3D Printing PLA on a flexible metal build plate

3D Printing PLA on a flexible metal build plate

By D. Scott Williamson

I love 3D printing.  I’ve designed and printed hundreds of models on the Replicator 2 and have developed many useful skills and techniques. The Replicator 2 has a non heated polycarbonate build plate with MakerBot emblems laser cut into one side and the other side is frosted.

I don’t care for having the MakerBot logo in relief on the bottom of my prints so I print on the frosted side of the platform.

These are rafts but I don’t like having the MakerBot logo embossed on my work.

This worked well for hundreds of prints but eventually, scraping the prints off the platform smoothed the rough surface and parts started sticking harder and harder to the build plate.  Ultimately they stuck so hard that the force required to get a spatula or razor under a part started cutting grooves into the build plate.

Two of the most common 3D printing problems are related first layer adhesion to the build platform…

If the first layer does not bond well enough it can result in corners lifting especially for broad parts on unheated platforms.  In the worst cases the part breaks completely free from the platform partway through a print leading to a stringy mess, wasted time and filament, and in rare cases the PLA can stick to and damage the insulation on the print head.

Catastrophe! Lifted corner, parts broke free, filament everywhere, and damaged thermal insulation on the print head.

If the first layer bonds too well to the build platform the part or platform may be damaged when removing the part.  When using blue tape, it may not be possible to completely remove the tape from the part.

I started using blue painters tape and Aqua Net hairspray on the build plate.  I found this combination to work well with PLA, though I’m not sure how necessary the hairspray is.  The problems are that the tape is damaged when removing most prints so needs to be reapplied frequently and can be difficult or impossible to remove from the bottom of finished parts.

Blue tape stuck to part.

Sometimes it’s impossible to remove all the tape residue.

Blue tape doesn’t last long and requires sticky messy maintenance.

I considered a heated build plate, and glass or metal build plates when the idea occurred to me to try to use a flexible metal build plate. I conducted several experiments using a cable chain model that is challenging to print due to fine detail and thin parts that need to bond well in the first layer.

Experiments

Experiment #1: Aluminum flashing with 2 coats of hairspray dried with heat gun and held by binder clips

First I tried aluminum flashing with hairspray.

  1. Measured and cut the aluminum on a paper cutter and nibbler to perfectly fit the build plate

    Aluminum flashing on roll with Replicator 2 build plate

    Rough cut aluminum.  Use gloves, sheet metal is sharp.

    Cut aluminum to size on the paper cutter.

  2. Rolled flat

    Thin sheet aluminum was curled and needed to be flattened.

    Rolling the thin rolled sheet aluminum flat with pipe on foam.  A towel could have also been used underneath the material.

  3. Cleaned with alcohol to remove oils/grease

    Cleaned the sheet aluminum with ammonia and alcohol.

  4. Coated one side with a thin film of hairspray, let it dry, and applied a thicker coat of hairspray

    Two coats of Aqua Net hairspray applied.

  5. Dried the hairspray with heat gun

    Used heat gun to rapidly dry the hairspray.

  6. Clipped aluminum to build plate with binder clips at the edge

    Aluminum plate clamped to build platform.  (The clip in the upper right corner is about to get knocked off.)

  7. Leveled the build plate to account for the thickness of the aluminum plate
  8. Printed a test

    The PLA bonded weakly to the platform and the parts detached easily in the second layer.

    The upper right and lower left clamps had to be moved because the print head knocked them off.

The nozzle interfered with some of the clips and knocked them off.  The PLA did not adhere to the build plate.  Failure.

Experiment #2: Aluminum flashing with wet hair spray and binder clips

Aluminum flashing with wet hairspray yielded the same results.  Failure.

Experiment #3: Aluminum flashing with glue stick held by clips

Using the back side of the same build plate I used a generous layer of glue stick.

  1. Using the back of the cut aluminum plate from Experiment #1
  2. Coated the plate with glue stick
  3. Clipped aluminum to build plate with binder clips at the edge where the nozzle would be less likely to interfere with them

The PLA adhered wonderfully and the print turned out great.
When done I removed the aluminum plate and was able to remove the print by bending the plate – Success!
But the aluminum does not lay flat and the part left dimples in soft thin aluminum plate before letting go.  I need a stronger material.

Experiment #4: Steel sheet with glue stick held by clips

I scrounged around and found a stiffer steel plate salvaged from a magnetic children’s book many years ago.  I did not try to cut the steel plate to fit the platform because I don’t have a shear and did not want to dull my paper cutter cutting steel.  I can cut the steel on the metal shear at Workshop 88.

  1. Using paper towels, I cleaned the steel plate with ammonia to be sure to remove oil or grease, then with alcohol, and finally with tap water
  2. Coated the plate with glue stick
  3. Clipped steel plate to build plate with binder clips at the edge where the nozzle would not interfere with them
  4. Leveled the build plate to account for the difference in thickness between the aluminum and steel plate
  5. Printed another test

The print turned out great!

But the plastic clips that hold the build plate to the printer are raised causing the plate to be irregular and warped and not flat against the polycarbonate platform beneath it.

