reprapmagazine issue 3 lowres

8/12/2019 RepRapMagazine Issue 3 Lowres

1/83

WWW.REPRAPMAGAZINE.COM

Practical exploration of 3D printing

Interviews - Pro tips - Reviews DIY projects - Tutorials - GalleriesJune

2013

Issue2

REPRAP MAGAZINEWWW.REPRAPMAGAZINE.COM


Interviews - Pro tips - Reviews DIY projects - Tutorials - GalleriesMarch

2014

photoby

KristofVrancken

Issue3

Paste extrusion - Hot-ends review - QuadRap buildConceptFORGE - Morgan instructions


2/83

REPRAP MAGAZINEWWW.REPRAPMAGAZINE.COM

Happy anniversary!

This issue marks the passage of the rst year of existence forour magazine. We know its been a while since our last issue but be-tween these two issues there was one special event that took many ofour eorts to organize and it took place lastSeptember in the TCT + Personalize show inBirmingham.

It was a great way to meet some of ourreaders, but when it was over we resumedback the eorts and here it is the third is-sue, with two features articles: Hot-Ends and

Paste extrusion.

In this issue we focus theIn the workssection on the work of Nicolas Seward andhis project Concept Forge. For those whohave been following his work you will see amore personal perspective of his works.

Enjoy this issue and stay tuned as we are working on deliveringsome news this year!

Paulo GonalvesEditor

Web:

www.reprapmagazine.com

Magazine team

Developers

Paulo GonalvesGary HodgsonRichard HorneViktor DirksScott Latheine

Contacts

[email protected]

[email protected]

To the readers

We want to have a close re-lationship with our readers. Forthat we encourage you to par-ticipate in this project. Send usphotos of your best prints andyour setup for possible publica-tion to our email [email protected]. Also take part at the dis-cussion at the http://forums.re-

prap.org/list.php?305.

To the contributors

This is an open magazine,and for that we encourage youto submit your articles for pos-sible publications to our email [email protected] you are also a developer of atool that RepRap users use youcan also send an email to bein our database for future con-tacts.

Independent

We are 100% independent.The manufacters of the prod-ucts featured do not determineour content nor our opinions.

Our mission

Reprap Magazine

RepRap Magazine


RepRap Forums
http://reprapmagazine.com/https://www.facebook.com/pages/Reprap-Magazine/291706147625408https://plus.google.com/+RepRapMagazinehttp://forums.reprap.org/list.php?305http://forums.reprap.org/list.php?305http://forums.reprap.org/list.php?305https://plus.google.com/+RepRapMagazinehttps://www.facebook.com/pages/Reprap-Magazine/291706147625408https://plus.google.com/+RepRapMagazinehttps://www.facebook.com/pages/Reprap-Magazine/291706147625408http://reprapmagazine.com/


3/83

Contents

This issues team

Follow us, and get in touch, atwww.reprapmagazine.com

4 Community newsBook review

Web-resources Print to improve life competition results Campaign for a universal lament spool

standard

11 In the works Nicholas Seward works

17 TCT Show+PersonalizeRepRap Community hub review

Feature26 Paste Extrusion41 Hot-Ends review

60 DIY Factory

QuadRap

66 Beginner space

RepRap Morgan visual instructions

Shauki Bagdadi Dries VerbruggenRichard HorneGary Hodgson Paulo Gonalves
http://reprapmagazine.com/http://reprapmagazine.com/http://unfold.be/pages/projectshttp://quadrap-3d-printer.blogspot.de/http://unfold.be/pages/projectshttp://reprapstyle.blogspot.pt/http://garyhodgson.com/http://richrap.blogspot.pt/http://reprapmagazine.com/


4/83

Community news

Open-source lab will focus its readerson building his own laboratory hardware toreduce research costs. The book will startby introducing the readers to a brief resumeof what is Free and Open-source Software(FOSS) and from there takes it to the freeand open source hardware (FOSH) andpoints some of the potential when they are

combined.

Divided in 7 chapters, the rst beingthe introduction mentioned before, it takesits readers progressively on the benets ofsharing and the open licensing (chapters2 and 3 respectively). These chapters areinteresting enough for even those who havebeen part of the Open source movement forsome time now, but are really interesting

for people who are just starting to join thismovement.

The reader will be introduced to theseveral licenses and chapter 3 will end talk-ing about the Continued IP Challenges andits relation with open-source licenses anddevelopment.

After this rst 3 chapters of more theo-retical content we get to the hands onpart of the book! First we will start withthe open-source micro controllers, with theavailable family and then focus on Arduinowith a practical example the polar bearopen-source environmental chamber (chap-ter 4).

Getting to chapter 5 we reach theRepRap project. This chapter shows theassembly of a printer, a Prusa Mendel reen-gineered primarily by Jerry Anzalone atMichigan Tech. After this chapter, the authorfocus on Digital designs and Scientic hardware taking us through out many dierentscientic tools created with the things re-viewed on the previous chapters and thebook ends with some thought on the Futureof Open-Source Hardware and Science.

Overall this is a book which isfocused on the development ofLaboratory equipment, mainly us-ing open-source hardware such asRepRap 3D printers and Arduinomicrocontrollers. It will guide theinexperienced reader making himcomfortable to embrace this tech-nological and social advances in a

very practical way, resulting in verysignicant cost reductions for re-searchers and teachers.

OPEN-SOURCE LABby J.Pearce

Author: Joshua M. PearceRelease Date: 07 Nov 2013Imprint:ElsevierPrint Book ISBN :9780124104624eBook ISBN :9780124104860

Pages: 300Dimensions: 235 X 191

4REPRAPMAGAZINE
http://store.elsevier.com/product.jsp?isbn=9780124104624&pagename=search


5/83

Community news

Design and print bracelets.

Nervous System has a newonline app that lets you custom-ize three bracelets templates and

download them as .STL or print us-ing their services.

The Kinematics app is the onlyone with the option to download.STL les, but there are other inter-esting apps available if you want touse their printing services.

New community sharing site.

OpenBuilds has unveiled theirnew platform for opensorce designsharing with a focus on OpenSorcemachines.

It has several categories ofmachines going from 3D printers,to laser, vacuum forming and scan-

ners. It is also possible to uploadother projects and the site includesa forum.

Repetier frmware new tool.

Repetier has a new online toolthat allows its users to congurate

Repetier rmware in a much moreuser friendly enviroment.

Like they mention on the web-page: In this congurator we tryto hide all information and settings,that are not relevant for your print-er, based on previous answers.

http://n-e-r-v-o-u-s.com/kinematicsHome/

http://www.openbuilds.com

http://www.repetier.com/rmware/

5REPRAPMAGAZINE
http://n-e-r-v-o-u-s.com/kinematicsHome/http://www.openbuilds.com/http://www.repetier.com/firmware/http://www.repetier.com/firmware/http://www.openbuilds.com/http://n-e-r-v-o-u-s.com/kinematicsHome/http://www.repetier.com/http://www.repetier.com/firmware/v091/http://www.openbuilds.com/http://n-e-r-v-o-u-s.com/kinematicsHome/


6/83

Community news

Print to improve life quality!

Competition Results

RepRapPro Huxley hardware Kit.Shipping included.

1 Prize Winner

Name: Benjamin LeclercAlias: proto.user

Customizable hand crank device.With a splined output shaft like on atractor, and a curvy crank.

We asked our readers tocome up with new designs toimprove the life quality, andhere are the results of thischallenge!

A special thanks goes to those who

participated in our rst competition andalso for the competition sponsors.

6REPRAPMAGAZINE
http://www.reprappro.com/http://www.reprappro.com/http://www.reprappro.com/http://www.reprappro.com/


7/83

Community news

Plastic2Print speciality plasticsbundle. Shipping included.

RepRapWorld bundle.Shipping included.

3 Prize

Name: Dave AkkermanAlias: Wavetracer

A shopping-trolley coin and a clip tohold your shopping-list.

2 Prize

Name: Roelof GrootjansAlias: RoelofG

Automatic window opening robot.

7REPRAPMAGAZINE
http://www.plastic2print.com/http://www.plastic2print.com/http://www.reprapworld.com/http://www.reprapworld.com/http://www.reprapworld.com/http://www.reprapworld.com/http://www.plastic2print.com/http://www.plastic2print.com/


8/83

Community news

Faberdashery four 50 vouchers

worth of plastic.

4 to 7 Prize

4 - Gert Galjoen (ggaljoen) Kitchen robot add-on

5 - David Schulze (unasone) Belt Clip

6 - Peter Fisher (stonesher) Modular Water Screw7 -Benjamin Minerds(PuZZledDucK)

White cane

4

6

5

7

8REPRAPMAGAZINE
http://www.faberdashery.co.uk/http://www.faberdashery.co.uk/http://www.faberdashery.co.uk/http://www.faberdashery.co.uk/


9/83

Community news

Coils and coils of lovely 3D printinglament, Im usually surrounded by them,I struggle with them and often spendtime and energy mounting, clamping,switching and feeding them into variousprinters.

There has to be a better way. (letsplease all nd a better way)

The problem

Every single one is dierent, bare coilsor reels, some are even dierent being fromthe same manufacturer.

Just to point out here, this is only a tinyselection of 3D printing lament reels avail-

able, these are the ones I happen to stillhave around me. Even the long list onRepRap.org of lament suppliers is a frac-tion of whats actually available.

