It’s the start of a new year for UBC Rapid. We have new projects planned, several new members, and another year’s worth of funding just came in! We’d say it’s time to give everyone an update on our progress over the summer and into the start of this year.
What has happened so far this year?
The biggest thing we would like to report is that we have finally written a governance document for our team! We defined member expectations, how people are elected to team positions each year, and several policies for our team are now written down and defined. Work on the document was done over the course of this past summer, and has been fully voted in and agreed upon by all members. We’re quite enthusiastic about this going forward, as we hope the structure it provides our team will facilitate a more efficient team this year, and increase member retention.
Additionally, last week we got the results from our application for funding from PAF (the Professional Activities Fund). We did very well on our application for this year: we were awarded $ 2,539.33. With an additional $500 in funding from APEGBC, we have a total of about $3000 to spend on team projects this year, which is fantastic.
We would also like to thank our new sponsors that have worked with us over the past year. Although all of our sponsors this year are listed on our sponsors page, we’d like to take the time to talk about them on our front page! The first to offer us sponsorship at the beginning of this calendar year is Zaber. They were founded in 1997, and has it’s offices right here in Vancouver, 30 minutes away from UBC. The company hires many students in the UBC Engineering and Science Co-op programs, and many of it’s full time employees are UBC Alumni. Zaber makes precision linear motion devices, some of which can be used in the construction of rapid prototyping devices. We have received several pieces of hardware from Zaber, including a box full of stepper motors, which are incredibly useful for every single one of our projects. In addition we received a $500 in funding that went towards project funding towards the end of last school year.
Our next new sponsor is IGUS. Primarily a plastics company, IGUS makes bearings, bushings, and several varieties of linear rail systems. They also make a unique type of filament for 3D printing, which is apparently suited for printing bushings and bearings. Although we have yet to try this filament yet, we look forward to seeing results with it, as it allows for a few more types of parts to be 3D printed, putting us a step closer towards a 3D printer made entirely out of printed parts. Check the filament out here. IGUS has kindly provided us with some free parts which we used in our latest 3D printer.
The last major thing to happen this year is that we have been running workshops for new members of our team again. We have done a couple sessions dedicated to bread-boarding and soldering, and have done a safety and lab orientation session as well. In the breadboard and soldering sessions everyone made TV-B-gones, which are devices whose sole purpose is to turn any TV on or off. They look a little bit like this:
After the circuit is breadboarded (pictured above), it is then soldered onto protoboard, and then in future sessions we will be using Solidworks and 3D printing to create custom enclosures for the device.
Whilst this week we ran no training sessions for our new members, we are hosting a Solidworks session next Thursday. We hope to see many of the new folks having fun and working on projects soon!
What do we plan to do over the course of the year?
With updates out of the way, it’s time to get onto the more interesting stuff: what we plan to do this year. So far our plans are thus:
- Last year we were able to convert one of the liquid handling machines we had stored for a couple years into a CNC milling machine. We are just starting to try out actual cuts with it now, and are slowly learning how software for milling works, as it’s a little more complicated and involved than 3D printing is. We hope to get this prototyping device fully functional and producing parts for future projects by the end of the year.
- Two new printers were also built, and we have parts around for a few more this year. We are undecided with what model our newest printers will be.
- We are also continuing work on a Steriolithography printer, which 3D prints parts out of a UV curable resin. This makes very small parts well and with a high degree of fidelity.
- Continue work on the filament recycler, perhaps getting to the point where we can dump a bunch of pellets in and walk away, having it completely reliable and automated. One of our new team members Tim Branch has taken interest in working on this and we’re excited to see what gets done!
That’s it for now. Look out for another update towards the end of term!
In my last post, I introduced a method to recycle plastic using a household blender and liquid nitrogen or dry ice. While I was pleased with the results, I wanted to further improve the consistency of the particle sizes.
To improve size consistency, a soldering iron was used to create an ad-hoc filter in the side of the blender by melting small holes roughly the size of the commercially available plastic granules. The idea is that once a granule has been blended until it is acceptably small, the blender will eject it through the filter and into a collector. While I had concerns about the filter not working or getting clogged/jammed, this did not end up being an issue. The filter worked great!