If you remove a print by flexing the steel it pops right off but it is still possible to dimple the steel this way.  The dimples can easily be gently pounded out with a broad hammer with the steel on a flexible surface like a neoprene mouse mat or a towel.  Parts firmly attached to the steel are easily removed using a spatula and/or a razor so dimpling turned out to be a non-issue.

Experiment #5: Steel sheet with glue stick held by magnetic sheet

To get the platform to lay flat on the platform I attached a sheet of flexible rubbery plastic “refrigerator magnet” material originally intended to be a furnace vent cover to the platform with double sided tape.

  1. Cut magnet to size
  2. Attached magnet to build plate using double sided tape and pressed it flat using a rubber roller
  3. Aligned the steel plate with the platform and when laid flat the magnet holds it firmly and flat
  4. Leveled the build plate to account for the additional thickness of the magnet layer
  5. Coated the plate with glue stick
  6. Printed another test

Best results yet!

Excellent first layer adhesion with fine detail.

The build plate is flat and level, firmly attached to the platform in the center without using clips which makes it very easy to insert into and remove the plate from the printer.  The finish on the bottom of every part is smooth and shiny, far better even than when printing with a raft.

Even though others reported using a coating of glue stick up to a dozen times, I found reused glue stick not to adhere well.  Adding layers of glue builds up, so every couple of prints I wipe the plate down with a wet paper towel before adding a new layer.

I’ve added alignment markings to the build plate to help install the plate consistently, to help center parts, and to help apply glue only where it’s needed for each print.

 Examples:

Markings help guide application of glue and placement of parts in MakerWare software.

Printed right where expected, with a beautiful first layer and finish quality on the bottom without a raft.

Extremely challenging pinhole lens print

This pinhole lens is .2mm thick and each hole is printed separately with 2 shells then the rest is filled in, if any pop off the build plate it will stick to the hot end and gather the rest into a blob of plastic.

This folding phone/tablet stand (http://www.thingiverse.com/thing:692523) is a favorite model in my house.  It’s a hinged phone/tablet stand that prints fully assembled. If you look carefully you can see that there is some slight curling. The plastic is pulling upward at the corners and even though it has not detached from the build plate it is deflecting it slightly up off the magnet.

The finish quality of the base is comparable to printing on glass.

Not every print has been perfect though, this is another phone/tablet stand that started to curl. The print head caught one corner and moved the entire build plate on the magnet. You can see it continued to print offset before I stopped it. It is impressive that it moved the whole build plate without detaching from the platform. I was able to reprint this model successfully. Only a heated bed or chamber can really prevent this issue entirely, but a stronger magnet may require more force to move.

Future work

  • Cut the build plate(s) to size on the shear at Workshop 88. The build plate is still larger than the platform and must bend to go over at least one of the platform holding clips.
  • Cover the entire platform with magnetic material.  The current magnetic material does not cover the entire platform, it is what I had on hand.
  • Find stronger magnets.  The print quality is wonderful but it is still possible for corners of large parts to lift the platform off the magnetic base while staying attached to the platform.
  • After I started printing on steel I found PRINTinZ’s flexible build plates.  I haven’t used them but check them out! http://www.printinz.com/printinz-3d-printer-plates/

Thanks for reading, and good luck with your 3D prints!

D. Scott Williamson
Compulsively Creative

CAD CAM tutorial

CAD CAM tutorial
by D.  Scott Williamson

This tutorial will show you how to use Computer Aided Design and Computer Aided Manufacturing or CAD CAM tools to create and preview a Gcode file of the Workshop 88 logo that can be run in a 3 axis CNC Mill.

Background

There are 5 main types of machine operations

  1. Engrave (follow path): The tool tip will follow the 3D path provided.
  2. Profile: The tool edge will follow either the inside or outside contour of a path down to the specified depth.
  3. Pocket: The tool will remove all the material within a contour down to the specified depth.
  4. Drill: A drill routine will be executed at each point location.  Drill routines come in 2 flavors:
    1. “Peck” used with drill bits, drills to successively deeper depths liftig the bit out of the work regularly to clear chips from the flutes.
    2. “Spiral” used with endmills that are a smaller diameter than the finished hole.
  5. 3D relief: The tool tip will remove material above a 3D surface usually specified in a 3D model or a 2D height map image.  There are two main modes:
    1. “Waterline” similar to inverted pocket operations where bulk material is efficiently removed outside the 3D model to a number of stepped depths resembling waterline in a topological map.  Typically used in a first pass with a large roughing bit to remove the bulk of the material.
    2. “Raster” moves the tip of the bit smoothly over the model in a raster pattern.

Gcode is a “numerically controlled programming language” which is why a Gcode file extension is typically .nc.  It is a human and machine readable text file.  You will rarely if ever need to look at or edit the Gcode.

Overview

This tutorial will demonstrate Engrave, Profile, and Pocket operations, which are the most popular.

There are 4 steps to this tutorial:

  1. Create a .svg file containing paths needed for machine operations
  2. Create machine operations
  3. Export Gcode
  4. Simulate, visualize and validate

Continue reading

THOTCON 0x5 Badge Revealed

Link

photo1 (3)

THOTCON  is the annual, small venue, hacking conference based in Chicago IL, USA. THOTCON is a non-profit, non-commercial event looking to provide the best conference possible on a very limited budget.

