User:Spiritdude

From RepRap
Revision as of 04:45, 8 January 2013 by Spiritdude (talk | contribs) (Replicability)
Jump to: navigation, search

My name is Rene K. Mueller, got interested in RepRap a few years ago but didn't follow up as the first models were expensive and too complex to build and use - this changed and since about December 2011 got interested more thoroughly again, thanks to several Kickstarter projects, such as Printrbot.

Contributions

  • svg2scad, converting Inkscape .svg (2d) paths to .scad polygons for further processing with OpenSCAD.
  • Public RepRap Family Database (Google Spreadsheet), raw data of RepRap Family (to post-process and present later e.g. to draw alike the RepRap Family Tree)
  • MakiBox, wiki-page started and basic information (not my invention, just realized it wasn't in the wiki yet)
  • FFF illustration (see also File:FFF.svg)
  • trying to put some structure and compareable numbers for each RepRap:
    • Printed vs Non-Printed Items (calculate a RepRapFactor)
    • Printing-Size / Building Volume
    • Material Cost (just the items)
    • Cost (assembled)
    • Precision (position/printing)
    • Speed (position/printing)

Favourite RepRaps

Prusa Mendel

High Building Complexity

  • Mendel aka "Sells Mendel", classic (1x Z-axis motor), Z-axis is X/Y-way stabilized

Not recommended to build since simpler versions exist:

Medium Building Complexity

Low Building Complexity

  • Printrbot, loosly based on Prusa Mendel, very simple, Z-axis in the open (one end fixated)
    • Wallace, parametric version, Z-axis stabilized using a horizontal rod to connect both Z rods.
    • Portabee, Z-axis in the open
  • Bukobot, alike Printrbot but with aluminium extrusions (very rigid) and Z-axis is X-way stabilized
  • FoldaRap, with aluminium extrusions, Z-axis is X-way stabilized, foldable

All-In-One

  • FoldaRap, power-supply & controller nicely integrated, speciality: foldable (easy to transport)
  • FoldaRap2, version 2 with scissor-like X-platform, and HBot XY, very small outer dimension vs build dimension
  • MakiBox, case with power-supply, controller, all in a box, speciality: pellets to filament converter built-in but does not feed the printer extruder direct
Round Bed

Unusual Setup

  • Rostock, beautiful setup and movement, yet, lot of motion on the bearing for XY movement of the head, also weight/inertia of the weight stresses more than cartesian setup.
  • PiMaker, rotational bed, Z & R head:
    • advantages: less inertia of the bed compared to XZ head / Y bed ala Prusa Mendel → higher printing speeds
    • disadvantages: variable printing resolution depending on radius, developer aims to patent the setup


Considerations

Rigidity

Reprap-complexity.png

Stabilized Trapez & Trapez

The Mendel based RepRap's are very rigid as the illustration shows, the MendelMax does best and triangulates the top pretty much, whereas the Prusa Mendel lacks one element so the top points are fairly triangulated and therefore stable - with slight possiblity to skew X-way and have vibration.

Stabilized Bridge

Prusa i3 (2012/11) and MendelMax 2.0 (2013/01) both look very alike, and reduced the trapez to what I call "Stabilized Bridge"; it is interesting how MendelMax and Prusa Mendel both in their newer generation become so similar.

Bridge

Bukobot and FoldaRap got rid off or let go some rigidity but tried (successful?) regain some of it by using aluminium extrusions instead of threaded rods like the Prusa Mendel - yet, from a geometric point of view that letting go complexity introduces Y-way vibrations, in particular Wallace which uses rods as framework, yet very appealing in its simplicity.

Two Sticks

The other simple ones, like Printrbot and Portabee sacrifice the rigidity of the Z-axis by skipping the stabilization further: Z-axis is in the open (only one end fixated), possible source of vibration to X-way and Y-way, even though the printing head / bearing provide some X-way stability at the height. Therefore at higher Z-elevation Y-way and X-way vibrations may decrease the position and printing precision, in particular at high speed printing and heavier printing heads (e.g. multi extruders).

Finger

This variant become implemented with the Printrbot Jr, unlike the illustration the actual implementation has two smooth rods for the Z-axis, and due smaller printing volume the risk of X-way vibration has been reduced. Eventorbot on the other hand provides larger printing volume, and the frame is a thick steel extrusion and the bed moves X & Z-wise and the head moves Y-wise only (as with this orientation).

Eigenvalue

Ideally, the printing base is moved only Y-way (in case of these RepRaps considered here), but in real world will shake the entire construction - perceive the RepRap as a complex music instrument in vibration with its own resonance (aka eigenvalue) - any part which can vibrate will vibrate. It is therefore desired to reduce the "open ends" which likely will vibrate and triangulate them to maintain precision in position and printing at any point within the printing or building volume - and for obvious reasons: if framework is sufficiently rigid it allows to increase the printing speed further with little regards of mass of the extruder.

There are several upgrades available, e.g.

