How to Improve Quality in Motion Control

Six Sigma methods can be a powerful tool to increase the quality and productivity of processes involving motion control.

By Kevin Frantz, GE Fanuc Automation North America Inc. September 1, 2001

All Six Sigma projects begin with team building and training,including frequent meetings between management and shop-floor employees.

I n any discussion about quality control, you are bound to hear the words Six Sigma. Other terms, such as “Black Belt” and “CTQ” (critical to quality) may be familiar, but how exactly do those words, and the Six Sigma quality process, apply to your motion control process?

Six Sigma has achieved dramatic results within companies like General Electric, yielding an annual return of over $100 million in quality improvement savings. But how can you make it work in your business, and where do you start?

Your first task in applying Six Sigma to a motion-control process is to understand its purpose, which is always to optimize quality and throughput.

Six Sigma is a tool that can unveil subtle problems that plague all processes, silently stealing productivity and quality. Like all processes, motion-control applications are not immune to production downtime and quality control issues. Slow machine set ups, product positioning problems, equipment failures, out-of-spec parts can all contribute to the problem. You may not know the exact cause, but you can see the negative results: low production counts, lackluster quality, customer rejection, and lost money. Six Sigma could be the solution to those problems.

Striving for Six Sigma soldering

O ne of the most surprising and rewarding projects for motion-control quality at GE Fanuc was initiated in the mid-1990s, shortly after the corporate-wide commitment to Six Sigma by parent company GE. The program continues to evolve today-attesting to the enduring power of Six Sigma.While evaluating fluxing system options for a Six Sigma wave-soldering project at GE Fanuc’s Charlottesville, Va. manufacturing facility, engineers tested and selected an “Opti-Flux I” spray-fluxing system from Ultrasonic System Inc. (USI, Amesbury, Mass.) to resolve flux issues that hampered printed wiring board (PWB) production for PLCs. As a result, USI became part of GE Fanuc’s Six Sigma team, and later launched a new, second-generation spray system, Opti-Flux II, with an innovative GE Fanuc motion-control solution that improves soldering quality.

Servo-based control of the spray head traverse, supplied by GE Fanuc, allows Opti-Flux II to independently adjust flux flow rate and traverse speed for precise application of flux.

Placing motion under control

Control of flux deposition in the wave-soldering process is essential to the quality soldering of through-hole components on PWBs. Amount of flux applied to the board is critical to the no-clean wave soldering process, as the flux must be completely consumed during soldering, so no residue remains for cleaning.

In the past, PWB production at GE Fanuc involved dousing the boards with large quantities of water-soluble fluxes followed by a cleaning system at the end of the line that removed residue and excess flux. Once this process was put under Six Sigma’s microscope, Measure and Analyze activities (see main article) revealed that the dousing process used excess flux, produced solder-joint defects, caused frequent setup delays, and required additional cleaning after soldering. Thus, the Six Sigma project’s main goal was to implement a method for accurately applying just enough flux to ensure solderability of components and eliminate the extra step of cleaning the PWBs.

Opti-Flux I used USI’s patented nozzle-free ultrasonic spray technology coupled with a synchronized traversing head that enabled a single, virtually uniform flux coating of the PWB. Effective through-hole penetration was achieved even at low flux volumes; and one-step setup Quality soldering (continued) allowed operators to quickly enter the recipe for the PWB flux application, resulting in virtually no downtime.

Ongoing process

Ultrasonic System’s Opti-Flux II spray-fluxing system helped boost GE Fanuc’s Six Sigma project and printed wiring board production at Charlottesville, Va. by effective through-hole penetration even at low-flux volumes.

While Opti-Flux I improved the company’s fluxing operations, the pneumatically controlled traversing head could induce slight variations in flux depositions across the wiring boards. After GE Fanuc approached USI and discussed the variations inherent in Opti-Flux I, USI joined the Six Sigma team.

By applying the same Six Sigma steps that GE Fanuc followed-Define, Measure, Analyze, Implement, and Control-USI was able to resolve flux variations of Opti-Flux I by implementing a GE Fanuc servo-motor-based solution. This optimized control over the speed of the traversing head and ensured uniformity of the flux application.

Next-generation Opti-Flux II uses a belt-drive linear actuator with a servo motor to traverse the spray head, allowing independent adjustment of flux flow rate and spray-head traversing speed for precise application and wide flux-position range. It replaces the air-actuated cylinder traverse method of Opti-Flux I.

Operators control motion by entering deposition specifications into an Operator Control Station in micrograms/in.2, after which the system automatically adjusts flux flow rate and traversing speed of the spray head.

With the motion-control enhancements of Opti-Flux II, each PWB now receives one non-overlapping, uniform coating of flux as it passes over the spray station. This new level of control was attractive not only to GE Fanuc’s Six Sigma project, but has also become a competitive advantage for USI, as Opti-Flux II is meeting with great success within the electronics assembly industry.

Since implementing the Opti-Flux II system with servo-motor control, GE Fanuc has realized several benefits. These include improved plated through-hole soldering, resulting in 75% fewer solder defects and 45% reduction in per-board flux usage, as well as lower maintenance costs and set-up time.

While moving closer to the zero defects goal, GE Fanuc’s pursuit of Six Sigma continues to evolve and reveal new, surprising variables, as well as help the company repeatedly “solve for x” at the plant.

