Q. Why is process control important for plating operations?
A. Greg Arneson: As with all manufactured products, one needs repeatability for product quality and company profitability. Process control for plating operations depends on simultaneously coordinating control of time in tank, sequence, temperature, concentration, part/tank cleanliness and, if rectified, quality of the DC current, setpoint versus time, and total amp-sec/part.
Some plating processes are process control critical—for example, printed circuit boards, plating on plastics and titanium colorizing—while others are relatively forgiving, such as chrome on steel, aluminum anodizing and pickling.
Peter Gallerani: The goal of surface finishing process control is to manage sources of variation in the finishing process so that high-quality product is continuously and reliably produced. Like all manufacturing processes, finishing processes are subject to predictable and unpredictable (random) fluctuations. Predictable fluctuation can be managed by first understanding the key process parameters and second by measuring and controlling these parameters. Random fluctuations are inherent in any system, and one of the objectives of effective process control is to understand that random fluctuations happen and avoid over-reacting.
Effectively measuring and controlling the variation of key process parameters establishes a capable process which produces few defects. Capable processes are required for long-term sustainability in a competitive market because there is little room for waste. Ultimately process control is about cost control.
Art Kushner: The term “process control” has many different definitions and can vary dramatically from one company to another, so I think it’s important that before we discuss process control we define what we mean by process control.
Wikipedia provides a simple definition: “Process control is a statistical and engineering discipline that deals with architectures, mechanisms, and algorithms for controlling the output of specific process.”
In my opinion, the typical electroplating operation does not approach this level of sophistication. Yes, there are plating operations that are coming close to this definition, but the typical shop functions at a much simpler level.
Even if we take the idea of process control at the simpler level, which I would define as trying to keep process variables within set values, process control should play a very important role to the modern electroplater.
Why is process control important? The simplest answer is the electroplater must have reproducible results from batch to batch of parts. Reproducible results can include such things as deposit thickness and hardness, solderability and so forth. If we don’t have controls in place to obtain reproducibility we will end up with rejected parts, which have to be stripped and reprocessed or (worse) scrapped. This costs money and time and generates waste, and in my opinion is the most critical aspect of process control. The electroplating industry in North America has contracted and must compete with operations around the world. If our costs are higher than similar operations in other parts of the world, business will migrate to those areas. This is exactly what has happened in the last 10 years!
Joelie Zak: Comprehensive and well-designed process control plans are important because they provide for consistent plating production. Process control can help to quickly troubleshoot and implement corrective action when needed. It helps to reduce reject rates on parts being plated, as well as extend and/or minimize the amount of chemistry being used. In essence, process control is also a means of pollution prevention.
Q. How has process control historically been handled? What are the advantages and limitations of the “traditional” techniques?
A. Arneson: Tank process parameters, including temperature, level, pH, conductivity, dosing and voltages, used to be controlled by lab technicians, operators and/or stand-alone controllers. Some still do it that way, but for an every increasing number of sites it is done using PLCs for data input/output, with process data fed in real time to a computer for analysis. Consolidating all the data with each part as it is produced enables platers to maximize both quality and profitability.
Titrations are normally done manually even today. But, when titration is combined with automatic dosing, required titration frequency decreases and the titration, plotted on a scatter graph, results in much tighter concentration control. There are automatic titration systems available, but for most platers the return on investment is still not there due to both initial and maintenance costs.
Gallerani: For many facilities, process control is about choosing processes with the widest possible operating range and using all of that range. This may produce acceptable quality, but costs (materials, water, wastewater, labor, energy, and so on) are not optimized.
Traditional facilities rely on skilled and experienced operators and engineers to troubleshoot out of control processes. Process procedures are often written to minimally comply with customer requirements as opposed to effectively instruct operators. Undocumented “tribal” knowledge can be characterized as “the ability to troubleshoot”—remediation as opposed to preemptive action and control. A more robust approach requires documentation of “tribal knowledge” in well-crafted procedures and rule-based, automated corrective actions.
Kushner: For many shops, process control consists of testing one’s process tanks on some irregular basis, usually when something goes wrong. For others, process control means visual inspection of finished parts or the process tanks themselves before determining that an addition should be made to a particular process tank.
