Making off-highway equipment, such as construction or agriculture machines, more efficient continues to be a priority after the final phase-in of the U.S. EPA Tier 4 Final emissions regulations.
Exa’s fluid dynamics simulation, based on the Lattice Boltzmann technique, finds a place in helping engineers with thermal management of equipment.
“Some of the companies we work with are running out of credits and unfortunately still struggling with Tier 4,” explained Tristan Donley, Technical Director, Off-Highway Heavy Vehicles North American for Exa, a supplier of software for computational fluid dynamics (CFD). These companies continue to balance thermal and noise requirements due to their upgraded engines and aftertreatment devices, for example. “Others are now focusing on cutting the additional cost from components that enabled them to meet Tier 4 Final, but drove up the cost of their machines,” she said.
A key element of this packaging is optimally managing thermal flows in systems that must be ever more efficient and smaller. That is where Exa’s flagship software PowerFLOW enters the picture. PowerFLOW uses a relatively new technique in simulation, Lattice Boltzmann CFD. Without going into mathematical details, it claims advantages over the more common CFD based on continuous Navier-Stokes formulations, according to Donley. Using Lattice Boltzmann allows PowerFLOW to use a computational mesh with larger cells, or voxels as they term them.
“It is easier to resolve complex geometry and is inherently transient. Our voxels can be larger than the feature we are trying to predict the flow around,” she said.
While the method is also useful for aerodynamic flows, off-highway engineers tend to use it mostly for aeroacoustics and thermal management. “The priorities for our current customers is using optimization techniques for package layouts. They want to know space claims for different components, such as heat exchangers or how to orient the engine to protect electronics,” she said. “They also want to know where to place grilles and louvers to make sure they get enough airflow through their engine compartment.”
She notes that some of these companies had never used any type of CFD software prior to designing systems for Tier 4. There were so many challenges that their tried-and-true design techniques of the past—what she calls “tribal knowledge”—was not working, creating an opening for using advanced simulation like PowerFLOW’s Lattice Botlzmann CFD. She also notes the increasing use of various optimization techniques, like multifunctional objectives and trade-off techniques that balance competing priorities for an optimum solution.
Integration and digital twins
At PTC, provider of CAD and simulation software as well as Internet of Things (IoT) solutions, “integration” is an important trend. The company’s flagship CAD software, Creo, is tightly integrated with its finite-element simulation package, Creo Simulate, for performing structural, thermal and vibration analysis.
“There are many geometric design changes in the lifecycle of a design, and typically if you make that change you have to start the analysis process again,” explained Jose Coronado, Product Manager for Creo Simulate. “With Creo Simulate, this is not the case. Whenever you change the geometry because you had to change the design requirement, then the downstream settings are updated automatically.”
He also touts that it is easy to use, with an interface designed for engineers without a deep background in advanced finite element analysis—no PhD required. “The user interface is common with Creo to make it consistent,” he said.
While Creo Simulate adapts to design changes, Coronado points out that the use-case scenarios for setting initial conditions and loads often remain assumptions. How much weight will be in a bucket, how fast will the machine be driven, or how many hours a day will it be used are derived from spotty observations or educated guesses.
“Better is to establish a digital twin by integrating data from a physical product with a digital representation of that product. This will give us greater insight into the product’s state, performance and behavior,” he said. “We can use real loads in our simulations instead of assumptions.”
Connected to embedded sensors, the company’s ThingWorx product will facilitate transmitting real data to the designer of, say, a backhoe. The data will show if operators are lifting 2 tons or 5 tons, operating it 12 or 8 hours a day, and so on. This provides “performance based analysis” that can reproduce critical circumstances.
“Not only for a family of products, we can even get down to a particular serial number and determine how it is being used versus another individual product, continuously re-evaluating assumptions,” explained Coronado. “Think of ThingWorx as the real-time aggregation engine of the real sensors to close the loop.”
Holistic solutions
Ravi Shankar from Siemens PLM Software offers a comprehensive view of the problems off-highway equipment manufacturers face. Machines need to increase their effectiveness in terms of greater loads, but they also need to improve durability and fuel economy. Operator comfort is equally important.
