is part of the Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Sitemap


Articles from 1997 In June


Optoelectronics demands the very best

The moldmaker often is the most critical member of a concurrent engineering team. This can be the case even if the project involves production of some kind of simple widget. Imagine if the project involves production of fiberoptic lens arrays out of Amoco's Radel PES; lens arrays with 22 optical vanes that are .0098 inch +/-.0002 thick; spaced .0098 inch apart from one another along a body; each lens with two optical faces at 90¡ connected by a true elliptical periphery whose position has to be held to within +/-.000098 inch. A moldmaker's role in a team concurrently engineering a precision part isn't just critical - it's super critical.

This optical lens array for fiberoptic connectors features 22 optical vanes .0098 inch thick, with a total tolerance buildup of .0002 inch. Nominal part wall stock is about .047 inch. A .020-by-0.47-inch. modified chisel-style fan-edge gate and a .05-by.05-inch trapezoidal runner are used. It is molded in a 334F mold at 694F barrel temperatures in a PES light-transmitting resin formulated to withstand 680F secondaries.

The tooling issues for the production of parallel array fiberoptic lenses were covered in a paper presented at the Molding '97 conference and expo (March 24-26 in New Orleans) by Timothy R. Erwin, technical advisor, engineering, of Pitney Bowes Plastic Components Operations. IMM took you on a guided tour of Pitney Bowes' Plastic Components Operations in Danbury, CT a few years back (See February 1995 IMM, p. 78), and discussed how P-B has used the team concept, both internally and externally, to help it add custom molding work to its captive molding operations. At Molding '97, Pitney Bowes' Erwin let everyone know that a dedicated team of experts can challenge long-standing ideas of just what plastics injection molding can and cannot do. In this case, the team included Pitney Bowes and a moldmaker - Apex Machine Tool Co.

As you might imagine, Apex is a special kind of moldmaker. Privately owned, it makes high-precision ±.0001-inch-tolerance fixtures, gauges, dies, and molds. It was started in 1944 by James Biondi Sr. One of his sons is CEO today, another is president. Apex occupies a 44,000-sq-ft plant in Farmington, CT, 5500 sq ft of which is dedicated to design. ISO certification is expected by summer's end, and two seats of Moldflow will be installed by the end of the year. It employs about 27 designers, three programmers, and 85 machinists.

Precision molds and multiple mold programs for molding machines less than 300 tons are its specialty. The company manufactures approximately 30+ molds annually, accounting for about a third of its $17 million in annual sales. The molds manufactured at Apex are super-high- precision, guaranteeing high volume for the most demanding parts imaginable. It has a long-established tradition of refusing to be the second best in anything it does. "Our best sales force is our customers and our competitors," says Apex's John R. Bruno, vice president of plastics.

Just listen: "I wanted that mold to run the first time. The mold was delivered on time, and the first shots were assembled into functional test units." That's Pitney Bowes' Erwin talking about the lens array mold. These lens arrays are used in a fiberoptic connector developed by IBM's Watson Research Center, 3M's fiberoptics laboratories, and Lexmark's Plastics Technology Center. (Pitney Bowes ran into the job quote on the Internet after Lexmark put it out for bid on the World Wide Web.) Technical product development of parts with such demanding specs is becoming routine at Pitney Bowes. The company saw the job as being ideal for its 50-ton Technoplas Super Injection Molding (SIM) machine.

Unfortunately, lead times for the special kind of tooling this special machine would have been 14 months if the mold had been sourced through Technoplas headquarters in Japan. IBM allowed only 10 months, maximum, to develop the lens array. "You always have the time to do it right the second time, but never enough to do it right the first time," Erwin jokes.

Fortunately, Technoplas had previously worked together with Pitney Bowes and Apex to develop Apex as a U.S. source of SIM tooling components. Technoplas SIM machines operate through closed loop cavity-pressure feedback control, using vacuum-tight molds of exacting dimensions. Mold plates must be flat to within .0001 inch, and the mold bases have to open and close with .0002-inch slide fits to be vacuum-tight. Erwin says such molds require "gaugemaker workmanship."

"When Bruno saw the drawing . . . to say he was excited would be an understatement," Erwin recalls. "Apex leapt at the chance to work on this project."

"We have a great engineering group that can look at things objectively," adds Bruno. "Ho-hum programs aren't our forté. But when we see something like this, associated with companies like Lexmark and IBM, we want to be involved; this was leading-edge technology."

"I was surprised at how easy it was to mold that part. It was getting it out of the mold that was the tough part," Bruno explains. The part filled with no difficulty on the first trial. However, it stuck in the mold due to the lack of sufficient ejection. The product, as designed, allowed for minimal ejection area creating a less-than-desirable condition.

After a brainstorming session among Pitney Bowes, IBM, and Apex, and a part concession allowing for more ejection surfaces, Apex engineers went to work. They redesigned the optical blade inserts and the ejector area. It was decided that precision gauge blocks could be used as the base material. But they'd have to be capable of holding sizes in three dimensions to within ±.00005 inch. After drilling, rough machining, CNC jig grinding and lapping, the best polisher in the shop, working hours on each insert, polished the elliptical surface, removing the final .0001 inch (±.000098 inch). The result? Yield was increased from a few usable parts up to 1000 high-quality lenses, just in time.

