Subsystem Build and Test

Team Vision for Subsystem Level Build & Test Phase

During this phase the 80/20 stock was finally obtained and our goal was to setup the frame of the printer. A lot of pieces had to be machined and there was a lot of work that needed to be done in order to put together the outer frame. Our goal for the end of this phase was to have the frame setup with the motion axes on and the stepper motors wired up so that an XYZ motion demo can be done. During our efforts to put the frame together we discovered that there was a difference between the Parco aluminum we ordered and standard 80/20. That difference had to be milled off in order for the T-nuts to fit. After that, the outer frame was put together and the motion axes are being put on.  Additionally, during this phase the modified extruder assembly was to be tested utilizing the windshield wiper motor acquired from Monsterguts, during this testing it was learned that the plastic 3D printed motor coupling to the auger did not provide enough rigidity and strength to be utilized and custom coupling was machined. 

Subsystem Level Test Results Summary

Due to the nature of our Customer and Engineering Requirements, not all requirements can be confirmed with certainty at this time through our defined Requirements and Test Plans. The vast majority of our requirements call for a completed model. As an alternative, the team has developed preliminary testing to ensure we are on a path to be able to test our final product against the previously established Test Plans. Below is a high level overview of this testing. Further details for each test can be found in their respective section below. 

Frame & Print Bed

Slight Difference in Parco Brand

Originally we decided to use 8020 however the face cost of 8020 would bankrupt our budget by the time the parts arrive. To mitigate this we attempted to use an 8020 distributor. After a kerfuffle with the purchasing office, we purchased Parco aluminum extrusion from CS automation. One problem we notice after the part arrived is the Tslot grove for Parco and 8020 differed slightly. For 8020, the slot had a slight divot near bottom dead center. For Parco, the slot had a slight ridge. For normal T slots this would not cause issues, however since some connections required a ball spring slot the slight ridge caused interference that did not allow it to fit. After milling some groves we were able to assemble the frame.

Frame Building

Once we got rolling the parts fit together nicely. Nico & Tyler worked on creating the custom parts with help of other members from time to time. The picture of the frame is from Februrary and we make a lot of progress every day. The custom parts are fitting well.

Custom Parts Progress

Due to the unique construction of the printer a few parts were made.

The first one was the z axis to frame upper connection. This piece was made form a 2" L bracket stock that we had in the cubby. its designed was inspired by the previous teams connections for the x and y axis. The same design could not be used due to the different ball bearing housing sizes.

The second and third part that needed to be machined were the z-axis ball screw support brackets.  These parts totaled in 4 parts, 2 of which being the mirror image of the other 2 (the finished two parts pictured below). The 4 parts all had the same external dimensions meaning that they were all milled from 3in x 3in x 1in 6061 aluminum stock,  This allowed all four parts to machined to size at the same time by orienting all the pieces in the same direction and utilizing facing operations to bring them down to the correct and and y axis dimensions.  Once the facing operations were complete and vise stop was used to reduce the number of times the edge of the part need tot be found and the thinner thickness shelf were machined and the two lard through holes were cut utilizing a hand programmed protract program to save time. Once this was complete the two tapped holes were drilled out on the mill out and then tapped by hand.

To simplify the machining progress the parts were modified from the original design to remove the triangular cut out as it served no functional purpose,  The original isometric drawing and modified isometric drawings are pictured below.

        Z-Axis Ball Screw Support Version 1.0                                            Z-Axis Ball Screw Support Version 2.0                

The fourth part that was custom machined was the top bracket support for the z-axis ball screw named Top Z-axis Ball Screw Support.  Two of these parts need to to me machined.  After analyzing the version 1.0 drawing it was realized that the part could be significantly simplified by creating a a custom L-bracket to adapt the bracket provided in the ball screw kit rather than machining a completely new bracket.  The Top Z-axis Ball Screw Support Versions 1.0 and 2.0 are pictured below.  The version 2.0 was machined from 3in right angle aluminum extrusion.  When a test fit was done it was realized the side of the bracket with only two holes on it needed to be slightly modified.  These modifications included slotting the two holes and removing excess material at the end of the face.

                   Top Z-Axis Ball Screw Support Version 1.0                                            Top Z-Axis Ball Screw Support Version 2.0                

The final part that was machined for the frame was few L-brackets.  These parts were machined simply because it was realized that we had not ordered enough and we had excess stock material and to save on the cost of buying brackets 4 were created utilizing the excess right angle aluminum extrusion stock.

Extruder

Obstacle 1: Extruder cannot push concrete through nozzle

The extruder was not able to print with the 1/2" square nozzle. The concrete would force its way upwards, past the auger instead of through the nozzle. This lead us to believe that an auger with a narrower gap between it's outer edge and the inner diameter of the PVC shell might give the concrete nowhere else to go but through the nozzle. Without the nozzle the extruder was capable of extruding the concrete without any issues. It extruded concrete as fast as we could feed it at a hydration level of 17% water.

Proposed Solution: Reduce the radial gap between the auger and the PVC shell. Reprint and test.

Obstacle 2: 3D Printed PLA Motor Coupler failed 

A new auger was designed and printed that narrowed the gap between itself and the PVC shell to 1mm radially. This caused there to be some friction and scraping between the auger and PVC tube. We tested the new design despite of these concerns. When we tested with concrete the 3D printed PLA motor coupler failed. The design used nuts screwed on the wiper motor shaft that fit into the 3D printed coupler which fit around the square tube which ran through the auger. These M6 nuts didn't have enough surface area to transmit the torque and cammed out. 

Proposed Solution: Print the PLA motor coupler in PETG. Test with concrete again.

Obstacle 3: 3D Printed PETG Motor Coupler failed

A new motor coupler was printed, but this time in PETG with the hopes that it would be slightly more robust. While the part did last slightly longer, we experienced the same issue as with the PLA part. We half expected the 3D printed coupler to fail, it was actually designed as a "mechanical fuse", but unfortunately it is failing prematurely. We will need to rethink this part of the design.

Proposed Solution: CNC an aluminum motor coupler in the machine shop. This coupler can directly thread onto the wiper motor and should be way stronger than a 3D printed part. The auger was filed down at the points it rubbed the PVC shell. Test with concrete again.

Resolution:

The aluminum coupler was machined and tested in the extruder. This coupler solved the issue of the printed parts failing during use. With this new part, the extruder was able to consistently extruder 16% hydration mortar mix as fast as we could feed it. The extruder also initially showed promise of extruding through a nozzle, yet for an unknown reason was only able to periodically extrude. We will have to continue diagnosing this issue, but at this point the team is confident the extruder will function without a nozzle.

The CNC machined aluminum part can be seen in the images below. V1 is on the left and V2 is on the right. V1 was designed to go around the square tube, while V2 was designed to fit inside the square tube. The drawing for the V2 part can also be found below.

The wiper motor power supply was wired up and an enclosure was created to ensure the safety of the user during operation. This was used all throughout testing and proved to be perfect for the task.

Duet

No changes with Duet. In regards to stepper motor wiring, finding a crimper has been a small pain since MSD office doesn't have crimpers in the size we need (or at all). An attempt was made to try and crimp using pliers but that was a failure as the resulting crimps wouldn't fit into the connector very well. Anthony has reached out to EVT in order to borrow one for a few hours to complete the wire harness.

Risk and Problem Tracking

Our updated Risk Management Tracking. Below is a high level overview of what we have addressed the last few weeks. 

Functional Demo Materials

Include links to:

• Pre-read
• Video of Extruder Test
  
  

Plans for next phase