The closest thing we have to a stand-ard reel is based around the use of 3mmplastic welding rod/coils for automotiverepair industry, these very large coils havebeen used with 3D printing for many years.

They have a lot of benets, but also plentyof negative aspects. Mostly being heavyand big whilst also using more than 300g ofABS to make each one.

Universal 3D printing lament spool standard 2014Richard Horne campaign with the support of the 3D Printing Association.

Dierent coils and reels

9REPRAPMAGAZINE


10/83

Community news

The Question?

Would it be possible for manufacturersto have a set lament standard for spoolsize, shape and mounting?

Can the 3D printing community helpdevelop that standard?

Can we make it Eco-friendly and recy-clable?

Does anyone want it?

Lets nd out. If you want to get involved(and please do), a discussion thread on the

RepRap Forum has started here.

Start discussions wherever you want,and spread the word, lets make somethinghappen.

Ideas spark more ideas!

We can see that many reels alreadyhave a 50mm mounting hole diameter, so

this looks like a reasonable starting point.

The inner diameter, bigger is better,but you need to allow enough room forlament without also having a very largeouter diameter. A size of around 120mmfor the internal seems to work well for mosttypes of plastics and lled materials in both3mm and 1.75mm sizes.

A diameter of 220mm for the outer will

allow for a typical quantity of lament de-pending on the width of the spool, 75mmshown here.

Cardboard spool concept, 1Kg spool or3 x 250g based on the same specication ofmounting hole and overall size.

The plastic dust cover/case is an item Iwould also like to see available (or 3D print-ed) - make it easy to mount on / in a 3Dprinter and rell with the recyclablecardboard lament spools.

How you can help

Ask your lament supplier if they cansupport the standard, if they say no, askthem why and let people know the reasons.

If you manufacture a 3D printer, what doyou require from a lament spool?Sketch or design something, say whatwould work for you and your 3D printer,whats not working now and how you solveit (or not).

The 3D Printing Association is providingsupport for this project and also looking forfeedback and ideas to help with this cam-paign, both from lament manufacturersand everyone in the community.

Cardboard spool concept

1REPRAPMAGAZINE
http://forums.reprap.org/read.php?262,324209,page=1http://forums.reprap.org/read.php?262,324209,page=1


11/83

Many in the Maker community are always look-ing for new and interesting projects to follow andperhaps take part in, and so this section gives abrief look at up-and-coming projects from in andaround the RepRap community.

It should be noted that unless otherwisestated these projects are most denitelyworks in progress and not ready for gener-al consumption.

If you like to live on the cutting edge then theseprojects might be for you.

by Gary Hodgson

In

thewo

rks


12/83

In the works

With a Bachelor of Science in Mechani-

cal Engineering from the University of

Arkansas, which is soon to be complementedwith a Master of Arts in Physics, Nicholas

is no stranger to mechanical design. How-ever the feedback and positive critiquevia the forums helped dene and develop

the designs, a fact he happily relates: the

RepRap community has drastically changed

the direction of these projects multiple

times. I can not take full responsibility for

the Wally or Simpson. They truly belong to

the community.

The Wally and Simpson printerdesigns of Nicholas Seward showshow eective open-source, rapid

prototyping can be. Within a monthof introducing his experimental delta

design on the RepRap forums [0] thedesign had been discussed, modi-ed and improved into a workingmachine already able to producevery respectable prints.

Nicholas motivation drew from early

experiences with his rst printer, an LC

Printbot, which he received as a birthdaygift from his wife. Whilst admitting that the

Printbot is great, he immediately recog-nised ways in which he would have designedit dierently. Once the machine was running

he slowly started tweaking the printer toremedy the various issues he encountered,resulting in a machine quite dierent to the

original. This process, of tackling problems

and tailoring the design to meet his own

needs, was instrumental in driving him togo on to produce his later work, he says,In retrospect, without the initial problems I

probably wouldnt have been as motivated

to take on all that I have.

Simpson after initial prints

1REPRAPMAGAZINE
http://garyhodgson.com/


13/83

In the works

The positive reception of the printerdesigns led to thoughts about starting a

business. Rather than rushing out and cre-

ating a Kickstarter campaign straight away,Nicholas decided, quite sensibly, to test the

waters by oering 20 beta units of eachprinter. The creation and delivery of these

units is still under way, but the experience

has already taught him that any furtherattempt at printer sales would be a bad

business move:I could print custom parts

for people for three times as much money.

I am now targeting a single arm SCARA

design that can be completely laser cut.

I have it mostly designed in my head.

He also admits that he sucksat pro-duction and took ak from some in the

community after being behind in the betarun and yet still producing a modication to

the Simpson design, codenamed LISA.

To alleviate the burden Nicholas intro-

duced discounts for those wishing to receivethe beta kit without printed parts, or who

would print parts for others in the scheme,reviving a long-held RepRap idea

of print it forward. The dicul-

ty and feasibility of small-scalemanufacturing is highlighted

as this beta production run willactually cause a slight loss for

Nicholas. Ultimately the biggest

loss is that of his time takenaway from working on new

designs. His goal now is to honourthe commitment so he can get

back to being a mad scientist

as he puts it.

The mad scientist personahas resulted in a profusion of

derivative designs. Alongside

several interesting mechani-cal variations of the Simpson

mechanism, which ultimately ledto the GUS Simpson, there has

since been a threaded-rod based delta bot

(LISA), which followed the incredibly rapidprototyping pattern of the original Simpson,

by being turned from idea to reality, andproducing its rst print, in only three weeks

[1].

Another is the wildly experimental

design called the Sextupteron [2], whichintroduces six degrees of freedom to the

hot-End. The key limiting factor to the lat-

ter is the electronics, something he hopesto remedy in the near future. Another

Simpson derivative is BOB, which stands forBoltless and Bearingless, an attempt to

reduce the vitamin count as much as pos-sible. Nicholas goes on to explain, he willbe a novelty but he will be all plastic except

for the hot end, wiring, steppers, string,controller, and possible the bed. I want to

explore ways to actually print the bed out of

higher temp plastic or print on a surface likea table that the printer sits on. I am viewing

this more as a learning experience thandesigning a serious printer.

Simpson

1REPRAPMAGAZINE


14/83


15/83

In the works

other robotic creations [3], but also givesan insight into what tools and techniques

are required for taking on such projects,

from a variety of backgrounds and experi-ence levels.

He laments that, whilst there are

several free to use CAD programs, thereare no viable FOSS (Free and Open Source

Software) options, particularly

web-based ones, and this is some-thing he would like to tackle.

The biggest problem in mybook is modeling software

and that is followed by theavailability of tools. If I canprovide FOSS web enabledmodeling software that can

plug into my parts catalog andgo with the distributed manufactur-ing then I remove both of those hur-dles to collaborative OSHW.

All of these ideas are certainly ambi-

tious, touching on the many facets ofmodern digital manufacturing, and theviability of such a far-reaching project is

open to debate, as Nicholas himself says,who knows what will happen. These could

all be pipe dreams..

But the fact that such projects can

be considered, promoted, discussed, and

championed within the maker community

hints at the exceptional opportunities avail-able to us.

[0] http://forums.reprap.org/read.php?178,206458,page=1[1] http://www.youtube.com/watch?v=N7XVxtgHHhY[2] http://www.youtube.com/watch?v=2kXdsU2bBp0[3] http://youtu.be/vnXVperYk3o http://youtu.be/0NXDAEupfOA

RepRap Wally

1REPRAPMAGAZINE
http://forums.reprap.org/read.php?178,206458,page=1http://forums.reprap.org/read.php?178,206458,page=1http://www.youtube.com/watch?v=N7XVxtgHHhYhttp://www.youtube.com/watch?v=N7XVxtgHHhYhttp://www.youtube.com/watch?v=N7XVxtgHHhYhttp://www.youtube.com/watch?v=2kXdsU2bBp0http://www.youtube.com/watch?v=2kXdsU2bBp0http://www.youtube.com/watch?v=2kXdsU2bBp0http://www.youtube.com/watch?v=2kXdsU2bBp0http://www.youtube.com/watch?v=2kXdsU2bBp0http://www.youtube.com/watch?v=2kXdsU2bBp0http://www.youtube.com/watch?v=2kXdsU2bBp0http://www.youtube.com/watch?v=2kXdsU2bBp0http://www.youtube.com/watch?v=2kXdsU2bBp0http://www.youtube.com/watch?v=N7XVxtgHHhYhttp://www.youtube.com/watch?v=N7XVxtgHHhYhttp://forums.reprap.org/read.php?178,206458,page=1http://forums.reprap.org/read.php?178,206458,page=1


16/83

The RepRap Magazine is open for your suggestions for content thatmight interest the community.

Its just as simple as sending an email to start the conversation orpost on the community forum.

We are waiting for your contributions!

[email protected]

Do you have a cool project to showcase?

A new develpment that works great?

You just went to a RepRap community event andhave its review to share?

Follow us, and get in touch, atwww.reprapmagazine.com
http://reprapmagazine.com/http://reprapmagazine.com/http://reprapmagazine.com/http://reprapmagazine.com/http://www.think3dprint3d.com/


17/83TC

TSHOW

2013

The TCT Show + Personalize 2013 took place inSeptember, and our magazine had the opportunityto be part of it due to our involvement in organizingthe RepRap Community Hub.