Cut your losses

Let’s say you run a shop that manufactures weather-stripping products, and out of 5,000 strips cut and drilled daily, 500 strips are rejected because the cut lengths and drilled holes are out of specification. Losing 10% of production to scrap, rework, and repair translates into higher labor and materials costs to satisfy customer demand, and an increased risk of missing delivery deadlines.

So, how do you begin to correct this loss and apply Six Sigma? Actually, by having identified the problem as lost production, you also identified the place to start: the cutting and drilling area.

All Six Sigma projects begin with team building and training. For GE Fanuc, before the Six Sigma workshop begins, team members from our staff and from the customer’s staff are selected to be part of a Six Sigma team. GE Fanuc’s team must have a Master Black Belt (MBB) trainer, a project sales manager, and a salesperson. The MBB is responsible for training and teaching the group Six Sigma concepts and for facilitating the workshop. The project manager and salesperson organize the program logistics and maintain focus of program goals. As the customer, your team would include skilled workers, supervisors, managers, etc.-your experts on the process, which, in this case, is the cutting and drilling of components.

Following the DMAIC roadmap for Six Sigma projects, GE Fanuc divides the project into three steps: Define and Measure; Analyze; and Implement and Control.

Step one, Define and Measure , combines two major objectives. First, the Six Sigma team, outlined above, is trained in the tools and concepts of Six Sigma. Second, the tools are applied to the specific customer problem, cutting and drilling out of spec.

All Six Sigma projects start by identifying and defining a “problem statement.” Developed in the Define phase, this statement defines the defects caused by the problem, identifying opportunities or benefits that can be achieved once the problem is resolved, calculating the Six Sigma baseline, establishing a Sigma target, and finally, calculating the financial impact. In the case of lost weather-stripping production and quality defects, a problem statement might look something like this:

“Part Number 1 Weather Strip is producing scrap due to out-of-spec cutting and drilling and is impacting costs, quality, and delivery of `My Company.’ From January 1 through June 1, scrap represented an average of 10% of daily production rates. This causes missed schedules and increased overtime. The goal for this project is a scrap level of 0.75%.”

Six Sigma dictates that process data must be measured before any improvements can begin. The team starts by creating a high-level process map followed by a detailed process map, which are used as reference points throughout this and subsequent phases. This is the stage where Critical to Quality measurements are identified to quantify performance. The Measure phase also includes identifying Key Process Input Variables (KPIVs) that affect CTQs. A Cause-and-Effect matrix is often used to identify the most important KPIVs. Lastly, a measurement system is identified to quantify results.

In a cutting and drilling process (photo, next page, CTQs would likely include holding tolerances-for example,

Analyze your measured data

All data collected from the Measure phase are then analyzed using statistical methods. The goal of step two, the Analyze phase, is to determine which variables are correlated with the defect studied, and then analyze the interactions among variables. From these data, the team can make inferences about which variables are important to the process, what the interaction is among the variables, and possibly determine specifications for the variables. So, during the Analyze phase, you might determine that quick starts and stops are causing the rubber material to stretch and bounce back during the cutting and drilling steps, thereby throwing tolerances out of specification.

At this point you may also determine that the solution lies in more accurate acceleration and deceleration control of the servo motor. This leads directly into the third and final step, Implementation and Control -the culmination for all data collection and analysis, as it applies to all the team’s conclusions and to specific recommendations that will improve and control the process. This step is typically quite large in scope and may be repetitive as data analysis and improvements are performed incrementally.

To obtain more accurate start/stop control, you might decide to implement a servo motor with a digital drive that allows user setting of different acceleration and deceleration rates. As a result, stretching of the rubber material during cutting and drilling becomes less likely, with significant reduction in scrap rates due to out-of-spec dimensions.

Implementation and Control extend beyond immediate remedies to include control of the process over the long run. For many applications, this step includes a control plan and some form of “watchdog,” such as statistical process control (SPC). It may also include an update on pertinent ISO Standards and additional raining, depending on the complexity of the Implementation phase. So, while Implementation and Control comprise the final Six Sigma step, they are not the end of the process. Actually, they’re a launch point to begin proactive management of the process and direct its outcome.

Critical to quality (CTQs) measurements for a cutting and drilling process include holding tolerances for length and hole-to-h ole dimensions. KPIVs might include acceleration and deceleration rates.

Solving for X

While the motion-control example fits nicely into each Six Sigma step and has a simple solution, not all Six Sigma projects are so compliant. In fact, many solutions need to address areas outside the process that are not always apparent from the start.

For instance, you may have found the rubber weather-stripping material was substandard or the storage temperature following production was causing the rubber to change shape, or employees required a higher level of training to program and operate the machinery. Any number of unforeseen variables can impact product quality between the time the raw material is gathered by the supplier until the time it reaches your customer’s hands.

That’s the power of Six Sigma-it helps find and resolve these unknown variables. With an experienced and committed team, Six Sigma can optimize any motion-control operation and lead the way to near-perfect quality control.

Kevin Frantz is a Six Sigma leader at GE Fanuc Automation (Charlottesville, Va.).

For more information, visit www.gefanuc.com or go online at www.controleng.com/freeinfo .

Further Reading

F or more on Six Sigma methods, see Control Engineering , Jan. 1999 (pp. 62-70) and Mar. `99 (pp. 87-90, 103). A sidebar in the latter article relates Six Sigma methods to ac motor and drive production.

Comments? E-mail Frank Bartos at fbartos@cahners.com