In some cases, rather than looking at the process, the plater blames the base material, the manufacturing process before electroplating, or some other reason that can best be described as peering into a crystal ball to come up with an explanation. Practitioners of this approach believe the main advantage is reduced costs (less record-keeping, less testing and so on), but this view is very short-sighted. This approach to process control will eventually result in increased costs, dissatisfied customers, and non-reproducible results.
Zak: Historically, the term “process control” in the metal finishing industry has been synonymous with “plating bath chemistry control.” However, looking at the plating process as a whole, there are a multitude of other factors that go into controlling the process in addition to just solution analysis. These include the cleaning cycle(s), racking techniques, rinsing techniques, filtration, equipment preventive maintenance and part testing, among many other items. Implementing a plan to regularly review and document these parameters is now considered process control.
Regarding the analysis of bath chemistry: modern technology allows us to continuously monitor many bath parameters, such as pH, temperature, and so on. Periodic analysis can be done in-house by most facilities. However, there are some areas of process control that are often neglected in facilities because they require manual testing. For example, periodically testing for metal impurities, or conducting frequent Hull cell analyses.
Q. How does process control differ between, for example, a plating job shop with manual lines, and a captive shop running automated systems?
A. Arneson: With ASC’s software there is very little difference. We have developed a solution for manual lines/tanks using the operators instead of automatic hoists. The operators are sent move commands for all loads in process, and tank process controls are just like those of an automatic line. The system uses our dynamic scheduler, and we fill out the plating “recipes” as you would with an automatic line. Each operator is equipped with a wireless PDA, which receives commands from the dynamic scheduling computer.
The operator gets real-time “move load” commands and responds to the computer via the wireless PDA once each load is moved.
We often add distributed PLC I/O at each line or tank so the computer, once notified the load has been deposited in the proper tank, automatically controls, monitors, alarms and logs parameters in each tank. This solution has allowed several shops to more easily meet NADCAP requirements for aerospace components.
Gallerani: It really doesn’t, except that an automated system makes extensive data collection and analysis for continuous improvement easier.
Kushner: Process control methods do vary with the type of plating operation. A captive plating operation that plates thousands of widgets each day can utilize more sophisticated process control methods than a job shop that has a number of plating processes and plates many different parts in a given period of time. The large captive operation can use statistical methods to help improve the process. A job shop plating many different types of parts using different plating baths has a more difficult time utilizing statistics.
That said, there are things that all types of finishing operations have in common. I look for written procedures for each type of plating, record-keeping for each process tank (testing results, history of additions, history of tank dumps and so on). I prefer to see day-to-day testing performed at the facility and not sent out to a laboratory. Testing should be done on a regularly scheduled basis, not just when something is wrong. Test results should be recorded in a notebook or spreadsheet and should be readily accessible.
Another very important aspect of process control is daily monitoring of the various pieces of equipment used on a plating line. Are all pumps working? Are tanks properly agitated? Are rinses functioning? What about filtration systems? If equipment is not functioning properly you will not get satisfactory plating. This is plain common sense.
Zak: The process control approach a particular facility takes depends on many factors, not just the type of finishing facility. Larger, high-volume facilities tend to invest in automating process control as much as possible, because even a small time lag between a problem occurring and implementation of corrective action can cause significant production losses—lines being shut-down and/or defective parts. A manual operation may not have these types of issues, but may have issues related more to operator error. In either case, a good plan with plenty of cross-checks is essential.
Q. What can plating shops do to economically boost process control capabilities?
A. Arneson: Whether you’re running manual or automatic lines, add automatic tank process controls. Also add true dynamic scheduling wherever it’s appropriate.
Gallerani: Start with the basics. Deciding which variables to measure is as important as how they are measured and requires experimentation. Furthermore, deciding how often to collect data is critical to both minimizing process variation and optimizing the cost of process control.
Process information must also be effectively managed for efficient and effective process control. The following are critical to effective process control and require no capital investment:
- Understanding the difference between critical and non-critical process control parameters
- Understanding true activity-based process costs, and setting control parameters to manage costs, not just comply with customer requirements
- Documenting process knowledge
- Developing effective process procedures
- Knowing the capability, accuracy and precision of analytical methods you’re using—you can’t control the process with an out-of-control laboratory.