By Paulo Gonalves


18/83

TCT Show + personalize 2013

We have only one wordto describe the two daysthat lasted the TCT Show:

Amazing! When we rst started preparing thestand for the show the rst thought we hadwas how can we ll out so much space,but after a couple of months of hard workand a few minutes after the opening of theshow we realised we could have used a lit-tle bit more space!

The reality is that the RepRap Commu-

nity Hub was swamped with people duringthe two days of the event making reallyhard for everyone on the booth to step outjust enough time to grab a cup of coee,and for that we want to thank to all ourreaders who visited us during the show.

Such success couldnt be achievedwithout a great team that got involvedmaking the RepRap community booth reallyinteresting and probably one of the boothsthat printed more stu during the show.

At the date of the TCT Show the maga-zine was yet to complete its rst year ofexistence, but it was truly inspiring to realiseduring the conversations we had with ourreaders that so many of the members ofthe RepRap community already knew aboutit.

MAG

A special thank you to the TCT Show + Per-sonalize for making it possible to do theRepRap Community Hub.

The view from inside the booth

1REPRAPMAGAZINE
http://www.tctmagazine.com/http://www.tctmagazine.com/http://www.tctmagazine.com/http://www.tctshow.com/http://www.tctmagazine.com/


19/83


20/83


Tony and Roland brought two oftheir machines, the M90LC. This ma-chine is a laser cut version of theMendel90 that features a full encloserthat comes really helpfull for thoselarge ABS prints.

I have been following their blog forquite a while and it was really interest-ing to see their new printers printinglive and we can say that a lot of work isput up into the details of this printer.

www.think3dprint3d.com

Above: Tony (left) and Roland (right)with their new printer.

Left: Roland talks about the M90LC

M90LC printer

2REPRAPMAGAZINE
http://www.think3dprint3d.com/http://www.think3dprint3d.com/


21/83


I have to admit that the Active3Dwas one of the very pleasants suprisesI had during the show. Personally I didnot know anything about their printerbefore the preparation of the showstarted and was very curious to see it

in action, and once I saw it I have tosay it produced really great prints.

Their printer named ARA-X hasa very strurdy frame, a Perspexbuild surface and a build envelop of270x220x220mm.

www.active3d.co.uk

Active 3D printer - Ara-X

Active 3D print in laywood lament Active 3D printers on display

2REPRAPMAGAZINE
http://www.active3d.co.uk/http://www.active3d.co.uk/


22/83


23/83


Hammish Mead

Havent you heard of the Opensource Time Machine yet? WheelHamisnh is the creative mind be-

hind this cool project.

OTM is an open source watchmodule that has a strong focus onusing o-the-shelf components, andit looks great!

github.com/hairykiwi/OTM-02

Slic3r

We also had the pleasure to be able to have Al-lessandro present during the two days.

It was very interesting to talk with the man behindone of the most used pieces of RepRap software. Hepresented some of the features he released on theversion 1.0 of his slicer Slic3r, and always had onemore advice for the stand visitors.

http://slic3r.org

NopBotShop

Well the name Nopbotshop may be stranger toyou but make no mistakes they are a major refer-ence in RepRap developments.

It was a pleasure to have Mary and Chris (Nop-head)on the booth. They brought the Mendel90printer that Chris designed from scratch, and if youcheck his blog (Hydraraptor) you can see for yourselfthat there are not that many printers out there socarefully designed and tested from the ground-up.

http://hydraraptor.blogspot.pt

Left to right: Hamish andAlessandro

Left to right: Chris, Maryand their Mendel90

NopBotShop Mendel90

Hammish 3d printable watch (OTM)

2REPRAPMAGAZINE
https://github.com/hairykiwi/OTM-02http://slic3r.org/http://hydraraptor.blogspot.pt/http://hydraraptor.blogspot.pt/http://slic3r.org/https://github.com/hairykiwi/OTM-02


24/83


They had a stand right next to uswhere they showed their printers and

their range of produts and solutions.Like us their stand was also packedwith visitors during the show.

They also had a presence on theCommunity stand where they demon-strated how to build a Huxley kit fromscratch. True be told they spend mostof the time multitasking between theHuxley assembly and answering hun-dreds of questions. It seemed thatthe visitors did not want to give himenought time to build the printer!

http://www.reprappro.com

Richard Horne was not only in the booth rep-resenting the magazine and showing is 3DR print-ers but was also on the main stage giving a veryinteresting presentation about his path as a makerinvolved in the RepRap Community.

Also on the show, and right next to us wasthe team from RepRapPro with the stand alwayscrowded like the Community Hub which made thatback corner the most alive place in the show.

Ian building a Huxley printer

Richard Horne on stage.

http://www.youtube.com/watch?v=KT5frWkqaW8

RepRapPro Team

2REPRAPMAGAZINE
http://www.reprappro.com/http://www.youtube.com/watch?v=KT5frWkqaW8http://www.youtube.com/watch?v=KT5frWkqaW8http://www.reprappro.com/


25/83


And for 2014?

A great show is being prepared forthe 2014 edition. This year it will be a 3

day show, at NEC Birmingham from Sep30th - 2nd October.

The 2014 Show Highlights Include:

-200 Exhibitors covering the entirespectrum of 3D printing and additivemanufacturing machines in action as wellas supporting technologies and softwarefor scanning, digitizing and inspection;

and for design and manufacturing.

-Three-day conference programmefeaturing a wide range of internationallyrecognised keynote presentations fromindustry leaders, evangelists, commentatorsand super users who shine the spotlighton the technologies and signpost howthey will develop in the future. ConrmedKeynotes to date include Wilfried (Fried)

Vancraen, CEO, Materialise; DrHans Langer, CEO, EOS; andBre Pettis, CEO and JennyLawton, President, MakerBot.

-Industry Focused Seminarsessions include Inspection,Digitising & Metrology; Soft-ware for Product Development;and Additive Technology & 3D

Printing Introduction.

-Start-Up Zone -a dedicatedarea for start-up companiesand entrepreneurs looking toshowcase their products andservices to the growing market.

-RepRap Hub - a dedicated space forthe makers and hobbyists out there toshowcase recent developments from thecommunity as well as the chance to net-

work.

-Brightminds UK - giving 300 schoolchildren the chance to get hands-onexperience of CAD, 3D Scanning and 3DPrinting technologies in a dynamic work-shop setting. Held in partnership with 3DSystems Corporation and Black CountyAtelier.

The show is free to anyone, all youhave to do is to register here:

http://www.tctshow.com/register.html

2REPRAPMAGAZINE
http://www.tctshow.com/http://www.tctshow.com/http://www.tctshow.com/


26/83

The road to better paste extrusion.

By Dries Verbruggen

photo

byU

nfold


27/83

FeaturePaste Extrusion

I started working on paste extrudersat our design studio, Unfold, in late 2009because we were pursuing the idea of 3D

printing ceramics by extruding tiny coils ofclay, a process very similar to a traditionalcraft technique called coiling. Its a form ofsemi-industrial craft production that projectsthe past history of traditional techniquesinto the future. Because of its ability to pro-duce such ne layers, new forms are possi-ble.

The Reprap was born with a paste ex-truder.

When we began we were still rmlyin the Darwin age (the rst-generationRepRap) and there were barely any com-mercial RepRap-derived kits available. Therewas a fair amount of talk in the RepRap wikiand forum on paste extruders, mainly as amethod for printing support material, con-ductive materials for circuits and motors,and hot-end ceramic insulators. One of the

goals for RepRap v2 (later born as Mendel)was to have true multi material capabilities

A paste extruder is a type of ex-truder that extrudes material in theform of a paste rather than melt aplastic lament. Typically this is acold process where the printed

object solidies by air-drying.Examples of paste materials includeceramic clay, silicone, resins, play-doh,cookie dough, bio-material, icingsugar, etc. One could even printchocolate or wax with a modestly-heated extruder.

This may sound like its only

useful for hobbyist applications andnot for serious use. But have heardabout all the research now goinginto 3D printing biological materials,organs, cells, and tissue? Basically,paste extruders!

photo by Unfold

photo by Unfold

Unfold blog: unfoldfab.blogspot.com

Unfold website: unfold.be

2REPRAPMAGAZINE
http://unfoldfab.blogspot.pt/http://unfold.be/pages/projectshttp://unfold.be/pages/projectshttp://unfold.be/pages/projectshttp://unfoldfab.blogspot.pt/http://unfoldfab.blogspot.pt/http://unfold.be/pages/projectshttp://unfold.be/pages/projects


28/83


so it could print more of its own parts. (Asmentioned, the RepRap project started witha paste extruder before lament extruderstook precedence. If youre interested in theback-story, check out the announcement onthe RepRap blog from March 23, 2005.

The photo of the original prototype a smallsyringe pump - is no longer there, but youcan nd it using the Internet Archive Way-back Machineif youre deeply curious.)

The general interest in paste extrusiondied down quickly in the RepRap communitybecause such exciting progress had beenmade in Fused Filament Fabrication. FFFproved far superior for producing mechani-cally-sound parts. Also the need for supportmaterial was reduced by designing partsthat could be printed without any support.Eventually people lost interest in the devel-opment of inferior paste extrusion.

