Multibody simulation drives dynamic system design
Multibody Dynamics (MBD) simulations are now a vital tool for today’s engineers to help understand and improve the performance of a vast range of dynamic systems, helping them to address many ongoing challenges in the design and development process.
Francois Barral, director of Simulia Multibody Systems Industry Process Consulting at Dassault Systèmes, explains: ‘Product development teams face increased pressure from global competition and government regulations, which require product development organisations to create new, energy-efficient, safe and reliable products in ever-shorter time frames. Products are also becoming more complex, with electronic and software controls combined with the mechanical system. To meet these many demanding challenges, organisations are increasing their use of MBD simulation, which helps them replace physical testing with virtual prototypes.’
MBD simulation can be applied to almost any branch of mechanical engineering, but it is regularly used across automotive, railway and wind-energy industrial sectors. And these multibody systems must often work across a broad range of applications and scenarios, all of which can affect their functional requirements.
For example, an automotive braking system may need to work across numerous vehicle types from many manufacturers. In construction, a machine may need to excavate different types of material or undergo rigorous safety tests. Japan-based crane manufacturer Tadano, for example, uses multibody dynamics to help accelerate and validate its R&D process, while providing a safe environment for operators.
Barral explains: ‘A crane boom that is equal to an arm of a crane can cause boom-crane deflection and load wing when it attempts to lift a load. To understand the physical property of a load swing and how cranes are used at construction sites, it is essential to consider the control method of a load swing. However, designers are not allowed to enter construction sites where cranes are at work due to "hazardous areas".’
Using Dassault Systèmes’ Simpack multibody dynamic simulation software, Tadano could simulate large-scale 3D vehicle models, reduce a company’s manufacturing prototype costs and time for rework ‘significantly’, according to Barral. ‘It was crucial for engineers to generate and analyse a physical model for structures based on accurate design information. Simpack is now being used to develop a model for major types of rough terrain cranes that are a mainstay product of Tadano.’
MBD is even helping our washing machines run more efficiently. Recently, Altair MotionSolve was used to simulate the motion dynamics of the washer drum, including the balance ring and the suspension system. An idealised motor model was used to define the various spin speeds of the washing machine during agitation and spin cycles. Combined with Altair’s other simulation and modelling tools, consumer appliance company Mabe increased the capacity of its washing machines by 35 per cent, improved the energy factor by 24 per cent and water factor by 52 per cent, while simultaneously reducing the cost per cubic foot by 10 per cent and the product development cycle time by 25 per cent.
EnginSoft’s RecurDyn solution also helped Toyota’s Forklift division, which was having difficulties predicting the behaviour of the vehicle when carrying loads and driving through obstacles. The truck was modelled in RecurDyn, including flexible bodies to model the forks and the built-in tyre module, to accurately predict the behaviour of a fully-loaded truck driving over obstacles.
Danujan Sivanesan, a senior project engineer specialising in multibody dynamics at Enginsoft, says: ‘RecurDyn reliably analysed the critical conditions normally understood during the testing phase and computed the dynamic loads between the forks and the loads accurately using RecurDyn Geo-contact technology. This helped the customer reduce their product development time, enabling them to iterate through multiple designs and improve the performance and quality of their product.’
This range of applications and interested industries continues to constantly expand. ‘Markets are starting to adopt this type of technology now to design better products and have not traditionally been operating in this space before,’ according to Chris Harduwar, vice president of Business Development at Maplesoft.
‘[Engineers are] exploring rigid and flexible multibody dynamics as a way to assess the interactions on components such as loads, tensions and torques and develop the specifications and limitations for a machine to operate,’ he says. ‘At Maplesoft, our customers often work to improve performance with components that fit into a larger machine or address issues that span multiple domains.
All industries have ‘some common and representative problems that can be addressed and solved through the use of MBD simulations,’ according to Jim Ryan, VP of Model-Based Systems Solutions at Altair. ‘MBD can really help improve the positioning to make sure that parts move as intended – one example is with those systems involving precisely controlled movements such as active suspensions and precise pointing systems such as telescopes and radar.’
MBD simulations can also help engineers calculate the sizing forces and torques needed to move the system with the correct actuators. You can also manage accelerations and peak forces, reduce wear and avoid friction- or contact-induced damage or calculate loads to minimise stresses, fatigue and, ultimately, failed operations or broken products.
MapleSim, for example, is used to develop high-speed robotics, including custom components between the joints or using motors that have not been defined yet. Engineers need to understand the torques and forces that are applied to those motors across the full design and including the interaction of the movement of all the components.
MapleSim was recently used to design a six-degree-of-freedom cable robot and help understand the entire system’s behaviour. ‘Using MBD, engineers could understand how fast the robot can go, how much payload it can handle, and how to optimise the design given the operating conditions,’ says Harduwar.
Engineers are still faced with challenges when running MBD simulations. Methodically constructing very large models where potentially hundreds of components may move separately from each other is one such issue. Ryan explains: ‘It is quite easy and common to model such connections incorrectly, in which case the simulation might not proceed, or it might proceed but yield incorrect results, consistent with the CAE engineer’s much-used axiom: garbage in, garbage out’!’
