Learning Objectives
By the end of this section, students will be able to:
- Comprehend design challenges around software.
- Learn about alternative file formats to the .STL file.
- Acquire knowledge about adopting AM within industry.
- Understand the design maturity phases according to a maturity model.
As stated previously, a new technology does not come without its caveats. Successful designs in AM acknowledge and address both technical and non-technical risks to implementation. This section discusses risks to design software users, and development on the horizon to mitigate them. Also discussed is the cultural resistance to AM that impedes adoption, and how to strategically address those. Phases of AM design maturity are defined and matched to a specific design optimization technique.
Software Interoperability
From a design perspective, perhaps the largest challenge in industrializing AM revolves around the overall difficulty of moving from a CAD concept to a software code that an AM machine can interpret. The software interoperability is defined as the functionality of different programs to exchange information, share files and use the same protocols. Within AM, this process flow path is what is known as the AM digital value chain. This digital value chain difficulty is typically more problematic with metal PBF and DED processes as these processes can incur more thermal distortion, require additional post-processing steps and cost more to fabricate. Therefore, the software landscape necessary to perform all of these functions is convoluted.
The landscape of disparate software programs that don’t communicate well with each other is the result of many things, from developers rushing products to the market, to lack of investment by private investors to evolve or maintain products. This results in working with an assortment of Minimal Viable Product(s) (MVP). MVPs can leave consumers at risk.
Furthermore, because the AM industry developed so rapidly, individual software startup companies focused on single basic functions. For example, startup software ‘X’ focused on build planning while software company ‘Y’ focused on build process simulation, while software company ‘Z’ focused on cloud-based systems that focuses on digital encryption of AM data.
Given such a market, the AM software landscape is an assortment of large and startup software companies attempting to meet consumer needs, while larger well-established companies continue to acquire startups. However, after acquisition, true integration is a major time investment, so, often, modules of software or simply the MVP of the startup is repackaged with a large company’s branding. This results in the AM designer being flooded with software solutions from companies both large and small for various steps of the process, with varying degrees of overlap from product to product. The designer is left with questions such as: Which software has more functionality? Which software is more reliable? Which software is the least expensive? Can’t I simply have one software to represent the digital value chain from concept generation to part fabrication?
Recall from earlier in this chapter the tables that showcase the different software companies that offer AM toolsets. You can see that a single designer could utilize up to 6 different types of software to get from DfAM concept through hardware inspection! This is a significant challenge for interoperability of digital files transfer and huge annual software maintenance costs to maintain.
From an interoperability standpoint, imagine that you are a designer and you create a metallic AM Laser Powder Bed concept where you want topological optimization, cellular features, and smooth surfaces in your 3D CAD program of choice. After realizing that your 3D CAD software doesn’t create these feature types easily, you need to export to a standalone software to create these features so you import the 3D CAD definition into the standalone software to create these features separate from their native format. It is important to acknowledge that at each step there are risks of unintentionally altering the part model when entering a new software different than the native 3D CAD program. Fortunately, leading developers are working on answers to these problems, mainly by providing platforms that allow users to remain in native 3D CAD throughout the digital chain. The table below outlines the existing risks to users, and acknowledges the solutions being developed today.
Software Process Function | Risk | Development | |
---|---|---|---|
Step 1 | CAD → Top Opt | Top opt software doesn't import geometry correctly | Rolled into one CAD environment, no import needed |
Step 2 | Top Opt → CAD | CAD software must change 'dumb' solid model correctly | Rolled into one CAD environment, no import needed |
Step 3 | CAD → FEA | If part features too complex, FEA sofware may not mesh correctly | Rolled into one CAD environment, no import needed. Special meshing and smoothing algorithms are being developed |
Step 4 | CAD → FEA | FEA sofware may not mesh correctly on 2nd attempt | Rolled into one CAD environment, no import needed. Special meshing and smoothing algorithms are being developed |
Step 5 | FEA → CAD | .STL file may be too large depending on triangle resolution | Rolled into one fluid CAD environment that better handles high resolution models. Special meshing and smoothing algorithms are being developed |
Step 6 | .STL → Build Plan Software | Build planning software may struggle with large .STL file size | Rolled into one fluid CAD environment that better handles high resolution models. |
Step 7 | .STL fixing/support structure | Support material may not be generated correctly | Rolled into one fluid CAD environment with intelligent, parametric support structures. |
Step 8 | Build plan software → Process Sim | Process sim software may not capture all post-process effects | Further developing known failure modes and simulation algorithms via machine learning. |
Step 9 | Process Sim → CAD | Pre-compensation may not be accurate, displacement of features inaccurately in CAD | Further refining compensation given empirical thermal data out of the machines. |
Step 10 | CAD → Build Plan Software | Support material may not be generated correctly | Further developing known failure modes for robust support generation. |
Step 11 | Build plan software → Process Sim | Process sim software may not capture all post-process effects | Further developing empirical data capture and analysis for improved sim. |
Step 12 | .STL file → Build Plan Software | Build planning software slicer inaccuracy | Always possible |
Step 13 | Build plan software → AM PBF machine | Data storage and configuration management of data | Solutions are in development to provide intelligent storage with reporting for increased user visibility |
After following the digital workflow, the loop must be closed with feedback or empirical data from physical part inspection. The inspection process is a light scanning machine that uses a different software to analyze the dimensional conformance to the original 3D CAD definition. If fortunate, the part will pass dimensional inspection and you will not need to further refine the part. If it is the first time attempting to fabricate your part, chances are that you will need to modify it somewhere in the aforementioned digital chain.
