Over the last two weeks, DefDist has been testing modifications of the reinforced AR lower design that was originally found at a site called Thingiverse and posted by a user named HaveBlue. That site will no longer host firearms-related files. We’ve been busy launching DEFCAD.org over the past few days as an alternative. Files we collect and those from other users can be found at DEFCAD from now on.
DefDist has now printed six lowers in three different technologies and four different materials. Our original print was in Objet ABS-like material. The other pieces were printed in SLA UV resin, Objet VeroClear, and ABSplus.
The materials sheets for each show that they roughly approximate ABS thermoplastic in their yield and impact strengths. Print times ranged from just over seven hours (Objet), to twelve (SLA, FDM). Additional cleanup time is a factor not accounted for in those ranges.
The pieces each feel and behave differently based on their material and mode of printing. Objet materials give much more than our ABSplus and SLA resin pieces, but the ABSplus pieces mostly come across as brittle. The SLA pieces definitely have the most hardness. But to the modifications and build process…
In the last few weeks, three designers and even more opiners have had their hands in the making of this seemingly modest current revision. Important areas of improvement include the thickening of the buffer tower and really the back half of the piece above the safety selector switch. Most of the improvements we wanted to make from our first review have made it into the current version. But for the modified trigger guard (whose shape was happily provided by a friend), each change has yielded improvements in the stress simulations we’ve run in SolidWorks.
New displacements in the buffer tower were ultimately reduced by 39% from the original file (down from .1824 mm). Stresses were reduced by almost 50%.
There are still points of high stress where the threads intersect the edge of the buffer ring, but by removing just a few revolutions of thread you get a large improvement.
After we had achieved these results in the software, we began printing an assortment of lowers to compare materials and technologies. Let’s do a little assembly.
Our working files all are meant to be tapped and threaded in 1/4”x20 now. The assumption is that a coarser thread is more accommodating of hand grips across the range of the most available print materials.
Assembling the ARs on site. The Five Seven upper is no longer necessary after our improvements. We are building completely for and around the .223.
Candidates one and two-
Results were challenging in new ways. Our highest round count to date is 80+ with an SLA resin lower. We’ve had two FDM printed pieces fail in totally different ways in similar builds, so back to the drawing board there.
But with the more rigid pieces, our new failures have been consistent. A split forms at the rear take down pins through to the hand grip. This split happens across materials and printers.
These splits happen usually after two or three magazines. Often after rapid fire. Our initial thoughts are that the volume of the material above the takedown pins simply isn’t sufficient and must be increased by building outward and swept along the profile of the back of the piece. We also noticed tiny steps leading to the buffer tower beginning right above the pin holes. This is a natural place for the split to begin.
The split again.
We’re satisfied that the problems in the buffer tower are for now sufficiently resolved for at least 100 rounds. This new failure will require designing around stress simulations where the load is directed downward at the back of the piece. These harder pieces cannot flex well, a property that we think is still an exploitable feature in FDM ABS or even PLA prints.
Though we’re seeing a budding user base and parts list at DEFCAD, we will not yet release our modified working files. We want to make one more major revision based upon our current failure point and then test FDM prints from two popular hobby printers before putting a development file out into the wild. At that time we will likely issue a DEFCAD and Pirate Bay release simultaneously. This will be a good practice run for releases to come and will also allow us to experiment with our current license drafts.
That’s it for DefDist in 2012. We have quite a bit of housekeeping to finish, and we’d like to undertake a more comprehensive strategic planning. Thank you for your interest in the project, and we hope you’ve had a Merry Christmas. We continue to believe the future for human liberty is different from the “plans” and prescribed GDP tweaks of neoliberal baby boomers. On that note, I’ve heard Joe Biden will be leading a committee to discover a palatable way to ban the future sale of the battle rifle.
Dedicated to Joe and co. then-