Detailed Design

Team Vision for Detailed Design Phase

At the start of this phase, the team hoped to complete the development, design, and testing of a pumping system, dosing system, mixing system, extruder, and frame. Each subsystem was in a different development state at the beginning of the phase. The goal was to drive the design of all of these subsystems to completion. There was additional design feasibility work to be done during this phase, that came in the form of prototyping of the mixing system. By the end of this phase, ideally, we planned to have all integral subsystems designed and tested. However, our prototyping of the mixing system design showed that the design had too many flaws to function correctly. The work required to design and validate a mixing system that would meet our engineering/customer requirements would take too much time away from our other subsystems.

One key decision made this phase was the choice to move towards a manual-mix printer design. The user will mix the concrete on their own outside the printer, then feed it into the printer's reservoir. The user will then have to refill the reservoir whenever needed. The extruder design also saw significant progress in this phase, from the extruder system to the nozzle design. Several tests were done on the prototype extruder design and we have progressed far enough to where we are confident that with a few updates the extruder will function as needed.

Progress Report

In the preliminary design phase, concrete testing was performed, and concrete type and hydration levels were narrowed down. The testing pointed to Sakrete Type N and Type S as both being viable concrete options. If concrete is needed to test a sub-system this type will be used at the manufacturer specified water ratio. The concrete recipe testing will be put on pause. We believe enough testing has been put into the concrete testing, and therefore it is low risk to pause our concrete selection for the time being. Attention will be directed towards other subsystems to ensure time is spent developing accordingly.

Additional testing of the mixing system was completed in this phase. This consisted of a total of two tests that determined the feasibility of the preliminary design. Last phase, the team tested a prototype system consisting of PVC, a metal auger, and a drill. This experiment tested if this system could mix properly.  The previous test provided us with some adjustments and improvements to make to our system. The second test was performed during the detailed design phase. This included improved methods of providing concrete and water, testing with more material in the system at a time, and larger diameter auger to fit snugger in the PVC piping. Results directed the team that major development would be needed for this concrete mixing methodology to work. The team has made the unanimous decision to pursue alternate options. This will be discussed more in-depth below in section "Volumetric Mixing and Dosing System" and "Pumping and Hopper Mechanism".

The team began deeper investigation of where the previous team left the design of the printer frame and Z-Axis. This is to prepare for the large purchase of 80/20 needed for the frame and to review/validate their design. Frame brace connectors have been redesigned as the previous teams design seems inadequate for the applied loads. The brace design will provide the frame with a more secure and secure interlocking system. Extrusion lengths have been adjusted as not all in the CAD model were correct.

Lastly, during this phase, the team began deeper investigation of the duet board and started testing it's functionality with the extruder system. The current extruder design is powered by a Nema 17 stepper motor, so the duet was used to control this stepper during testing. This provided the team members an excellent change to familiarize themselves with the board, it's firmware, and GUI. Adjustments were made to get the extruder system and duet board updated and ready for testing. This will be discussed below in section "Extrusion Mechanism" and "Duet Board".

Concrete Recipe

Prototyping, Engineering Analysis, Simulation

During this phase, a test was conducted to narrow down the water level options of the two potential cement choices. This test was conducted by following the cements recommended measurement process. This included measuring out the desired amount of cement needed and recording the weight. Then calculating what 20% of the cement's weight is. Then that weight will be measured out in water. The cement and water is then mixed to create the cement solution. See test plans in next section for more information. 

Moving forward from this point, it will be assumed the final recipe will be either Sakrete Type N at 20% water to cement or Sakrete Type S at 15% water to cement. Further testing will be conducted after development of other sub-systems have been completed.

Test Plans

Volumetric Mixing and Dosing System

Prototyping, Engineering Analysis, Simulation

Prior to our detailed design phase, we thought it would be possible for the Concrete 3D printer to utilize a volumetric mixing system to produce concrete on demand. This system would require a method to properly mix concrete while accurately dosing the correct water to cement ration. While this design would be the most elegant and perhaps the best design for a production level concrete 3D printer, through feasibility testing it was determined to be far too complex to be within the scope of this project if all the other customer requirements were to be met.