There is a noticeable difference between the plastic that was recycled with and without a filter. The green plastic was recycled without a filter, the blue and purple plastic was recycled with a filter. The red plastic is commercially available granules for comparison.
While blending, the granules shot out of the filter and into a cup as soon as they were small enough to fit. Mixing was sufficient to prevent clogging of the filter and this did not appear to be an issue. I am very impressed with how this batch of plastic turned out, the filter has made everything so much more consistent.
Recycling plastic for use in plastic extruders such as the Filastruder isn’t a new idea, and attempts have been made by a number of groups over the past few years to make recycling as accessible as possible. Presented here is a proposed method for grinding down old printed plastic into a size suitable for use in open-source filament extruders such as the Lyman Filament Extruder or the Filastruder. This following process was developed by UBC Engineering Physics student Scott Lawson.
By cooling a polymer below its brittle transition temperature prior to blending, we solve a number of problems encountered by previous recycling efforts such as the RecycleBot. Once cooled, a common household blender can effectively grind plastic into sizes compatible with filament extruders.
Importance of recycling
Most plastic extruders on the market can convert small plastic granules into plastic filament, but it would be nice if there was an easy way to make the granules. Ideally, we could grind down old prints or perhaps other sources of plastic like water bottles to make 3D printing filament. At ~$30-40 per kilogram, commercially available plastic filament is expensive compared to the raw price of bulk plastic.
Using liquid nitrogen, we cool PLA plastic below its brittle transition temperature to reduce the energy absorbed before failure. Unlike room temperature polymers, cold polymers can be easily chopped and blended into sizes compatible with filament extruders (see the Theory section below for details). The target temperature of approximately -20C (-4F) to -40C (-40F) is nothing compared to the blisteringly cold temperatures of liquid nitrogen (-196C / -320 F). A more commonly available substitute for liquid nitrogen is dry ice, which has a temperature of -78C (-109F).
The blender used is a Magic Bullet donated by Carly, one of UBC Rapid’s team members. As far as blenders go, the 250W Magic Bullet is not especially powerful and is weaker than many common household blenders. This really shows how much easier it is to grind cold plastic. The first step is to cool the plastic by submerging it in liquid nitrogen. Once cold, the liquid nitrogen is strained out and the plastic is placed in the blender.
Grinding the plastic takes about five minutes until a fine consistency is achieved. It’s a good idea to periodically filter out the smaller pieces and re-blend the larger pieces until everything is the right size for filament extruders.
The final product is shown against commercially available granules for size comparison. Note that while many of the green pieces are even smaller than the red granules, there are some larger pieces that would need to be filtered out and blended again prior to use in a filament extruder. For this trial, everything was grinded in a single run without stopping to filter out the smaller pieces.
While grinding is easy to achieve with household equipment and liquid nitrogen / dry ice, this procedure is dangerous for a number of reasons and should be carried out with extreme caution and only with the proper training and safety equipment. At cryogenic temperatures, both metals and polymers are below their brittle transition point. The blades of the blender become very brittle and could shatter, cause serious injury or death.
Blenders are not designed to be operated with dry ice or liquid nitrogen, and you should never pour liquid nitrogen or put dry ice directly into the blender. This will almost certainly break your blender and can cause glass to shatter from the sudden temperature change. Place cold plastic into the blender but not liquid nitrogen or dry ice. We do not endorse attempting any kind of plastic grinding using a blender.
Never place liquid nitrogen in a sealed container. It will explode. Liquid nitrogen is constantly boiling off and if placed in a sealed container, pressure will build until it explodes. Always follow proper cryogenic safety procedures.
But where can I get liquid nitrogen or dry ice?
Depending on where you live, it may not be hard at all to obtain liquid nitrogen or dry ice. In fact, even some Walmart locations sell dry ice. Liquid nitrogen has to be stored in a vacuum insulated Dewar flask, or a vacuum insulated thermos. Almost all universities and cities will have suppliers for liquid nitrogen and dry ice.
