Quality Scan: Focus on What Works to Improve Quality and Functionality
In a perfect world we wouldn't have to be concerned about manufacturing variables. Parts wouldn't ever be too big or too small, too hard or too soft, too long or too short, too thick or too thin, and our assemblies would always work exactly as they are designed to.
But we don't live in a perfect world, and the parts we deal with every day are all of those things and more.
There are two traditional answers to this problem:
1) Specify tighter and tighter tolerances or
2) Inspect everything.
Both approaches work, but costs mount quickly.
Fortunately, there is a third alternative--measuring critical part parameters or functions during assembly, and using this information in a closed-loop system to manufacture better parts at reduced cost. It's based on the realization that products are made to provide a function.
How well and how reliably they perform that function is the best measure of quality. In short, stop focusing on how well things fit, and start looking at how well things work.
Take, for example, a simple pressing operation used to install studs in a wheel hub. The force required to seat the studs varies from stud to stud and hole to hole due to manufacturing variables. Conventional wisdom says to set the press at the highest force needed to seat a stud, and then simply hit them all that hard to make sure they're seated. Result: a lot of hubs will be scrapped because the insertion process overstressed them unnecessarily, and distorted them beyond the flatness tolerance.
Replace that "dumb" hydraulic press with a "smart" electromechanical press equipped with sensors to monitor the insertion process and a controller able to decide when the operation is successfully completed--regardless of the amount of force required--and you have an entirely different solution. By setting upper and lower limits for the process variables, such a system can weed out those assemblies likely to fail in the field, while producing a much higher percentage of "good" assemblies (even allowing the use of less-expensive components with much looser tolerances).
You can apply the same principle to functional assemblies such as automotive hood latches. These are typically riveted together with a "dumb" press (of some kind) providing the force. Given the normal variations in rivet hardness and latch-component thicknesses, a lot of these assemblies end up in the scrap bin because they simply don't work.
But, if you use a smart press with appropriate monitoring and control technology, the moving parts can be cycled while the rivet is pressed. When the force that's required to move the latch reaches the specified value, the operation is complete--again, regardless of how much force is required. The result is near 100% production of good latch assemblies.
Another example, this time in the hydraulics industry: Pressure relief valves are calibrated by adjusting the preload on a spring that holds the poppet down against system pressure. This typically is done by pressing a retainer down on the spring.
If you simply press all the retainers to the same point, spring variables will produce a fairly wide range of actual pressure relief settings in apparently identical valves. For 1000-psi (6.9-MPa) valves, a 200 psi (1.38-MPa) variance is fairly typical and, even then, scrap runs to 6% or more.
Using a smart system, though, you can monitor fluid pressure while the retainer is being pressed. When the specified pressure is reached, the operation stops--regardless of retainer position. Result: a consistent pressure variance of only 62 psi (427 kPa) and a reduction of scrap to near zero.
Results of this kind are common when manufacturers adopt technologies that allow the process to adapt to variables rather then trying to control every variable down to the last micron. At the heart of this change in mindset is a measure-and-press, press-and-measure philosophy. In other words, you test it as you make it.
You often hear the phrase "we must work smarter not harder" applied to employees. But this concept needs broader applicability: it must apply to assembly processes as well if you are going to be competitive in the world market.
Changing your focus from what fits to what works is the first step.
This article was first published in the June 2005 edition of Manufacturing Engineering magazine.