|External view of degreasing system, as seen at a recent trade show.
Choosing a new cleaning process can be a challenge for any production engineer. Any new cleaning process is expected to be fast, safe, sustainable, versatile, and affordable, and to clean more with less. Industry regulators are constantly imposing new air-quality rules, new shipping, handling, and storage regulations, and new waste treatment rules. Some locations may even be more difficult than others, with a lack of water or rising energy costs, all of which can limit a company's cleaning options. But fortunately, production engineers can apply a procedure (referred to at MicroCare® as a "cleaning scorecard") to select an optimal cleaning solution even when faced with these conflicting demands.
It is possible to keep track of scores for different cleaning solutions. Many think that the scoring begins and ends with the cost of a cleaning machine or drum of cleaning solvent, and that the lowest-priced machine or lowest cost/liter of cleaning solvent is the best choice. In truth, there are many other factors that are much more important than the cost of these two items.
As in an automobile, where fuel economy is measured in terms of miles/gallon, smart engineers need some form of score or "cleaning index" to measure the economics of cleaning electronic parts. Perhaps the most practical cleaning index is the total cost per part cleaned, which focuses on how to clean a product at the lowest total cost. Such an index does not just examine the cost per liter of cleaning solvent but on the total cost per part cleaned, which is where profits will be made or lost. Of course, depending upon the environment, some customers may perform to examine the cost of cleaning per day or the cost of cleaning per shift as alternative ways to compare cleaning costs.
Cleaning machines come in many different sizes, although comparing them may not make size in inches or feet, but more in terms of "production capacity" or "cleaning capacity." In terms of size, it is this cleaning capacity that must be matched to a production facility to determine the best fit. A "cleaning fit" can be measured in a number of ways. The least precise, but often effective, method is to measure the number of units that must be cleaned in a given time period, such as 300 printed-circuit boards (PCBs) per day or 500 pieces per week. Another approach is to estimate the total surface area of all parts to be cleaned, which is a useful method when dealing a large variety of parts with different shapes and sizes. However, it is important to note that very small or fragile components may have different cleaning requirements than larger pieces, and the same cleaning methods may not be applicable for a wide range of different shaped and sized components.
|Degreased PCB drips into tank.
Once required cleaning capacity as been determined, different types of cleaning technologies can be evaluated. For example, bench top cleaning machines are slow, but compact and inexpensive compared to faster, larger, and more expensive high-volume cleaning systems. Different cleaning options can be evaluated by preparing a batch of typical products to be cleaned for each cleaning system manufacturer to run them through their systems. Each cleaning system manufacturer should be capable of producing a brief written report that describes the process, solvents, temperatures, times, and results.
Testing the capabilities of the different cleaning systems is the easiest part of the selection process. The most difficult part is computing which system produces clean parts at the lowest total cost. It is possible to assemble a "shopping list" of the major costs, although the list can change depending upon each cleaning project.
Part of estimating the costs of cleaning electronic assemblies with different systems is to estimate the average productivity of different cleaning systems in terms of assemblies per hour. This is crucial to computing the cost-per-part-cleaned since operating and labor expense are usually determined as hourly costs. Comparisons of different cleaning systems can start with the cycle time, which is the duration of one complete cleaning cycle, including loading and unloading parts. For example, a cleaning machine that cleans 20 boards simultaneously during a 40 minute cleaning cycle has a throughput of one board every 2 minutes or an average throughput of 30 boards/hour. Cleaning machine manufacturers should be able to provide benchmark numbers of their systems for comparison.
|PCB gets ultrasonic cleaning.
Cleaning system throughput can be affected by many factors, including loading and unloading times. Water-based machines often cycle quickly when cleaning large parts with simple shapes, while vapour-based systems are faster when cleaning parts with tight spaces and components with many voids. Throughput has a dramatic impact on operating expenses. For example, DuPont reports that in a small, modern vapor degreaser, "normal" solvent losses are about 0.118 lbs./ft.2 of vapor area per hour of operation, or roughly one pound of solvent lost per day (larger and more-efficient machines have lower loss rates). In stand-by mode, this loss drops 75%. To compute total cleaning costs for a system, a fairly accurate estimate of the throughput requirements are needed for cases of significant cleaning material losses.
