Using simulation to determine the root cause of product failure

Using simulation to determine the root cause of product failure

Your product is out in the marketplace. Your design team has done the best job it can. The shop guys have put their all into making it. But it’s failing. Warranty claims are escalating and putting pressure on the bottom line and you, the person tasked with figuring out what’s wrong and keeping it from ever happening again. When you ask an executive what keeps her awake at night, it’s usually some variation on this scenario. You can start, today, to get control of these issues and take preventative action for the future.The good news is products that fail under warranty are often sent back to the manufacturer, so you can start examining what went wrong. As you try to figure it out, you might use the “Five Whys” technique from Six Sigma: Keep asking “why?” each time you discover something or make an assumption. There’s no magic to the number five; it just seems to be a rule of thumb. This technique lets you peel away each layer of symptoms to get to the root cause of a problem. Did the customer use the product in an unanticipated way? Was the material inferior somehow? Are the failures structural? Are they all happening in roughly the same place?Once you’ve got a good handle on the root cause, try to duplicate the demonstrated behavior with simulations. The first question your boss is likely to ask is: “Can we trust these results?” Your answer is easy: Yes. That’s because you’ll keep tweaking your simulation setup until you predict failure exactly as it happened. Once you’ve got that, by changing the loads, boundary conditions, meshes and other parameters, you’re ready to insert simulation into your design process. Rather than asking what happened after the fact, you’ll be looking at what might happen. That’s proactive; predicting failure modes so that you can prevent them, with design or manufacturing changes, before the product ever gets to your customers.Your boss’ next question is likely to be: “Can we get the results in time, so that we can make design changes without affecting our release schedule?” This one is tougher because it takes into account how often the design changes, your expertise with simulation and how much compute capacity you have at your disposal. Your mileage may vary, but the answer is “yes” because so many other companies have created workflows that use simulation as an integral part of the design process. Perhaps you don’t simulate every iteration of every design; perhaps you use cloud CPUs for peak demand — there are ways to make this work. If it’s important enough, you’ll find a way.The next question is, “How do we get the data for our simulations?” You’ve already got most of it, if you have the CAD model for the product, component or assembly. You’ll need to know a bit more about the use case. For example, how much centrifugal force is exerted in a washing machine tub filled with wet towels? What’s the tensile strength of the steel in the support bracket? You can find those things out without too much difficulty.“Can we do this?” This is really the biggest question and it revolves around using simulation on every project, to anticipate problems and correct them before they get into production. Simulation is just a tool, and you can use it well or badly. If you set up the wrong boundary conditions or underestimate the forces, your results may be correct for the simulation but are incorrect when compared to real life. You need to learn how your simulation solution works; it’s easier in an integrated CAD/CAE environment. The integration also means that data is never translated, never leaves the design environment, which makes it possible to simulate early in the design cycle and then often as the design nears release. That enables the team to innovate while improving quality and performance.Your boss’s last question is likely to be a comment: “But this stuff is expensive!” The reply here is easiest of all: Software plus training plus hardware is cheaper than building all of the physical prototypes your team currently puts together. You may never eliminate all physical testing, but replacing even a fraction will save money. But this is rarely a money question when you dig a little deeper. It’s about brand image, customer satisfaction and knowing you did everything you could to make the product a success.Editor’s Note: What to learn more about PTC’s simulation strategy? Be sure to read this interview with the director of PTC’s simulation partner strategy, Mark Fischer. Also, be sure to read Geoff Hedges’ blog about how American Wave Machines is making waves with PTC simulation software. About the Author: Monica Schnitger is the founder, president and principal analyst of Schnitger Corporation. She has developed industry forecasts, market models and market statistics for the CAD/CAM, CAE, PLM, GIS, infrastructure and architectural/engineering/ construction and plant design software markets since 1999. Ms. Schnitger has extensively written and spoken on these topics for technology buyers, investors and developers, and has worked with clients both large and small all over the world.Original Post By Monica Schnitger | Published: Apr 7, 2014

Leave a Reply

Your email address will not be published. Required fields are marked *