Additive manufacturing offers freedom of design and the ability to make tools and end-use parts in a more efficient and sustainable way. By printing parts that are close to final shape, this process reduces the amount of subtractive waste that is generated for enhanced efficiency and sustainability. It also uses less energy to produce the part than traditional molding and offers opportunities for recycling excess material or parts to promote circularity. 

Large-format additive manufacturing (LFAM), a subset of additive manufacturing, is growing in popularity. As LFAM usage expands, there are increasing concerns about what to do with large parts and scrap materials when a project is finished, other than landfilling (Figure 2). 

Initially, LFAM was used primarily to print one-off parts for prototypes or demonstrators, so there was little material to be reclaimed and recycled. As LFAM applications diversify to include end-use parts as well as tooling, jigs and fixtures, there are more shavings and end-of-life parts that need to be disposed of — ideally in sustainable ways. Manufacturers that use LFAM want to be responsible, but currently there is no established process for re-using LFAM scrap and parts. To advance this issue, SABIC and Local Motors completed a study to verify the viability of using LFAM regrind in manufacturing processes. This study, which looked at the printability and mechanical properties of reground LFAM parts and shavings, demonstrates that extending the useful life of this material is indeed possible. 

Reusing LFAM Parts and Shavings

SABIC and Local Motors studied the option of regrinding LFAM printed parts and machine shavings and the effect that would have on reprocessing the material in a LFAM printer. Reclaiming material this way is considered mechanical breakdown, in that the grinding process shortens the reinforcing fibers in the final pellets, which can reduce overall mechanical performance. In addition, when a polymer is put through multiple heat cycles (grinding, re-pelletization, re-compounding, etc.), this cumulative heat history breaks down the polymer chains and can degrade performance. However, mechanical recycling has long been used in plastic injection molding with success; reground sprues, parts and runners can be incorporated back into the process with acceptable results when good housekeeping practices are followed.

Based on the companies’ experience with LFAM processes, SABIC and Local Motors were aware that LFAM printers can be more sensitive to inconsistencies in material feed than injection molding and standard extrusion processes, due to the need to optimize print layer uniformity. Given the small barrel sizes and short residence times typical of LFAM machines, thoroughly mixing reground material in the barrel would be more difficult than with injection molding and standard extrusion processes, potentially leading to greater inconsistencies in the material. With that in mind, the two companies decided to have the regrind material pelletized before conducting the blending and printing study. 

The challenges associated with grinding LFAM parts and shavings stem from the large part sizes involved and the possible contamination with other elements. There is a limit to the size of a part that can fit into a grinder. Also, foreign materials could be present and potentially contaminate the regrind. These substances could include metal from machining and milling processes; build sheet residue attached to the part; and in-use contaminates (e.g., cement dust, adhesives, Bondo) that might adhere to the part or be mixed in the shavings. Separating out contaminates and cutting down parts into acceptable sizes would be necessary to make the material more suitable for conventional grinding processes.   

This study focused on SABIC’s LNP Thermocomp AM compound, a carbon fiber-reinforced grade for large-format additive manufacturing. (The material has been used for autoclave tooling, among other applications.) Local Motors cut down the LFAM parts into 12-inch-square blocks, making them easier to grind. An outside company was contracted to grind up the blocks and re-pelletize, package and ship the material to SABIC’s Polymer Processing Development Center in Pittsfield, Massachusetts, for printability and mechanical performance studies.

Evaluation of LFAM Regrind for Reuse

There were two parts to the evaluation of the reground material: 1) printability and 2) mechanical performance. To begin the evaluation, the reprocessed Thermocomp AM compound and virgin materials were manually blended to produce 100-pound samples containing 0, 15, 25, 50, 75 and 100% reprocessed content. Both types of material were in pellet form for ease of blending and feeding into the LFAM machine.

Printability Evaluation

To assess the printability of the material samples, a throughput test was performed, and a single-wall hexagon was printed for each sample (the hexagon is SABIC’s typical test part geometry for processing and material characterization) using SABIC’s Big Area Additive Manufacturing (BAAM) machine from Cincinnati Inc. Each printed part was examined and compared to a control part printed from 100% virgin material to evaluate surface appearance, consistency and print defects.

Throughput Test

A throughput (purge) test was conducted with each of the materials. After setting the temperature melt profile for the resin and purging for 10 minutes to ensure the equipment had reached thermal stability, the material was purged in various settings and locations to cover all rpm (revolutions per minute) conditions (100, 200, 300 and 400 rpm). Purge piles were allowed to cool and then weighed. During each purge step, the material was observed for melt fracture or other extrusion anomalies as it exited the nozzle. 

The throughput test showed that material flow was relatively consistent among the six samples, with one exception related to the 400 rpm purge test rate. Due to the viscosity shift of the material, which occurs as increasing amounts of regrind are added, the sample with 100% regrind was able to complete the 400-rpm purge test. With lesser amounts of regrind (i.e., 0 and 15%) the machine torqued out, preventing the testers from completing the 400-rpm purge test. Because the 25, 50 and 75% regrind loaded samples appeared to print in a similar way as the 0 and 15% samples, the 400-rpm condition was omitted for them. The increased flow seen with the 100% regrind sample prompted the attempt of the 400-rpm condition, resulting in a successful outcome.      

Printing of Single-Wall Hexagon

Once a temperature profile was selected, as determined by the above purge pile procedure, a single-wall hexagon was printed for each sample. At the start of printing the machine was further dialed in by adjusting print speed to achieve a bead size of 0.4 inches wide by 0.15 inches high. The hexagon’s sides measured 22 inches wide by 15 inches tall (Figure 3). To keep a few variables constant, all hexagons were printed with the same machine setup, with air cooling on the part and tip. The same temperature profile and slice/G-code file were used for all the prints.

Mechanical Performance Evaluation

Testing of each sample’s mechanical properties was performed using the single-wall hexagons. To obtain specimens for tensile and flex testing, shapes were cut from the hexagons with a Z-X orientation and an X-Z orientation using a waterjet cutting machine (Figure 3).