The STEM course at Mass Academy is taught by Dr. Crowthers. In this class, each student performs a six month long research project focused in either science, engineering, or mathematics. This project culminates in a school-wide science fair in early February which judges and peers from various colleges, universities, and organizations attend. The projects begin in A term with research and brainstorming, and throughout B and C term, more research, experimentation, and data collection is performed. Scroll down to learn more about my project!
Exploring Interlayer Bond Strength in Consumer-Grade 3D Printing
3D printing is a commonly used type of additive manufacturing, but printed parts have been shown to have an inherent anisotropic weakness in the vertical build direction, caused by a less-than ideal bond formed between extruded layers. Fused filament fabrication (FFF) 3D printing is thought to be unsuitable for commercial applications because the planar construction method used seems to result in weaknesses in the vertical build direction. The objective of this project was to explore the existence of this anisotropic weakness, and to test novel ways of increasing the interlayer bond strength. In an attempt to combat this anisotropic weakness and strengthen the bond between layers, physical changes were imparted on the surface of a specified layer in a test specimen designed specifically for this experiment. The layer was manipulated by increasing the surface area through scoring of the plastic with a metal blade, and the next layer was extruded upon that changed surface. A uniaxial load was applied to the printed test specimen which broke at the specified manipulated layer, and the force used to break the part was recorded. In contradiction to the hypothesis, the strength of the bond between layers in the modified parts was seen to be lower than the strength in unchanged printed parts. These findings show that physically manipulating the surface of 3D printed layers by solely scoring the surface with a blade is not an effective way of increasing the bond strength between said layers. This method of increasing bond strength was simple to attempt, and although it did not work as expected, in the future this method could be combined with other procedures to develop a way of increasing bond strength.
Click this link to access the Literature Review, Project Proposal, and Project Notes.
Engineering Need: Fused filament fabrication (FFF) 3D printing is not suitable for commercial applications because the planar construction method used can result in weaknesses in the vertical build direction.
Engineering Objective: The objective of this project is to engineer a printing method that will increase interlayer bond strength so as to eliminate the weakness of planar construction.
FFF 3D printing, the most common method of printing, extrudes filament layer by layer to build a part vertically. This method of printing has its benefits, such as the ability to create more complex geometry than traditional subtractive manufacturing processes. But 3D printing in this method results in weakness in the z-plane because the bonds between extruded layers are very weak relative to the tensile strength of the material itself (Rane, 2018).
The most common cause of poor bond strength is the temperature of the material when the bond is created. The print head moves relatively slowly, so by the time the next layer passes over a specific point, the layer below has already cooled more than desired. The optimal temperature for bonds to be formed is at a material’s glass transition temperature, when the molecules are most willing to intermix while still holding their structural shape (Ali et. al, 2020). If the layers were to remain at a temperature closer to their glass transition temperature, then the molecules of the layers would have a greater chance of intermixing, which would result in more effective layer bonds and stronger overall parts.
Engineers often have to be mindful of this constraining weakness in layer bonds, as using parts in situations where high stress is present in the layer direction is not feasible. The weakness between layers not only limits the possible applications of the parts but also requires engineers to account for this in setting up the printing of these parts. More work must be put into orienting the part on the printing platform so that the layers are not in the same direction as the stress being applied, and this change can also result in the need for extra support material. Print settings can be one way to address this problem, but they are imperfect and one benefit comes at the cost of another. For example, increasing the flow rate of the print will result in higher part strength, as the contact area between layers is larger (Rane, 2018). In exchange for this increase in strength, a higher flow rate will cause a decrease in the print’s geometric quality and surface finish, and the print will also take longer. A different way of increasing bond strength while still maintaining geometric proportions and acceptable print time is needed.
Some ways of increasing interlaminar bond strength are post-processing methods. An example of these methods is applying heat to the parts after being printed. The heat in the right quantities will partially melt the layers and cause molecular intemixing, making the parts and layer bonds stronger. Some examples that have been experimented with include the use of infrared lamps, concentrated lasers, and atmospheric pressure plasma. Post-processing has been shown to result in a significant increase in part strength, but this process is time-consuming and takes added effort after the part is printed. Addressing the issue of layer bond strength during the printing process would remove the added work required for post-processing methods, and also cut down the time and effort needed to create stronger parts.