The RepRap potential as a serious 3Dprinter has grown exponentially. Today thereare dozens of professional RepRap-derivedcommercial 3D printers and kits kickstartingthe home 3D-printing revolution. This explo-sion of public interest in 3D printing has hadthe further eect of diverting eort fromAdrian Bowyer original noble goal of a fullyself-replicating machine, and towards the(still quite noble) goal of developing pro-

fessional-grade 3D printers that are more

sturdy, standardized, easy touse, and maintenance-free.

I was never personallyinterested in the challenge

of developing hardware, andfrankly I lack the skills to doso. What excites me is thenew form language creat-ed with the tool and its appli-cations and implications. Outof necessity Ive continuedresearching paste extruderson and o, and Ive learnedto enjoy it.

After a few years of relative neglect,lately Ive seen a revival of interest in pasteextruder development. My sense is thatthe low-hanging fruit in the development ofextruders, frames, and drive mechanismshas been pretty-well picked, so develop-ers have begun to look for new challengingventures, especially in materials, fabricationprocesses, and frame design. This is veryexciting and hopefully we will soon makelarge strides in developing RepRap paste

extruders. And this is necessary because -as I have experienced by mainly using themin the last three years - the issues are morechallenging than they initially seem.

In this article I want to give astate of paste extruder develop-ment from the perspective of anon-engineer, go over the main

principles with their pros andcons and show examples of var-ious types of extruders.

In this article I want to give a state ofpaste extruder development from the per-spective of a non-engineer, go over themain principles with their pros and consand show examples of various types of ex-truders.

Syringe pumpphoto by Adrian Bowyer

2REPRAPMAGAZINE
http://web.archive.org/web/20070904115030/http://blog.reprap.org/2005_03_01_archive.htmlhttp://web.archive.org/web/20070904115030/http://blog.reprap.org/2005_03_01_archive.htmlhttp://web.archive.org/web/20070904115030/http://blog.reprap.org/2005_03_01_archive.htmlhttp://web.archive.org/web/20070904115030/http://blog.reprap.org/2005_03_01_archive.htmlhttp://web.archive.org/web/20070904115030/http://blog.reprap.org/2005_03_01_archive.html


29/83


1) Stepper-Driven Syringe Pump

As the name implies, a Stepper-DrivenSyringe Pump (SDSP) uses an electric mo-tor to push a plunger down a syringe bar-

rel. This is sometimes called Direct Drive.Here we have various options. The Fab@Home Model 1 Syringe Tool uses a priceylinear stepper. Other current models usethe more standard rotational stepper withsome extra mechanics like a gear train.Early examples of this approach include theFab@Home Model 2 Syringe Tool, the abovementioned syringe pump by Adrian Bowyer,Zach Hoekens Frostruder MK1, and Viktors(VMX) Syringe Tool. More recently theres

Davids Paste Extruderand the UniversalPaste Extruderby RichRap.

The concept is simple enough! Wantto empty a syringe? Just push the plungerdown with a motor! This simplicity probablyaccounts for the popularity of this approach.Brute simplicity has some benets, butthere are some major drawbacks as well.The rst drawback of syringe designs isthat the extruder becomes very bulky. Thetotal extruder height needs to be at leasttwice the syringes length to make room forthe extended plunger, then add even morefor the mechanics. A 60cc syringe with theplunger extended is 30cm long. Add in theextra drive bits and the nozzle, and your ex-truder can easily end up being 40cm high.

Various approaches have been usedto slim down the design. For example,

RichRaps Universal Paste Extruder pulls abelt over the plunger with a geared steppermotor. Unfortunately its tiny 10cc syringeis barely enough to print a cookie! Bonsai-Brainhas produced a 20cc version, but touse his own words: The construction is quitemonstrous! It requires so much torque thata fourth gear is needed. But the syringecapacity is an important factor. 60cc is oneof the largest standard syringe volumes, butit takes 40cc of ceramic clay just to print a

small coee cup. Anything larger requires amore capacious syringe or multiple syring-es, meaning a vastly larger and more complex

extruder. One could try swapping out syringesin mid-print or move the syringe out of theextruder by design (something I will dis-cuss), but thats not the most workablesolution.

Another vexing problem is that syringedesigns vary a lot by manufacturer. Anydesign that depends on a specic syringeprobably wont work with other sizes orbrands.

Frostruder MK1photo by Bre Pettis

2REPRAPMAGAZINE
http://www.fabathome.org/wiki/index.php?title=Fab%40Home:Deposition_Tools#PREVIOUS_VERSION:_Model_1_Syringe_Toolshttp://www.fabathome.org/wiki/index.php?title=Fab%40Home:Deposition_Tools#PREVIOUS_VERSION:_Model_1_Syringe_Toolshttp://www.fabathome.org/wiki/index.php?title=Fab%40Home:Deposition_Tools#Model_2_Syringe_toolhttp://wiki.makerbot.com/frostruder-mk1http://forums.reprap.org/read.php?180,18770http://reprap.org/wiki/Davids_Paste_Extruderhttp://www.thingiverse.com/thing:20733http://www.thingiverse.com/thing:20733http://www.thingiverse.com/thing:26047http://www.thingiverse.com/thing:26047http://www.thingiverse.com/thing:26047http://www.thingiverse.com/thing:26047http://www.thingiverse.com/thing:20733http://www.thingiverse.com/thing:20733http://reprap.org/wiki/Davids_Paste_Extruderhttp://forums.reprap.org/read.php?180,18770http://wiki.makerbot.com/frostruder-mk1http://www.fabathome.org/wiki/index.php?title=Fab%40Home:Deposition_Tools#Model_2_Syringe_toolhttp://www.fabathome.org/wiki/index.php?title=Fab%40Home:Deposition_Tools#PREVIOUS_VERSION:_Model_1_Syringe_Toolshttp://www.fabathome.org/wiki/index.php?title=Fab%40Home:Deposition_Tools#PREVIOUS_VERSION:_Model_1_Syringe_Tools


30/83


Some designers have proposed usingsomething like the Bowden approach (rstpopularized by Ultimaker) in which lamentis pushed down a long PTFE tube. Follow-ing this approach, paste or clay would be

plunged into a tube and fed to the nozzleon the print head. This allows the extruderto be mounted o to the side, reducing theweight and complexity of the nozzle car-riage, but it adds some friction and thusrequires more torque. Nevertheless, withsuch an approach its theoretically possibleto use syringes even as large as a 310cccaulking gun cartridge.

Another drawback of the Syringe Pump

design is that its not really a good idea tocontrol your extrusion by pushing your en-tire volume of material from behind. Thisapproach may work ne in a small 10ccsyringe, but it becomes unwieldy whenwe try to scale up to 100cc and beyond.

This is akin to pushing your whole roll oflament from the end of the roll insteadof feeding it from the extruder. As wescale up to larger syringes the diam-eter of the plunger gets larger and it

becomes much harder to extrude smallamounts with precision.For every mil-limeter of linear plunger travel, a doublingof the plunger diameter results in a fourtimes the amount of extruded materialor a quadratic increase, so any imper-fections in the drive mechanism arenoticeably multiplied. This also increasesthe start and stop time, and we see alot of oozing when scaling past 20cc.

It takes substantial force to pushpaste through a syringe, and the moreviscous the paste the more force it re-quires. With lament you can use moreheat to reduce viscosity, but we cantdo this with paste. Paste needs to havesucient viscosity at room temperatureto be stable for a printed object. Sinceit takes longer to set, it must be able to

maintain its shape and not collapse underits own weight. This is especially importantin printing taller objects. We experiencedfrequent mechanical failures while testingRichRap=E2=80=99s Paste Extruder with

Precious Metal Clay (an ideal material forthe small syringe) and our own 60cc design,which is comparable to the Frostruder MK1.Gears came loose, steppers lost steps, beltsbroke, and syringe retainers failed underpressure.

By now you may be wondering iftheres anything positive to say about theStepper Driven Syringe Pump. Actually, yes!Foremost, a syringe pump is perfectly volu-

metric, a vital characteristic of any pasteextruder design.

Fab@Home - Model_1.1 SyringeTool-photo by Fab@Home

3REPRAPMAGAZINE


31/83


Volumetric means that a speciedamount of rotations from the motor leadsto a predictable and repeatable amount ofextruded material. Just as in a lament ex-truder, you can easily calculate the amountof material extruded per step and control

your extrusion pretty nicely. The use of astepper in the design means you have fullcompatibility with existing electronics andsoftware. RichRap Universal Paste Extruderis an excellent way to get started with pasteextrusion and do small tests with variousmaterials. But as weve discovered in ourtesting, for any serious printing of multipleor larger parts, we need to look for a dier-ent approach.

One potential way to improve the Sy-ringe Pump (which no one has implementedas far as we know) is to use a fairly smallpermanent chamber with a plunger (like asyringe), perhaps 5-10cc in volume, thatcan be quickly relled from the side. Oncethe plunger has pushed all the material outof the chamber, printing pauses, the nozzlemoves o the print, the plunger retracts,the chamber is relled from a large capacitypressurized container, and nally it does asmall starter extrusion before resuming.

RichRap - Universal Paste Extruder

photo by RichRap

Davids Paste Extruderphoto by David Mckenzie

VDX Paste Dispenser 2photo by Viktor Dirks

3REPRAPMAGAZINE


32/83


A valve on the feedline would switch betweenextrude and rell. With some clevernessand two alternating pistons it could evenprint continuously. Such an approach couldretain the straightforward volumetric princi-ple while keeping the size of the print-head

small and the control precision high.