Since dry ice and liquid nitrogen can be a hassle to obtain, it may be wise to save up plastic and grind a larger amount once in a while.
Grinding down polymers into small granules is more difficult than it sounds. The difficulty arises because polymers can absorb large amounts of energy before failure, due to plastic deformation. A typical polymer stress-strain curve is shown below.
Past the yield point, the long horizontal region on the stress-strain curve represents plastic deformation. During plastic deformation, the polymer will absorb large quantities of energy and it will become stronger as the polymer chains align. Grinding plastic with a common household blender can be difficult because the blades will become very hot and the plastic will absorb large amounts of energy.
However, by lowering the temperature of polymers below their brittle transition temperature, we can dramatically lower the amount of energy absorbed before failure. The graph shown below demonstrates the difference between brittle and ductile failure of metals, and a similar result is observed with polymers.
For ABS, the brittle transition occurs between -20C (-4F) and -40C (-40F). Below this temperature, grinding the plastic into granules becomes much easier. Each polymer has its own brittle transition temperature, although typical polymers will be brittle at -40C (-40F).
The RepRap Gen7 Board-AVR1.5 (http://reprap.org/wiki/Gen7_Board-AVR_1.5) is fairly robust. Unfortunately, the stock design doesn’t feature accessible heat sink mounting posts for the Pololu DRV8825 drivers and there are only 4. Additional drivers can be mounted via a “shield”/”extension board”, but further thermal management is unwieldy.
We respun the board and remade it in-house, for use on a MendelMax 2.0 system. It now supports 5 drivers, (X, Y, Z + dual extrusion), and has mounting holes for an overhead heat sink+fan that we will fabricate and mount.
A pull request has been submitted to the RepRap github, but, in the meantime, the design files are available here:
A few weeks ago we were asked to print a wind tunnel model for UBC’s Formula Racing team. Since a smooth surface finish was paramount, we printed the model in ABS to permit vapour smoothing.
Presented here are some general techniques for achieving fine surface smoothness, even when the printed item is of mediocre quality.
Printing the Formula team’s nose cone marked the first time using ABS in our printer, so we encountered some calibration errors and the end result was of poor quality and contained noticeable artifacts.
Here’s how the printed nose cone turned out. There are some obvious artifacts in the print, including z-axis ripples, perimeter holes, overheating and overhang issues.
The worst of the problems can be seen more clearly in the next picture.
The general approach to finishing these parts involves repeatedly sanding and vapour smoothing the ABS parts using acetone. I recommend this RepRap blog post for an introduction to ABS vapour smoothing.
The first step is to sand the surface of the print with emery cloth or sandpaper. Start with higher grit (300+) sandpaper for prints that are of high quality, and lower grit for parts with more obvious artifacts.
When sanding, you will start to build up ABS powder on the surface of the printed item. Instead of removing the powder, the trick here is to try to coat the surface of the print with it. This will improve the effectiveness of acetone vapour smoothing.
By first building up a layer of ABS powder on the surface, the vapour will condense onto the print and dissolve the outer layer of powder, which will then fill in small cracks and voids on the surface. The high surface area of the ABS powder means that it dissolves faster than the rest of the print, so surface detail can be retained to a greater extent.
Repeated acetone vapour treatments can be time consuming, so acetone can also be applied directly to the powder with a paintbrush.
After several iterations (or just one with good initial quality), a smooth part is obtained. Here’s how the nose cone turned out.
As you can see, the nose cone is now significantly smoother than before. Most of the artifacts are gone, and the print has adopted a glossy sheen even when dry. The part could be finished even further, but this was sufficient for the Formula team’s needs.
On a semi-related note, after careful calibration the printer no longer encounters issues printing ABS, as demonstrated by one of the most recent prints.