"This was a true developmental program, real concurrent engineering, with all of the parties involved understanding it to be a collective enterprise," Pitney Bowes' Erwin summarizes. "Everyone was focused on success." Such teams will continue to push the edge of the technological envelope, and in so doing, propel injection molding into challenging 21st century markets, like the emerging field of optoelectronics. And moldmakers like Apex will play a critical role in teams creating such new niche markets, those that demand the very best from all the super-critical team members.

The Troubleshooter, Part 15: Flat, full parts

This article continues ourseries of troubleshooting reports from one of the leading on-the-spot problem solvers in the molding industry. Bob Hatch is manager of technical service and customer support for Prime Alliance, the Des Moines-based resin distributor. Before his present assignment, Bob managed a molding operation for 25 years.

A customer sent me a box of parts and runners and asked for help in keeping his parts flat and eliminating short shots. The parts were round, a little larger in diameter than a silver dollar, with raised areas on the outside edge in a couple of places. They were edge gated and ran in an eight-cavity balanced runner mold with a cold sprue. The material was 33 percent glass-filled nylon 6/6. The customer also asked for help in running his parts fully automatic.

The parts each had a notch cored out on one side where it could be gated (Figure 1). The current gate was on one side of the notch, and there was a sharp corner on the other side that was trapping air because it was the last place to fill. Short shots were fairly common because the mold wasn't vented very well.

Figure 1. The customer's parts were edge gated with a notch cored out on one side that could be gated into. The problem was a restricted flow path, poor venting, and the wrong kind of gate.

The problem as I saw it was a restricted flow path, a mold not vented sufficiently, and the wrong kind of gate for running the mold automatic. I started to correct his problems by sizing the runner system. Since I figured we would end up with either a chisel gate or a curved tunnel gate, I recommended a full round runner be used because it works best with all kinds of gates.

The Diagnosis

The diameter of the runner that feeds the gate needs to be sized for the material. In this case, a .200-inch subrunner was about right. The main runner needs to be larger in diameter, so a .250-inch main runner took care of that.

The small end of the sprue, or the sprue O diameter, needs to be bigger than the main runner diameter, so a .312-inch O was a good fit. That allowed the nozzle orifice to be drilled out to .290 inch, or we could have used a replaceable nozzle tip of that approximate size. I generally don't use a reverse taper nozzle with glass-filled nylons, so I didn't have to worry about the inside orifice diameter or draft angle of the reverse taper nozzle.

The bigger sprue diameter won't necessarily slow the cycle down if the barrel heats are kept on the low side of the material manufacturer's recommendations. With nylons, I use the correctly sized sprue orifice not only for keeping pressure on the melt in the runner system, but also to keep the nozzle from freezing off. After sizing the runner, I usually move to the gates, but since we had to change the entire gate design I decided to save the gates for later.

Next in line was venting. On round parts, I like to use continuous or perimeter venting. In this particular case, a vent depth of .001 inch was about right for the glass-filled material. That is, of course, if I kept the barrel heats on the low side of the material manufacturer's recommendation, which is now achievable, since we have removed all the flow restrictions from the nozzle, sprue, and runners.

With the vent depth at .001 inch and the entire parting line vented, all that was left to do was to make it self-cleaning. We did that by going out .040 inch from the parting line, dropping into a .040-inch-deep channel to a racetrack, then venting the racetrack to atmosphere. Polishing the entire vent lip in the direction of airflow to an A1 finish made it self-cleaning.

In addition to the parting line vents, we needed to be sure the runner was vented. We vented down the sprue puller and at the end of each runner section. The only difference between part vents and runner vents is the depth of the vent. Runner vents are deeper than part vents because I like to feel a little bit of flash on the end of each runner to prove I am getting rid of the air.

We made the runner vents .003 inch deep for this material. The width of the runner vents is equal to the diameter of the runner being vented. We vented the runner by going out from the parting line of each individual vent with a land of .040 inch to a deeper channel to atmosphere just like with the part vents. Again, we polished the vent lips to make them self cleaning.

With proper venting, the injection speed could now be increased to a fast fill to achieve good fill and pack conditions, all without having to raise the barrel and nozzle heats.

The Gate

With everything else done, it was time to figure out what kind of gate we needed to run the parts automatic and keep them flat.

Since the part didn't have enough of a side wall to put a subgate in, we decided to stay with an edge gate. We could have used a rectangular subgate if we had had a little bit of side wall to work with, but it didn't work out that way. We decided the chisel gate would be the ticket for keeping the parts flat and also provide for automatic ejection. All we would have to do is put a sprue picker on the machine. This would also reduce the operator involvement by at least 50 percent, maybe even 75 percent, which helped in quoting.

Ejection

Because we were using stiff materials, we decided to use an ejection delay system with the chisel gate (Figure 2). Basically, with a chisel gate you have an ejector pin on both sides of the gate, one close to the gate on the runner side and one close to the gate on the part itself.

Figure 2. Automatic degating of an edge gate for rigid materials. Here an ejection delay system was used.

The ejector pin on the runner side is cut off about 1/2 inch short, so plastic will fill in the hole and make it into an undercut to hold the runner in place during ejection. When the ejector plate starts to move, the ejector pin on the part side pushes the part away from the gate and shears it flush with the edge of the part. The gate breaks cleanly at the part due to the notch sensitivity of the sharp corners and edges of the chisel gate design. The ejector pin on the runner side of the gate will begin to move after the ejector delay space has been taken up. At this point, the runner is ejected out of the undercut and lifted up and out of the way by the sprue picker.