The preliminary design for the dosing system of the printer's  volumetric mixer utilized an adjustable flow water pump for the dosing of the water. For the dosing of the dry concrete a spinning wheel would fill with a known amount of concrete mixture per revolution and then be dumped into the mixing chamber of the volumetric mixer as seen in Figure 1.

Figure 1: Preliminary Dry Concrete Dosing Mechanism Sketch

The preliminary design of mixing portion of the volumetric mixer utilized a horizontal auger design as can be seen in Figures 2 & 3.

Figure 2: Preliminary Mixing Chamber (the inlet for the dry and wet concrete components is the circular hole on the left side and the outlet for the concrete mixture is at the end of the reducer on the right)

Figure 3: Preliminary Mixing Chamber Exploded View

This design was tested in Test Plan S4-Mixing System Feasibility Part 2 and found to not be an effective method of thoroughly mixing concrete.  This failure to mix could be attributed to the mixing chamber design but it could have also been attributed to the method material was being inserted into the prototype mixing chamber. We were unable to find an effective method of gradually adding unmixed material into the mixing system at a constant rate. 

After this failure and acknowledging that it was likely going to take a lot of design work and feasibility testing to get the mixing chamber and dosing system to completion it was decided that volumetric mixing should be abandoned, and a manual premixed concrete mixing method will be pursued. For consistent mixing either a standalone concrete mixer will be purchased, or a concrete mixing drill attachment will be used. 

Test Plans

Pumping and Hopper Mechanism

Prototyping, Engineering Analysis, Simulation

After researching pump designs and their uses it was found that a peristaltic pump would likely be our best option to move concrete from a reservoir to the nozzle. We found that with properly defined design parameters  it would be possible to have a peristaltic pump that could produce a continuous flow of concrete.

Key Design Parameters are as follows:

• • Occlusion: Target should be between 10% and 20%
  • Pump Diameter
  • Roller Diameter
  • Tubing Abrasive Resistance
  • Tubing Inner Diameter
  • Tubing Bend Radius
  • Tubing Wall thickness

One of the most important parameters in defining the design of the peristaltic pump is occlusion. Occlusion defines the space between the roller and the wall of the pump housing.

Figure 4: Occlusion Calculation Formula

Given this equation it is clear that tubing thickness will be very important in defining the internal layout of the pump.  The team came to the conclusion that it would be beneficial to find a low cost tubing to exist within the pump in order to minimize the cost of replacing the tubing inside of the pump. The tubing within peristaltic pumps tends to wear quickly due to the crushing action that is necessary for the pumps to function.  The pump tubing will then have a fitting on the inlet and outlet to meet with a more abrasive resistant tubing to go to the reservoir and print head respectively.  This abrasive resistance will be important in order to minimize the wear on the tubing by the concrete flowing through it.  Due to the amount of research and calculation that was needed to begin the peristaltic pump design a reservoir design has yet to be established. The reservoir as well as a complete pump design will be completed before the start of the spring semester. This is addressed in our Plans for Break section.

Drawings, Schematics, Flow Charts, Simulations

Figure 5: Preliminary pump internal assembly.  The central drive shaft is hex shaped in order to constrain the shaft to the roller assembly.

Figure 6: Preliminary roller design.  Only central part of roller will be in contact with the tubing to provide the required squeezing force. The flanges aid in keeping the tubing in place.

Test Plans

The next phase will include testing to further develop the pump design. A specific test that will be completed for the peristaltic pump will be crush force test on selected tube samples. This test must be done in order to properly spec out a motor to ensure that the torque of the selected motor will be able to crush the tube with material in it. Tubing bend radius will also need to be determined in order to ensure that the housing is designed to fit the tubing. Final testing once a prototype is built will include pressure and flow testing in order to test the pressure and flow rate generated by the pump. 

