Tuesday, 29 October 2013

Innovation in 3D print

Finally here are some really interesting innovative 3D printing projects. Both are linked by the use of employing micro processors in non conventional ways for 3D printing. The Peachy printer uses a sound card to drive a mirror and the other is some schoolboys from South Africa who are have written an app to print from a mobile phone to an extruding 3D printer that they built.

What is so great about the Peachy using a sound card is that the sound card is a highly developed piece of kit that sends small electrical impulses to an electromagnet that drives the speaker mechanism. It is such a neat fit for a device to power a tiny mirror. The elegance of this design solution elevates it to that exalted and over-baked description - awesome.

The Peachy also has a fantastic solution to the moving the build platform using a drip fed saline solution to support the body of resin. This means that the layer thickness can  be very tightly controlled.

This is a 3D printer with no motors! No vibration, no noise, energy efficient, does not require custom made electronics. This is revolutionary.  Well done Rylan Grayston and team.

In the second project, two 15 year olds, Pieter Scholtz and Gerhard de Clercq, from South Africa, created an app to slice and send data direct to their 3D printer from their mobile phone. Now I am not sure how unique this is. I know that some vendors have been talking about accessing 3D printers from mobiles and via html but you have to ask why would anyone ever want to do this? Well these guys have realised that in Africa mobile phones are widespread whereas computers are not. A good reason to build an app to 3D print from a phone. True most mobile phones in Africa are probably not smart phones but it would seem more likely that mobile device usage will develop where PC usage may not.

There has been some discussion of using 3D printers to make items in remote locations and this helps to progress this cause a little further on its way. Additionally with apps like Autodesk's 123D Catch a mobile could potentially capture scan data and then (possibly with some cloud data clean up) print it -  to allow on the spot reverse engineering.

Friday, 25 October 2013

BIM - 3D Printing - Hurdles to overcome

BIM has the one key ingredient needed for 3D printing and that is, 3D.  However, to get to a useful 3D print there are three main hurdles to overcome. 

1. Scaling

Because BIM is not a free-for-all modeler, objects that may be too thin to print at model scale may not easily be edited.  This usually leads to some re-modelling work outside of the BIM software to get to a 3D print ready state.

The image below shows the same data printed at 4 different scales.  Each one needed some editing to work at its particular scale. 

Model House printed at 1:100, 1:200, 1:500 and 1:1000 scales
Despite several attempts by software geniuses, it is still not possible to get a good scaled model from an automated process.  At Lee 3D we manually edit the file spending a good deal of effort to keep the detail clear and the design intent unaltered.  On a large dataset this can be time consuming.

2. Excess Data

A BIM model can become large and unwieldy when exported to a 3D printable format.  3D print data loses all of the object oriented efficiency of a BIM or CAD file.  For this reason the hinges on each of the doors and the text inscribed on each of the hinges can make a file very large indeed.

Where it is possible to export key geometry a good deal of time and megabytes can be saved.

3. Unresolved Geometry

Revit has an aversion to unresoved meeting points where structural parts should join.  Where details have not been fully designed, Revit carefully leaves all parts poised in readiness.  The consequence of this is that many structures contain large numbers of floating joists, beams and columns.  The hurdle is a double one as not only do all parts need to be extended a few mm, but it is very important that none are missed to avoid the model falling to pieces.

The image below illustrates the kind of thing that frequently occurs in Revit.  The beams would not quite reach the columns as this has junction has not been resolved.

In conclusion while BIM is great for 3D printing, there can be some hurdles to overcome in getting a good result.  Excess and unresolved data would seem to be the easiest hurdles to overcome.  How scaling can be addressed so that data can be reliably printed is still unclear.

For more information on 3D printing for AEC visit www.lee3d.co.uk

Thursday, 24 October 2013

BIM - 3D Printing - Introducing the Infomodel

BIM needs 3D printing in the same way that spreadsheets need pie charts.

BIM is a technocratic virtual construction that tends to create distance between its technical participants and the man-in-the-street world outside.  BIM needs interfaces between the information model and the stakeholder.  To do this BIM needs to break out of the virtual computer model into physical reality.  I am sure this is going to be challenging for technocrats that live and work in virtual space but it needs to happen if their work is to be appreciated, understood and made use of to its fullest extent.

