Evaluating 3D printed parts – if you can, show off the part
Like all GrabCAD Challenges, the Oxygen Valve Splitter Challenge presented a difficult problem: create an oxygen valve splitter that allows for independently controlled flow rates. As the valve splitter is meant for “low-resource environments,” participants had to design their part with certain considerations in mind (speed, minimal post-processing, and printability on a basic FDM machine). In addition to earning the top prize, the winning part has the potential of saving lives.
With over 130 designs submitted, our large panel of judges worked hard to determine the winning entries. While in the ideal world all Challenge participants would have been able to test their designs by printing them out, it is understandable that not everyone has access to a 3D FDM Printer just yet. Luckily, some of the judges including Victoria Au (3D4MD) and Adam Arabian (Assistant Professor of Mechanical Engineering, Seattle Pacific University) were able to print out the finalist entries to test the devices. Here, Andrzej Stewart at the Hi-SEAS IV mission, explains the methodology his team used to evaluate entries.
Perhaps I can give a bit of insight into the preliminary judging we used at HI-SEAS.
First, a bit of background. Sheyna Gifford and I are crewmembers on a year-long NASA-funded Mars mission simulation. We have limited supplies here: a limited amount of ABS, and a single Lulzbot AO-101 for printing. Due to the parameters of the mission, we're isolated, and can't readily obtain more ABS or another printer at a moment's notice. Additionally, before the mission, none of the crewmembers had any 3D printing experience - we've been learning over the last five months while we've been here. So, in a sense, we're a simulation of what the real designs will have to go through - manufacture from limited supplies, by personnel with limited means and training.
We had 130 eligible entries to evaluate in preliminary judging, and were instructed to narrow down to 20 designs. We looked at every design submitted, taking time to understand how each one worked. The first thing we did here at the HI-SEAS habitat was look at whether the design fit the judging criteria: Could it split flow? Could it control each flow independently? Note that, for accommodating multiple tube sizes, we allowed for the model to be scaled - the SCAD format allows this to be done very easily. If the design did these things, then we next evaluated it on the ability to print without support material.
Based on those criteria alone, we were able to get down to a selection of 24 designs. From those, we did a second round of evaluation to narrow down to 20, based on amount of material used and effectiveness of the design. Note though, that this was 24 designs, not 24 designers...some designers had multiple designs that fulfilled the requirements. Ultimately, 16 *designers* were chosen. Their 16 designs are all printable, and I feel that's a good first step - if the design is impossible to manufacture, it doesn't matter what else it's designed to do.
From here, we're printing all 16 finalists here in the dome. We'll be judging how well each design prints, how durable it is, and how well it performs the intended functions. We've got our eighth print on the bed as I type this. We've already had a few break under light handling - for me, that's going to be a disqualifier, as you can't rely on something that'll break easily in a medical emergency.
A bit of feedback, please take this constructively...I was surprised by how few designs were submitted with photographs of an actual printed piece. Those designs that did tended to be viewed favorably by us during preliminary judging, as those photos proved printability. Testing and identification of problems is an important part of the engineering design cycle - I believe the designers printing out and testing their own designs helps them identify flaws, allowing for improvement.
After the basics: A 3D printing handbook
There's virtually no end to 3D printing resources for beginners. But what to do after you've mastered the basics is a little less clear. Let us help with this free eBook.
About the author: (Abraham Gladstone)
All posts by Abraham Gladstone