2) Direct Air Pressure Extruder

The Direct Air Pressure Extruder (letscall it DAPE) is almost the exact opposite ofthe Stepper Driven Syringe Pump. Insteadof using a mechanical method to push downa plunger DAPE applies air pressure directlyto the material. This extruder design is radi-cally simpler and lighter than the method

previously described. The top of the syringeis sealed with a pressure cap which connectsvia a narrow tube to an air compressor.Air ow is then controlled using a solenoidvalve (a uid or gas valve controlled by anelectromagnet) to start and stop the pres-sure.

Using this method the extrusion canbe started and stopped almost instantane-

ously. While the extruder design is light andsimple, it does require a supply of com-pressed air. Most homes unfortunately arenot equipped with central air pressure, soan air compressor needs to sit next to themachine.

This approach was rst proposed byZach Hoeken for Makerbot Frostruder MK2and in all fairness (and to stick with ourdigging into RepRap history for traces of

paste): Adrian Bowyer designed a verycrude paste extruder that uses a balloonlled with material in a PET bottle pressur-ized with a bicycle pump! You can nd thisrather amusing 2007 Support Extruder 1.0on the RepRap wiki. So thats the majorbenet, a simple and lean extruder.

At Unfold we still use DAPEs all thetime because of their relatively straightfor-ward simplicity. We designed a couple of

iterations of the extruder for the Bits from

Bytes Rapman which became very popularwith paste extrusion experimenters becauseits print-bed is limited to Z movement. Youcan nd these designs on Thingiverse; justsearch for Claystruder. Mechanically thisdesign is also far easier to scale up. Justextend the syringe or go for a larger syringeand cap.

RepRap Support Extruder 1.0photo by Adrian Bowyer

3REPRAPMAGAZINE
http://reprap.org/wiki/SupportExtruder_1_0http://reprap.org/wiki/SupportExtruder_1_0


33/83


British ceramicist Jonathan Keep, for exam-ple, has shown that a DAPE with a 310cccaulking gun syringe can be installed directlyon a DeltaBot XYZ carriage.

As you might expect, DAPEs tend toperform very poorly in the controllabilityand predictability of the extrusion process,and in their compatibility with RepRapelectronics and slicing software. The rstproblem is that there are so many vari-ables involved in getting a repeatable andpredictable ow-rate. The most obviousvariables are air pressure and material vis-cosity, which combine to produce a certainow rate. Then you must consider plunger

friction, the changing material level in thesyringe, etc.

Flow rate = 3D Material Viscosity + Air Pres-sure.

So if your materials viscosity changesonly slightly you need to compensate forthat by raising or lowering pressure to keepa steady ow. A static input of air pressuregives a wide range of ow rates depend-

ing on the material. Compare that to the

static speed of your Stepper Driven Extruderwhich regardless of the type or viscosity ofpaste gives a predictable and continuousow rate.

Many ideas are being considered to ad-dress this chaotic system. Firmware couldbe updated to work with the parameters ofair pressure systems and compensate forsome known eects. But this may requirereplacing syringes with each successiveprint because a plunger acts dierently in aused syringe, and generally the consisten-cy will never be exactly the same on eachprint and throughout the entire process.

Today we can control for these variables

with an industrial setup, but not in RepRapstyle.

A possible x is to create a closed con-trol loop with an electronically-regulatedpressure valve monitoring ow at the nozzleand adjusting pressure to achieve the de-sired ow-rate. The components required todo this arent the cheapest parts available.Alternatively it could adjust the print speeddynamically based on the ow rate within a

certain workable range, not too fast/slow.

photo by Unfold

3REPRAPMAGAZINE


34/83


Unfortunately, cheap ways to measure theow-rate of a high viscosity material whichdont obstruct ow havent yet been found.Most of us now actively printing ceramicsuse the very simple Direct Air Extruder, but

since the printer has no way of knowinghow much comes out of the nozzle at anygiven time, we all have to babysit our ma-chines and adjust air pressure as-needed.Our attention to the process makes thismore craft than 3D printing.

The incompatibility with exciting newelectronics, rmware, and slicers is also anextant issue with DAPEs. Theres no step-per to control, only electronic valves, and,

potentially, digital pressure valves and owsensors. But never mind that, because theextruders physical control to stop and startthe air pressure is highly unreliable anyway.Each time the syringe is depressurized andquickly returned to the same pressure,the ow rate is likely to be dierent, so

repeatability is low. One solution would beto maintain a constant pressure level andthrottle the material ow at the nozzle.

There are various industrial valves that dojust this, and these could be easy to repli-

cate. F@Hs valve tool does this by using ano-the-shelf valve between the syringe andthe nozzle. The vintage RepRap SupportExtruder 1.0uses a similar method.

Honestly, most people using DAPEs(mainly for ceramics) have stopped usingelectronic valves altogether and designedaround the problem by creating objectsthat can be printed in one continuousaction without the need for the extruder to

jump from one place to another. Pressureis plugged in at the start of the print andreleased at the end, again a very hands-on approach. As with the Stepper DrivenSyringe Pump, If you dont mind manuallycontrolling the extrusion process, DAPEs area great way to start exploring paste extru-sion, especially for printing ceramics andother large volume materials.

3) Positive Displacement Extruders

So we want the best of bothworlds dont we? We want to move thesyringe out of the extruder, and a steppermotor would be lovely for compatibility, wewant a pure volumetric or metered systemand we dont want to control the extrusionby pushing all our paste material back andforth but rather control the ow close to the

nozzle.

Enter the volumetric or positive dis-placement pump. Frustrated by the rst twoprinciples as a starting point for an extruderworthy to sit next to your high qualitylament extruder I started to look at thevarious ways in which industry, and morespecically the dispensing industry, tackledthis problem and started digging throughcountless websites and books.

photo by Kristof Vrancken

3REPRAPMAGAZINE
http://reprap.org/wiki/SupportExtruder_1_0http://reprap.org/wiki/SupportExtruder_1_0http://reprap.org/wiki/SupportExtruder_1_0http://reprap.org/wiki/SupportExtruder_1_0http://reprap.org/wiki/SupportExtruder_1_0http://reprap.org/wiki/SupportExtruder_1_0


35/83


From wikipedia: A positive displacementpump makes a fluid move by trapping a xedamount and forcing (displacing) that trappedvolume into the discharge pipe.

So the key here in the denition isthe trapping a xed amount part. Eachrotation of the pump mechanism traps aspecic volume of material and sends it tothe exit of the pump, the nozzle. If a step-

per motordrives this, we have a system inwhich we know that X amount of motor stepsresults in Y amount of extruded volume.

You place this pump as close as possible tothe nozzle and let it pull in the materialfrom a feed line in exactly the same wayyour lament extruder pulls in the lamentright before the nozzle.

The only dierence is that we donthave a nice roll of lament but a messy blob

of paste in a reservoir. So there is the needfor an external reservoir. In case of a runnyuid like a molten chocolate one could usea gravity feed system (the material ows inthe pump by gravity when the reservoir isabove the extruder) but more often than notwe will need to pressurize the reservoir. Theeasiest way is by air pressure. Yes, that un-controllable thing we rejected above. In thiscase we apply more than enough pressure

to get the material to feed into theextruder, which will take over fromthere and control the ow rate.

Lets look at a couple of positive

displacement pump principles thatare often suggested for a paste ex-truder.

The peristaltic pump is a type ofpump where the pumped mediumtravels through a tube and nevercomes into contact with the pumpmechanism. The pump consists of aset of rollers that squeeze the tubeand push the medium forward. It

shows a lot of pulsation in the outputat each contact point between roller andtube. This can be solved by having multipletubes with alternating wheels. IoanFesteudesigned a very elegant, compact peristal-tic extruder design with dual channel feedlines.

Unfortunately neither Ioan or me man-aged to get it to work with thicker pastesand literature will tell you that the peristal-

tic pump is only suitable for low viscosityuids and there are several other issueswith it too.

Ioan Festeu Peristaltic Paste Extruder 2photo by Ioan Festeu

3REPRAPMAGAZINE
http://www.thingiverse.com/thing:6756http://www.thingiverse.com/thing:6756http://www.thingiverse.com/thing:6756http://www.thingiverse.com/thing:6756http://www.thingiverse.com/thing:6756http://www.thingiverse.com/thing:6756


36/83


Another system is the gear pump,commonly found as an oil pump. Two lock-ing gears rotate in a sealed housing and theuid is pumped on the outside of the gears.While this could work ne for certain self

lubricating materials (like oil but may bechocolate too?) its terrible when trying todispense a ceramic clay which is lledwith abrasive particles that easily get inbetween the large contact areas betweenthe static and moving parts and grind thepump down.

One of the systems that caught myeye during research was the auger pump,an Archimedes screw extruder, often used

for dispensing small shots of solder paste inPCB assembly lines. Its billed as capable ofdispensing high viscosity, abrasive, particlelled materials with ultra precise control.

Auger pumps works like this: The ma-terial is fed from a continuously pressurizedexternal reservoir into the top of the pump.The pump consists of an auger screw ttingperfectly in a cylindrical housing which atthe bottom ends in the nozzle. The pressure

on the syringe is just enough to feed thematerial in the pump where it will hit thetop of the auger and stops there becauseof the increased friction caused by the nar-rower size of the uid path along the screwthread and ultimately at the even narrowernozzle end. The screw is actuated by a mo-tor and this rotation forces the materialdown the screw thread, creating a pressuredierence and the material ows out of the

nozzle. The ow rate is controlled by theRPM of the motor.