3D printing offers BIM the ability to democratise design information.  Physical models allow untrained individuals to visualise and relate information spatially.  It offers a direct link from BIM model to physical artifact. 

In particular, colour 3D printing allows the overlaying of design information on to design geometry. 

New kinds of communication model need to be envisaged for communicating BIM models through the media of colour 3D printing.  Image mapping, keyed colour and shading, labeling and symbols can all be applied to present 3D printed infomodels.

Introducing the Infomodel

We have printed some great models in our time but we have not seen enough use of labeling and annotating of models.  To illustrate this, we made this infomodel to show the kind of results that could be achieved.

To find out more about colour 3D printing, visit: www.lee3d.co.uk

Colour 3D Printing for Marketing

Unique, customised, full colour 3D prints from Lee 3D add impact to events and presentations. 

Unique Gifts
3D printing allows you to make short runs of unique objects to promote your brand. 

3D Prints of Aldwych

Customised Promotional Items
We can 3D print customised objects for individuals creating a positive brand experience for your clients.

Full colour 3D printing is a must to stand out from the crowd.

We currently have a large body of colour 3D print work on show at the Design Museum (until) Jan 12 2014) where we have sponsored Adam Nathaniel Furman, Designer in Residence. 

For more information on how you can use this process to produce unique, customised colour sales and marketing tools call Lee 3D on 07563 243 891 
Visit www.lee3d.co.uk to find out about colour 3D printing

Thursday, 17 October 2013


The first series of colour 3D prints designed by Adam Nathaniel Furman for the Design Museum. All parts printed and finished by Lee 3D.

Yantrament 01
Shown in a custom made acrylic case.

Yantrament 02

Yantrament 03
Big Boss

Yantrament 04

Yantrament 05

Yantrament 06

For enquires about purchasing these pieces please contact the artist at adamnathanielfurman@gmail.com. To see the accompanying blog created alongside the parts visit - http://identity-parade.blogspot.co.uk/

For more information about colour 3D printing visit www.lee3d.co.uk

Friday, 11 October 2013

What is so shameful about plaster?

In the last few days Asda (Walmart in UK) started advertising a service to 3D scan and 3D print figurines. The press release for this seems to include the description of the process as "The shape is then recreated in 3D form by spraying ceramic fluid in thin layers to build up a solid object." - see both Guardian and Daily Mail.


I need to buy this machine!

Or do I?

It turns out a UK 3D Systems reseller is boldly stating today (11th October 2013) that it is selling Projet Colour Ceramic 3D Printers.

Now I am pretty sure that ceramic is a sexier material than plaster but is this in any way factual?

The 3D Systems MSDS for the powder that they now call VisiJet PXL Core (formerly zp151) states that between 80 and 90% of the material is Calcium Sulphate Hemihydrate. CaSO4·~0.5H2O is more normally known as Plaster of Paris.

I do not know know if the Asda press release and the UK 3D Systems reseller are connected (although the UK reseller sells both the 3D printer and the 3D scanner shown on the Asda 3D printing promotional video) but something is clearly going on here. ZPrinters now known as ProJet x60 are to my knowledge plaster based 3D printers - our ZPrinter 650, by contrast to the Asda machine, sprays water based binding fluid from printheads on to plaster based powder bed.

So what is ceramic? 

A ceramic is, according to the fount of all knowledge Wikipedia, an "inorganic, nonmetallic solid prepared by the action of heat and subsequent cooling."

The last part of this definition is important. Ceramics are formed by the transformational action of heat and subsequent cooling. I can find no connection to ceramics and plaster except that plaster moulds are used for ceramic slip casting.

The ZPrinter works by spraying a water based liquid (mostly it is water) onto a plaster based substrate (mostly it is plaster of Paris) and the part is created by in the first instance as a result of the water + plaster reaction.

As a final note the 3D Systems website Cubify is selling parts made in ceramic as Ceramix. These do look like real ceramic parts and 3D Systems do not make any claim to sell a ceramic based 3D printer to produce these parts. It is possible, though I have no specific knowledge that Cubify's Ceramix parts are printed on a legacy ZPrinter using something like ViriClay from Viridis.