The keys to the success of a chisel gate are the placement of the ejector pins and the delay space below the head of the runner-side ejector pin.

Of course the sharpness of the chisel gate design utilizes the notch sensitivity of the material being used to ensure that the gate will break cleanly away from the edge of the part. For the depth of the gate, I figured 90 percent of the wall thickness, which gave us .090 inch for the wall of .100 inch. The width of the gate for a part this size is twice the depth, so it would be .180 inch wide.

With the runner, sprue, and nozzle sized correctly and the chisel gates in place, we put the mold back in the press to see how it was going to run. With the flow path opened up we could lower the barrel heats to 540F from the previous 580F setting.

We were able to increase the injection speed now that the mold was properly vented. We raised the mold temp from 150F to 180F to help bury the fiberglass and get rid of as much molded-in stress as we could.

The results were great. The parts were coming out flat, the cycle time went down from 32 to 24 seconds, and the operator,replaced by a sprue picker, was given a different job.

For quoting purposes, the molder still costs the job out with 25 percent

of an operator since someone has to pick the boxes of parts up from the conveyor belt, but the machine is really running in fully automatic mode.

TROUBLESHOOTER'S NOTEBOOK
Part: Round, the diameter of a silver dollar, with a raised area on the outside and a notch.
Material: 33 percent glass-filled nylon 6/6.
Tool: Eight-cavity balanced runner mold, cold sprue.
Symptoms: Parts not running flat, short shots occurring, needed help running fully automatic.
Problem: Restricted flow; poor venting; wrong kind of gate for running automatic.
Solution: Subrunner resized to .200 inch, main runner resized to .250 inch; changed to chisel gate; increased size of sprue O diameter to .312 inch; drilled out nozzle orifice to .290 inch; made vent depth .001 inch; vented entire parting line and made it self-cleaning.
Result: Parts started coming out flat; cycle time went down from 32 to 24 seconds; labor decreased.

Analyzing plastics with FEA: Part 6

Sinks and voids can be the nemesis of an otherwise well-designed part. But how does a designer know beforehand that the part he or she is creating on a CAD screen will cause major headaches at the molding press?

According to Shrikant Oak, an analyst at injection molding process simulation specialist Feamold Inc. (Troy, MI), the creation of sinks and voids during the molding process is a 3-D phenomenon that can be detected using the finite-element method. However, few analytical studies are published in this area. Why is it so difficult to simulate voids? All analysis packages assume that flow in the thickness direction is minimal, so they don't account for it. If they did, compute times would be excessive. Unfortunately, the pressure causing sinks and voids is in the thickness direction. So to minimize sink and void creation, analysts at Feamold track temperature across various points of the part, then find out how to keep the flow path and gate unfrozen for the longest time.

IMM asked Oak to share a recent project for Siemens Electric Corp. that brings this type of analysis to light. He explains how the project began: "Siemens' Ben Reginella came to Feamold with a slight problem. Prototype tools for a resonator, molded from glass-filled polypropylene, were turning out parts with voids at a thick T-section supporting a locator tab," he recalls. "Typically, analysts have been unable to capture sink and void problems using CAE, but with the help of C-Mold filling analysis software and our own in-house expertise, we were eager to try."

According to Chrysler's John Van Hout, project engineer for Large Car Platform, the resonator is a portion of the air induction system that helps attenuate noise generated by the engine. This part experiences short-term temperatures up to 250F under the hood. "The initial concern," says Van Hout, "involved the structural integrity of the locator tab. We performed an FEA to verify that the design would meet our requirements, then asked our supplier Siemens to implement that design."

During the prototype stage, however, Siemens found voids at the critical section that weakened the tab itself. That's when Feamold received the project. "Our objective was to evaluate options to minimize the potential for voids at the base of the thick locator tab, and to determine the optimum gating design for minimum part warpage," adds another analyst, Ross Nordin.

Main criteria used to compare part warpage characteristics included bowing in of the part rim and lifting or sagging of the part away from and toward the mating half. Analysis was performed using two gates, each 60 mm wide, for a total width of 120 mm along one edge of the part.

Feamold analyzed several different designs, varying the gate location along the long edge of the part and along the locator tab flange. However, the design already in production at Siemens ultimately prevailed to avoid the cost of recreating prototype tools.

Siemens provided part geometry to Feamold in the form of an IGES file. A wireframe model of the part was created, and topological attributes of thickness and shape factor were assigned to it. The design incorporates two tunnel gates at two ribs along one side of the part (Figure 1). Pressure required to fill the part was approximately 9000 psi.

Process Specifics

In the final, optimized design, nominal wall thickness of 2.5 mm was increased to 3.5 mm close to the gates for better packing of the locator tab base, ultimately reducing the potential for voids. In this case, Oak says, the uneven wall stock did not have an adverse effect on part warpage.