Quality Control

There are many factors regarding quality:

  • linear printing integrity / structural skew (misaligned framework)
  • framework rigidity (vibrations within framework)
  • surface linearity (unevenness of material dispension, chosen slice height)
  • resolution (testing what the limits of resolution are)
  • material (printing gaps / bridges)
  • etc

Q: How to measure easily quality (verifiable and quantitative)?

A:

  • See Calibration for a start
  • Linear Printing Integrity
  • Framework Rigidity
    • simple and fast printing model which stresses the axis which expose possible vibration affecting the positioning of the extruder
    • different models, e.g. stress-01-ybed-xzhead.stl or stress-printrbot-original.stl, stress-printrbot-jr.stl etc
  • Surface Linearity
    • simple and fast printing model which reveals surface linearity: several low volume (fast printing) large surface items with different steppings (e.g. different slice height for a Y-bed XZ-head RepRap); optical measurement, e.g. take photo of surface with a light source near perpendicular (right angle) so ripples cast a shade, then use program to analyze the dynamic range and contrast of that photo: uneveness of the surface.
  • Resolution
    • Grid test print object, array of otherwise uniform grids of decreasing wall thickness (2, 1, 0.75, 0.5, 0.4, 0.3, 0.25, 0.2, 0.15, 0.1, 0.05 and 0.01 mm)
  • Material
    • Printable hole test array, array of holes, from 1 to 30 mm in diameter as printed, designed to test the ability of a 3D printer to handle voids accurately.
  • Other Tests

Specific Fixes

Printrbot

Z-Wobble

It has been debated among RepRap developers intensively and Z-wobble occurs due

  1. wobbly threaded rods (it seems majority of threaded rods have wobbles)
  2. fixation on both ends.

The suggested solution is: give the threaded Z-rod some space for X/Y motion, and let the other end of the Z-rod in the open, only fixate one end at the stepper motor (see Prusa i3).

Inheriting Precision

A more general thought is about how improve or at least maintain precision from one generation of RepRap to the next, and by all means avoid degenerative quality of RepRaps printing themselves.

Q: What kind of forms or mechanisms improve, maintain or worsen precision?

A: For now it seems not to have been an issue, as

  • motion precision is usually provided by smooth rods (all 3 axis), so linear motion precision is maintained
  • case is usually not printed (e.g. threaded rods, or aluminium extrusions)

Replicability

Whether a threaded rod vs aluminium extrusion frame is closer or farther from RepRap I can't tell yet - it depends which part likely will be easier to be printed in the future.

Q: Has been there any tests made with printed struts/rods/bars instead of steel rods, aluminium bars or laser cut plates?

Beiwagerl (fromerly known as '(Prusa) Air 2 Mini')

A: Yes, listing projects:

Regarding printing RepRap parts with PLA, Josef Prusa (2013/01/06) wrote: Few of my friends left their printers made out of PLA outside or in car on the sun, rendering the printers non functional. 50°C (120°F) can be enough., so it's recommended ABS even though it's less recycable than PLA.

Replicability Factor

I propose a RepRap Factor, printed parts (pieces & volume) in regards non-printed parts (pieces & volume), for example (made up numbers):

  • rrf = (1 / (p + np) * p)
  • rrfpieces = 1 / (30 + 220) * 30 = 0.12 = 12%
  • rrfvolume = 1 / (20cm3 + 50cm3) * 20cm3 = 0.286 = 28.6%

A pure or 100% RepRap or rrf has 100% printed parts, 0% non-printed parts.

To compare weight seems not so informative, e.g. stepper motors are heavy, yet, a well designed strut with little infills but good inner structure weights little, so the information of the replicability to compare weight isn't so good. On the other hand does weight represent also usage of resources (matter), heavy metals as used by stepper motors require a long supply chain, whereas a PLA filament likely shorter one - that wouldn't be an indicator of replicability but resource usage in general. There are further considerations required how to qualify RepRaps further, keywords: "greenness" (quite broad term), "ecological footprint", etc.

Pellets

Pellets are available at industry level, where usually filament is made from. Several projects have aimed to provide filament production:

  • FilaBot, started as Kickstarter project in 2012/01 and since then has grown significantly.
  • MakiBox aimed to be an All-In-One, e.g. you fill in pellets (1-2mm large pieces of material), it creates filament internally for the extruder - ideal for recycling, yet as of 2012/12 they did not manage to resolve to combine the pellet extruder with the printer extruder reliable, so the pellet extruder makes filament on a spool, and then later is used for the printer extruder
  • Recyclebot

Recycling Printing Material

With the growing printing activities also the issue of recycling arises, how can printed material recycled:

ABS (Acrylonitrile Butadiene Styrene)

Used for luggage, water pipes, LEGO, etc.

  • softens at 90C, melts at 105C, extruded at ~180C but needs to be ~240C to bond strongly to itself
  • recycling:
    • regrind parts (of the same color) into pellets and re-melt them into filament again (degrading quality?)

PLA (Polylactic Acid)

Used for packaging, organic-waste, bio-degradables, tea bags etc.

  • softens ~50C, melts at ~160C, extrudes and bonds well at 180C-190C
  • recycling:
    • reprocessing (see Loopla)
    • composts slowly (with large surface within weeks, otherwise 6-12 months)
    • regrinding ???