Adding One-Time Costs
For a fair comparison, all the costs for each cleaning system must be assembled into a standard unit of measure cost, the total cost per part cleaned. This is simple enough for consumable materials, such as cleaning solvent and electricity, but should also consider initial acquisition costs. Such costs include the actual cost of the machine, freight, site preparation, and setup costs. A new cleaning system may also require building renovations, ventilation enhancements, electrical upgrades, and water-treatment subsystems. When developing plans for a new cleaning system, a facilities manager, health and safety people, environmental people, fire-safety people, and the production team should be involved in those plans for a realistic estimate of up-front costs.
The funds required for a new cleaning machine can be tabulated with a software spreadsheet, with calculations of cost-per-month for the equipment easily converted into a cost-per-part for cleaning costs. Part of the equipment expenses are the cost of the space on the factory or production floor required by the cleaning machine and any support systems for the cleaning equipment. The space required is usually defined as a multiple of the physical size of the cleaning machine.
Aqueous and semi-aqueous cleaning systems typically require more floor space than solvent-based cleaning systems. They also can require water-treatment facilities that can be as large as the cleaning systems themselves. (Note: an excellent review of aqueous cleaning technologies and required water treatment systems can be found in the book, The Handbook for Critical Cleaning, by Kanegsberg & Kanegsberg, published by CRC Press, 2001, in the chapter "Aqueous Cleaning Essentials.") Aqueous systems also have slower cycle times, so more space is needed for work-in-progress, supplies, conveyor systems, and access aisles. As an example, an aqueous cleaning system with 200-square-foot footprint tied up 1400 square feet of floor space with ancillary systems, for a 7X floor space factor. For a vapor-based cleaning system, a 4X floor space factor would be a more reasonable rule of thumb.
Determining Labor Costs
Any labor costs should be based on the "fully loaded labor rate" for those technicians who will operate the cleaning machines. Their time required to maintain the equipment, including solvent levels and waste treatment systems, should also be included in cost estimates. In addition, the costs of training, of maintenance technicians, and any chemical safety training should be included in estimations of labor costs. If turnover is a problem, funds can be added to labor cost estimations to cover quarterly supplemental training.
Any cost estimates for cleaning systems must include funds for maintenance, since large systems can have complex maintenance problems. Aqueous cleaning systems can develop complex problems, in part due simply to the large size (such as 30 feet long) and number of moving parts in these systems. Aqueous cleaning systems also have complex water-treatment and recycling processes which must be maintained and sustained for the life of the machine. Also, the alkaline additives that boost the cleaning power of many of these systems can coat the system's interior surfaces and cause additional maintenance problems. However, vapor-based systems are not maintenance free; filters must be checked and replaced, solvent in the degreaser must be boiled down, and sludge at the bottom of the system must be occasionally removed. This usually results in the loss and disposal of approximately 10% of the solvent in the machine on a quarterly basis.
Calculating Indirect Costs
With the unwanted costs of litigation, the various materials used in cleaning systems must be properly handled. For example, n-propyl bromide (nPB) is an excellent cleaner but, due to significant toxicity issues, it must be handled properly; an old, leaky degreaser, for example, can expose employees to dangerous levels of this solvent. The tradeoffs between cleaning effectiveness and health risk can be expensive and must be considered for each application.
Environmental requirements can also play a role in calculating the costs of cleaning systems. For example, in California, tight volatile-organic-compound (VOC) regulations suggest that water cleaning might be optimal. But the high costs of water and electricity may make that choice less affordable today.
Planning for the future in terms of cleaning systems can be difficult. For example, in the 1990's, the electronics industry migrated from vapor-based cleaning systems to aqueous technologies, enjoying years of successful cleaning of through-hole and surface-mount-technology (SMT) assemblies with those aqueous cleaning systems. But as the densities of BGA circuit boards have increased, many electronics manufacturers are shifting back to vapor-based cleaners to more easily clean dense circuitry with less electricity and without water.
Contact: MicroCare Corp., 595 John Downey Drive, New Britain, CT 06051 800-638-0125 or 860-827-0626 E-mail: firstname.lastname@example.org Web: http://www.microcare.com