Kushner: One of the biggest problems is is that many plating operations have a poor “handle” on the basics of process control. What a shop must do first is determine where it is with regard to process control. A shop may think it is following strict process control because it monitors a few parameters of a particular plating process, but in reality it may have missed the most important aspect, which in many cases is people. In other situations a shop may have standards/procedures in place, but they are not followed.
The best place to start is to audit your process control measures. The audit should be performed by an unbiased individual. This may cost some money, but the shop may discover that it has all the elements for improved process control in place, but just doesn’t have them implemented properly.
Zak: Shops should conduct periodic reviews of their process control plan, perhaps even having the review done by an outside source. The plan and review should cover all aspects of the process, including:
- Process line hardware: Evaluation of tank structure and configuration, rectifiers, bussing, motors, hoist mechanisms, piping, air feed lines, filtration devices and preventive maintenance schedules.
- Process line chemistry: Review of current bath specifications, analytical and monitoring techniques and addition techniques.
- Pretreatment processes: Review of part cleaning and surface prep prior to plating.
- Process line efficiency: Evaluation of plating turn-around time, common plating defects and failure analysis.
Q. What about the future? What’s coming down the line in terms of plating process control hardware, software or management techniques?
A. Arneson: Process control hardware and software continue to decrease in cost, but travel and labor cost increases will outpace the decrease, leaving a net increase in control costs.
A majority of plating lines in North America are manual lines, but the way to profitability is and will remain automation. For job shops with multiple substrates and processes, putting them all in one line and under a few automatic hoists with scheduling software will save on floor space, manpower, and wastewater treatment costs. For manual shops, converting to automation with scheduling will allow flexibility needed to do in-process mechanical gaging, color quantification and so on.
Currently, hoists are over-designed with wasted mass and therefore operating costs and motor sizes are not economically optimized. Expect companies to start using tools such as finite-element analysis to design automatic hoists that can pick up closer to their weight. Current systems can handle hoist weight:load ratios of ~4:1 at best. Platers will do the same for superstructure, which will reduce cost/foot of support superstructure materials. The combination of reduced hoist size versus load and reduced superstructure materials/running foot will bring the cost for automation within the reach of companies that might consider automated plating lines too expensive.
Gallerani: Hardware and software will become less expensive and more robust.
Kushner: I think the biggest strides in process control will be in the area of management techniques. In my experience, a great number of screw-ups and rejects are due to the humans involved. Management quite often doesn’t really work to get the operators to buy-in and adapt the procedures used in a modern plating facility. Operators are under-trained and in many cases don’t understand what they are doing. The solution to this problem is better training of operators and getting them to be “owners” of what they are doing.
No doubt there will be better process control hardware and software, but if the people monitoring it are not better trained, process control will still be a hit-or-miss proposition.
Q. Anything else you feel our readers should know about plating process control?
A. Arneson: The current trend of adding surveillance cameras in cities and buildings will take hold in the wet process market. These cameras have no moving parts and they record automatically at 10–30 frames/sec. They can be programmed to download images from before, during and after a user-selected event—for example, an audible noise, disturbance of user-selected pixels or a specific PLC input.
For automatic lines, the cameras will provide visual forensic files for any event for the entire line, including rack crashes, hoist crashes, operator interventions, chemical adds and so on. They can do the same for manual lines, recording the actual load going in and out of each tank along with a date and time stamp on each image. This will allow hard visual proof of both the sequence and time in tank each load received without any scheduling computer.
Kushner: The most important thing people in our industry should be aware of is that good process control doesn’t necessarily mean you have all sorts of computers, bells and whistles, and automated testing methods. It means you truly understand your process and follow the process control steps on a daily and continuous basis. This means doing the chemical tests as required, confirming that operators are following the written procedures for each type of part/process and, most importantly, demonstrating that senior management believes in and is supportive of good process control.
Zak: A plating line is constantly in a state of flux, due to changes from part configuration, drag-in, and changes in production. Being able to control as many parameters of the line is essential for consistent production output.