Interesting quote, Vik Olliver January 7, 2013:

I'm the guy that first used PLA in RepRap 3D printers as far as I know. It was done for environmental reasons.

HDPE (High-density Polyethylene)

Used for milk jugs, bottles, bottle caps, water pipes, etc.

  • melts at 124C - 135C, extruded at ~180C but needs to be ~240C to bond strongly to itself
  • recycling:
    • regrinded and remelt, degrading quality (downcycled), details???
 

LDPE (Low-density Polyethylene)

Used for trays, plastic wraps, slides, etc.

  • melts at 95C - 115C
  • recycling:
    • regrinded and remelt, degrading quality (downcycled), details???

PET (Polyethylene Terephthalate)

Used for bottles, etc.

  • melts at 220C - 280C
  • recycling:
    • regrind and remelt, (degrading quality?)

Nylon

  • melting point 260C - 500C
  • recycling:
    • regrind and remelt, (degrading quality?)

References:

Recommended:

Alternative Printing Materials

Homemade Bio Plastics

To keep the idea of open hardware also apply to the source of material to be printed, I was looking for ways to make bio(-degradable) plastic from common available ingredients (short supply chain), e.g. from cornstarch, water, glycerin and vinegar (e.g. Make your own bioplastic) - and experiment thereby also with softer and bendable material to make elastic printed forms (e.g. soft parts of a wheel or ribbon) or foamy parts (extreme lightweight). Some critical thoughts (The Guardian, 2008) and here on bio plastic in general.

Q: Any experiments with direct ingredients to mix to make pellets or filaments or feed the extruder direct?

A:

Making RepRap a CNC Machine / Hybrid / 2D Plotter

Switch the extruder with a small milling head; consideration: strength of stepper motors and material to cut.

Q: Any projects?

A:

Cost of Printing

It was pointed out at User:Traumflug user page, that printing itself costs:

  • material
  • electricity
  • wear of printer (parts/material)

printing a set for a RepRap takes up to 10 hours.

Q: Are there any more substantiated numbers how many resources are used (usage of material, electricty) and therefore cost of printing of a RepRap?

A: Measure volume to be printed in cm3, assume a certain filling (e.g. 50%), given layer thickness (0.1mm-0.3mm) with speed (e.g. 30mm / s) determines the duration of the print.

  • timeprint = ( volumeform (cm3) * filling ) / layer thickness (cm, e.g. 0.01cm) / layer thickness (cm) / speed (cm / s)
    • e.g. ( 10cm3 * 0.5 ) / 0.03cm / 0.03cm / 1cm/s = 1851 s = 0.51hr (or 30min 51sec)
  • electricity: power (kW) * time (hr) * priceelectricity kWh = costelectricity
    • e.g. 0.2kW * 0.51h * 0.25 Euro = 0.025 Euro
  • material: volumeform * filling * costper cm3 = costmaterial
    • e.g. 10cm3 * 0.5 * 0.035 = 0.175 Euro.
    • ABS density 1025kg / m3 = 1.025g / cm3 => 35 Euro (1kg filament) / 975cm3 = 0.035 Euro / cm3

First few layers a heat-bed may be used, drawing most energy (~3x than the motors). So, the costs in this example make up 87% for material, and 13% for electricity. Wear and resulting maintenance are neglected for now, but come in effect at longer duration of printing (e.g. weeks non-stop printing).

See also 3D Printer Calculator

Recycling 2D Printers to RepRaps

Conclusions:

  • older printers (e.g. Inkjets) have strong stepper motors than newer ones which also often have servo motors (unuseable)
  • 40-53 Ncm torque recommended for RepRaps (1-2A minimum), NEMA14 like with 12-15 Ncm seem to work too (from another source)
  • 200 steps per evolution prefered, for Z-axis lower steps possible

I have disassembled a Canon multi-functional machine (scanner & inkjet printer), and to my surprise found only one small (flat) stepper motor for the scanner, but otherwise ordinary DC motors with optical encoder feedback for the printing part.

Dynamically Adjustable Size of Nozzle Hole

The precision of the printing is highly dependable on the size of the nozzle hole e.g. 0.3mm or 0.1mm - usually the precision of positioning is much higher 0.02mm or so and therefore neglectable in this consideration. It would be interesting to consider a nozzle hole which can be dynamically adjusted in size: for outline lines (becoming surface) go 0.1mm, and for the interior 0.3mm or even bigger for structural strength.

Q: Any projects which has considered or implemented that?

Batch Printing

With BotQueue (networked printing) the batch printing has become more in the focus.

Q: Are there any project to print in batches: clear the bed after the item(s) are printed?

A:

  • Automated Build Platform (ABP) by Makerbot (2010/09) - it seems to have some issues (no leveling mechanism, no way to tighten belt, belt made out of plastic instead something with higher melting temperature).

Useful Resources

In general G+ has become a place to connect with developers, in particular the G+ "3D Printing" Community.

Further Considerations

Some articles by me:

Contact