When wrestling with vexing issues such as product complexity, lightweighting, advanced materials and new manufacturing methods, today’s manufacturing engineers increasingly use high-fidelity simulations to visualize solutions to these challenges. The latest simulation software can offer clues to improving product design and performance with high-end CAE packages and NC simulation software that help determine not only the best designs, but the most efficient ways to achieve them.
New manufacturing processes like additive manufacturing (AM) present different problems for manufacturers, and simulation can determine upfront what solutions work best. Many simulation software packages now offer additive-specific simulations that can help illuminate the layering processes by which additive products are built, while also visualizing how to make traditional subtractive metalcutting processes more efficient.
In discrete manufacturing, product development engineers face enormous challenges, as anything from autos to aircraft to massive ships and heavy machinery contain millions of lines of code that program systems with electronics geared for the Industrial Internet of Things (IIoT).
“The larger trends that we see are the increasing complexity of products,” said Ravi Shankar, director, global simulation product marketing, Siemens PLM Software (Plano, TX), while noting the increasing level of electronics in mechanical components. “What’s driving that is the digital twin and model-based engineering.”
Autonomous vehicles and drones are the latest example of complex systems requiring simulation software systems, Shankar said. “We’ve also seen the focus on automotive fuel efficiency and emissions, with the increase in lightweighting and use of generative design,” he added.
Siemens recently introduced a solution for self-driving cars that incorporates its Simcenter simulation solutions with new sensor technology. At Siemens’ US Innovation Day in March in Chicago, the company unveiled the system, using Tass PreScan virtual sensor imagery with the Mentor DRS360 platform that automates development of algorithms for sensor fusion and processing.
“The first trend is that physics-based world models and physics-based sensor models can be created,” said Martijn Tideman, director of products for Tass International (Steenovenweg, The Netherlands), which was acquired last year by Siemens. “These models make high-fidelity artificial sensor information that can be generated as if it came from real cameras, radars and LIDARs. When you feed these artificial data to in-car processing units, such as the DRS360, you can evaluate the in-car hardware/software without driving a single mile.”
Another key is making sure that the automated driving simulation software runs on high-performance clusters (HPCs), Tideman said.
“These clusters can run many simulations in parallel, which speeds up the virtual evaluation/validation process. You want to be able to drive a million virtual miles over the weekend.
“Automated driving simulators need to be connected to a wide range of other hardware and software modules, for example map-importers, to automatically generate virtual road networks or vehicle dynamics simulation tools to make sure the vehicle responds correctly to control actions,” he explained. “Interfaces between simulation tools are increasingly being standardized,” such as with FMI/FMU (functional mock-up interface/functional mock-up unit), Tideman added.
Several key enabling technologies are affecting the way simulation is being applied in manufacturing. “While there are numerous technical developments that continue to evolve and improve, a handful have the potential to fundamentally disrupt where, how and by whom simulation capabilities are used,” said Seth A. Hindman, senior manager, product strategy and management, manufacturing, construction and production at Autodesk Inc. (San Rafael, CA).
“Moore’s Law has continued to hold true far longer than most would have imagined. With continuous advances in processing power, incredibly powerful hardware is accessible at a very low price. Combined with the development of extensive fiber-optic networks, companies can be connected to external computational resources that exceed the speeds of their own intranets,” Hindman said. “Burst capacity, elastic compute and configurable HPC [high-performance computing] are enabling companies to fundamentally change the way in which they engineer and manufacture their products by running enough analysis to truly understand how their product will perform in numerous applications and environments.”
Manufacturers are also moving away from mesh-dependent analysis, said Hindman.
“The ubiquitous nature of 3D design data has continued to press the demand for simulation tools that are no longer mesh-dependent and that can benefit from associativity with the native design data. Not only does this increase productivity in the general workflow, it enlarges the audience that can benefit from simulation capabilities. The natural down-pressure is to create solutions that are more robust, more intuitive and that break from traditional CAE requirements. The long-term potential is that this enables a simplified interaction with solutions and automation of onerous tasks.”