And as an even more final note. Axiatec developed a cold ceramic finish to spray onto ZCorp parts. The notion of a cold ceramic precipitated heated debate among ceramicists and the general consensus was that a ceramic is not the physical characteristics of the material but the a product of a transformational process that includes heating and cooling. The Axiatec product looked and felt cold and heavy like a ceramic - Plaster parts feel like plaster parts.

Below is an image of a colour 3D print made at Lee 3D using 3D Systems plaster based powder in a ZPrinter 650 colour 3D printer. Part of a series of work on show at the Design Museum and on show until January 12 2014.

Plaster and proud of it.

(As of March 2016 - the UK 3D Systems reseller continues to sell the Projet x60 range of plaster based 3D printers as ceramic.)

To find out more about Quality Colour Plaster based 3D Printing visit www.lee3d.co.uk

Wednesday, 9 October 2013

7 reasons why plaster beats Nylon for making architectural concept models.

Nylon is better than plaster. Right?

Here are 7 reasons why plaster is better than Nylon for making architectural concept models:
  1. Nylon parts (SLS) cannot show the surface features achievable with plaster. For 2 reasons, first the economics of SLS bureaus dictate they print with the thickest layer they can get away with and second the laser spot size physically prevents printing features smaller than about 0.5mm. This makes the plaster based machine better for printing architectural facade models at scale.
    Part printed on ZPrinter 650 at Lee 3D showing 0.3mm surface detail.
  2. Plaster parts do not shrink or warp. Nylon parts are created at melt temperature which means they shrink as they cool. If the cooling is not even the parts warp and distort. Plaster parts do not warp. Model makers who know the technology have strategies for dealing with warp but it can be a problem when a model arrives for a meeting and the walls are bowing. 
  3. Plaster printing is probably 4 to 5 times faster than Nylon. That is 3 to 4 days that could be better spent designing. We often print same day jobs for clients in plaster which would simply not be possible in Nylon.
  4. Plaster printing can be full colour. Nylon can be dyed and that looks really nice - but it is a single colour. 
  5. Plaster is less wasteful. Now here is contentious issue. I really do not know the facts but if you think that a typical refresh rate applies to an SLS build of 25 - 30% then there may well be a good case for this assertion. Refresh rate means that 25 - 30% of the powder must be replaced after each build either used in a part or thrown in the bin. With the plaster system we reuse pretty much everything that does not go into a printed part.
  6. SLS is very sensitive to geometry. With the ZPrinter you can print multiple, intersecting and touching shells and sharp triangles with no loss of quality to the part. With SLS these can cause double passes of the laser that mark the surface of printed part. If you print a part with sharp triangles these appear on the surface of the part even if it should be flat.
  7. As a consequence of #6 it is much easier to model for a plaster printer. Ever tried getting rid of sharp triangles? 
In summary I would like to remind the reader that these points are made in reference to making architectural concept models.  A concept model by its nature is made quickly at a very early stage in the design process when it is often desirable to show simplified design geometry. Conversely are many reasons that model makers use SLS parts for making presentation models.

Certain technologies fit certain tasks. Nylon is not better than plaster and neither is plaster better than Nylon. Each material has its own characteristics which inform the choice of technology. I have deliberately not included cost in these points but this too will need to be part of the equation.

To find out more about Lee 3D  visit www.lee3D.co.uk

Thursday, 3 October 2013

TCT Show 2013

Each year I come away from TCT with different feelings about the industry.

This time around after being bombarded all year with media interest and social media mania, you could be forgiven for thinking that 3D printing was in a state of perpetual revolution. 

In fact TCT carries the same core professional and production machine manufacturers selling pretty much the same line up of technology at it has for the past 5 years. Seeing this was somewhat reassuring, the ground rules still apply, the industry is evolving, public perception is changing.

On my list of objectives for visiting TCT was to take a closer look at the Blueprinter and Mcor Iris machines.

These 2 machines are could not be more different but they both present technologies that can potentially upset the status quo. 

The Blueprinter uses a thermal printhead to fuse plastic powders. This technology is clearly aimed at the entry level (for now) of the SLS part market. Unlike SLS the lower temperature Blueprinter promises full recycling of materials in each build. Unlike SLS it uses a relatively inexpensive thermal printhead instead of a laser and so the cost of the systems is very much lower.