The application of heat, specifically local heating, has been shown to increase the interlayer surface bond and overall part strength (Duty et al., 2017; Ali et al., 2020), but how would a physical change affect part strength? A way of mechanical interlocking has been used in large-scale concrete 3D printing, and positive results have been recorded (Khoshnevis, 2017). In this study, the researcher created square indents at regular intervals along the layers, and when the next layer passed above, the concrete was forced into the open space below. This method of creating divots or indents in a layer, which are then filled by the next layer extruded, could have the desired impact in small-scale, consumer-grade printing as well. Or, another way of changing the surface texture, such as scoring or scratching the layers could increase surface area and thus increase bond strength.
The focus of this project was to investigate the effect of introducing a physical change to the layers of a 3D printed part, specifically the change in behavior between two adjacent layers of the part. The test specimen used was specifically designed to fracture at a prescribed location between two known layers, this being done by designing an exaggerated notch in the test specimen in Solidworks. The test specimen was then printed in PLA material on an Ultimaker 3 printer. Using Ultimkaer's Cura slicing software, it was possible to pause the printing process at the exact layer of the part's notch, and impart the change to one (and between two) specific layers.
When the print paused, a clean and sharp razor blade was used to cut grooves into the surface of the part. These grooves were handmade and effort was put into ensuring their equal spacing and depth, however due to the imperfection of humans, this could be a cause of discrepancy in the data. Two diagonal sets of grooves were cut into the layer, at 90 degrees to each other. Once this manipulation was complete on all the necessary specimens, the print head was heated again and the print was resumed, with the next layer being extruded directly on top of this manipulated surface.
The aim of manipulating the layer was to increase the surface area of that respective layer, resulting in a larger area of contact between the two layers at which the break would occur. Once the prints were completed, a load was applied to the specimens, which broke at that specific layer, and the strength of the single bond at that layer could be measured. This experimental data was compared to a control group which was identical to the experimental group, minus the surface manipulation. The differences in force applied before failure were compared to obtain results.
Figure 1: Graph of force (N) versus time (s) showing the load applied to four test specimens of each group. Red lines show data from the control group while blue lines show data from the experimental group.
Figure 2: Table displaying force applied (N), time elapsed (s), and displacement (mm) of each test specimen of the control group. Force values are the force applied by the testing machine, and displacement is the distance the part stretched before breaking.
Figure 3: Table displaying force applied (N), time elapsed (s), and displacement (mm) of each test specimen of the experimental group. Force values are the force applied by the testing machine, and displacement is the distance the part stretched before breaking.
Figure 4: Image captured in Solidworks of the test specimen's geometry and unique dimensions. All measurements are in inches.
Figure 5: Image of a printed test specimen. Printed on an Ultimaker 3 using silver PLA at 0.4mm layer height.
Analysis + Conclusion
The data collected manipulating the surfaces in this way, by solely scoring the plastic with grooves inserted using a razor blade, shows that the overall part strength is decreased when scoring is introduced. The control group, to which no change was introduced, had an average fracture strength of 734.0 N. The experimental group, in which two-directional scoring was introduced, had an average fracture strength of 581.1 N. Based on these average group strengths, there is observed to be a 20.83% decrease in overall part strength when the surface of a layer is manipulated in this way.
The hypothesized results of this project would demonstrate that changing the physical properties of a 3D printed layer’s surface will increase the strength of the bond between layers. The data collected suggests that this hypothesis is incorrect when the surface of a part is manipulated in the ways tested. When the parts were scored using the razor blade as previously described, the resulting bond of the layers adjacent to the scoring was less than the bond between the same layers with no scoring. However, some portions of the layers were seen to bond to the grooves as expected, which shows promise in the idea that increased surface area is a plausible way of increasing bond strength. When only manipulating the surface as tested in this experiment, it is clear that the bond strength decreases. It is possible, however, that if this physical method of manipulation is used in combination with the application of heat or pressure, physical manipulation in some capacity could increase part strength. Heat application, in various different forms, has been explored and extensively tested, and positive results in part strength have been noted. If one of these temperature-based methods were to be added on to the currently tested methods of this experiment, the combination of the two could very well have a joint positive impact greater than their individual results.