The long-standing perspective within CAE to bring simulation “upfront” is now being replaced by the idea of objective-based analysis, Hindman added. “While upfront simulation is powerful, it is still reliant on a traditional convention of testing what you’ve designed vs. driving the exploration of designs that meet the objectives that you’ve defined,” he said. Autodesk is currently introducing what it calls generative design into the engineering market, Hindman added, which has just become available in Autodesk Fusion 360 Ultimate.
“Our generative design technology enables objective-inspired designs to be created by the system, which facilitates widespread exploration of the design space, enabling insight-based trade-offs,” Hindman said. “The principal task of engineers has historically been to create a design that works. With generative design, every outcome successfully meets that base requirement, which means that decision-making and trade-offs are elevated to the level of core business initiatives.”
Additive developments continue to excite the manufacturing industry, and many simulation developers have recently released either new or enhanced additive-specific versions of their simulation software.
For example, on April 19 simulation developer Ansys Inc. (Canonsburg, PA) released its new Additive Print and Additive Suite solutions that deliver simulation for metal AM processes. The solutions are said to enable users to print lightweight, complex metal parts and analyze microstructure properties and behavior. Ansys said this will help reduce AM costs by limiting design constraints, reducing waste and shrinking print time.
Ansys’ complete additive simulation workflow lets customers test their product designs virtually before printing a part, according to the company. The software incorporates simulation prior to the printing process, which lets engineers design, test and validate the performance of a part at the design stage and greatly reduces the high cost of physical trial and error.
AM is a game-changer for manufacturing, said Brent Stucker, Ansys director of additive manufacturing. “Medical devices can be produced with patient-specific geometries. Spare parts inventories for many components will be a thing of the past, as replacement parts can be produced when they’re ordered,” Stucker noted. “Products that operate in extreme environments, such as in the oil and gas industry, can be produced with new [more durable] hybrid material compositions. The geometric complexity offered by AM means that dozens of components can be integrated into a single component that is lighter weight and higher performance.”
Stucker noted than in AM, designers and machine operators who typically have not been engineering simulation users now must understand a complex printing process.
“We’re seeing that machine operators want more than just educated guesses when it comes to predicting if a part can be built successfully,” he said. “They need to quickly understand how a particular machine setup will result in part distortion before and after removal from supports, and whether excessive distortion might cause the powder spreading mechanism to hit the part [what’s known as blade crash]. The situation is similar with designers doing Design for AM [DfAM]. They want to know if the part they have designed will print successfully, and if it can, what the properties of that component will be.
“Simulation puts the power for understanding the additive process into the hands of designers and operators,” he added. “That is why we have developed Ansys Additive Print to be a stand-alone print process prediction tool—so that a non-engineer can use the tool within a few days.”
While simulation software has been successfully used to ensure that a designed structure will stand in-service conditions during field operations, it is also constantly challenged by complex loadings, materials, and physics, noted Subham Sett, director, Simulia Strategic Initiatives, Dassault Systèmes (Paris). “Simulation software is now trending toward providing multiphysics and multiscale solutions providing predictions for every aspect to accelerate industry growth. For example, in automotive industry, simulation software is being used to solve multiphysics problems from multibody dynamics, noise and vibration, crash-worthiness to unsteady flow, as well as multiscale problems from material design, multiscale material up and down scaling, and substructures.”
Adding the ability to predict the shape resulting from the AM process ranks as one of the most exciting new developments in manufacturing simulation, Sett said. “Taking into account the scan path, material properties, machine and laser properties, we can accurately predict the part deformation due to heat and gravity while the part is being printed,” he added.
Simulating the additive process enables builders to more accurately predict and control the process, preventing errors from occurring in the layer-by-layer additive part build process.
Siemens’ Shankar said simulation in additive processes, which Siemens added with its NX Additive module last year, enables manufacturers to predict manufacturing outcomes.
“As you create the layers, you have to know many things: How long does it take to cool? What are the voids [the empty spaces or pockets] in the part? Simulation can help address the voids and also the residual stresses in the product,” he said. Simulation also will help manufacturers know how to standardize processes and understand how the product will perform.
Siemens currently is working on issues such as additive processes with phase changes and how parts cool. In some cases, HPC is being used, due to the computation-intensive nature of those simulations. “Computations can often be paralyzed. If you’re trying to solve large models, it’s [HPC] attractive,” Shankar said.