Feamold used the following conditions for the analysis, which mimicked conditions during mold trials:

  • Material: Thermofil P6-20FG-2153 (20 percent glass-filled PP).
  • Pressure: The pressure distribution followed the path of melt front advancement, and decayed linearly with flow length. The pressure along the two long edges was identical, thereby reducing the potential for the longer edges to warp unevenly.
  • Melt front advancement: Fill pattern was balanced. Melt front contours advanced uniformly parallel to the short edge.
  • Temperature: Melt temperature was 440F, with a fill time of 3 seconds. The temperature distribution in the part ranged from 392 to 445F at the end of fill. Melt temperature in the thicker areas close to the gate remained high (due to the increase in wall stock).
  • Volumetric shrinkage: Shrinkage distribution in the part ranged from 2.5 to 8.5 percent at the time of ejection. Potential for part warpage increases with increasing shrinkage gradients in the part. Therefore, ideally, the shrinkage distribution in the part should be uniform, and as low as possible.

The shrinkage level (magnitude) observed in the analysis was higher than ideal. However, the shrinkage gradient was low, indicating that the part would shrink uniformly. Uniform shrinkage results in uniform deformation, that is, less warpage. Shrinkage can be reduced further by increasing packing pressure. The packing pressure used for the analysis was 4000 psi, resulting in a maximum shear stress in the part of approximately 28 psi, well below the acceptable limit of the material. Oak adds, "For this particular part, packing pressure can be increased up to 8000 psi, without seriously over-stressing the material."

Evaluating the Design

Two criteria were used to determine the effects of gate design and location. First, analysts checked these factors for their potential to create voids at the locator tab base. Second, they were analyzed for their contribution to part warpage.

The base of the locator tab (where it joins the part wall) has very thick sections for structural integrity. The part wall leading to this thick section originally had a nominal wall of 2.5 mm. The thinner part wall would freeze early, thereby cutting off the (packing) pressure to the thick section, while the core of the thick section remained molten. This would allow the thick section to solidify under low or zero pressure, resulting in voids in the section.

Analysis showed that the void problem could be eliminated by allowing longer packing time and/ or increasing packing pressure for the thick sections. Moving the gate close to the thick section increases the packing pressure at the thick section.

Increasing the thickness of the part wall between the gate and the thick section also allows the part wall to remain fluid longer, and therefore allows the thick section to be packed longer. Cooling time is proportional to the square of thickness. Therefore a small increase in the part wall thickness allows it to remain fluid for a significantly longer time.

Oak explains that the potential for sink and voids cannot be explicitly identified by the analysis software. Instead, the temperature traces were examined at various points along the flow path leading to the thick sections, as well as sampling points in the thick section itself. The temperature traces at these sampling points are color coordinated; the trace color and the marker color are the same. In the first iteration, a single gate was located along the center line of the short edge of the part.

Void Creation

To understand how voids are created in thick sections, Oak offers the following explanation.

During the injection phase, plastic temperatures are lowest at the mold walls and hottest at the center. More of the cross section begins to freeze and shrink as pack and hold pressure is applied. This pressure should compensate for shrinkage by pushing more material into the areas that are contracting. But if pressure is too low or is not applied long enough, sinks and voids are created. Packing pressure has its limits, however, and cannot go beyond a safe range for the material. Packing time can be varied more easily.

Melt velocity also has a significant effect on temperature. The velocity determines how much convective and shear heating will be experienced. Also, if velocity decreases, the residence time may increase, thereby allowing greater cooling due to conduction. When melt velocity drops to zero (at freezeoff), there is no more positive contribution to the heat load; instead, the negative contribution is increased, resulting in a drastic change in the thermal balance and a sharp drop in the temperature trace at that location.

Both pack and hold pressures and residence time contributed to voids in the resonator. As the path feeding the thick section froze off early, the thick sections were allowed to cool under no or low pressure. Cooling occurs from outside-in, that is, the outer skin freezes first and the core is the last to cool. As the outer layers cooled, they shrank, pulling the fluid core towards them, resulting in voids.

Packing Out Voids

Several additional iterations were performed to analyze design options to pack out the base of the locator tab. Optimum results were achieved when a single gate was located along the horizontal flange of the locator tab. However, this gate location was not practical, since it could cause leakage or maintenance problems in that region. Instead, a long edge gate along the short edge of the part was chosen, which reduced the distance from the gate to the locator tab to transmit greater packing pressure.

In this case, Oak recalls, the reduction in flow length alone was not sufficient, because it did not eliminate the problem of early freezeoff of the flow path leading to the tab. To allow the flow path to remain open, the part section in that area was thickened iteratively. Optimum results were achieved when the green area was thickened to 3.5 mm.

A single long edge gate required a thick secondary (feeder) runner tapering down into a thinner film gate, along one edge of the part (Figure 1, p. 37). Based on feedback received from Siemens' Reginella, analysts felt that the presence of the thick secondary runner would increase the heat load in that area, and that this was undesirable. To eliminate the need for a secondary runner, a split runner scheme was utilized (Figure 1, p. 37).

Temperature traces taken at the sampling locations, using the final

recommended design, showed that the flow path leading to the base of the locator tab remained fluid much longer, thereby allowing better packing.

Taming Warpage

Feamold compared warpage characteristics for three resonator designs. To do this, a reference (datum) plane was first defined:
  • The first point is fixed in all three directions.
  • The second point defines the positive (local) X axis.
  • The third point defines the local first quadrant, and can move only in the X-Y plane defined by the three points.

Assume that the blue and green outlines represent the original and deformed part shapes, respectively. To measure part warpage, the deformed part could be placed on a flat surface, and the distance from the flat surface of any point on the part can be measured as displacement or deformation.