The easy re-use of pre-constructed CAD assemblies (3D geometries) to retain the relative starting positions of all the components in a multibody system, together with all the mass and inertia properties, is another challenge to accurately predicting the motion.
‘Even though multibody dynamics modelling has improved across several decades, it can still be challenging for engineers using MBD to properly model 100 per cent of the complex physics involved, due to such phenomena as part-to-part contacts and friction, especially involving flexible bodies and elastic parts of the system like belts, ropes and cables,’ adds Ryan.
Mounting system complexity puts engineers in a difficult position when it comes to MBD simulations. Harduwar says: ‘The engineers are inevitably left with a trade-off between the level of realism or fidelity of a model, compared to the computational and human engineering effort of defining the elements within the simulation.’
The motion control of the manipulators in a robotic arm, for example, is ‘particularly challenging’ to simulate, according to Harduwar. ‘This is where system-level modelling becomes more effective – where equations of all the systems and components are combined and handled by a simulation tool.’
As product complexity continues to increase, this situation will continue to impact engineers in the near-term where trade-offs are often made. Harduwar explains: ‘The challenge is doing something at the right fidelity level that will take the appropriate amount of time to get the right level of results.’
This is where system-level modelling using MBD has advantages over finite element analysis (FEA) or computational fluid dynamics (CFD) analysis. Using system-level modelling and multibody dynamics, different types of equations are used where each component and the interactions between components are described. This is a more efficient way for engineers to understand the behaviour of certain systems in response to the force acting upon it.
Barral explains: ‘These integrated capabilities enable designers and engineers to collaborate on replicating real-world scenarios and analysing multiple design options quickly and accurately. MBD simulations can help engineers study sub [systems] and complete systems, optimise critical components and systems earlier in the development cycle, avoid costly last-minute changes during physical testing and achieve overall efficiency and performance goals.’
As such, MBD is not only a more computationally efficient way to simulate a dynamic system, but also to simulate the rate at which engineers can get answers on a whole system when using system-level modelling.
Visualisation tools are vital for MBD simulations; enabling engineers to analyse results quicker, thanks to the animations generated. Ryan says: ‘People expect to see full 3D geometries realistically moving on their screens – and they can with MBD. And in the spirit of ‘a picture can be worth a thousand words’, it is also true that ‘an animation can be worth 1e6 pictures’.’
Visual animations, for example, can help engineers understand rates of change for the system as it moves. ‘Especially when limits or targets are involved, it could be essential to know if actual values for velocities or accelerations are exceeding those limits or targeted values – for example, with a cruise-control system in a car that is set to make sure the car doesn’t exceed 100kph,’ Ryan explains.
Plot and graphs are helpful, while complementary visualisation aids are ‘essential in those cases where motion-related quantities cannot easily be perceived or even accurately measured through animations alone,’ according
to Ryan. ‘For instance, an animation typically will not show the time-history of specific forces in the system to enable an engineer to spot the peak value of a force and also the time at which it occurs... which could be vitally important to the product’s design.’
Advancing electrification will continue to impact mechanical simulations and MBD is helping ease this transition. Ryan explains: ‘For years, as the products that companies make have continually grown – and accelerated – in complexity, we have seen an increased need to perform simulations of smarter electro-mechanical products involving all kinds of logic and feedback control systems.’
This is ‘especially applicable’ to systems in motion, according to Ryan. “This is to help ensure that they move as intended – especially when safety considerations are important such as when developing something like antilock braking systems (ABS) and traction control systems (TCS) for different types of ground vehicles.’
MBD is also increasingly being used earlier in the design and development lifecycle. Previously, control system engineers and multibody dynamics engineers would develop their separate systems in isolation and then try to put them together at some point later in the development cycle – maybe as late as the first prototype stage.
But this is still ‘much later than desired,’ according to Ryan, who adds: ‘Fortunately, we are now able to help companies break down the silos that have long existed between engineers and engineering teams who are focused on different disciplines – electrical versus mechanical versus controls, for instance. These streamlined workflows are now helping companies to minimise bottlenecks in their product development cycle so that they can continue to accelerate time-to-market.’
AI and machine learning are complementary technologies that are really finding their feet in the MBD realm. Ryan says: ‘Leveraging artificial intelligence/machine learning (AI/ML), especially using neural networks to create reduced order models (ROMs) of subsystems to accelerate simulation throughput, is another trend.’
‘We’re also seeing greater interest in digital twins in general and finding VR and AR applications to manage and monitor real-time operations of fleets of machines,’ Harduwar adds.
Environmental concerns will continue to push MBD simulation adoption, a trend that will affect not just MBD but the entire simulation landscape. Barral explains: ‘Trends in sustainability, green energy and autonomous – and driver-assisted – vehicles are really important factors in the increased use of multibody simulation.‘Moreover, industries are subject to irreversible ‘trends’ such as climate-change mitigation redefining operations. Climate issues are accelerating quickly and, with global regulation pushing for a more circular economy in line with market expectations, aligning to the sustainability agenda won’t be easy.
‘However, the first step to reaching this goal is creating products in a fully collaborative, virtual environment. Once you have all of this information available in a single immersive experience, it becomes so much easier to visualise your data in the right context and then apply it for the best results.’