Adoption of AM
Since DfAM is such a disruptive way of design thinking, it is often difficult for companies to change best practices and standards that have been set in stone for many years. Similar to Lean Manufacturing Systems and Six Sigma practices that require cultural changes to occur before becoming effective, DfAM behaves in a similar fashion. The availability of designers that have the DfAM way of thinking is a challenge and it takes time to acquire the appropriate software, train employees, set up capital investments, and conform to regulations and standards. Since re-architecting a product from scratch is the essence of DfAM, each DfAM project can take many months to accomplish with many stakeholders involved in meetings to discuss concepts. Often this additional labor cost and time is overlooked by executive leadership seeking to cash in quickly on sensationalized organic designs that impress customers. So, it is essential to have knowledgeable executive sponsorship when undergoing such projects.
In addition, it is helpful to broadly educate the benefits of DfAM within companies so that everyone can share the vision of enabling new product designs. And because there is change, there will be skeptics. After all, the designs don’t look anything like what has been sold in the past, there is limited information for servicing complex, organic designs, and more data is needed to prove such processes as DED hold up against traditional weld repair. Doubts will persist, but agents to change will continue to conduct tests and draw data-driven conclusions to overcome such concerns. Recognize that you aren’t alone in your journey to DfAM implementation and that the best approach is to prove your designs through testing and simulation.
Aside from the change management aspect, one of the biggest challenges remains in the lack of awareness of AM designing within companies.
Additive manufacturing is a relatively new fabrication process on the historical timeline of fabrication processes. Sand casting and investment casting has been on the earth for thousands of years. Machining has been around for a few hundred years. Additive manufacturing has been around for merely a few decades. As such, one must appreciate the large amount of legacy knowledge existing in traditional fabrication methods and the relative lack of knowledge with AM.
Because of a relative gap in knowledge in AM processes in industry, let alone how to design for it, many industrialized companies are in different phases of their AM design journey, which are characterized using a maturity model. A design maturity model helps engineers and executives alike effectively strategize and streamline their AM efforts.
For example, the Powder Bed Fusion Maturity Model progresses in five phases, increasing the value of AM contribution with each successive phase.
- Phase 0: Rapid prototyping and tooling
- Phase 1: Direct part conversion
- Phase 2: Part design integration
- Phase 3: Structural (topology) optimization
- Phase 4: Full system optimization
In Phase 0, most companies see the benefit of using AM for rapid prototypes and tooling in the form of polymer tooling and design aids meant to showcase design intent. Limited value is gained by a company focusing solely on the fringe.
Next, in Phase 1, companies eventually try to run case scenarios where they attempt to build existing products the company already makes with AM. There are situations where value can be gained from this approach. Specifically, this benefit would include difficult to source parts and/or quick response part fabrication. However, often this approach doesn’t favor AM from a business case standpoint as AM machines can be quite expensive to operate and the cost of an AM part can be pricey compared to the exact same part made from a conventional manufacturing process. In recognizing more inherent value in merging the design of conventional parts/features together in a single AM part, the company begins to understand more of what AM’s design capabilities are and moves to the strategy of Phase 2 – part design integration.
Realizing that the parts that were merged in Phase 2 are now quite heavy, perhaps they could be lightened using advanced topology optimization software to produce an organic-looking structure that is both structurally sound and lightweight. This leads to the epiphany of Phase 3, topology optimization. After a company becomes proficient using advanced topology optimization or generative techniques, then they begin to realize that new products should be designed from scratch using the design freedom AM affords to enable products that offer full product differentiation from any conventional designs used in the past.
This brings us to Phase 4, full system optimization. From this point, new products are architected keeping AM as the primary fabrication method using system level requirements to drive optimized product design to a new potential.
The maturity model phases are useful in navigating the adoption of AM, as well as defining a company’s or design engineer’s strategy for using AM. It should be considered a helpful tool for all levels of an organization, especially when trying to encourage adoption of the technology.
Acknowledging the industry current and future states, especially in terms of software tools is critical to success with AM. Maintain awareness of risks and development on the horizon as the technology swiftly evolves, so as not waste time on dated workflows. This practice in conjunction with AM design maturity phase analysis will help ensure adoption and success with AM technology in a lasting way.