If the deformed part is kept on the table, so that any one of the corners of the deformed part coincides with the original location of this corner, then this corner acts as the local origin for reference, and the flat surface defines the reference plane.

The anchor plane is defined by three nodes (Figure 2, p. 38):

  • The local origin for reference.
  • Two nodes, along with the original, that define the reference plane (flat surface).
Figure 2. (Left) Definition of anchor plane and (below) sampling points for part warpage comparison.

To measure deformation characteristics, sampling points were chosen by the analysts along the rim of the part, and the nodal displacement values at these locations were recorded.

The first criterion was to measure the tendency of the part (top) edges to bow in. All designs had comparable warpage characteristics. There was no significant difference between the three designs, indicating that the proposed design had ample tolerance to allow minor changes if required during tooling.

Secondly, Feamold measured the displacement in the Z direction, the tendency of the part to lift away from (positive) or sag toward (negative) the mating half. Again, all three designs showed similar results, as shown in Figure 3.

In terms of absolute deflection, the bowing in of the long edge was the greatest deflection observed in the part. Therefore, analysts concluded any improvement in this characteristic would result in the greatest improvement in part quality. - Michelle Maniscalco


Statistical Know-How

When it comes to interpreting analysis results for temperature and stress, Oak explains, single point minimum or maximum values can be misleading. For example, consider a case where 99 percent of the part is at a uniform temperature, say 500F, and only one tiny point may be at 200F. Typically, such local discrepancies may be ignored, since they do not affect the process as a whole. Rather than basing your results on these single points, the following data are also useful:
95th percentile value.
This is the value associated with 95 percent (by volume) of the part.
5th percentile value.
This is the value associated with 5 percent (by volume) of the part.
Average.
The numerical average.
RMS deviation.
Root Mean Square deviation from the above numerical average.

The characteristic of a good design is that the 95th percentile value should be within the acceptable processing/design range, and the RMS deviation should be as low as possible, indicating low variation or gradients in the part.

U.S. moldmakers need to get on the stick

Moldmakers in the U.S. can no longer rest on their laurels and expect OEMs to ignore the favorable lead times from off-shore mold shops. That's the message three OEMs gave molders and moldmakers meeting in Coronado, CA for the annual SPI Western Section Conference.

Oscar Miramontes, engineering manager in El Paso, TX for the Philips Components division of Philips Consumer Electronics, told attendees, "We need to reduce mold design and fabrication time from 16 to 20 weeks to 8 to 12 weeks. For that, moldmakers need to introduce soft mold technology to not only reduce lead times but to reduce mold costs and provide flexibility for engineering changes."

He also notes that moldmakers need to improve their CAD/CAM systems by upgrading software in order to maximize concurrent engineering technology. "We still have a long way to go with respect to concurrent engineering and we need more compatibility," he says. "We need to be able to speak the same language."

Miramontes says companies such as Philips are caught in the middle - on one hand being driven by the consumer and on the other, being pulled by suppliers. "The electronics industry is trying to accelerate others," he says. "We demand a lot from plastics and from our toolmakers."

Faster lead times on molds is a fact of life, yet Miramontes says he's experienced various levels of resistance from U.S. moldmakers. "Asian moldmakers responded," he notes, "and U.S. moldmakers had to catch up because they're behind in lead times."

Because of shorter product life cycles, many OEMs have asked moldmakers to look at soft tooling and to rethink how OEMs can achieve quantity requirements. Spencer Barnes, plastics engineering manager for Hewlett-Packard's Vancouver, WA operations, says that, for example, maybe it's better to build 10 P-20 tools to mold 100,000 parts each than one large, hardened steel mold to mold a million parts. Barnes pointed out that the printer market window is two to three months.

"It is indeed the market that drives us and we must provide refreshed products," says Barnes. "That means steeper ramp times, lots of P-20 tools, some single cavity to start, then behind those come the multicavity, hardened molds. Moldmakers must react to this need."

Tony Matlock, plant manager for Rain Bird's Camsco Manufacturing in Azusa, CA, says his company comes out with a new product every six months within its various business units. An average of 10 to 60 molds is required for a typical program, and redesign during mold build is a critical area for the company.

To minimize tool rework and engineering changes, moldmakers need to find ways to simulate the function of the parts because of the research and development that takes place during the product's build, says Matlock.

One moldmaker in attendance made the statement that OEMs will eventually return to have all their mold requirements built in the United States because U.S. moldmakers still produce the best molds in the world, expressing the belief among many moldmakers that offshore tooling is of poor quality.

In disagreeing with that, H-P's Barnes re-emphasizes something he told attendees of another plastics industry meeting recently. "Hell no, we won't be back," he says emphatically, "because we aren't getting junk from offshore sources."

Matlock says U.S. moldmakers need to rethink the way they do business in today's global environment. "The tooling industry needs to be operated from more of a business philosophy than from the standpoint of a craft," he says.

Magnet sticks with the mold - really

Some people have a hard time trusting it at first. "They just couldn't believe that a magnet could hold the mold up against the platen," says Craig Powell. He's the molding manager of Pyrotherm, a division of Nypro Clinton in Clinton, MA. He's talking about the 18 pairs of magnets used to hold the mold in the press - sans bolts, clamps, or any other support.