Frame and Motion System - Z-axis

Prototyping, Engineering Analysis, Simulation

An informal test for stepper torque strength was conducted to verify linear speed of the Nema 17 stepper motors while under load. This test used an existing XY axis motion assembly and orientated it upright, so that it created motion normal to the floor. A 30 lb weight was temporarily attached to the carriage and, using the duet for control, was moved upwards. The carriage moved with considerable speed up the acme rod and did not appear impeded by the test. With this knowledge we know that a smaller stepper motor with our ballscrews/acme rod combination can lift at least 30 lbs and we can use smaller motors if required during testing. We plan on using a stronger motor that is also geared to increase torque since speed will not be a limiting factor. The printer will move very slowly and intermittently in the Z direction because of the layered nature of 3D printing, so a slow but torquey stepper motor should work perfectly for our Z-axis.

Drawings, Schematics, Flow Charts, Simulations

Figure 7: CAD assembly of previous team's printer

Upon closer inspection of the previous team's CAD files we found that many parts didn't fit correctly and other parts needed redesign. When first inspected there were many parts clipping each other. In an attempt to rectify this, extrusion lengths were updated and assemblies were recreated. 

Figure 8: Custom brackets that are found in the previous team's design.

After reviewing this design, the team concluded that these custom brackets needed redesign. We want to avoid custom fabrication if possible and strength of these brackets is a concern. The cross sectional area between the vertical and horizontal mounts is minimal. These brackets are responsible for holding the entire Z-axis motion assembly in place. We have decided to find an off-the-shelf 80/20 part to replace these. This should reduce cost, fabrication time, and increase rigidity.

The CAD assembly will be redesigned to have correctly sized parts as multiple updated components. The initial BOM for the frame is below. It is subject to change slightly and shall be finalized before the next phase.

Extrusion Mechanism

Prototyping, Engineering Analysis, Simulation

We continued refining the extruder design that was initially created by the previous team. We are treating the previous teams design as our first prototype and thus we performed feasibility on this design to determine whether it would work for our purposes.

Drawings, Schematics, Flow Charts, Simulations

Below is the prototype design for the concrete extruder. This extruder uses standard PVC parts, various 3D printed parts, square metal tubing, and a Nema 17 geared stepper motor. All of these components are easily replaced off-the-shelf parts. Areas of improvement are the auger design, stepper motor selection, nozzle design, and bearing system.

Figure 6: CAD models showing the extruder design. Cross Sectional view shows the internal auger and bearing system.

Identified Areas of Improvement:

Auger Design: Current auger design does work, yet the fit to the PVC wye could be improved. Diametrically the auger should be designed with a reduced gap between the ID of the PVC and the OD of the helical thread. The length of the auger should be increased to fill the length of the extruder. A more robust auger will be needed for the final design, as the 3D printed PLA auger isn't going to last long when operating with concrete.

Stepper Motor Selection: The current stepper motor doesn't provide enough torque to extrude concrete through the nozzle. We will perform torque/extrusion testing to determine what amount of torque will be necessary for extrusion. This will change the design of the stepper mounting block colored in blue.

Nozzle Design: Current nozzle prevents the flow of concrete. While a larger stepper motor may solve this issue, it is in our best interest to also investigate the nozzle design to make improvements. The current design has a high angle of approach leading to the outlet. This is likely helping to impede concrete flow. We will redesign this and test it during our torque/extrusion tests.

Bearing System: Current bearing system consists of 3D printed parts. We will need something more robust for our final design. Need to look into off-the-shelf bearings that will fit our design, they will need to be double shielded to prevent the inlet of moisture and abrasive concrete. May want to look into a plain bearing with no rolling elements to increase lifespan.

Test Plans

The extruder test plans span a few tests over a week period. The initial test was to see if the extruder design that the previous team put together would work, and it did not. The failure was due to the Nema 23 that the previous team purchased for the extruder; it was not able to provide anywhere near the required torque to extrude concrete. The solution was to swap out the Nema 23, which had a pull out torque of about 0.5 Newton-meters for a geared motor that had more torque capabilities. We had a geared Nema 17 on hand, with a gear ratio of 5.18:1. This was for use with the Z-Axis but we borrowed it temporarily to test with the extruder system.