Siemens PLM believes the industry is in the early stages of seeing an impact of sensors, data collection, and analytics on the design process beyond just predictive maintenance. Here, a prototype is instrumented as part of a test campaign. (Siemens PLM)
“They need to ensure their operators are not stressed too heavily to ensure smooth operation and efficiency,” explained Shankar. “These requirements often work in conflict with each other, for example increasing load capacity means larger components that negatively affects emissions and fuel economy, as does air conditioning.” These often force a look at different types of hybridization, including both electrical and hydraulic. These are more difficult to design than a pure mechanical system.
The solutions Siemens now offer are just as comprehensive and growing, expanding through acquisitions. To their existing tools in simulation and data management, they have added 1D and 3D systems simulations expertise through its acquisition of LMS, and CFD and design exploration through CD-adapco. Predictive analytics tools have come along with Camstar and its Omneo product.
Making sense of how simulation, IoT and predictive analytics all fit can seem confusing. “There are really three ways of thinking about Big Data and predictive analytics,” he said.
First is integrating sensor-based data with physics-based simulations, using the digital twin concept. This means augmenting test and test data protocols with virtual sensors and bringing that data back into the early stages of design and engineering.
A second way is in the postprocessing of data, for instance by using its LMS Test.Lab technology. “For example, to use data from multiple sensors that measures durability for critical types of off-highway equipment,” he said. Think of aftertreatment devices or critical joints. “There are reams and reams of such data, but you need to postprocess that data using math techniques and convert them to some key performance indicator useful in the design process.”
The third way is to view simulation optimization as another form of data analytics. Shankar believes Siemens’ HEEDS multi-domain optimization tool is really a form of Big Data analytics that uses hundreds of simulations—reams of data—to arrive at optimum solutions that meet given constraints. Multi-domain optimization, or MDO, combines the results of multiple simulation types, say engine, heat transfer, vehicle dynamics, into a single, results-oriented model. Sorting through the multiple, combined runs lets the software suggest the best solution that meets constraints.
Off-highway often looks at hybridization using hydraulic accumulators as well as mechatronics. CAE simulation from providers such as Siemens PLM is becoming a critical tool in developing these complex machines.
“These are all valid ways of combining sensor data with physics-based simulations and test. Twenty years ago, some thought that CAE would replace test, but in some ways, test is just as important as CAE in the context of predictive analytics,” he said.
Simulation and results
Makers of off-highway equipment rely heavily on classical proven methods of design and the company’s legacy knowledgebase. These are viewed as reliable and dependable.
“Currently CAE simulation is becoming a more useful tool in a world dominated by traditional methods of equipment design,” said Venugopal Ravula, Program Manager of Altair. “Many of the success stories we have had with heavy equipment companies have been in reducing weight and costs, and developing the methods to manufacture the biomimicry designs—large fabricated components.”
The key to success in reducing weight and cost is in using CAE for optimization, according to Ravula. Design and topology optimization tools are premier components of Altair’s software offerings. Across all industries, educating engineers in the proper use of optimization is needed.
“Often, customers do not know what they want,” he said. “They might come to us through our consulting service asking about shape optimization, but once we expose them to other tools, like multi-domain optimization, they can see the benefits.” He also stated that the industry is moving towards increasing use of MDO.
Topology optimization, where the shape of a component is suggested through automatic means, is one of Altair’s specialties. A recent case study the company shared in an interview points to some of the peculiarities and advantages that future engineers need to know. Using its OptiStruct tool, Altair helped Liebherr redesign a crane boom that resulted in a component that was 20% lighter yet could lift 400 kg (880 lb) more. The resulting shape is not one that “tribal knowledge” or design books might produce, looking more organic with odd spacing of reinforcements. Other examples he showed included castings of tractor transmissions that reduced mass by 10%.
Ravula also notes that IoT, Big Data and data analytics are becoming more important in predictive maintenance, a field Altair is also active in.
Source: sae.org