How CAD eases the build/buy dilemma

Jan. 9, 2014
How Park Manufacturing Makes Retaining Clips for Wheels on Plastic Garbage Cans
About the Author

George Radcliffe is engineering manager at Park Manufacturing in Melrose Park, Ill.

We usually make custom tooling for the wireforms and strip metal end products we manufacture, but when a job requires a lot of machine time or needs a special function, we might also design and build our own machine for it.

For instance, we recently started a project that's become one of our best jobs — making retaining clips for the wheels on plastic garbage cans. Originally, we weren't sure of the quantity of clips needed, so we made tooling for our GRM80E multi-slide forming machine from Bihler and then designed a dedicated machine as production ramped up.

SEE ALSO: System Design Tools Make Simulation Real

The GRM80E gives us a nice platform to build onto, but it's costlier to use, and if we fill it up with 2,000 hours of making one part, then we can't use its flexibility for something else. So when we're going to make a lot of one part, it's better to offload that work onto a single-purpose machine. To decide when to make this switch, we do a cost-benefit analysis. With this clip project, we determined we could design and build our own dedicated machine for about one tenth of the total cost of buying and customizing a second GRM80E machine.

We were an early adopter of Autodesk Inventor, and we use it to do our CAD and simulation in-house. For the single-purpose clip machine, we took the 3D model of the tooling we were using on the GRM80E and designed our new machine around it. We started with the tooling mount geometry and slide motion requirements and then worked outward — all the way back to the frame. The result is a multi-slide machine, but it's our design, dedicated to just one family of parts.

Our design process always starts with the finished part specifications that our customer wants and works back from the customized forming and cutting tool geometry to the more general, off-the-shelf drive components it will require, including cams, linear slides, rotary actuators, gear or belt drives for power transmission, and ac motors or servos. Finally, we add a control system that includes the user interface, sensors, drives and controllers needed to make the whole system robust, reliable and easy to use.

We also use Inventor's positional representations. Most of our metal-forming processes are naturally "interference" designs, with different components occupying the same space at different times. With positional representations, we can step a machine through every point in its forming cycle, so we can be sure the correct cam timing and geometry are built into the design. By the time we're on the shop floor assembling the physical machine, there's almost zero need for rework.

The clip machine was a successful project, and we gained a lot from creating and verifying the complete digital prototype in Inventor before we cut or even bought any steel. We were able to analyze almost every part of the machine prototype in software and make precise digital refinements to avoid costly, inefficient fixes at build time.

Because we do so many one-off designs, having a digital prototype like this gives us a lot more confidence, especially as we take on more complex projects, such as our dedicated clip machine. So far, it's made millions of clips and is still running strong, which is a real testament to the quality of the design. In fact, it probably wouldn't have been possible without using Inventor software. This is because physical prototypes require a lot more tinkering once they're in place on a shop floor, and this leads to more time and expense. If we couldn't make a good digital prototype, the difference could even tilt our decision to build a machine back to having to buy a more-general, less-dedicated machine, or it could mean we just couldn't do some of the more complex projects.

However, with our machine designs, we've also learned not to jump into using the most advanced electronic automation components just because they're new or easy to design around. For example, we could have used several types of servo actuation throughout our clip machine, but we just didn't need the flexibility of programmable servo motion on every action. Instead, our clip machine synchronizes 10 different mechanical motions with just two electronically linked servos, one of which drives a timing belt though a series of steel cams. This system might have been more difficult to design, but it makes the machine faster, cheaper and more reliable in the long run, and that's what counts.