Simulating a 3D printed part

- [Speaker 1] Simulating apart before it's built is a pretty mature process for some types of manufacturing. If you're going to machine a steel bracket of a well understood alloy, you can design it in a CAD program. If you know the forces it will encounter fairly well, you can probably be reasonably confident that your simulation will reflect whether the part will stand up to its intended use. The combination of simulation and testing should converge pretty quickly to a functional part. But for 3D print, a few more steps might be involved. In this movie, we give an overview of the types of simulation available for 3D printing and where in the process you might use each one. Since issues affecting the final part can arise at all stages of any manufacturing workflow, you need software procedures to ensure quality at each step. Simulation of a 3D printed part might need to occur at various points during the workflow of creating one. A 3D print starts as a digital model with a CAD file. We might use a simulator of the properties to end use part in that CAD program, perhaps with some simplifying assumptions. Generative design, a technique for automating the design of parts might be used at this stage. The process of slicing apart into layers requires a user or automated system to make a lot of decisions about optimal ways to print the part. Finally, we might capture data about the printers it's fabricated as well as some testing and other validation of the completed part. Simulation outputs might be compared against this real world result to improve the simulation, or to iterate the design of the part. Mechanical simulations of finished parts are often built into CAD programs. If the focus is on how strong a part is or how much it will deform, we use what's called finite element analysis or FEA. It's hard to analyze how a force will impact a complicated shape. However, it's relatively well understood how a force will interact with a simple shape like a cube. This type of analysis breaks the part and relevant forces on it into tiny volume elements. A program malls stresses on each element by solving a number of complex equations. The computer simulation may start at one end of a part, figure out how much force is being applied to the side of one volume element and work its way across the part. Some general purpose CAD programs are adding extensions to handle these issues for 3D prints. Various ad-ons and other programs are emerging in the ecosystem around these big programs too. End-use part modeling software is often used with estimated average or perhaps worst-case book properties to ask qualitive questions like what if I change materials? What if I change requirements on the part? What type of defects can I tolerate? These hypothetical simulations find good boundaries of a design space. They typically assume an ideal build process and homogeneous materials, which, as we'll see, can be problematic for 3D prints. If you want to learn the details on how to simulate a part in CAD even before we add the complexities of 3D printing, you need some additional coursework. There's a LinkedIn learning fusion 360 simulation course, and two solid works courses, solid works, simulation for finite element analysis and solid works advanced simulation. These are all aimed at the general case of simulating a homogeneous part that can get you started. - [Speaker 2] We've said that standard modeling software in mainstream CAD programs may need enhancing to accurately model a 3D printing. Let's explore why that is. Prints created with filament will be stronger within a layer than between layers, and will have infill instead of being solid. Resin prints might cure unevenly evenly after printing. So even though the printing process is isotropic, the prints may not be. Metal prints may build up internal stresses during printing or have internal voids. Any print might warp during the printing process. Parts are often complex pieces that would have been assembled or fabricated other ways. This means that the simulation needs to know about the details of slicing. Alternatively, slices will integrate varying degrees of knowledge about the details of the physical print. In either case, these models can range from simple geometrical checks to complex multi-physics simulations. Check out the other videos in this course, they go into more depth on various aspects of simulations that model a slicing or fabrication stages of a print. Simulation has the most value in building an expensive and time-consuming part such as a metal one. These parts may be able to leverage older simulations like welding ones. Plastic filament parts tend to be low cost, but hard to simulate. It may be more cost-effective just to make a test part itinerate a little. However, there is a place even filling in with printing for simulation to optimize slicing and orienting a print, particularly for large prints. - [Speaker 1] When the 3D printing process crosses over into physically making and testing a part, simulation data can be a valuable basis for comparison with the real part. As the part is flawed somehow, that simulation might give insight into what to change before we try again. Testing and validation may compare data about the fabrication or final state of the part to what was expected from a simulation. In this movie, we walk through the various places where simulation is used in the 3D printing process. Just because a part is born digital, it doesn't mean it's easy to simulate accurately. You can get into more detail on how simulation capability is evolving along with 3D printing and some of the other movies in this course.