I designed a rst extruder using thisprinciple nicknamed the Claystruder 2, Au-ger Paste Extruder which was based on asmall 25 disposable plastic auger extruderfrom an industrial dispensing system. Therst tests with clay were fantastic, this wasexactly what was needed. It behaved verymuch like your standard plastic lament

extruder. Turning the stepper on resulted(after priming) in an immediate ow of clay

proportional to the speed. Stopping thestepper resulted in an immediate stop andno material passed the auger form the con-tinuously pressurized syringe.

We did some initial test with clay,chocolate and potato mash and thoselooked all promising. But after some furtherextensive testing with porcelain clay thesystem became less and less reliable andodd symptoms started to occur: ow ratebecame unreliable during a print and ad-

justment was needed, material leaked pastthe auger when the motor was not turning.Closer inspection of the auger conrmeda suspicion, that the delrin plastic augers

were being worn down by the ceramic ma-terial and this happened already after afew minutes. For 125 one can get a steelversion of the auger but even with that thehousing degrades quickly.

Claystruder Augerphoto by Unfold

3REPRAPMAGAZINE
http://www.thingiverse.com/thing:28018http://www.thingiverse.com/thing:28018http://www.thingiverse.com/thing:28018http://www.thingiverse.com/thing:28018http://www.thingiverse.com/thing:28018http://www.thingiverse.com/thing:28018


37/83


With even more budget one couldbuy a complete stainless steel disposablepackage or even a ceramic one but thosego at the averages selling price of a Mendelkit. Plus they all look geared towards dis-

pensing small volumes of material and notcontinuous extrusion in the amounts weneed, they are under dimensioned andtherefore pushed beyond their limits. Therewas another observation that was againstwhat one needs in a reliable predictablepaste extruder.

I did a couple of tests to see what hap-pens when you change the air pressure onthe syringe that feeds the material into the

auger and, obviously, below a certain pointthe pressure would drop too low to feed theauger but above that point the output ofa xed speed turning auger would vary inrelation to the input pressure. Denitely notas much as when you would use a DirectAir Pressure extruder but still more thanenough.

What this means is that a change inmaterial viscosity would still result in a

changing ow rate at the nozzle. In spring2013 an engineering student camped acouple of months at Unfold to work on hismaster thesis on paste extruders, and hewanted to look again into the auger. WhileI was skeptical, there were two reasons forhim to pursue: the auger is the most com-monly used extrusion system in the industryand his professor made him do so. The maingoal was to over dimension the augers for

the purpose so that wear and tear would bekept at a minimum even with plastic printedparts. The augers were factors larger thanthe disposable ones and could turn at muchlower speeds.

He started with using drill bits as arst test, something a team from BauhausUniversity Weimar also tested some timeago but both without much success. Thethread of a drill bit is to steep for an ex-

truder, material ows through it by air

pressure alone. We build around two dozenprototypes with many dierent variablesand bench top testing them all to map outthe relation between RPM, inlet pressureand outlet owrate. We measured this by

running the extruder for one minute andweighing the resulting amount of extrudedmaterial with a precision scale.

After two months of researching andpainstakingly testing we managed to im-prove the auger but never had meteredoperation. There was always a relation be-tween material viscosity, inlet pressure andthe resulting owrate in the charts.

Another issue observed and maybespecic to clay (which is a suspension) wasthat the owrate decreased in a veryconsistent manner over time. When takingapart the extruder an almost dried out lumpof clay could be found at the end of theauger clogging the nozzle. This happenedtime after time and the nal conclusion isthat this is due to backow of water that isbeing pushed out of the clay owing backup the sides of the auger. While the auger

pump sounds simple, its main operatingmechanism is pressure dierence build upand that is a challenging thing to work with.

We found quotes in books that statedthat while auger extruders are successfullyused in industry, various aspects of it arenot yet fully understood. The conclusionwas that one could probably build a workingpaste extruder based on the auger but that

you would need to carefully tune all theparameters for each dierent material. Nota one size ts all (or most) approach. Thisapparently works ne in a controlled singlepurpose industrial setups and indeed Iveseen research labs 3D printing successfullywith 3000 auger dispensers, but as basisfor a versatile DIY RepRap extruder, not sosure.

Before this renewed eort in experi-

menting with augers I found this quote

3REPRAPMAGAZINE


38/83


which still holds:Positive displacementimplies that a specic volume of materialis displaced within a specic mechanicalactuation. Positive displacement is notinfluenced by changes in temperature or

viscosity. The Archimedes screw pump isvery consistent, but this consistencydepends on the viscosity and flow charac-teristics of the material. Archimedes screwshave been marketed as positive displace-ment pumps, but viscosity metering pumpsis a better denition.

True positive displacement pumpsshould produce the same ow at any givenspeed no matter the discharge pressure.

And this is where the last system comesinto view, the Moineau pump. From thesurface this pumps looks pretty much thesame as an auger pump but the geometryof the auger (the rotor) and housing (thestator) are a bit more complicated. In aMoineau there is no continued path downthe rotor but several sealed cavities be-tween rotor and stator which progress downwhen the rotor is turned, hence its othercommon name: progressing cavity pump.

Every cavity has a known volume so witheach rotation a specic volume is beingdischarged from one or more cavities. Theintricate geometry ensures that the cavitiesalternate to avoid pulsations in the extru-sion.

The only inuence the input pressurehas, is that it needs to be high enoughto feed the material in the pump but not

to high that it would leak past the rotor-stator seal and this is often a wide range.Other benecial characteristics is that theMoineau is very well suited for pastes andparticle lled slurries. So its designed forextreme applications like in the oil drillingindustry but also very well suited for lessdemanding applications.

I never thought it would be possible(at least not for me) to design a printable

Moineau pump but early 2012 Tomi Saloposted a Moineau stepper extruder on

Thingiverse designed in OpenSCAD andbased on the basic parametric rotor andstator geometryby Emmett Lalish. Tomisdesign is a gravity feed design and notsuitable for working with materials underpressure. I made a quick and dirty updateto seal all the parts and stepper shaft andtested the updated extruder with clay. Inthe initial test, which you can nd on

YouTube, the extruder performed very con-sistent with instant start and stop andresponsive ramping.

The extruders drive mechanism and

shafts failed rather quickly though becauseof the substantial torque needed to pump

Tomi Salo Moineauphoto by Tomi Salo

Unfold Moineauphoto by Unfold

3REPRAPMAGAZINE
http://www.thingiverse.com/thing:7958http://www.thingiverse.com/thing:7958http://www.thingiverse.com/thing:7958http://www.thingiverse.com/thing:7958http://www.thingiverse.com/thing:16741http://www.thingiverse.com/thing:16741http://www.youtube.com/watch?v=YWDGTfmd7IAhttp://www.youtube.com/watch?v=YWDGTfmd7IAhttp://www.youtube.com/watch?v=YWDGTfmd7IAhttp://www.thingiverse.com/thing:16741http://www.thingiverse.com/thing:7958http://www.thingiverse.com/thing:7958


39/83


a viscous clay slurry. Months later, again aspart of the engineering thesis and after thedisappointing work on the auger,we pickedup again on the Moineau design, which ismy favorite. The length and parameters of

the rotor changed based on the analysis ofan industrial Moineau pump in order to geta better seal. This is crucial because nor-mally the rotor is stainless steel and thestator a very durable rubber and so the twoseal very well. This makes that it dispenseslow viscosity material like water with thesame ease as a thick clay. In a plastic 3dprinted version you will never archive thistype of sealing but we also dont need thelow viscosity capabilities, youll use high

viscosity pastes. Making the rotor longercreates more cavities and seals.

Various other improvements takenfrom literature were sketched out to reducepossible pulsation. The same one minutebench top extrusion tests were done withthis 2nd update of the Moineau and theresults were vastly more consistent versusthe auger based design. At this point theMoineau still needs a lot of improvement

but that is mostly on the mechanical partof the design, the working principle itselfseems capable to oer what we need.

What needs to be improved is the me-chanical strength of the coupling betweenthe rotor and stepper. Because the rotormoves in an eccentric manner, there needsto be a exible coupling shaft and this shaftploughs through the material in the inlet

chamber. This is now a very basic threeprong coupling that has the tendency towobble up and down and frequently breaksbecause the torque is too high. Also relatedto the torque, a standard nema 17 coupleddirectly to the shaft is really on the edge.Sometimes it works sometimes it doesntand your stepper will skip a lot. Either a sizeup stepper like the nema 23 needs to beincorporated or a geared stepper. So basical-ly the whole extruder needs to be hardened

for high torque operation, thats work thatstill needs to be done.

As indicated in the rst section of thisarticle, the Stepper Driven Syringe Pumpand Direct Air Pressure Extruder both havetheir merits but dont have the potential forgrowing into a multifunctional software driv-

en and volumetric extruder that operates ina similar manner as our lament extruders.

Because of its excellent volu-metric continuous mode ofoperation, the Moineau ProgressingCavity pump principle is by farthe best design to focus devel-opment on today and a lot of the

initial community work is alreadythere to start with. Lets hopeyou join in!

photo by Unfold

3REPRAPMAGAZINE


40/83


41/83

For the last year I have been using andexperimenting with as many dierenthot-ends I could get my hands on. 2013was a strong year for Hot-end and noz-zle development. As we move awayfrom building hot-ends out of plasticmaterials that can melt to ones that canoer extremely higher temperatures,more interesting materials will be tried

in RepRap 3D printers.

