The economic benefits of automating high-volume work are readily apparent. But for shorter runs and frequent product changes, a more strategic approach is called for. System flexibility is key to automation amortization.

In today’s global marketplace, processors in high-cost production environments are increasingly called upon to manufacture parts of greater complexity, albeit often in smaller production runs. Simultaneously, they need to automate in order to remain competitive.

In this manufacturing environment, successful automation is often determined by what equipment is invested in upfront. “The more prior information the processor provides [on what parts may be molded], the better suited the equipment will be for a variety of applications,” says Jim Schmitz, applications engineering manager at Wittmann Battenfeld Inc. (Torrington, CT). “Processors should install robots with sufficient grippers or vacuum circuits for the most demanding application conceivable,” he notes. Sufficient input/output (IO) capacity should also be built in for tasks such as conveyor indexing and box filling. “Once a system is out in the field, adding options can be more difficult,” says Schmitz.

Says Scott Kendrick, engineering manager, Sepro America (Pittsburgh, PA), “The key is to size the robot to the press rather than a specific application or part. This would give you the flexibility to handle any part that is likely to be molded in that particular machine.” If an oversized mold may be used (one for a part that extends beyond the tie bars), this needs to be taken into account by making sure the robot has the vertical stroke and wrist motions necessary to manipulate the part through the available opening.

Integration and compatibility with third-party equipment such as conveyors is also an important consideration, according to Schmitz. “Who is responsible to integrate the cell when using equipment from multiple vendors should be discussed up front,” He says. “For example, a lot of robots use either sinking or sourcing for their switching methods. It’s a whole lot simpler if the entire system uses the same switches.”

Programmability

Schmitz also suggests processors should opt for robots that are programmable, meaning they can be configured for a variety of tasks without having to install new hardware. Naohisa Kawaguchi, planning section manager at Harmo Co. (Nagano, Japan) concurs. The company’s I Series controller for its HRX robots enables a molder to add an automation step in as little as 15 minutes without the operator requiring any special programming skills. “The controller can also communicate with other peripherals such as the dryer and mold-temperature controller to automatically change settings if a mold is changed,” adds Kawaguchi.

“A control system that is easy to program and can store and recall a range of different part handling recipes is important,” says Sepro’s Kendrick. “This will minimize changeover time, and you can even add a binary interface between end-of-arm tooling (EOAT) and the robot so that the necessary programming for each set of tooling is automatically downloaded when the tooling is mounted on the robot arm.”

Yushin Precision Equipment (Kyoto, Japan) also advocates ease of use as a key parameter for successful automation. “Changing the mold is the most time-consuming step in product changeover, and whatever else you can do during this time in an automated manner that reduces or eliminates the manual burden will contribute to productivity,” says Takeshi Okamoto, manager of systems engineering at Yushin. This includes changing the settings of peripherals such as the mold-temperature controller.

However, Sepro’s Kendrick says there is very little demand for robotic EOAT changing. “It is not usually necessary to go to the expense of automating tool changes since quick disconnects and automatic program loading make manual changing of EOAT easy and fast,” he says.

Automation-ready molds

Mold design is another area where foresight can make or break successful automation. “In many cases, basic design changes can make automation much less expensive, and most of these changes, if discussed early enough in the project, can be made with no additional cost to the mold,” says Schmitz. These include issues such as how the mold will hold onto inserts, mold centerlines that match packing tray centerlines, tray capacities that are exact multiples of cavitation, and part orientation. Centerline and orientation differences can be accommodated using pneumatically-actuated expanding/ collapsing and ‘wrist flip’ EOAT, but extra costs are entailed.

Adds Sepro’s Kendrick, “During the mold design process, it is easy to add features and capabilities that ensure smooth, precise robot operation. For instance, locator blocks can be added for accurate placement of inserts, and pins or blocks can be matched to the molding machine platen to ensure that the mold is installed in exactly the same position every time and the stored robot positioning programming does not need to be fine-tuned.”

Another critical issue is repeatable part control when the part needs to be captured by the robot in precisely the same way every cycle, especially if it is going to be positioned for secondary operations outside the press. “This is where the Sepro ‘Y-free’ function becomes very important,” says Kendrick. “Parts are normally in their most stable position while they are still on the core or ejector pins in the cavity of the mold. With Y-free, the robot EOAT can get a grip on the part and then signal for ejection. Then the strip arm floats freely with the part as it is ejected from the mold. Once the part is completely free of the ejectors, the robot drives take over to remove it the rest of the way from the mold.”

The Flex Cell

ABA-PGT Inc. (Manchester, CT) is one processor actively implementing automation to secure business that was formerly out of reach. The company’s ‘Flex Cell,’ developed jointly with Wittmann, is flexible enough to allow for molding of one product with one type of insert and/or packaging tray, and then switch over to a completely different product within a few hours. “The upfront costs to the customer are low enough to provide them with payback over the span of a few hundred thousand pieces as opposed to a few million pieces,” says Tim Vale, Director of R&D at ABA-PGT. “Designing an automation cell around a clearly defined product is not always that easy, but it is certainly a degree or two easier than trying to design an automation cell around a bunch of parts that don’t exist yet,” Vale comments. “Coming up with the parameters was a challenge because you were making an educated guess on what type of parts you might see in the future based on what was quoted in the past.”

