Feed Throat Design on Single Screw Extrusion 3D Printer
Malachi Alvarez, Zijie Liu, Brandon Sielski, John Torresani
University of Wisconsin, USA
Abstract
Single screw extrusion printers allow a wider variety of materials to be printed using an FFF style printer. An issue with the screw extrusion FFF printer is that the pellet material gets stuck in the throat while feeding into the screw. To solve this issue testing was done to determine the reason for the pellets getting stuck. Testing showed that the pellets were getting stuck because they were getting above the glass transition temperature and melting together. Because the sticking was not due to a geometry issue, the throat needed to be redesigned to cool the throat. The throat redesign introduced a fitting for compressed air that would cool the pellets as well as keep them moving to prevent melting and sticking. The resultant design stopped the pellets from sticking and allowed the printer to print smoothly.
Introduction
Additive Manufacturing (AM) techniques have risen to the forefront of engineering given their capacity to reproduce intricate design features that would otherwise be difficult to attain through traditional subtractive manufacturing processes [1]. One of the branches of extrusion additive manufacturing is screw-based extrusion systems, which can use the material in pellet-form that is too brittle to be made into filament. In our project, the printer that will be optimized has both a screw extrusion system and a filament extrusion system, as shown in Figure 1. The benefit of this type of printer is that it can extrude two materials, one of which is used as a support structure.
Figure 1. The single screw extrusion 3D printer.
However, the pain point of employing this 3D printer to generate any model is the discontinuous feeding due to the clogging issue of pellets in the feeding zone. The purpose of this project is to improve the performance of the feeding zone of the single screw extrusion printer to avoid printing failure due to the discontinuous extrusion.
Funnel flow such as the flow of pellets through this throat is typically obstructed either due to ratholing or bridging. Ratholing is the formation of a central column over the exit hole whereby all material is resting along the walls and not flowing through the hole. This is usually encountered in fine-grain and/or cohesive materials that tend to stick to each other or that have high inter particular friction. Bridging, on the other hand, is the formation of an arch over the exit hole whereby a dome supports the material in the funnel instead of allowing it to flow. This can occur with particles of high cohesiveness or with particles that can lock together geometrically [2]. Bridging through geometric locking may cause interrupted flow in this application as plastic pellets usually have edged geometry and are relatively large.
There are two additional hypotheses for what the fundamental reason is behind the pellets getting stuck. One is that the surface finish of the printed throat is not smooth enough, which gives more friction during the pellet feeding. The other is that the design of the throat is not suitable for pellets with sharp shapes which can catch on surface features. The goal of this paper is to release the fundamental reason for the clogging issue and report an effective design of the feeding throat to solve this issue.
Manufacturing Considerations
As our work is based on the design optimization of the throat, the original design was received directly from Fused Form, as shown in Figure 2. To determine the cause of the pellets getting stuck, our work will be mainly separated into two steps, as shown in Figure 3.
Figure 2. Original Design of the Feeding Throat from Fused Form.
We will print the part with a Stereolithography (SLS) printer first. The reason for it is that the parts printed by SLS are chemical bonds instead of mechanical ones generated by extrusion-based AM technologies, like Fused Filament Fabrication (FFF) [3]. The chemical bonds result in a better surface finish. Additionally, the black color of the current design’s FFF filament makes it tough to analyze how the pellets flow through the throat. The transparent material will be selected to contribute to the observation and analysis of pellet flow in the throat. If the first proposed solution works, we will focus on how to install the throat from SLS on the single screw extrusion printer.
Our second plan is to improve the design of the throat. The feed flow observed from the transparent SLA part will play a big role in this step. More specific information will be provided after the performance of the first solution.
Figure 3. Flow Chart of the Initial Plan
Preliminary Work
One concern with using the SLA printer was resin being trapped inside the model. A preliminary model was printed at the 30% scale to see whether the resin would be trapped on the Formlab printer at Makerspace. The resin was not trapped between the supports and our object as shown in Figure 4.
Figure 4. a) The first printed part of the 30% scale from the SLA printer; b) The side view of the part with the inside support structure.
Another concern with the SLA printer is that there were a couple of supports inside the throat which were hard to remove and would easily disrupt pellet flow. Another throat was printed in a different orientation designed to eliminate the internal throat supports, as shown in Figure 5.
Figure 5. a) The second printed part of 30% scale from SLA printer; b) The side view of the part with no inside support structure.
The second prototype was sliced so that the main tube path was perpendicular to the print platform. This eliminated the need for internal structural support because it reduced overhangs in the geometry that typically require support structures. Further work includes printing a full-size throat using the exact same orientation and determining if the structural strength is the same.