As a ZPrinter user I am attracted to this system as it is powder based and so it has all of the advantages of a self supporting system. The reality is that 3D printing machines are complex beasts and the road from proof of concept to production machine is a long and perilous one. Like SLS the printer heats the chamber to close to melt temperature. As heat is applied via the thermal printhead some energy is used to fuse powder but some heat energy will flow into the printed part and surrounding powder. Compensating for this additional localised heat imbalance on the next pass of the thermal printhead must be a significant challenge to the designers.

The Iris, full colour paper based 3D printer is built on the back of the single colour machine that is well built and probably pretty reliable. 

Adding the colour element may prove a mixed blessing. There is a difference between a single colour sketch model and a colour presentation model. You can get away with a lot of defects in the former while the latter will be seen in a much more critical light.

My hunch with this machine is that it needs some really good post processing development to transform the raw parts. 

If I were to make a criticism of ZCorporation when they developed the ZPrinter range, it would be that they were too willing to let customers do all of the innovating in the post processing realm. Their engineers were content to focus on the machine and the raw parts - they did not look closely enough at the process and the products that came out of their machines.

Well, I guess that is what we have built our businesses on.  Pre and post production to make parts that are better than the next bureau...


Walking out of TCT into the adjoining Sensor & Instrumentation show, we found an array of colour 3D prints on the Infratec stand. These are massively scaled up versions of their micro assemblies - an interesting reverse on the usually massively scaled down buildings and city models that we produce at Lee 3D. These models were apparently 3D printed somewhere in Germany. www.infratec.de/en.html

3D prints that serve a purpose always seem more interesting than demonstration pieces.

To find out more about Lee 3D visit www.lee3d.co.uk

Saturday, 14 September 2013

How It's Made - Live!


... in quiet contemplation...

... is often the reaction to seeing a 3D printer operating for the first time.

Recently, seeing a video of a Mold-A-Rama machine in action, I immediately recognised the same fascination that a 3D printer induces in the onlooker. Is this an instinctive human reaction to seeing machines making artifacts? Is this what humans do when they see new tools?  

The Mold-A-Rama, for those that do not know it is "miniaturized injection plastic factory" apparently often housed in tourist attractions. It basically blows some molten plastic into a mould. The mould releases the part and it is swept into the vending trap for collection by the customer.

Moulds of US presidents, dinosaurs, zoo and aquatic animals are available themed to the venue, though each machine makes only one object, depending on the mould that is fitted. Oh and you can still buy reconditioned Mold-A-Ramas!

So the question is what made this an attractive proposition? Why not just sell shoddily moulded toys from a factory at the other end of the earth that can produce 10,000 parts in a day - at a very small fraction of the cost.

While there are certainly some similarities between the Mold-A-Rama and the 3D printer there are also some differences. The Mold-A-Rama makes collectible items, strange but true. 

There are limited molds available and they are available at different geographical locations. So to get all of the pieces you need to go to each machine. A 3D printer on the other hand may make plastic toys but every machine with access to the files can produce them. 

The Mold-A-Rama is also able to deliver a result in about a minute, where as a 3D printer is very likely to take an hour or more for the same sized product.

For more information on Mold-A-Rama visit mold-a-rama.com
For more informatino on the Cube 3D printer visit cubify.com
For more information on Lee 3D colour 3D printing visit lee3d.co.uk

Monday, 2 September 2013

The Making of Totem

Totem was designed by Adam Nathaniel Furman for the Design Museum, Designer in Residence programme. All 3D prints were made at Lee 3D, plaster moulds were made at CP Ceramics.

3D prints being removed from bed of powder

Finished 3D prints

Close ups 

The Assembled Parts of Totem

After 3D printing, the parts were sanded to make patterns for slip casting. once fired the cast parts shrink by 15%. 

Cast parts of Totem in the moulds. 
Used 3D printed patterns can be seen on the shelf above.

The Glazed Totem

The exhibition of 3D printed forms designed by Adam Nathaniel Furman will be on display at the Design Museum,  from 4 September until 12 January 2014.

To find out more about 3D printing at Lee 3D visit www.lee3d.co.uk

Wednesday, 28 August 2013

The Making of Kitschpot

Kitschpot was designed by Adam Nathaniel Furman for the Designers in Residence programme exhibition at the Design Museum, 2013. 