The Flex Cell concept, which includes the molding machine, robot, and automation system for presenting inserts to the molding process and a similar system for handling the finished product (tray pack or bulk pack), is owned by ABA-PGT. The only costs to the customer are about $2 more per hour to the machine rate to cover the added cost of the automation and the cost of product-specific components. A typical EOAT for the Flex Cell costs anywhere from $15,000 to $25,000, depending on the complexity and number of cavities in the mold. The total cost of the pallets for the inserts is about $6000.

Feeding of inserts can be the most costly component to automate and in the Flex Cell, it is semi-automated. An operator is still necessary to place the inserts on the pallets, but this requires less than 15-30 minutes every two hours.

Fully automated insert presentation to the robot involves feeding (usually a bowl feeder if parts are supplied in bulk), orientation of the insert (either mechanically or with the aid of a vision system), sometimes inspection (either mechanically with a go/no-go type of fixture or with a vision system) and placing the inserts in the correct spacing to match the EOAT and mold (possibly with a small pick-and-place robot). Such an investment might be justified for millions of parts but, “For lower volume, it’s hard to justify not only from a cost standpoint, but also in terms of floor space,” says Vale.

Arburg (Lossburg, Germany), however, says the modular construction of its systems enables it to feed various insert types. It has successfully automated production of a knurled wheel, where a threaded rod and threaded bolt are automatically inserted. The range of parts covers five different sizes of threaded bolt and also five different threaded rods. These are provided in bulk and inserted in a four-cavity mold by two separate feed systems. “The inserts are picked up individually so that the process is repeated four times per cycle. This reduces set-up work on the feed system,” says Holger Kirschmann, project engineer at Arburg. The insertion module features quick-change insert holders so that the actual functional elements do not require changing.

Automation involving thermoset resins also brings its particular challenges. Inserts, for example, may be preheated to 170°C.

“Removal modules must be heat-resistant,” notes Arburg’s Kirschmann. “Sensors, wiring harnesses, and hoses are in particular need of protection. Further, the gripper should not be located in the vicinity of the mold for too long, as it could be damaged by radiated heat in the event of a fault or downtime.”

Logistics may also come into play when considering automation, as evidenced in automation of toothbrush production and packaging by an Arburg customer. On a five-component machine, the basic toothbrush bodies have bristles inserted and are encapsulated in four different colors in a single operation, enabling the production of 16 toothbrushes in four different color combinations during each cycle, before being automatically packaged.

“Complete time from manufacture and packaging through to shipping—with four different colors per packaging unit—could be reduced from several days to a few hours,” says Herbert Kraibühler, managing director of technology of Arburg.

Considering manual

In developing economies, the rules most certainly differ on account of low labor costs—monthly wages are still in the $100 range in many countries—and numerous MPW plant visits in Asia have confirmed that manual insert placement and product take-out remain firmly entrenched, particularly for low-end products. At most, robots may be adopted for insert presentation and take-out, but operations such as part trimming and packing are still manual in the vast majority of cases.

In the West, “Generally speaking, we might consider manual insert loading into the tool for volumes less than 40,000 per year,” says ABA-PGT’s Vale. However he cautions that there are various other parameters that are difficult to quantify and can work against the manual option. “On weekends and holidays we would normally shut down a job as it is difficult to find people willing to work. Now, with the Flex Cell we can keep molding parts until the inserts run out.”

Another concern is mold damage due to operator error. “In some cases, the cost of tooling repairs alone could justify the cost of automation,” he says.

Vale also points out, “Outside of cost savings, there are product quality improvements to consider as well. The more repeatable automated process results in better dimensional and material property consistencies compared to a manual process. Also, the automated process reduces the chance of part damage during handling.”

Upstream automation

Processors also have numerous automation options when it comes to drying material and feeding it to the press. “[For large applications,] we are trying to get people away from beside-the-press drying and into central drying,” says Dan O’Donnell, Midwest regional manager-material handling systems, at Wittmann Battenfeld USA. “Dryers dedicated to material, rather than machine, coupled with manifold and conveying systems give you a lot of flexibility as to what material can run in what machine.”

Osamu Tsuda, deputy GM of the international operations division of Matsui Mfg. Co. (Osaka), says that locating dryers in the material room rather than the shop means you don’t have to worry about heat being generated. “A lot of molding in Japan is carried out under controlled environmental conditions,” he notes.

In Japan, high stainless-steel prices have made the pricing of silos prohibitive and forced a rethink of the best way to feed large machines. “These days, processors are more likely to empty flexible containers into a storage tank with capacity of four hours’ worth of material located in a material room perhaps 10 meters away from the press,” says Takayuki Shiba, General Manager Kawata Mfg. Co. (Osaka).

For small-part applications, Wittmann Battenfeld is promoting its portable drying cart with convey equipment in a single unit. The cart can deliver dried material to a small just-in-time loader right on the throat of the injection machine and can be configured to serve multiple presses.