So last November UBC Rapid moved from the basement of the Chem/Physics Building to the basement of Hennings. The room that we moved into has many names, including:
- “High-Bay Area”
- “High-Head Space”
- “The Rapid Room” (we like this one for obvious reasons)
- “The Chamber of Secrets” (according to one of my friends double majoring in Physics and Astronomy)
Regardless of what the room is actually called, it’s ours. Before we moved in, the room had fallen into disuse, and people had been using the space largely to store things. When we arrived everything was disorganized, dirty, and not very suitable for a student team. Initially we cleaned the space up a bit, mostly to a state that we could use it to build Maxwell, our MendelMax V 1.5 printer. However, we managed to finally get the team in to properly clean everything up over the last week of summer break, and to prepare for the fall. We had a lot of work to get done.
The full photodump for people who don’t like reading is here.
First we planned how we wanted to organize the room. We didn’t really have to move too many things, and since the start of term we’ve moved several things again. Our plan was originally this:
After deciding where we wanted to put things, we then decided to start with the messiest area of the entire space, which was the corner by the garage door. It originally looked like this:
It was mostly fully of scrap materials and other things we didn’t feel like organizing when we originally moved in. We got most of the materials cleaned up pretty well, by making a sheet metal/plywood sheet rack:
After about a day’s work, the corner was all tidied up. We still had to move the lathe further into the corner, and get the e-racing stuff out of there, but we wanted to get the corner clean so we could begin organizing the rest of the room. It’s really helpful to have floor space to move stuff in and out of while cleaning!
Above photo: Rapid Captain John Harvey happy to finally see cleanliness in the space. Up next was the “middle” of the room so to speak. This was undoubtedly the messiest part, mostly owing to the blue rack on the left. The plan was to eventually move the rack into the back of the room just to the left of the lathe also pictured in the above photo. That required the lathe be moved further into the corner, so for a day or two of cleaning we mostly cleaned everything around the blue rack. In the photo below we see John Harvey again, trying to get some mystery substance off of a small table. It took lots of elbow grease and harsh cleaner:
Later that day in the afternoon we got some of the machinists upstairs to help us move the lathe. Moving a machine that heavy is tricky without specialized equipment:
With the lathe finally out of the way, we could move the blue rack. We finally had a legitimate reason to use the “MAX 5 TONS” ceiling lift that came with the room. It was pretty exciting!
Move successful! Now to finish up the middle of the room:
The stuff in the bottom right corner belonged to the E-Racing team that used the space years ago, but never picked up some old equipment. Thankfully they came by and took what was theirs out of the space.
Almost done! Just have to clean up the area around the other door to the space, then we were done. There wasn’t any difficult moves or a bunch of boxes in the way, we mostly just had to put tools and other things back where they belonged. We also cleaned up the table where our Solidworks workstation sits:
Finally, we cleaned up the main tables where we sit most of the time. I don’t think I’ll see them this clean for a long time!
That finishes up the coverage of cleaning up UBC Rapid’s space. It isn’t the most interesting update, but it was really great for myself and my teammates to finally have a clean space. It will get messy again fairly quickly, no doubt. Stay tuned for more technical and interesting updates coming soon! We get our PAF budget in the next week, which means we can start ordering parts for our projects!
Last year, UBC Rapid member Scott Lawson published a video of magnetic putty consuming a neodymium magnet. The video became popular, and Scott found himself holding a check from YouTube.
What does a Rapid member do with windfall money? If you’re Scott, you finish the intensive ENPH 253 class first. Then you set out to make the best 3D-printer you can. Scott settled on the Mendel90 design, and got to work.
He put together Generation 7 electronics, milling the board in the Engineering Physics Project Lab. With a future Dremel-heat modification in mind, he chose easily-machined 6 mm-thick aluminium stock for the frame. In about a month, all the parts were together and Scott was printing.
Our printer is finally printing! After a lot of tweaking with settings and calibration, it is producing some absolutely amazing prints. Take a look for yourself:
Maxwell, our printer
Rowan making sure the first layer prints nicely
Link to Scott’s video of our printer printing a catapult:
A quick comparison of last year’s print (behind) and this year’s print (infront). Note the huge improvement in quality with Maxwell!
That’s all for now. More pictures of newer prints will be uploaded soon.
Photos courtesy of Scott and Dylan