3D printed Kitschpot

The original file is modeled and textured in Rhino version 5 and exported to VRML format. At Lee 3D we opened the file in Magics for checking before 3D printing on the ZPrinter 650 colour 3D printer.
Part sanded sitting on the 3D printer

Two sets were made. One was rubbed down and sent to CP Ceramics where a mould was made ready for slip casting the forms. 

As a 3D printer this image really appeals to me. Parts that come out of the printer are covered in plaster powder and need time to be revealed. These cast parts come out so clean and crisp it must be a joy to work with this process.

Slip cast Kitschpot still in the mould 

After firing, the cast forms have shrunk, which is a normal part of the ceramics process. The 3D printed pattern bears the mould makers marks used to ensure that the moulds will part easily and releasing the cast forms. The colour in the patterns is permanently faded by now, probably indicating the parts were wetted at some point in the mould making process. 

The Kitschpot pattern with offspring

Another attraction to casting from 3D prints is that many parts can be produces form a single mould.

Kitschpots sitting around waiting for glazing

The exhibition of 3D printed forms designed by Adam Nathaniel Furman will be on display at the Design Museum,  from 4 September until 12 January 2014.

To find out more about 3D printing at Lee 3D visit www.lee3d.co.uk

Monday, 26 August 2013

3D Printing for AEC – Both Feet on the Ground

For some years, those of us working in the field could sense that 3D printing was ready to metaphorically explode.  This was based on the fact that very few people actually knew of 3D printing and people’s reaction when first introduced to it was often one of wonder and amazement.  Working predominantly in the field of 3D printing for Architecture, Engineering and Construction (AEC) we naturally and perhaps naively assumed that when the boom in 3D printing came there would be a similar step change in 3D printing for AEC.

We were wrong.  The boom in 3D printing was a boom in consumer 3D printing that has unleashed the imagination if not unleashed the realities.  3D printing for AEC has steadily grown since 2005 when the Spectrum 510 (ZPrinter) was unveiled.  The 510 printed at 600 x 540 dpi with a layer thickness of 0.1mm.  It enabled us to create architectural maquettes overnight to a standard that professionals were willing to accept.

Since the first consumer 3D printer was made in 2007 by the RepRap project, there has been an exponential growth in both sales and manufacturers of these machines matched by media interest and hype.

At the same time, use of 3D printing in AEC has grown steadily but in no way exponentially.  So what is the continued growth in 3D printing in AEC actually linked to? Some factors that influence uptake of 3D printing are:

  • Improvements in technology
  • Cost
  • Increased use of 3D software tools
  • Increased competition
  • Greater awareness

Improvements in Technology
The technology most used for making building maquettes is the ZPrinter.  SLA and SLS are also largely used but usually by model makers.  FDM has never produced a good enough surface finish, while the Objet has remained too expensive and not quite up to the mark set by SLA.  SLS and SLA have hardly improved in the past 10 years both having been around for 20 years or more.

The ZPrinter technology has a variety of advantages; being quick, relatively low cost, aesthetically acceptable, full colour and easily accessible by small niche bureaus willing to spend time preparing architectural data.  Architects wanting multiple concept models at short notice make use of the speed of the technology, master planners make use of simple colouring for indicating zones, construction professionals use colour for indicating different materials or trades in a programming model.

Eight years after the Spectrum 510, the latest best in class ZPrinter (now a called a Projet by the current owners of the manufacturer) still prints at 600 x 540 dpi with a layer thickness of 0.1mm and though there have been significant steps made in raw strength and colour, the technology itself is essentially static.  This is evident in the number of 510s still in use in bureaus producing white models that are no different to the parts printed in the most up to date machine available today. 

Steady improvements in technology are consistent with the steady increase in use of the technology in AEC.

In London, bureau prices for 3D printing have remained fairly stable over the past 8 years.  Manufacturer prices have risen significantly and bureaus have frozen prices by hollowing larger models to a greater extent as the material raw strength has allowed.  Personally I began hollowing models as soon as I had software capable of doing so to make more projects viable to 3D print.

Meanwhile for those companies buying the machines for use in-house, the capital costs and consumables costs have risen.  Companies buying 3D printers have rarely bought software that enables them to efficiently hollow models and this combined with lack of internal demand it is likely that for most low use companies the cost of printing in-house is higher than that of using bureaus.

