Simulating Your Way out of 3D Printing’s Problems
Michael Molitch-Hou posted on April 28, 2017 | 5235 views

3D printing is quickly transforming from a method for prototyping and making tooling into a technology for manufacturing end parts. GE has just snapped up metal systems producers, and large manufacturers like Adidas and Oracle are using Carbon’s platform for end production. As 3D printing is incorporated into the larger manufacturing workflow, there will be plenty of issues to address.

In addition to workflow management concerns, such as those that SAP is attempting to solve, there is the problem of characterizing the additive manufacturing (AM) process. In many cases, AM systems can be difficult to control or predict, particularly when making a part the first time around. A component’s orientation and shape will completely change the nature of the printing process and the part’s ultimate performance. As a result, it may require multiple attempts to print an object before achieving the proper result.

Part distortion simulation of a manifold to be 3D printed on a Renishaw system. (Image courtesy of MSC.)
Part distortion simulation of a manifold to be 3D printed on a Renishaw system. (Image courtesy of MSC.)

One answer to these problems and more may be the use of simulation software, both for its use in characterizing the printing process itself and for predicting the performance of the part once it comes out of the printer. So far, there are few companies working to engineer simulation software for AM. One of the few is the recently acquired simulation company MSC Software, which has developed Digimat Additive Manufacturing (Digimat-AM) and Simufact Additive software for polymers and metals, respectively.

Simulating for 3D Printing

Both metal and plastic AM have their issues. With metal, parts can become distorted during the printing process and residual stresses may cause part or support failure during a print. Occasionally, a powder scraper will collide with a part and cause damage to the machine. Even successful prints must meet specifications related to porosity and microstructure.

In plastics, parts may warp or fail. With fused deposition modeling (FDM), bed adhesion may be an issue, with the initial layers of the print not sticking to the bed properly. For both technologies, the approach so far has been based on trialanderror, by modifying the print parameters or design until the print is successful.

To address these issues, MSC looked to physics-based simulations to optimize the materials, processes and structure for metal and polymer AM with the goal of enabling proper prints with the very first print. To do so, the company’s Simufact Additive and Digimat-AM software are designed to optimize the print materials, process and the structure of the designs themselves.

Both platforms use a layer-based voxel approach to simulating the printing processes, which provides a macro-level look that is much quicker to run than finite element analysis (FEA). As you can see in the video below, a digital model is represented as a stack of cubes enmeshed by an outer surface. As it simulates a printing process one full layer at a time, the software calculates the difference between the voxels as they will be printed and the part surface, determining how much the surface will be filled with the voxels.

Simulating for Metal

Simufact Additive is designed specifically for powder bed metal AM processes, such as selective laser melting and electron beam melting. For directed energy deposition processes, MSC suggests that its Simufact Welding software can be used.

Simufact Additive relies heavily on the concept of inherent strain, already utilized in welding, to determine the strain—creep strain, thermal strain, phase transformation strain and others—placed on a part during the metal 3D printing process.

The cantilever beam process used to characterize the inherent strain of a metal 3D printer. The image is taken from a webinar published by MSC and embedded at the end of this article. (Image courtesy of MSC.)
The cantilever beam process used to characterize the inherent strain of a metal 3D printer. The image is taken from a webinar published by MSC and embedded at the end of this article. (Image courtesy of MSC.)

To determine the inherent strain of a metal AM system, the operator fabricates three cantilever beams that are oriented horizontally, vertically and at an angle. The beams are then cut in half, causing them to bend upward. The displacement of the tip of the beam from the flat surface is then measured and the result is plugged into the software, which determines the inherent strain of the machine. The video below shows how a part would be distorted during a printing process.

Once this data is stored in a user’s database, it’s possible to run a variety of simulations, including how heat treatment will affect the part, part removal and the effects of hot isostatic pressing (HIP). With heat treatment, material properties are a function of temperature. Using the software’s built-in materials database, a metal’s performance characteristics—elastic modulus, young’s modulus, flow stress, conductivity, etc.—are incorporated into the digital model as it is virtually exposed to heat.

