. This process usually use low melting temperature metals and eutectic alloys. Eutectics are mix of two metals or more that can have a lower melting point than the separated pure metals. Since most metals have a rather sharp solid <-> liquid transition, the control of the behavior of the deposited metal can be tricky and the metal should be cooled rapidly to stay in place. Compared to other process, this is rather simple because the printed part does not require further processing. However the choice of metals is limited and most low melting temperature alloys are rather soft metals.
InkJet metal binding
The inkjet process (see it here and here) starts with a bed of metal powder. Layer by layer, a liquid binder is printed, by an industrial inkjet head, on successive layers of powder. After completion, the solvent in the binder is evaporated by air drying or a low temperature cooking. The remains holds the metal particles together. The part, now in the so called green state, is still fragile, but strong enough to allow for the loose powder removal. The parts are then heated to a high temperature to be sintered or impregnated by a lower melting point alloy. 
In the case of sintering, the metal powder is heated just below its melting point. The temperature is enough to allow for metal atoms migration on the surface of the particles. Because of the surface tension forces, this metal is more likely to migrate towards the contact point between the particles, fusing them together. Since the starting part is not fully dense, the resulting part remains porous and shrinking of the part can occur.
In the case of metal impregnation, the space between the particles is filled by a lower melting point metal. The capillary forces, very strong in small gaps, are responsible for the metal absorption. In everyday life, this is the equivalent of a sponge in a layer of water. The result is an almost void free part. Since the original particles are not melted, the part usually remains the same size throughout the process.
The basic process of printing is relatively simple, but the post printing process require time and precision.

Direct metal laser sintering (DMLS)
This process also starts with a metal powder bed. But this time, instead of binding the powder with a liquid, each layer of powder is directly molten by a laser beam onto the other layer. This result in a fully solid part, straight from the printer. Because the laser power can be adjusted and there is no other materials involved, many metals can be printed this way. However, high power laser doesn't come cheap and are very dangerous in household application.

Electron beam melting (EBM)
This process resemble DMLS, but this time an electron beam is used to melt the metal. Electron from an electron gun are given kinetic energy by passing in a strong electric field. Upon impact with the metal particles, this energy is transfered to those particles as heat. Because electron wouldn't go far in air, this process is made under vacuum. The vacuum also allows the melting of metals that would otherwise react with the oxygen in the air when melted.
Casting with 3D printed sand mold
This process is very similar to the InkJet metal binding, except that, this time, fine sand is used instead of metal and the printed part is the negative of the desired part. Once the sand mold is printed, the metal can be cast. This method allows the used of any castable metal. However some post processing is required to remove the sand and the channels used for casting and cooling shrinkage.

Casting with lost wax or sand mold made from 3D printed original parts
The idea is to simply print the desired part out of plastic or wax, make a sand mold out of it and cast it. However, the possible geometry with sand casting are limited to allow for the removal of the original part. Lost wax casting can be used instead, however, the burning of the plastic part can be a very dirty step. Although they can be very cheap, those methods require a lot of post printing processing and many equipment.

Here end our tour of the existing 3D metal printing technology. If you know of other or want to add some information, please share with us in the comments.

Monday, 30 April 2012

Hello everyone,

The start of a project is often critical to it's future. At the beginning, ideas and concepts are proposed and challenged. Choosing the wrong concept can lead to very big challenges. Since this is an open project we think that it could be interesting to get everyone involved. Do not worry, no special skills are involved. You can help us by simply answering the following survey.

Thanks for your help!

The 3D metal printing project

Hello everyone,

In this blog, we would like to document the design of an open 3D metal printer.
In the past years we have seen a lot of cheap 3D printers that print digital designs into various materials. Just think of projects like Fab@Home, RepRap, MakerBot and printrbot, to name a few. Most of them print parts using either ABS, PLA or PVA plastics. Other are more exotic, offering the user many choices of materials from silicone to chocolate icing (Mmm! Chocolate!) But what if one wants to print a part in metal? Right now, very few possibility are available for DIY metal printing. One can order from professional 3D printing services (usually not cheap) or can make a mold of their plastic prints and cast it (usually dangerous for the average user).  That's the reason why we seek to develop an user friendly, relatively cheap and safe way to print metals.

What do we seek to do in this project?   
Our goals can be divided in three objectives:
  1. Identifying a 3D metal printing technology that is simple enough for anyone to use, as low cost as possible and above all, relatively safe to use at home.
  2. Developing the basic process to allow for a good repeatable print
  3. Developing the tools (printer head and others) to allow everyone to do it. 

Why an open project?
Simple: Share and you shall receive!... Many heads worth more than one! ...
We believe that by sharing our ideas with everyone, some will be able to help us in return, thus achieving our goals (Having fun creating things) faster.