SLS prices have become more competitive with increased global capacity and the commodification of machine space. But this is still a four to five day process that most AEC professionals do not have, most of the time.

It is therefore difficult to make any case for cost influencing increased use of 3D print in AEC.
I should add a note here that competition with traditionally made models does have a part to play though not a large one.  3D printing can show geometry, some colour and can indicate surface textures but it really cannot compete with traditional models that show actual materials such as wood and steel.  3D prints can be less expensive than traditional models but as they serve a different purpose the overall effect is small.

Increased Use of 3D Software
In 2005 there were very few architectural practices with more than one or two 3D modelling specialists who were usually engaged in visualisations for early stage design work and competition work.  Most high end visualisations were outsourced and probably still are. Increased use of 3D software in university departments with 3D printers has underpinned the steady increase in use of 3D printing in architecture.

SketchUp has been around since 2000 and is one of the most widely used modelling tools for making AEC 3D printed models.  The reason for this is that it is good for making initial design models at a stage in the design process where 3D printing has most value in developing designs and as a communication tool in persuading clients.  Nobody can pretend that SketchUp is a tool made for 3D printing.  Often SketchUp models are difficult and time consuming to prepare but their very existence has contributed enormously to the number of 3D printed models made.  The question remains, has there really been increased use of SketchUp over the past 8 years?

BIM has long been seen by the 3D printing community as a double edged sword.  Because BIM is a fully 3D environment it creates favourable conditions in which 3D printing can thrive, but BIM models suffer from a need for editing.  The first kind of editing is where details are not resolved. It seems to be in the nature of BIM models that where construction details are unresolved, parts are left floating in space.  Someone needs go through the entire model and make mullions, cladding and structural elements touch so that the 3D printed model does not fall apart.  The second kind of editing that needs to be done is to remove information that is not relevant to the model.  3D printing at scale, BIM models are notorious for containing pointless information.  Doors with hinges and screws and the manufacturer’s name pressed into the hinge would not be an exaggeration.  At 1:500 a 50mm door needs to be thickened from 0.1mm to 1mm and we do not have any place in the file for ironmongery.

However the use of BIM has created roles throughout the construction industry that require professionals to work in 3D.  This has to be good for the future of 3D printing but perhaps has not had a significant effect in the past.

There has undoubtedly been an increase in use of 3D software and this has probably had an effect on increasing the amount of 3D prints made.

Increased Competition
With the downturn in the economy that occurred in the late 2000s, winning scarce jobs undoubtedly stimulated increased use of 3D prints to help bid teams to win new work.
Similarly reduced numbers in architectural practices led increased outsourcing of sketch models.

At the same time there has been less work about and consequently less money for experimentation and non-essentials.

In London, the increase in the number of bureaus offering 3D printed models has led to a greater number of industry professionals trying out the process.  This is not quite the same as the number of businesses becoming repeat customers and incorporating 3D printing into their workflow.

Increased competition has led to an increase in the number of 3D printed models being made.

Greater Awareness
As a consequence of intense media hype and a proliferation of 3D print based companies there are not many individuals that have not now heard of this miraculous new process.

It is often observable that when a project involves a 3D Print at an early stage, 3D printing is used extensively throughout.  My reading of this is that when developers see 3D prints they encourage their use. In many ways it is developers that will be responsible for demanding 3D prints.  I have written previously about the ability of 3D printing to democratise design communication and it is usually in the interests of developers to have their projects communicated clearly and to the widest possible audience.

Greater awareness has led to a wide range of customer expectations of what is possible and also is certainly responsible for increased use of 3D printing in AEC.

The real work that needs to be done in order to incorporate 3D printing into the AEC workflow is challenging.  One often needs to remind oneself that the Sydney Opera House was built not just without 3D printing but without the CAD packages we are used to today (computers were used for structural analysis). Architects have been designing buildings for many years without the need for 3D printing or CAD or BIM or computers.

However all of these tools have advantages that generally outweigh the disadvantages.  The fact is that 3D printing is not needed for all stages of every project.  In fact some projects just do not need 3D printing and never will.  There are many reasons for 3D printing to be used between conceiving and selling a completed building to its end user and it is for architects, engineers and construction professionals and all of the conditions listed above to be right in order for use of 3D printing to grow.

