Any time you print an FFF part, you have to decide what print settings to use. You may find yourself asking questions like: "How thick should the shell be?", "What infill density and infill pattern are appropriate?", and "What material should I use?" The settings you end up selecting have a significant effect on the performance of your part and can make it weaker or stronger and softer or stiffer.
Before SmartSlice, the only way to answer these questions was to (a) guess or (b) discover them through print-break-iterate cycles. Fortunately, SmartSlice takes the guesswork out of configuring FFF parts by using experimentally derived material data, simulation, and optimization algorithms to find the optimal settings for your part and its in-service application. This means the problems of yesterday are not the problems of today.
One question that our customers often ask us is about the minimum print settings used in our optimization workflow. They want to know what the minimum wall count, minimum top/bottom layer count, and minimum infill density are. The answer to this question is that our optimization routine uses 2 walls, 2 top/bottom layers, and 20% infill density as minimums. The reason we chose these values was because we came to a consensus that they are the minimum values one would choose when printing a functional part. We define a functional part as any part this has a purpose besides being decorative.
This begs another question: Can SmartSlice accurately simulate the behavior of parts printed with these minimum print settings? That is the focus of this validation study.
This is the 4th post in our validation series. The previous posts cover validations of:
Validating SmartSlice with Lightweight Parts
A simple lever was loaded at one end and the 2 cylinders on the other end were fixed. They were printed using 2 different materials: 3DXTech ABS and MatterHackers NylonG, which is a glass-fiber filled nylon. For each material, 5 parts were printed. Each part used the same print settings:
Figure 1 below shows a sliced view of the part. Notice how thin the walls are and how sparse the infill is. This is a "lightweight" part. As an example of how lightweight and thin these parts are, the infill pattern can be seen through the top/bottom layers as shown in Figure 2.
Figure 1. Sliced view of ABS part.
Figure 2. Pictures of printed parts showing the visibility of the infill pattern and demonstrating how thin the top/bottom layers are.
Each part was secured in a test fixture on a load frame (the test fixture is shown in Figure 2 of this blog post) and a load was applied to the part. A load cell and laser extensometer were used to measure the applied load and displacement and the resulting data was used to generate the stiffness of the part and the yield load of the part. The yield load refers to the load corresponding to the onset of yielding in the part. In other words, this is the load when the slope of the load-displacement plot becomes nonlinear.
In SmartSlice, the experimental loading and constraints were modeled by fixing the 2 cylinders and applying the load as shown in Figure 3. The Validate feature in SmartSlice was used to simulate this use case and calculate a minimum factor of safety and part displacement. From this data, the stiffness of the part and the yield load of the part were calculated.
Figure 3. Use case definition in SmartSlice.
How did SmartSlice stack up against the experimental results? Comparisons of the part stiffnesses and part yield loads are shown in Figure 4. For both ABS and NylonG, SmartSlice is in very good agreement with the part stiffness and is in very good agreement with the part yield load for ABS. The SmartSlice part yield load for NylonG is just 17% above the experimental value.
These plots also demonstrate the differences in performance between different types of materials. Filaments that have either glass or carbon fiber content are generally significantly stiffer and stronger than their unfilled counterparts. Of course there are some nuances. For example, fiber-filled filaments generally have the same or even lower interlayer bond strength compared to unfilled filaments, but that is a topic for a different blog post!
It is also important to discuss the differences in the amount of time it took to generate these results. The total time to print, test, and process the experimental data was approximately 20 hours. In SmartSlice, the total time to set up the use case, run the validations, and process the data was less than 1 hour. Time savings is one of the key values of SmartSlice.
Figure 4. Comparison of part stiffness and part yield load.
Minimize Print Time and Material Use with SmartSlice
SmartSlice makes it easy to find the optimal print settings for your parts. The optimization tool quickly explores the design space of key print settings and finds the settings that meet your part's strength and stiffness requirements while using as little material as possible. This study confirms that SmartSlice is able to simulate the performance of lightweight parts, printed using our minimum recommended print settings. In addition to finding the most appropriate print settings, SmartSlice can be used as a tool to quickly find the best material for your part and application.