To make the throat transparent, a clear resin will be used on the Formlab printer. The transparency of the part is not high enough for our purpose to observe the feed flow so it will be polished after the print. Some preliminary research on how to polish the surface shows that sanding the part initially at a rough grit all the way up to 2500-4000 grit can produce a clear part. Additionally, an acrylic clear coat spray can be applied to increase the transparency [4] [5].
An initial attempt at increasing the transparency of a part printed in clear resin was performed using 2500 grit sandpaper. Acrylic clear coat spray was unavailable, so super glue was used instead. Using a clear fluid to resurface the part provides a smooth exterior which improves transparency. Practically, the fluid should solidify and remain clear in order to provide similar funnel characteristics to the original funnel.
Figure 6. Left: Original Part Transparency. Right: Treated Part Transparency
Blocking Reason
From the literature review, the types of blockages in hopper/funnel systems are mainly divided into two types, namely ratholing and bridging, as shown in Figure 7a. The throat was removed from the printing head in order to investigate if either of these phenomena were to blame for the blockage, however, pouring pellets through the throat when disconnected revealed no signs of either phenomenon. An alternative hypothesis for throat blockage was that the pellets were sticking together in a near-melt condition due to elevated temperatures. A sample of pellets was placed in temperature conditions similar to those in the throat region. Upon removal, the pellets in the sample had fused, indicating that the primary reason for blockage was the pellets fusing together. The printer was run with pellets in typical operating conditions and the feeder was removed to reveal a mass of fused pellets at the feeding zone seen in Figure 7b. This confirmed that the blockage was attributed to elevated material cohesiveness which is most reminiscent of the ratholing phenomenon.
Figure 7. a) Schematics of Two Types of Blockages; b) The Blockage State on Botero’s Throat.
The pellets should be melting within the screw extruder, not in the feeder throat itself. Melting in the throat causes the pellets to get stuck together and coagulate into one mass which stops the flow of material to the extruder. There is no cooling around the throat or the top of the extruder, which causes the pellets to achieve near-melt temperatures. In a normal screw extruder, there would be a cooling jacket around the feeder to maintain solid-state feed material, and heaters along the extrusion path would then melt the pellets in a constrained volume as the material advances as shown in Figure 8.
Figure 8. Single Screw Extruder at Melting Section [6].
In order to observe the pellet’s performance in the feeding zone, the throat was printed with clear resin, processed to improve transparency, and installed on the printing head as shown in Figure 9.
Figure 9. Installation of the Throat Printed from Clear Resin.
Solutions to Blocking Issue
The initial reprint of the throat did not solve the issue, but it did provide some insights into what the real problem was. Analyzing the flow of the pellets and their properties showed that the reason for the clumping was that they were melting in the throat. Two methods are being pursued to solve this issue. The first method is adjusting the temperatures of the heating zones. Tests will be performed to determine how low the temperature can be in the first heating zone without preventing melting within the screw. The second method to further lower the temperature of the pellets without compromising the melting in the screw is to implement a cooling system. The cooling system would use compressed air that is already supplied to the screw and split it to also cool the pellets. Further testing will be performed to test both solution methods and determine if more design is needed. This process is outlined in the flowchart in Figure 10.
Figure 10. Process for testing two solution methods.
Adjust temperature profile
The material that we are using is ABS/BiTiO3 composites with 5% plasticizer by weight. The result of TGA shows that the glass transition temperature (Tg) of the composites is lower than the room temperature [2]. As the feeding zone requires the material to be still in pellet form, the temperature should be lower than Tg. Therefore, the original temperature profile, as shown in Figure 11a, is high enough to make the pellets deformed. The most direct way to solve this issue is to decrease the temperature in zone 1 to 0? and zone 2 to 120?, as shown in Figure 11b. However, zone 2 is close to zone 3 and therefore suffers heating due to proximity. Even set to 0?, the temperature at zone 3 still averages around 40?, so designing a cooling system is necessary. The set temperature of zone 2 could not be lowered due to clutching. If the temperature of zone 2 is lower than 120?, the materials would be solidified in zone 2.
Figure 11. a) The original temperature profile was set at different heating zones; b) The improved temperature profile for solving the early-melting issue.
Design a cooling system
Addressing the elevated temperature at Zone 1 is critical to preventing the pellet-form material from sticking to itself and forming a clogging mass. Cooling the entry point of screw extruder systems is commonly achieved with a cooling jacket. Given the tight geometry at the screw entry point, integrating an effective cooling jacket would likely require design changes to the existing components surrounding the throat.
As an alternative to a cooling jacket, a pneumatic cooling system was developed to take advantage of an existing feature of the printer unit. The main pellet supply for the printer utilizes a compressed air system to maintain consistent flow in the pellet hopper. A second hose could branch off of the air supply and serve a similar function at the screw entry point. This could place cold air directly into the hottest part of the throat and jostle the pellets within the throat to further prevent sticking. Figure 12 shows a diagram of the proposed solution.