With regard to additive, Autodesk continues to extend its portfolio of solutions, said Hindman. “The newest addition to the Netfabb portfolio is the inclusion of Autodesk generative design capabilities. As I mentioned previously, generative design enables objective-inspired designs to be created by the system to facilitate widespread exploration of the design space. A key ingredient of how the system can function as an active participant is that we’ve taught it to be additive manufacturing process aware, so it will generate results that are optimized for 3D printing.”
With the March release of Netfabb Ultimate, Autodesk introduced integrated process simulation capabilities in addition to the existing stand-alone Netfabb Local Simulation offering. Hindman said that both Netfabb versions now include: enhanced predictions of effects like trapped powder, hot spots/scorching and lack of fusion; additional process emulation like EDM part removal and the impact of heat treatment; improved performance in Autodesk’s optional elastic compute service; introduction of process analysis capabilities into Netfabb Ultimate; and streamlined capability to swap in simulation-driven compensated part preforms for the original geometry.
Autodesk also launched support for the direct energy deposition (DED) process in partnership with the company’s Netfabb and PowerMill (PowerMill Ultimate) portfolios to leverage expertise in multiaxis robotic controls.
“This adds another capability for predicting the potential of exceptionally large deformations and failures during high-rate deposition manufacturing,” Hindman said. “As companies embrace metal additive manufacturing, there are common perils that they must confront and overcome.” He added that the most common challenges of operating a metal powder bed fusion printer are: part deformation, warpage (the printed part is unacceptable and may damage the recoater); thermal stress-induced failures/fracture (part breaks during printing and may damage the recoater); support failures (so much stress is built up in the part that it breaks the connection between the build plate and part, making the part unacceptable and possibly allowing the recoater to collide with the part); and varying material properties (parts have visible defects or do not perform as anticipated).
For NC simulation and verification processes, AM remains a key development area with developers of systems like Vericut NC simulation software from CGTech (Irvine, CA), which has recently added a Vericut Additive module. “Additive Manufacturing continues to be one of the hottest trends in manufacturing, so simulation software endeavors to keep up,” said Gene Granata, CGTech Vericut product manager.
“Using simulation software capable of simulating the same NC codes that will drive the machines is the best way to protect CNC equipment and create a quality part the first time out in the shop.”
In composites, two of the latest trends that stand out are a focus on graphical display and accurately representing the workpiece, Granata noted. “Although improving graphics may seem like a cosmetic improvement, there are real engineering benefits from accurately predicting and visualizing a composite piece’s net shape. Manufacturers are starting to rely more heavily on software to predict a part’s final shape and quality,” he said. “With a high-resolution display of the finished workpiece, engineers can begin to interrogate features most concerning to them with greater fidelity.”
Simulation software provides important process evaluation and optimization tools that enhance shop productivity, he added. “New choices for creating ‘efficient’ toolpaths seem to appear on the market regularly, but how well do they really work? Simulation software—driven by post-processed G-code toolpaths—reveals the ‘truth’ in machine runtimes,” Granata said. “This aids NC programmers and manufacturing engineers judging different machining methods so they can ultimately select the most efficient methods for making their parts.”
Automation, machining optimization, and machine flexibility are key areas that Spring Technologies (Cambridge, MA), developer of NCSimul and Optitool software, is pushing in its simulation software development, said Silvere Proisy, Spring Technologies general manager.
As the market is demanding more automated processes, NCSimul is developing more automation, he said. “From the data sent by the CAM software into NCSimul, it’s been automatically verified, and the result of simulation is delivered as a final report to the programmer through e-mails. The users do not have to interact anymore with the software; it can be all running on a remote server.”
CNC machining optimization with Optitool is improved, he said. “It now offers two levels of optimization: one is air-cutting reduction, optimizing all approach and retract motions without changing feed rates in material; the second level is learning-mode optimization that regulates the material cutting feeds based on nine cutting parameters, such as chip thickness or chip flow, without compromising the original feeds and speeds.”
Machine flexibility is the third focus of development with its NCSimul 4CAM option. “Giving capability to a manufacturing company to change a job from one machine to another, in a few minutes and without having to reprogram the part in a CAM software, is what we call a revolution,” Proisy said. “It reads the initial G-code and re-writes the new code automatically.”
Connect With Us