Powell is the virtual leader of Pyrotherm's Nova project, a molding operation that uses 18 custom designed 30-ton Ferromatik Milacron electric injection machines. They're spread around the country in Clinton; Burlington, NC; Atlanta; San Diego; and in Puerto Rico, molding proprietary medical parts.

Nypro's Nova project uses 18 custom designed 30-ton electric Ferromatik Milacron presses, each of which is equipped with a pair of Pressmag magnets that hold the mold against the platen without clamps.

What makes them custom is the fact that each press's centerline is 53 inches off the ground, leaving room underneath for dryers, loaders, temperature controllers, and other secondary equipment. The machines are designed to save space (25-sq-ft footprint) and mold parts in short runs. "The design of the project was short-run injection molding," says Powell.

That's where the magnets come in. Powell needed quick mold changeouts to accommodate runs that lasted as little as three hours. With the traditional mold clamping system, he says, mold changes took at least 30 minutes, if not longer. He opted instead for the Pressmag, the magnetic clamping system produced by O.S. Walker Co., based in Worcester, MA. Nypro bought 18 pairs of magnets, one for each electric machine. "Now," says Powell, "changeouts take about 10 minutes."

No Clamps - No Fear

The Pressmag consists of two magnetic chucks, made of a permanent magnet metal alloy called Alnico. One is mounted on the moving platen, the other on the fixed platen. The Pressmag requires no backplate adaptations or other modifications to the machine or the mold, and is compatible with most machines, says John Daly, product manager at O.S. Walker. And without clamps or bolts to secure the mold, the magnet gives molders more surface area with which to work.

For molders who struggle with power outages, have no fear. The Pressmag does not require a continuous supply of electrical current to maintain its magnetic attraction. The electropermanent magnet requires an electrical current generated by a controller to magnetize and demagnetize only. "You don't have to worry about the electricity going out on you," Powell says. Daly reports he's not received a complaint yet from a molder who had a mold slip or move in any unwanted way on the magnet.

Still, it's no wonder that some molders are slow to give all of their faith to a 90-lb magnet holding a 400-lb mold. At Pyrotherm, Powell says that was the biggest hurdle his molders had to clear. "It was a real culture change on the manufacturing floor to suspend the mold with no visible support," he says. At first his operators mounted the mold with the magnet and clamped it to the platen. Then, they removed each bolt one by one until the mold hung by itself on the magnet and, says Powell, "they could see for themselves that it wasn't going to crash to the floor."

Powell says that once the molds are installed and aligned properly on the magnets, he's not had a single problem with slippage. He says the heaviest mold he's put in a 30-ton press is 450 lb - 50 lb over the limit for the tiebars, but not a problem for the magnet. He says the controllers also have a turnkey safety switch to prevent users from accidentally demagnetizing the magnet before the time is right.

For molders contemplating a magnetic mounting system, Powell counts three drawbacks, each of which can be overcome. The first is cost. Nypro spent about $5000 per machine for the magnets. He says that while this may seem like a lot, he counts his savings in fast and efficient mold changes, which are vital to the schedules he runs. The second drawback is the fear factor, which is just a matter of seeing and believing. "A lot of people are just afraid," Powell says.

Third, if you mount the magnets in an existing machine, press daylight will be lost to the thickness of the magnets - about 4 inches in Nypro's case. "When you add magnets to the platen, you're adding thickness to the platen," he notes. Powell says this may be of concern to custom molders who rely heavily on the flexibility that maximum daylight gives them.

Daly says that if you're worried that much about losing daylight, instead of mounting a magnet on the platen, O.S. Walker can make a new platen - one that has Pressmag built right into it. That way you don't lose space between the platens. This is especially useful for molders buying new machines. Daly says he's worked with several injection molding OEMs to offer magnetized platens for new machines at the customer's request.

One final note: Remember that magnets don't work with aluminum. For that you'll have to revert to clamps.

Molding a medical niche in Southern California

On paper, American Technical Molding (ATM) might look like a typical custom molder. It operates out of a modest 50,000-sq-ft shop in Upland, CA, some 35 miles and several traffic jams east of Los Angeles. The company employs about 60 people, runs 23 presses, and molds medical disposables and device components. And with $7 million in annual sales, ATM isn't on the verge of bursting into the Fortune 500.

But in person, the story is different. ATM's attention to detail and its small size help it remain nimble and efficient in a market that is increasingly challenging, competitive, and cost-conscious. ATM is a custom molder, but does not buy into the low-bidder mentality, preferring instead to produce quality parts - for a price. "We're generally not the cheapest ones on the block," says president Rocky Morrison, "but we are the most cost-effective."

ATM is also unusual in that right next door is its sister company and primary moldmaker, JK Molds, a 26-year veteran of Southern California and a direct descendent of Caco Pacific via owner Jack Kelley. The two companies have grown up together and have developed a close working relationship that is not common in the industry.

The "T" in ATM

As the company's name implies, the molding at ATM is technical. With a Class 100,000 molding room, ATM's operations are spic and span and relatively complicated. Typical ATM molds are loaded with undercuts, multiple pulling cores, slides, and threads of every ilk. This ability to run complicated parts well has earned ATM a reputation in the industry as not only a quality production shop, but a lifesaver as well.