So amid the intense hype surrounding 3D printing we need to keep our feet firmly on the ground and keep making models that serve a purpose in bringing buildings to completion.

For more information on 3D printing for AEC visit www.lee3d.co.uk

Tuesday, 13 August 2013

Anatomy of an STL File

Quick Summary of main points on this post:

  • STL files desicribe a mesh of triangles with no other information.
  • STL files to not contain unit information
  • STL files do not contain colour information*
  • Some applications always export STL files in feet or inches
  • 2 types of STL file - Always choose Binary as it makes smaller files.

STL is sometimes called Stereolithography Format or more properly Standard Tessellation Language. It has become the default file format for many single colour 3D printing applications. It was developed by Chuck Hull soon after he built the first 3D printer and the first specification was released in 1988.

Below is an example of the contents of an ASCII STL file that partially describes a cube drawn with one corner at the origin and with sides of length 1 unit. It is worth pointing out that there are two types of file STL format: binary and ASCII. They contain exactly the same information but the ASCII files will be much larger as they contain human readable text while the binary file is essentially a compressed version.

It can be seen that the contents of the STL file is very simple, it is essentially a list of coordinates that describe triangles and the normal that describes which direction each triangle faces. STL files describe a mesh of triangles with no other information. When exporting STL files, vector and nurbs based software may ask for a tolerance which is used to calculate how many triangles are needed to describe curved surfaces. Generally, the lower the tolerance the closer the triangles match the curve of the geometry and the more are required to create the form and so the large the file becomes.

One important thing to understand about an STL file is that it does not contain any information on what units are being used. The units used in any particular file will be determined by the software that was used to export the STL file. Usually the units will reflect the units that were used to model the geometry so if the model was modelled in meters the units will be meters, if the model was modelled in mm the units will be mm. However some software and export plugins that were written in the US may default all exported STL files to inches or possibly even feet.

  facet normal -1.000000e+000 -0.000000e+000 -0.000000e+000
    outer loop
      vertex 0.000000e+000 0.000000e+000 1.000000e+000
      vertex 0.000000e+000 1.000000e+000 1.000000e+000
      vertex 0.000000e+000 1.000000e+000 0.000000e+000
  facet normal -1.000000e+000 0.000000e+000 0.000000e+000
    outer loop
      vertex 0.000000e+000 1.000000e+000 0.000000e+000
      vertex 0.000000e+000 0.000000e+000 0.000000e+000
      vertex 0.000000e+000 0.000000e+000 1.000000e+000
  facet normal 0.000000e+000 1.000000e+000 0.000000e+000
    outer loop
      vertex 0.000000e+000 1.000000e+000 1.000000e+000
      vertex 1.000000e+000 1.000000e+000 1.000000e+000
      vertex 1.000000e+000 1.000000e+000 0.000000e+000


  facet normal 0.000000e+000 0.000000e+000 -1.000000e+000
    outer loop
      vertex 0.000000e+000 0.000000e+000 0.000000e+000
      vertex 0.000000e+000 1.000000e+000 0.000000e+000
      vertex 1.000000e+000 1.000000e+000 0.000000e+000

In the future the STL file format is likely to be around for some years to come but in time it is likely that the format will be replaced by something like the AMF (Additive Manufacturing Format). Intitially known as STL 2, AMF is being designed on the XML (Extensible Markup Language) platform which means that it is flexible and can contain both geometry and additional information such as multiple materials and colours etc.

Since writing this in August 2013 a consortium of industry leaders has been formed to develop a new format known as 3MF. AMF has been pretty much dropped even though it became a viable format. As of March 2016 we have still not received an AMF file or had any enquiry about this format. 

These new formats are attempting to address the needs of 3D printing, additive manufacturing, copyright issues and other rights to reproduce. It is going to be complex and less accessible - I for one appreciate the simplicity of the STL file. "Simple is efficient" as my engineering drawing teacher used to say.

*In fact binary STL can support colour, if we save an STL from Magics we can open it again in Magics preserving the colour information. However the important point is that all of the STL file exporters written for 3D modelling software like Rhino, Revit, MicroStation, SketchUp etc do not export colour.

For further information about Lee 3D visit www.lee3d.co.uk