By Richard Horne

2013, the year of the hot-end.
https://grabcad.com/library/e3d-kraken-quad-extruder-1https://grabcad.com/library/e3d-kraken-quad-extruder-1https://grabcad.com/library/e3d-kraken-quad-extruder-1https://grabcad.com/library/e3d-kraken-quad-extruder-1https://grabcad.com/library/e3d-kraken-quad-extruder-1


42/83

FeatureHot-Ends review

I decided to make this a general

overview of the dierent hot-endsand give you some feedback frommy real-world use of these. BecauseI have used them to print real partson a lot of dierent machines andover a period of a year, so manythings have also changed, even thehot-ends themselves, based on userfeedback have been further tweaked

and rened by the developers.

I also had dierent lamentsized, both 1.75 and 3mm and awide range of nozzles, so Ill makethe specication clear in the belowand in a future article these dier-ences will also be compared.

So for this overview if seemed fair toprovide general feedback and observa-tions rather than going into exact scienticmeasurements for this article. But now thatthings are starting to settle down with lots

of these newer designs, with a few of thetop performers I do plan to test dierentaspects like force to extrude, maximum owand ooze characteristics at various temper-atures, in a more controlled and repeatableway.

Some general tests you can perform onany hot-end - It almost goes without saying,but be careful these things get extremlyhot and molten plastic is not something you

want to get on your skin.

The two fnger extrusion test (2fnger)

You need to do this one hot (at mate-rial temperature) and after auto-calibration(M303) has been completed and settingsadded to rmware. I always like to see how well a lamentcan be extruded by only using two ngers

to push the lament directly down into thehot-end manually. This can tell you a lotabout the hot-ends ability to melt materialand the pressure required to extrude. Itsworth doing this test at various tempera-tures going up in 5 degree steps, both withand without cooling fans. I nd this espe-cially useful for 3mm hot-ends, dependingon how the nozzle has been drilled and thetaper, gaps and seals etc. this simple test

Girae - Voronoi Style, model by roman_hegglin- Voronoi models are great for testing the abil-ity of a hot end to print ne structures withoutooze.

4REPRAPMAGAZINE


43/83


can give yo a lot of indication how well itwill perform and what sort of drive systemyou are going to require on the extrudermotor.

The force required can of course beexactly and carefully measured for hot-ends

and nozzle sizes, but its good to feel howmuch pressure dierent nozzles take, youwill nd that some are almost impossible topush by hand and others ow and extrudequickly with only gentle down pressure.Much also depends on the material andtemperature, so perform like-for-like tests.

Plug test (Plug)

The plug test is another manual testto perform while the hot-end is at tempera-ture, I usually use PLA and allow it to heatto 190 Degrees C.

Perform the 2nger test above, butafter 3 seconds of pressure pull the lamentas fast and safely as possible out of thehot-end. allow to cool an take a look at themelted Plug of lament. This can help yousee how much of the lament is melting

(melt-zone) and if any steps in diametersize are visible due to parts not sealingush or double tapers that can cause swell-ing, increase torque and potentially cause

jamming.

Some hot-end side walls will grip ontothe lament and it will stretch as its being

removed, but you usually still end up witha plug of some description at the end. Thestretching point, length and plug can allgive interesting indicators to how the hot-end has been constructed and how well itwill process material.

Manual Ooze test (Ooze)

This is simply a test to see how wellthe lament relaxes and self oozes afterpressure is applied. Its well worth trying this test at vari-ous temperatures, just 5 Degrees C lowercan often make a big dierence to reducethe amount of oozing out of a nozzle.

I draw a line on the lament around60mm from the end (with a sharpie) thenperform the 2nger test above, and after3 seconds of constant pressure release thelament and observe both the marked lineon the lament and the material extrudingout of the hot-end nozzle. Some hot-ends will allow the lamentto decompress and the line will rise quickly,others will grip the lament and only slowlyrise or not move at all.

Pico PLA plugs Pico manual ooze test

4REPRAPMAGAZINE


44/83


element. Both of these have issues too, butthey were usually less powerful than mostof the cartridge heaters being tted now.

I have always limited the power avail-able to the hot-end in rmware. While thisdoes not take away the risk of a shorted

Mosfet applying full power, it will at leastlimit the energy going into the hot-endwhen in normal operation. It may takeslightly longer to get to temperature, butthis can also have benets for the tempera-ture control algorithm (PID loop) that is aimingto reach a target temperature without toomuch over-shoot.

If you do adjust the maximum power toyour hot-end do make sure you re-run theM303 Auto-tune again and enter the newvalues. You can get 24V 40W heaters, run-ning these at 12V is another option to lowerthe power. Ideally we need a ~12W 12VCartridge heater for most general 3D print-ing requirements.

A long melt-zone can show up as botha lot of spring (compression) in the meltedlament and also increased ooze from thenozzle end.

In my use of dierent nozzles it seemsbetter to have a shorter melt-zone and so

less spring /compression in the lament andminimal side-wall grip so when pressure isreleased oozing is minimized. You cant justhave a bigger barrel size as this usually pro-duces a plug at the thermal transition zoneactually causing oozing and other problemswhen trying to retract the lament duringprinting.

Materials

Wherever possible I have tried to usesimilar materials and batches for testing, Iuse PLA mostly, then Nylon, PET and lastlyABS. Not all materials are the same, so asthey say your mileage may vary

Wood/chalk/mineral lled laments,conductive and soft materials all bring theirown challenges and will be covered in afuture article.

Nozzle Heating - a word of caution.

Many nozzles not come with a cartridgeheater, these heating elements are usuallywell over powered for use in a 3D printinghot-end. The standard types mostly avail-able are usually rated 12V at 40W. This isa lot of power and can get your nozzletemperature well in excess of 400 Degrees C. Be aware that if uncontrolled full power

(usually +12V) is applied to a hot-end itsgoing to get really, really hot indeed. A 12V40W cartridge heater running at full powerfor just 15 minutes in free-air will gloworange then white hot, expand the metalcase and start to break-down the breglassinsulation. Itll melt almost anything it touch-es and make things catch on re.

Back in the early days of RepRap is

was common to t nichrome wire or use avitreous enamel resistor as a heating

Cartridgeheater

Running a cartridge heater at 12v, this is cur-rent limited 2.7A by my power supply, stillglowing red-hot and burning the insulation.

4REPRAPMAGAZINE


45/83


Extruder Drives

You have so many options for drivinglament down into a hot-end and that it-self would really require a separate article.

At the end of this I sum up the results but Ihave always found a geared extruder to bethe most suitable for most of the printing Ihave done over the years.

I just wanted to state here that I havetested lots of dierent types, and mademany of my own, both direct drive andgeared in one way or another over the lastfew years.

After testing and using everydierent type I could get myhands on, I come back to thestandard hobbed M8 bolt anda V groove 608ZZ bearing forthe idler as my drive method of

choice in most machines.

Example Direct drive extruder systems,very simple to assemble and service, butdo tend to have limited torque to drivethe lament and depending on the pinchsystem can be more prone to jamming orskipping lesser quality lament.

Almost all apart from the Kraken test-ing used a geared drive extruder. for theKraken you can use gearing but I found astandard Nema 17 with a direct pinch-wheelworked perfectly even at fast extrudingspeed.

Many dierent drive systems are in use forextruders.

Hobbed bolt

Direct drive extruder gears

4REPRAPMAGAZINE


46/83


J-Head

The J-head has been my default hot-end for almost all my RepRap adventures,I have a lot of experience with it in almost

all its revisions and I still use them today invarious models and adaptions.

Recently we have seen an explosionof J-head like designs, some are built to thesame or similar standard as the originaldesign (hot-Ends.com) and others fall wayshort of being even slightly usable.

J-heads do use multiple materials andtend to have more parts than other stain-

less or minimal designed hot-ends, but awell made and built J-head can run for yearsand process tens of Kg of material withoutany signicant issues.

They do have some limits and unfor-tunately people do tend to melt them more

J-Head MK5(dissassembled) -four core parts - Nozzle is mostoften Brass, but can be machinedin Aluminum too.

often than you would expect (see abovewarning about cartridge heaters).

J-Head MK2 -Still a good andsolid design, I have a stock of themthat only run ABS, no external cool-ing required, extrusion speed islimited to around 40mm/sec, butthey run ABS well at 245 Degrees C

J-Head MK4 -Longer and heavierthan the MK5 with similar printing

performance.

J-Head MK5 -Compact andvery light weight, great for Bowdenextruders running PLA, needs to befan cooled for best results.

4REPRAPMAGAZINE


47/83


One thing that I have never been ableto explain is that in all my printing with Ny-

lon the J-head always seems to give better

print results than any other nozzle I haveused to-date. I do not know if its linked

with the level of moisture in Nylon, or howthe PTFE sleeve transports the material, or

the level of compression, oozing or melt-zone. I have a feeling its a combination

of all these. So for Nylon I do prefer to use

a J-head MK4/5 over anything else at themoment.

Do watch out for cloned or badly man-ufactured J-head nozzles, I have had some

during the year with various issues. Somerequiring a massive amount of force to

extrude, due to an incorrect machined noz-

zle hole length and others with sealingissues, incorrect use of materials and poor

cooling capability giving a lot of oozing.