For the past 2 years Workshop 88 has been honored to design and produce the electronic attendee badges for the conference as a service to the local community.  The badge crew this year consisted of: Paul Reich, Bill Paulson, Karl Knutson, Zach Cassity, Russell Lankenau, and Rudy Ristich

This year’s badge was inspired by portable gaming systems from the past and featured  102 x 64 pixel graphic LCD screen and a push button interface. Once again, the badge features an Atmel AVR based microcontroller. The badge used nearly every byte of the 32k available SRAM on its Atmega32u4 chip.  The software consisted of a Break-out style game which participants could play to passtime, a complete schedule of talks and labs for the day long conference,  and the ability to patch into arcade panels hosted in the Hacker Village, and a few surprises for discovering inside.

photo1 (2) photo3

Just like the THOTCON 0x4 Badge, the 0x5 Badge is compatible with the Arduino open hardware programming environment and can accept standard Arduino shields. This means the badge can be easily reused and repurposed to power any sort of project. An improvement from last year’s badge is that no additional parts need to be added; conference goers can simply plug the badge into their laptop once burning a bootloader to reprogram it, encouraging easier exploration and badge hacking.

photo5

The badge is designed to be completely open hardware and software.  Workshop 88 would like to thank the open source hardware and software community especially: Arduino, Oliver Kraus and other contributors to the U8glib graphics library, Dean Camera for the LUFA Project, and last, but far from least: Twisted Traces, our local assembly partner in Elk Grove, IL.

photo4photo2 (2)

Workshop 88 will be holding a badge hacking contest throughout the month of May.  Judging will consist of a panel from Workshop 88 and the THOTCON crew.  Interested contestants can register on the badge website: http://badge.workshop88.com

Full details on the badge specifications and firmware will be released on May 1st in conjunction with the opening of the badge hacking contest.

Lots of lasers!

Originals0701Mark Edmonson recently donated a big box of pretty high quality used battery powered laser levels to us.  They’re in various states, from apparently completely functional to rather dead.

Each one contains 5 diode lasers, as well as some other parts.  There are fairly complete teardown notes and pictures here.

Parts0769Here are the parts I salvaged from one.

Thanks, Mark!

Massimo’s great talk

MassimoRachelJimTwo members of Workshop 88 went to hear Massimo Banzi’s talk on Arduino, open source hardware and more.  The talk was part of Ge Garage’s Idea Week.  He gave some great stories of the philosophy, joys and problems of putting the Arduino out as open source hardware.

Among many other insights, he described how the fashion industry – with no intellectual property protection – made a lot more money than the entertainment and music industries with all their DRM efforts.  He told of the value of the many iterations of Arduino and how a primary metric of its success was the time between a new user opening the box and getting a useful result.  We learned it was named for a bar where they held many design meetings.  It was a great talk.

SignedArduino0243-300Rachel also scored some excellent networking time with Massimo, including connections that will be very useful in her upcoming trip to the Rome Maker Faire.  Jim brought home a newly autographed Arduino that had run the dollhouse at Rachel’s New York Maker Faire booth.

(Thanks to Drew Fustini from PS:1 for the lead picture!)

 

 

THOTCON Badge Hacking Contest Begins!

We’ve all had a full week to recover from the THOTCON and B-sides activity here in Chicago and it is time to get back to hacking.    The badge that was distributed to THOTCON attendees was designed to be hacked and reused in your projects. In the spirit of badge hacking we’d like to announce our first badge hacking contest for the attendees of THOTCON 0x4.

The contest will start today and will run until 11:59 pm CT on Monday, May 27th 2013.

The rules are simple: In hacking there are no rules.

Although there are no rules your submission must be reproducible and should include:

  1. Video Demonstration of your Badge Hack
  2. Any applicable schematics for your hack
  3. Any code and compile instructions

In the interest of collaborative learning any requested information about the badge for the purpose of the contest will be shared with other contest participants. All contest submissions will also be archived on the official badge website.

There will be several categories we will judge against, you’re automatically entered to each category:

  • Best overall badge hack (make us say uhh!)
  • Most hackerish hack (what can you hack with the badge?)
  • Most unorthodox hack (does your badge now dispense cat food?)
  • People’s choice (the tubez chuze)

The prizes will be notoriety and some 3D printed randomness courtesy of the badge crew at Workshop 88.

The astute observer will notice that the pin outs on the side of the board fit the Arduino footprint for access to many of the ATMEGA128RFA1’s peripheral systems and compatibility with most Arduino shields.  The badge can be easily reprogrammed  via the unpopulated ICSP header with (at least) the following methods:

If you’re looking to hack your badge over and over again we have a few left over prototyping kits we were selling during the con and you can get them for $20 plus shipping by emailing us here. 

These include all the prototype rails and headers you need to use arduino shields  plus the passive components necessary to power the badge from a wall wart or other external supply.  The power system components are not necessary to reprogram or hack the chip.

Register here for the contest!