By Brian MacCleery, National Instruments and Stephen Endersby, Dassault SystÃ†â€™Ãƒâ€šÃ‚Â¨mes SolidWorks Corp.
Design failures may be the key to success in designing complex motion control machinery. Failing early and often may lead to rapid design success. Every mistake you catch in the early phases of development is a potential avoided catastrophe.
Any programming mistake or mechanical design flaw that causes a physical machine to crash, vibrate, fail early or underperform is a very serious matter. If those little monkey wrenches stay hidden until you build the first machine, a well-managed project can turn on its head. However, new virtual prototype tools are available today that provide a simple, safe environment to test your design and root out problems while your design is still just a virtual CAD model.
In a machine design, every decision has a ripple effect on other aspects of the design.
We all know that to avoid crisis at the end of the machine development process, it’s critical to find and solve the flaws early. But the problems most difficult to spot are often the ones that lie in between engineering disciplines—the no-man’s land between mechanical, electrical, software and embedded control systems. In a modern electromechanical machine, there are very few decisions that can be made in isolation, without the decision (or design change) impacting the other disciplines. For example, selecting the pulley diameter for a belt drive actuator in a multi-axis motion control gantry, a seemingly mechanical decision, has a ‘ripple effect’ into the electrical and control domains. This decision impacts everything from motor sizing, to the positioning accuracy and cycle time performance of the machine, as the chart indicates.
The Mechatronics approach
Getting today’s electromechanical machine designs right requires a multidisciplinary approach to development, often called a mechatronics approach, in which mechanical, electrical and controls engineers work side-by-side in a parallel development process. Many design teams are still using the traditional sequential “throw it over the wall” approach that works well when there aren’t many dependencies between the engineering domains, so is it surprising that many mistakes stay hidden until the physical machine is built?
The emergence of new virtual prototyping tools enables a multidisciplinary team to start collaborating as soon as the CAD model (the virtual machine) is created. Control engineers can start writing and testing the control software, electrical engineers can make well-informed sensor placement and motor selection decisions, and mechanical engineers can optimize the mechanical design using realistic force and torque data.
What benefits could you expect by adopting a virtual prototyping approach? A recent survey by the Aberdeen Group of more than 140 design companies found that mechatronics-oriented companies significantly outperform those with a traditional “domain segmented” development model. These “best in class” design companies are seven times more likely than the laggards to use virtual prototype simulations. These simulations capture multi-domain system level behavior so the team can iterate on the design before building a physical prototype. The top companies performed an average of 25 design iterations using virtual prototypes and six design iterations using physical prototypes. These companies consistently outperform the laggards in terms of hitting development cost targets, product quality goals, launch dates and more.
The top performing companies use multi-domain virtual prototyping simulations in a number of ways. First and foremost, virtual prototypes provide a way to improve communication and collaboration across engineering disciplines. Second, they help to increase the team’s vision into how well they meet the design requirements throughout the project. Third, they increase the team’s ability to predict the behavior of the system before the physical machine is built.
Motion system design
Design of a motion control system is a classic mechatronics challenge, and to address this challenge new virtual prototyping tools are emerging, embedded directly into the 3D mechanical CAD environment. For example, NI SoftMotion for SolidWorks connects NI LabVIEW, a graphical programming language, to the SolidWorks 3D CAD environment, enabling designers to test their CAD models by controlling the virtual machine with realistic industrial control software. This type of simulation is richer in information compared to a simple static screen capture or basic animation. It offers a way to demonstrate realistic machine operation, collect customer feedback, and communicate design intent within the design team. It also provides a wealth of velocity, force and torque data that empowers the multi-disciplinary team to make better, more informed design decisions.
Greg Radighieri, owner of Radical Engineering, LLC, explains how virtual prototyping is changing his machine design process. “When you have to wait for the physical machine it takes a long time to get to the testing and validation stage. With the virtual prototype you aren’t waiting a long time for the machine shops to complete the machine to start testing it. These tools provide a really slick solution. Let’s take a grinding machine. Early questions include, how fast does a tool need to be? Should it be a diamond tool? Your first design needs to tell you how much mechanical power it will need. The motor sizing and electrical solutions fall out of that.”
All models are wrong, but some are useful
The goal with virtual prototyping is not to have a 100% accurate model of the machine, since that is what the physical prototype is for; the goal is to quickly test your design, while evaluating the “ripple effect” of design decisions. If it takes too long to get results then the value of the information is significantly diminished. Rather than waiting months for an answer, you want a quick process of iterating and optimizing the design. To that end, virtual prototyping tools should reuse the investment you have already made when creating your CAD models. The inertia and mate information in the CAD model is used to create a mechanical simulation, which then comes to life when it is connected to the motion control software. The team’s investment in developing the control software for the virtual prototype is also reused when the same code is deployed onto the physical machine. All this virtual testing means you can make more design iterations, thus improving your design, while making sure the physical machine will work as expected.
Mechatronics-oriented companies that use simulation are significantly more likely to hit project management goals than laggards.
Machine design is an open-ended process where there are many possible solutions. Unfortunately, though, many are bad solutions and only a few are good ones. Many highly effective development teams look at product development as Thomas Edison did, a process of elimination. The famous inventor once wrote, “If I find 10,000 ways something won’t work, I haven’t failed. I am not discouraged, because every wrong attempt discarded is another step forward.” Virtual prototyping can help you eliminate bad solutions quickly.
How can we use virtual prototyping in the design of a motion control system? The starting point for any design is to define the attributes that make it desirable to your customer. For most machines, the classic desirable attributes include throughput (how fast can it produce?), downtime (how reliably can it produce?), and yield (what is the quality of production?). However, these are all bound by one powerful constraint—cost. The virtual prototype helps you analyze these factors again and again as you refine and add to your design.
A virtual prototype of a pick-and-place machine involving three axes of motion and a gripper arm.
When evaluating cost you really want to analyze the price of production (dollars per widget) rather than just the price of the individual components. In some cases, selecting a higher-price component will increase the throughput of the machine and therefore lower the resulting cost of production. These complex business tradeoffs lie at the forefront of any design project. For example, increasing the throughput of the machine may require a more expensive motor and control system, but result in a net reduction in the cost per item produced.
Solving these riddles still requires good old fashioned engineering savvy and experience. No computer tool is going to replace those skills anytime soon. What the tools can do, however, is replace “back of the envelope” calculations with solid engineering analysis. Some of the technical questions virtual prototyping can help to answer are shown in the illustration below.
Virtual prototypes can help you make better informed decisions.
Kent Wedeking, a mechanical design engineer with Fastek International Ltd., is beginning to use virtual prototyping technology to improve his machine design process. “The old way, I would have just thrown the mechanical design over the wall and said ‘hey, make it work.’ Now we can look at the whole system. Instead of sequential development, we can do concurrent design that includes the mechanicals and all of the motion control code to go with it. It’s much easier to make a change and respond to different things when it’s on the desktop in the virtual environment. (It could cut) the overall project timeline in half.”
In summary, virtual prototyping can increase the efficiency of your development team by providing the mechanical, electrical and controls engineers with valuable insight into how the proposed machine will perform. Virtual prototype testing can begin as soon as you have a CAD model of the conceptual design and it can be used throughout the development process. Before the “final” drawings are released to manufacturing, it can give both engineers and managers a higher level of confidence.
Dassault Systèmes SolidWorks Corp
To see virtual prototyping demonstration videos and learn more, visit www.ni.com/virtualprototyping.