Morrison says that in its infancy during the late 1980s, ATM survived predominantly as the molder riding in on the white horse to save the dying mold. "A lot of customers think of us as the company of last resort," he says. Morrison is helped by his unique perspective. He and manager for technical sales and marketing, Walter Gacek, used to work for medical OEM McGaw. Says Morrison, "We understand the OEM's side of the fence."

Early on, some customers brought to ATM molding projects crippled by problems with the mold, material, or design. Morrison says mold and material problems are relatively simple to fix. But that's not where the bulk of the problems rest. "Very often, the problems we solve are violations of Design Basics 101," he says. "The most difficult problems to solve are those that involve product design."

Still, ATM dug many of its customers out of very deep holes, in the process winning the production responsibility for some of these recovered projects. This helped ATM develop into what it is today: half problem solver and half production shop. And more and more, ATM is called upon by OEMs to help prove out parts in research and development. "A lot of customers don't even want people to know they're working with us," says Morrison.

ATM's devotion to quality first is contrary to some custom molders' need to be low price. Some potential customers, says Morrison, don't recognize that molding is a science, not practiced equally by all, and shy away from ATM's higher prices. But for customers who are willing to pay for quality, the ends justify the means. "The cheapest is not always the best," he says. "We like to think we do things the right way. Our motto is that we do what it takes to make the customer happy."

On the Floor

ATM's propensity for technical and complicated molding, says Morrison, led the company to Sandretto. There are 23 Sandrettos in the cleanroom, ranging in size from 35 to 300 tons. Morrison says that Sandretto's ability to perform programmable, sequenced core pulls makes the press ideal for his shop. And they're quiet. Says Morrison, "Listen to our shop. You don't hear molds banging."

Material includes all of the usual suspects for medical products:

PC, PVC, acrylic, PEI, polysulfone, PE, and PP. Volumes range from a high end of 40 million to 50 million parts per year, to a low range of 5000 to 10,000 parts per year. Morrison says ATM practices decoupled molding, transferring on pressure read from the injection unit.

Over the press, automation is the norm, with robots extracting parts into sorters nested in portable Hepa-filtered air systems to blow out potential contaminants and maintain purity. Sprues and runners are conveyed through the wall for granulation in bays outside the cleanroom.

One family mold running during IMM's visit produces parts for a drug delivery device. The OEM for the part requests a bio burden of three bacteria colony forming units (CFU) or less per part, a strict standard ATM has not yet failed. This is accomplished with a Hepa-filtered air system and a separator that sorts parts by cavity, directly into plastic bags for shipping.

Across the aisle, a 165-ton Sandretto is pressing polycarbonate bodies for a needle-free syringe in a 16-cavity mold. Beside the press is an elaborate maze of chutes and tubes, into which a robot is dropping finished parts. Gacek says the contraption is an ATM brainchild and creation, devised to separate parts by cavity. The need for this device was prompted by the fact that the tip of the syringe body is only .003 inch in diameter. The core pins in the mold are unusually small and susceptible to damage. The flawed products a damaged pin produces are difficult to detect by simple visual inspection and may not be discovered for several minutes or hours. By segregating parts by cavity, when a core is damaged, ATM can pull just the affected parts, reducing scrap. "If we know that the core pin in cavity eight is damaged, then we dispose of only the parts produced in cavity eight," says Gacek.

The proof of ATM's abilities is generated in the quality control room adjacent to the molding floor. Gacek says the lab authorizes all parts before production, and uses part weight throughout production as an indicator of the stability of the process. "Sink on a part shows up on a weight SPC curve before it's visible to the eye," says Gacek.

In this room ATM also spent $100,000 over the last year on a Smart Scope vision measurement and inspection system from Optical Gaging Products. Gacek says the investment was considerable, but the payoffs enormous. "A manual inspection of a particular drug part we make used to take 5 hours," he says. "This scope does it in 5 minutes."

Although ATM sends SPC data and charts with each shipment, the molder's high quality and good reputation allow many customers to take product without inspection, knowing it meets specification. "ATM is dock-to-stock for several customers," says Gacek, "without incoming inspection." As a result, ATM's scrap rate is less than .25 percent.

The Moldmaker Next Door

Despite recent trends toward rapid tooling, Morrison says ATM prefers to proceed cautiously with each molding project. His reasoning is two-fold: first, the human body hasn't changed much in the last few thousand years - medical molds subsequently tend to last longer and should be made to endure; second, rapid tooling is not magic, it's just fast. "When it comes to tooling, people think magic is involved," says Morrison. "There are still just 24 hours in a day. It just means toolmakers are working overnight and on weekends." So ATM primarily uses paper drawings for design and tooling, and takes its time to produce a quality mold the first time.

More often than not, when it comes to moldmaking for ATM, JK Molds does the job. Although ATM and JK are separately managed, they are jointly owned and enjoy a close working relationship. As a result, ATM qualifies and proves out almost every mold JK makes. And occasionally, says Gacek, a qualification turns into full-blown production job for ATM. And ATM gets first dibs on any customer who comes to JK with a molding job. The relationship between the two companies, says JK president Terry Colbert, allows each to act independently when necessary, but still provide the one-stop service that is vital to many customers. "We get the best of both worlds," he says. "It gives us a real competitive edge."

