One of the most challenging engineering projects of late was the rapid development of Vauxhall’s Vectra plant. Brian Tinham finds out how it was achieved
Last year’s speedy launch of the Vauxhall Vectra owes its success in part to one of the most sophisticated collaborative engineering projects the automotive industry has seen, involving robotic simulation, web-based simultaneous engineering and off line cell programming (OLP) with suppliers around the world. It enabled massive development at Opel’s plants in Rüsselsheim, Germany, and Ellesmere Port to ready production in just two years.
The project was ambitious right from the start. It required automotive welding plant and assembly line builder TMS Produktionssystem and TMS-South Africa to work together with engineering services company Design Group in the UK and South Africa to establish and prove the assembly line design across all engineering disciplines in a very tight time scale and budget. To achieve that, they used the Delmia digital manufacturing planning system for everything from generating 3D virtual layouts and robotic simulations, to programming the robots and providing the central working engineering environment, via the Internet, for all involved.
Massive savings
Nick Stanford, simulation manager at Design UK, which worked on the front wheelhouse robotic cell, while Design South Africa handled the rear wheelhouse, says: “It was a big factory floor project. Opel provided an initial cell concept, while we proved out the efficiency of the cell and the reachability of the robots at the same time as designing the tooling. We had to simulate all the welding processes, to prove the robot welding operations, check movements and reach, the tools and the rest before they started to bolt them down to the floor. We also had to write the robot programs offline and download them.”
In all there were about 150 robots to program and simulate in concert prior to installation and commissioning, and the whole process took around six months – about a quarter of the time it would have taken with conventional techniques. Sanford: “We saved about 75% of the lead time. Without the software we wouldn’t be able to ‘see’ what the robots were doing so problems would have to be put right at the pre-production manufacturing stage.”
And it’s worth noting that it wasn’t just a time saving. Sanford says decisions and changes were being made sometimes by the hour, “so the simultaneous engineering environment was essential. For example, the decision to use an extended reach robot, or move it, effects everything, including the procurement process, and we’ve got to agree performance for operations…” And he adds: “It also meant we saved about 25% of the engineering staff.”
Interestingly, Design hadn’t used Delmia before, having been a Robcad house. But as Opel specified Delmia for the OLP, it bought the Igrip and UltraSpot modules. Nick Herrmann, Design UK’s managing director, says he was reluctant to invest in £50,000 worth of software, but confirms that these are “superior simulation tools.”
Sanford comments: “Systems like Delmia and Tecnomatix do more and provide better presentation.” And he adds that Delmia in this instance was fast, easy to drive with just a week’s training, and ran on a PC, making it familiar and keeping the cost down. He also makes the point that with the tooling design done in Dassault Systemes’ Catia v5 CAD, everything, including Delmia, is in the one environment. “The virtual tooling went straight into the simulated robotic manufacturing cells on the Delmia system.”
Says Herrmann: “It is a manufacturing truism that the shorter the cycle times, the more equipment is required for any cell. Accessibility can then be a problem, but Igrip was able to verify clearances between 1 and 5mm. With the front and rear wheel house robotic cells completing 66 jobs per hour and with the average weld taking four seconds, the reachability and clearance tolerance figures had to be utterly trustworthy.”
Dave Kupferman, director of design automation in South Africa, adds: “This is the biggest project combining both simulation and OLP we have ever attempted and we simply would not have been able to undertake it without [this]. Even experienced eyes can analyse a robot cell and believe it will perform, but as Igrip configures each and every point, robot singularities soon become glaring.”