The 5.18:1 geared Nema 17 has a max torque of 2 Newton-meters. The extruder system with this stepper motor could extrude concrete well; however, it was not able to extrude any concrete once the previously designed nozzle was put into place. This informed us that the nozzle was a problem, as well as the required torque being above 2 Newton-meters by some margin. We did a quick redesign of the nozzle to have more tapered shape and ran the test again. The geared Nema 17 was barely able to push concrete out of the new nozzle with the motor skipping several steps. A measurement with a torque wrench showed that the required torque was less than 5 Newton-meters, because that is the minimum the torque wrench is able to measure. 

Although technically the tests were a failure in that the design does not work as is, they clearly outlined issues and areas that need improvement. We now know that the required torque is in the range of 2 to 5 Newton-meters. Thanks to the Duet web controller we also know the speed information of the stepper motor. The Nema 17 belongs to the Z-Axis subsystem so the extruder is going to need a new stepper motor, the only question is which stepper motor is optimal in terms of torque and speed for our needs. Some geared Nema 23 stepper motors would be more than able to handle the torque required but have very slow RPMs which would limit the extrusion rate. Several stepper motors online are missing their respective torque-speed curves so we have reached out to the manufacturer to get those so we can make an informed stepper motor selection.

Duet Board

Duet Connections

The Duet board will be the controller board used for all onboard motors and electrical circuitry. The main benefit to using the Duet instead of a different board is that the Duet has pre-built interfacing with NEMA Stepper motors, which makes it extremely simple to test and troubleshoot issues in the design. The interfacing software is a site called Duet Web Controller, which can be accessed through the wifi-module of the Duet. This allows us to test the Duet without having to write any code. A current list of connections is below.

There is a lot of remaining work to be done with the Duet. Now that the overarching design work is complete, the Duet subsystem can receive a lot more focus. Specifically, the custom configuration requirements the Duet needs in relation to our printer and the motion control. Smaller Duet subsystems will also be planned out in detail, such as the calibration system components and the user interface. The Duet offers a lot of capabilities and there are some interesting functions we can implement on the printer. 

Bill of Material (BOM)

Completed BoM (Google Drive Link)

Completed BoM (Excel download - may not always be updated)

Design and Flowcharts

All the feasibility testing done during this phase has informed the decision to get rid of the automatic mixing system architecture in favor of the manually mixed system.  The manual architecture diagram has been updated to reflect some of the changes of this phase, such as the peristaltic pump design transporting mixed concrete from the reservoir to the extruder design. Another change is the stepper motor to Duet board integration has been cleaned up since we now have a better understanding of where we stand in regards to stepper motors required. 

Risk Assessment

A live Risk Assessment document can be found here.

Design Review Materials

• Pre-read is linked here 
• Presentation is linked here

Plans for Break

Between now and the end of the break the team hopes to complete the peristaltic pump design as well as finalize some bills of materials. We do not expect to do much more testing or prototyping for the rest of the semester. In order to prepare for MSD II some subsystems will be worked on over the 6 week winter break. Anthony will be working on the electrical subsystem. As mentioned before, the Duet needs custom configuration files for the connected motors and since we will be having a new extruder stepper motor, we will need new configuration files. In addition, certain minor subsystems such as the user interface and the calibration system can also begin to get designed since we now have a good idea of the direction we are going with the project. Wiring diagrams can also begin to be drawn up but subject to change; they will not be needed until the frame is assembled. Any related code will be researched and by the start of next semester we will have a solid understanding of the scope of the Duet within this project.

Plans for Next Phase

As a team, we would like to be prepared for the start of the spring semester term. This includes preparation for MSD II and any left over tasks from MSD I. Below are individual tasks the team will work on to help our team achieve these goals.