From left to right, Simufact Additive simulating the residual stress of a part after manufacturing, after cutting off the base plate and after removing the support structures. (Image courtesy of MSC.)
From left to right, Simufact Additive simulating the residual stress of a part after manufacturing, after cutting off the base plate and after removing the support structures. (Image courtesy of MSC.)

With HIP, the same variables are taken into account and the physics of pressure is added. The software is able to determine the additional stress and distortions made to the part as it is placed under high pressure. This will be essential to understanding the density and porosity of the part as it is pressed.

After performing a number of simulations, such as the strain placed on a part during printing or the densification caused by HIP, a user can redesign a part to compensate for such issues. For instance, if it’s known that a component’s thin walls will deform while being printed, the walls can be made thicker.

A user might also decide to choose a specific orientation in the printbed to minimize stress on one area of the design over another. This might expand to the placement of support structures, which can be located minimally and only where they are most needed to reduce the amount of support removal after the part is complete.

After support removal, residual stress of a part oriented horizontally compared to a part oriented vertically. (Image courtesy of MSC.)
After support removal, residual stress of a part oriented horizontally compared to a part oriented vertically. (Image courtesy of MSC.)

To validate the software’s capabilities, MSC partnered with the Fraunhofer Institute for Machine Tools and Forming Technology and NTT Data. The partners produced several parts and compared them to simulations made in Simufact Additive, and found that such issues as part distortion and failed support structures were captured accurately.

In the future, the software will also be able to perform meso- and microscale analyses of a part before it is printed.

Simulating for Plastics

Digimat-AM was developed for use with SLS and FDM in particular. In Digimat-AM, stress-strain curves are calculated based on the material that is being printed. This is determined by inputting the ratio of the polymer’s matrix material to its filler material. Due to the software’s built-in materials database, it’s possible to calculate the mechanical, thermal and electric properties of the material, as well as the warpage, residual stress and porosity of the process.

Key to plastic AM (especially with FDM), is the Z-axis strength, which is usually much weaker than the X- and Y-axes. Digimat-AM is able to calculate part failure as it relates to orientation of the part during the fabrication process.

All of this information can then be used to inform one’s design. A part may be reoriented in the printbed so that the areas in which strength is essential are printed along the X- and Y-axes.  Support structures can be placed only where necessary to save material and minimize post-processing.

Failure indicator distribution before ultimate failure of the plenum under pressure. (Image courtesy of MSC.)
Failure indicator distribution before ultimate failure of the plenum under pressure. (Image courtesy of MSC.)

To validate the software, MSC worked with Solvay, which developed a 40 percent glass bead reinforced polyamide material, called Sinterline Technyl, for selective laser sintering (SLS). The behavior of the material was calculated with Digimat-AM using the filler-to-matrix ratio, and test coupons were printed that validated these calculations.

Then, a plenum chamber for a car engine was simulated and printed in Sinterline Technyl material. The goal was to understand at what pressure the part would fail. Again, the simulation matched up closely with the performance of the actual part.

Simulating for the Future of 3D Printing

Given the current reliability and unpredictability of 3D printing technology, simulation may be key to avoiding wasted time and money when producing a part. Until recently, however, there haven’t been many options on the market.

3DSIM is one of the few companies—along with firms like ESI—working on tools designed specifically for AM, but its products are only in beta. While 3DSIM is incorporating high-level calculations that its team suggests should take billions of years to perform without 3DSIM software, MSC seems to have leveraged its history with simulation to create a fast solution. Once other simulation providers find the 3D printing niche, MSC likely won’t be alone in the market, either.

As simulation software is increasingly adopted for predicting the printing process before it’s performed, we may also see new tools developed for incorporating closed-feedback loops into the machines themselves. This will make it possible not only to produce better parts the right way the first time around, but it will also enable 3D printers that correct themselves before falling outside of the proper parameters in the first place.

To learn more about Simufact Additive and Digimat-AM, head to the product websites. Below, you’ll find an in-depth webinar that covers both software, as well.

Recommended For You