Another interesting read for J-headbackground information is Brians J-head

and clone overview here.J-Head MK5

Overall opinion and general recommendations

1. Best for Nylon Printing (245 Degrees C @ 30-80mm/Sec) (1.75mm recommended)

2. Good for general PLA printing (190-220 @ 10-80mm/

sec) - Exceptional performance and print results at verylow PLA temperatures down to 155 Degrees C. - needsactive cooling of the PEEK

3. Do not overheat or the PEEK & PTFE can fail - Keep tounder 250 Degrees, insulate the heater block.

4. Industry standard Groove mount. R

evie

w

4REPRAPMAGAZINE


48/83


Prusa Nozzle V1

One of the rst single piece stainlesssteel nozzles - The original V1 nozzle wasright on the edge of thermal design. without

exactly the right amount of barrel cooling Ifound even the slightest insulation aroundthe heater block could cause problems withow when extruding PLA, leading to stallsor jamming. With too much cooling a similarissue could happen where material, espe-cially when being retracted could cool toofast and jam in the barrel.

The V1 design could also be driven atvastly higher temperatures (+280 DegreesC) to extrude PLA at amazing speed, I even-tually ran out of extruder performance to gomuch higher than around 180mm/Sec printspeed, but the nozzle was still capable offaster speeds.

When its setup well, this nozzle couldextrude at the fastest rate (0.6mm nozzle /3mm lament) using the 2nger test, andwith minimal pressure applied.

Overall opinion and generalrecommendations:

1. No longer available.2. Use the V2 nozzle.

Prusa v1 mounted

Prusa v1 nozzle

4REPRAPMAGAZINE


49/83


Prusa Nozzle V2

With revision 2 came stability and stillgreat results. The single piece stainlesssteel nozzle is retained and a heat-sink was

added to the barrel. With some level ofactive cooling (25/30mm Fan) this hot-endperforms in a much more controlled way.You dont need very high temperatures,(192 Degrees C is adequate for PLA) and itsstill capable of fast speed printing (70mm/sec)

This is a great nozzle for ABS printing.I had the best results with PET material,very clear prints can be retained even at

higher speeds (+60mm/sec)

* Because its light weight it would make anideal Bowden, but its only available in 3mmversions at the moment, this makes it greatfor most RepRap style machines, but any-thing that is using a Bowden feed like mostDelta printers will usually benet from using1.75mm material. So I have not been ableto implement it on a delta yet.


1. Best print results and performance with PET materials(220 to 235 Degrees C)

2. Great for ABS, even at high speeds,

3. Good for printing PLA fast (longer melt-zone), butmore ooze than some other nozzles at low speed PLA.

4. Industry standard Groove mount.Using a lower temperature and a little more extruder re-traction (I used 1.6mm) than normal really helps with theV2 nozzle, for me that resolved almost all issues I had withooze and ne angel-hair strings on slow-printed intricate

models using PLA.

Review

Prusa v2 nozzle

4REPRAPMAGAZINE


50/83


E3D Nozzle V4/ 4.1 and V5

E3Ds Nozzle consists of a StainlessSteel tube (thermal break), Aluminium heat-sink and heater block with a Brass nozzle.

V4 had only minor machining issueswhen rst released, these were quickly rec-tied in V4.1. This nozzle has become loved,trusted and used by many in the 3D printingcommunity.

The E3D nozzle is quite heavy due tothe size of the heatsink, it requires fan cool-ing for best performance and is ideal formost materials printing slowly (it has very

little ooze and a short thermal transitionzone) to moderately fast, for me the 3mmversion maxes out at around 140mm/Sec(@218 Dec C PLA) but with more extrudertorque it could go faster still.

Having a brass nozzle is a big bonus,you can easily clean and change nozzlesizes. Worked very well with ColorFabbWoodll material (0.6 and 0.8mm nozzles).

My Hot-end of choice for PLA printing.

E3D do make a dedicated Bowdenhot-end, this has a lightweight plas-tic pneumatic tting directly in thealuminum heatsink, or you can t aBowden adapter like the one shownabove.

Great for traditional geared anddirect drive extruder systems, Re-duced weight E3D hot-ends areplanned, these will then be ideal forBowden and Delta based printers.

E3D hot-End Cube almost nished print-ed PLA

-E3D Hot-End-190deg-65mm/sec

5REPRAPMAGAZINE


51/83



1.Best all-round for PLA precision printing, very well con-trolled output even at higher speeds.

2.Great for more tricky materials like wood, mineral lled,bendy and also higher temperature (PC, PC-ABS)

3.Good for ABS, Nylon and PET, (I never managed to get asclear results with PET as a J-Head).

4.Industry standard Groove mount.

Revie

w

E3D hot-End for bowden setup E3D hot-End assembled on a Gregs Geared extruder

5REPRAPMAGAZINE


52/83


53/83


Water cooling does make

for a quiet machine but at thecost of tting extra tubes anda cooling system.

Even if you want only asingle or dual head machine,I would still recommend theKraken if you want to experi-ment, for tting it to a ma-

chine like the Prusa i3 its easyand everything can be madeneat.

I ran the Kraken at fulltemperature for 4 days print-ing constantly, it never oncecaused a single print issue orskipped a step.


1. Its experimental! Great for PLA printing. (1.75mmBowden)

2. Great for multi colour prints and material experi-

ments.

3. Quiet and compact.

4. Has its own mounting system - printed mounts willusually be required to t to most printers.models using PLA. R

evie

w

Kraken assembled on Prusa I3

5REPRAPMAGAZINE


54/83


Pico Nozzle V1

I was an early Pico tester before therecent and successful Kickstarter campaign.

Its a very interesting nozzle, almosta work of art in its production. It sharessimilarities with both the E3D having aremovable Brass nozzle, and the Prusa V2with having a complete Stainless steel pathfor the lament.

Its quite dierent in operation to manyother nozzles and takes time to get the bestfrom it.

After using the Prusa V1 nozzle I wasconcerned about the use of an all stainlessdesign, I know from experience over-coolingof the stainless, combined with a non-spe-cic thermal break can cause all sorts ofpotential issues.

As normal I started the testing with PLA,and had some issues with the 2nger test,requiring a higher force to manually ex-trude. The Plug test also showed up a long

melt zone. I have to say this is the longestmelt zone I have ever seen on a hot-end,and I do not believe its to a benet for theuser.

Perfect and exact cooling of thestainless steel barrel was required for PLAoperation, after quite a lot of testing andtuning of fan position and power I had asolution that I could get working for PLA,but I have not managed to get high speed

printing. For reliable prints without barreljams I have xed it at under 30mm/sec forPLA printing. This is not at all what I expect-ed as with a longer melt zone you reallywant to be printing fast to minimize issueswith oozing and retraction based angel-hairstringing etc.

Pico bowden setup Pico on J-Head mount with cooling fan

5REPRAPMAGAZINE


55/83


Another issue with PLA was manifest-ing as extruder skipping but this seemed tobe during the rst layer printing. Increasedback-pressure from the PLA really seems tocause printing issues with rst layer prints,

where you are squashing the lament intothe bed and also on solid ll or with manylled layers.

I spent a very long time using thePico, trying to understand how best to useit for many dierent materials and wasnever really comfortable using it with anytype of PLA.

Wood lled materials did not extrude

well at all, too much force required evenwith a bigger 0.6mm nozzle tted. I did nothave a single successful print.

I only had one attempt at using PET ma-terial and that constantly jammed during therst layer (3mm Colorfabb_XT) but I expect

Taulman T-glase would be the same.

For ABS, its a dierent story, I havemanaged faster and good quality print

results, I found running at a slightly lowertemperature of 240 Degrees C and a 2mmretraction provided good results.

Bear printed in ABS with the Pico on aPrusa i3 using a Bowden fed geared extruder.

ABS is a material I do not work with all thatoften so the Pico is of limited use becauseI could not get as good results with othermaterials.


1. Good for ABS printing (3mm material).

2. Not recommended for Bowden based printing (due toooze and high extruder pressure)

3. Long melt zone but I only managed low speed withPLA.

4. Has its own mounting system - Standard groove-mount aluminium adapter is available.

Review

5REPRAPMAGAZINE


56/83


Undened PTFE/Brass Nozzle

Im not going to say all that muchabout this type of Hot-end apart from beingwary about trying to use it.

I did not have high hopes because itsusing PTFE for the thermal break and alsoas mechanical support.

The main issue with any nozzle thatuses PTFE as a mechanical element is thefact that it changes dimensions as it heatsup. This nozzle is 0.2mm so a very accuratedistance from the bed is required to get therst layer down well. If that distance chang-

es due to temperature you will have no endof problems trying to calibrate it.

The nozzle does come with metalstudding, but because the studs are alsotted to a PTFE disc this does not stop thethermal expansion issues causing a variabledistance for the nozzle. That alone makesit not worth the eort of trying to use thistype of nozzle.

What can I say, not recommended foruse in a 3D printer...

Future hot-ends and nozzles

We are going to see a lot more experi-mental mixing nozzles, lighter all metaldesigns and higher temperature materials,driving all sorts of new nozzle designs.

And remember you can still make yourown, its not that hard to do and good re-sults can be achieved with basic tools and adrill press.

I would b

reprapmagazine issue 3 lowres

Documents