Figure 12. Proposed air-cooling system
The printer has a switch-controlled air feed and pressure control system used for the main supply hopper shown in Figure 13b, so the initial test was done by allocating the hopper air channel to feed the throat instead. This allowed the cooling system to operate with upstream flow controls. The throat was designed with an embedded air channel ending in a low-profile tube hole at the supply end. An effective press-fit tube attachment point was achieved with a 2 degree draft shown in Figure 13a.
Figure 13. a) Tube-to-throat attachment method; b) Air control system and tube connection points.
Leveraging 3D printing allows for controlled airflow through the walls and interior of the throat and for the hose attachment point to be shifted back away from the screw where there is ample space to avoid extruder hardware. Figure 14a shows the air route through the embedded air channel. Controlling the airflow through geometry allowed for directed entry into the throat chamber. By creating two perpendicular and offset air entry regions it was hoped that the airflow would form a cyclone which rotated the pellets while they moved through it shown in Figure 14b. The idea is that swirling the pellets reduces inter-pellet contact and sticking and improves airflow and heat transfer from the pellets to the air.
Figure 14. a) Air flow through walls of throat structure; b) Idealized vortex formation from controlled air entry points.
Future Work
While the current setup does work, it can be optimized. Several parameters can be varied: air pressure, timing of air on/off status, hole size and placement, main air channel design, temperature settings, pellet flow control, and exhaust air routing. The air pressure, timing of air, and temperature settings are more easily regulated and changed. The hole size and placement, main air channel design, pellet flow control, and exhaust air routing involve changing the geometry of the throat and a complete reprint. For each of the changed parameters, a base printing test will be done to compare the changes.
The system should also be modeled with the pellet motion and airflow in the throat using CFD. This could reveal points of inefficiency and can be corrected with a change in the geometry of the throat. Heat transfer can also be characterized from the extruder to the pellets, and a change of geometry in the throat or an addition to the system (such as an insulator) could reduce the heat transferred to the pellets. The CFD could also determine the secondary effects of the airflow, such as how it affects nozzle extrusion.
Another concern is the routing of the air. The air was routed out of the pellet feeder to the throat. This could cause the pellets to not be fed properly, reducing the efficiency of the printer. Testing for this could involve timing how long it takes to fill up the throat with and without the air feeder.
Conclusions
FFF printing is a very useful technology and integrating single screw extrusion with a 3D printer allows for printing with different materials. This project was designed to fix a problem with the extrusion system caused by the pellets feeding into the extruder melting together. After testing showed that the temperature at the screw inlet was higher than the glass transition temperature a new throat design was created to cool the pellets. Changes were made to the heating zones to decrease the temperature of the pellets and the cooling features integrated with the throat stopped the premature melting and resulted in successful printing. Future work is needed to better understand the effects of the cooling jacket on print quality and characterize the flow of pellets through the throat.
References
- Ngo, T. D., Kashani, A., Imbalzano, G., Nquyen, K. T. Q., & Hui, D. (2018, June 15). Additive manufacturing (3D printing): A review of materials, methods, applications and challenges. Redirecting. Retrieved February 28, 2022, from https://doi.org/10.1016/j.compositesb.2018.02.012
- “Solve or Prevent Poor Flow,” Jenike & Johanson. https://jenike.com/solutions/solve-or-prevent-poor-flow/ (accessed Mar. 02, 2022).
- How 3D printing direction/orientation affects strength? FacFox Docs. (2019, November 26). Retrieved February 28, 2022, from https://facfox.com/docs/kb/how-3d-printing-direction-orientation-affects-strength
- Polishing Clear Resin for a transparent finish. FormLabs. (n.d.). Retrieved February 28, 2022, from https://support.formlabs.com/s/article/Polishing-Clear-Resin-for-a-Transparent-Finish?language=en_US
- Guide to clear 3D printing. Formlabs. (n.d.). Retrieved February 28, 2022, from https://formlabs.com/blog/3d-printing-transparent-parts-techniques-for-finishing-clear-resin/
- T. A. Osswald, Understanding polymer processing: Processes and governing equations. Munich: Hanser Publishers, 2017.
- G. A. Mazzei Capote, M. C. Montoya-Ospina, Z. Liu, M. S. Mattei, B. Liu, A. P. Delgado, Z. Yu, R. H. Goldsmith, and T. A. Osswald, “Compounding a high-permittivity thermoplastic material and its applicability in manufacturing of microwave photonic crystals,” MDPI, 28-Mar-2022. [Online]. Available: https://www.mdpi.com/1996-1944/15/7/2492/html. [Accessed: 02-May-2022].