ATM does have its own three-man moldmaking and maintenance shop in house - building one to two molds per month - but nothing compares to having a moldmaker right next door. JK, a Class A101 shop, employs 22 toolmakers, with an average of 20 to 30 years experience. The shop, which produces about 98 percent medical tooling, is a nice mix of new technology and traditional craftsmanship that's helped JK grow to a high-quality shop that's developed its own hot runner system and become a specialist in unwinding tools. "We don't follow fads and we don't buy anything brand new," says Colbert.

One thing new - but not brand new - is a centerpiece of the shop. It's a Roku-Roku graphite electrode machining center, one of the fastest in the industry and alleged to be the only one in operation west of the Mississippi River. With a maximum contour cutting speed of 315 inches/minute, the Roku-Roku is expected to keep JK's CNC center hopping.

The big eye-catcher at JK is its Shrinkmold system, a proprietary hot runner design that allows high-cavitation molds to run in smaller machines. The design reduces the projected area behind the cavities; this allows cavities to fit in a smaller space, reducing the overall size of the mold. It's also designed to hold higher pressures better, eliminating blowouts. JK also reports that the design allows fast color changes - in 10 to 20 shots. Of particular interest is a 16-cavity mold for a proprietary commercial product that uses the JK hot runner system. It ultimately will be half of a stack mold running in a 300-ton press.

JK also reports it made a 128-cavity mold with its hot runner system and that it, too, ran in a 300-ton machine. Although such a system appears to have good commercial potential, JK is playing its hot runner cards close to the chest. "It needs a track record before we get too far ahead with it in the field," says Colbert.

Back at ATM, the 16-cavity mold is being qualified for JK on a 300-ton Sandretto, with hopes of winning full production responsibilities. With all of its qualifying work, Morrison says ATM will add two machines later this year strictly for mold testing. Beyond that, he says, growth is controlled and steady. Morrison says ATM prefers its low profile and agility as a small, well-run, clean shop where everyone knows everyone else. "For custom molding, bigger is not always better," he says.


American Technical Molding
Upland, CA
Square footage: 50,000
Markets served: Medical
Annual sales volume: $7 million
Annual resin usage: 1.7 million lb
Major customers: Divisions of Abbott, Baxter, McGaw, Johnson & Johnson
Materials processed: PC, PVC, acrylic, PEI, polysulfones, PE, PP
Annual parts production: 250 million
No. of employees: 60
Shifts worked: Three shifts, five to seven days a week
Molding machines: 23 Sandrettos, 35 to 300 tons
Secondary operations: Ultrasonic welding, pad printing, UV-cure adhesives, assembling, annealing, hot stamping
Internal moldmaking: Yes
Quality: ISO 9002, Class 100,000 molding cleanroom, Class 10,000 assembly cleanroom

CIM promotes more than productivity

Century first ordered three Smart Box 1000 systems for evaluation. Immediately, Tom Karpovage spotted cycle time variations on one insert molding job. After using the data from the Smart Box and tweaking other inefficiencies, variation in total cycle time was reduced 26 percent in one day alone.

There's a coffee dispenser in the cafeteria at Century Mold Co. Inc. in Rochester, NY. Like others elsewhere, it serves coffee with lighteners and sweeteners. But it also serves cappuccino, espresso, and latte. Such a dispenser is a dead giveaway to management's commitment to quality of life at the plant. It also shows Century Mold's willingness to invest in trying something new to improve quality. Something new like CIM.

An engineering-quality-purchasing team from Century Mold visited a neighbor in its industrial park - Kodak. The team was interested in Kodak's well-publicized successes with its $1 million CIM system from Hunkar Laboratories. The visit seemed to be a much more cost-effective approach than spending a lot of money in an exhaustive vendor selection search. The company was looking for an upgradable system that could provide process monitoring, real-time process control, and SPC data gathering capabilities tailored to its needs. At the ISO/QS 9000-certified, $35 million, mostly automotive molder and moldmaker, Jim Gaffney, Century Mold's director of engineering and manufacturing, felt it was time to take a closer look at CIM. He and his team liked what they saw at Kodak.

But it's what a plant does with its data that makes it successful. The hope was that the new CIM system was going to help the company reduce scrap and cycle times, while improving outgoing part quality and press utilization at its growing design and manufacturing operations in Rochester and in Shelbyville, TN. But someone would have to be found who could implement CIM, someone who knew how to use all that data the system provides. Fortunately, Century Mold found the right person for the job in Thomas Karpovage.

Karpovage was 19 when he started with General Motors. He worked for GM for nine years in supervisory and quality engineering positions. He says he was attracted to Century Mold because of its growth and growth potential, and because of its commitment to proactive quality management. He started at Century Mold last summer, with no prior experience in injection molding of plastics. The company put CIM implementation into the hands of a young quality guy inexperienced in molding. Sound crazy? Read on.

Karpovage developed a business plan, then spent about two solid months familiarizing himself with manuals, reading books on the molding process, and learning the ins and outs of setup by working with folks on the floor. The initial purchase order was for three top-of-the-line Hunkar Smart Box 1000 systems for evaluation (about $15,000 each, including sensors, wiring, and installation). Hunkar's customer service engineers came in to conduct two days of training. The project started its trial run last December.

Karpovage went to work. "There's a lot of information in there. It was overwhelming at first," he recalls. "Once I started using it, it took me about two weeks to get comfortable with it, and I made a lot of mistakes. I must have worn out Hunkar's 24-hour hotline all by myself." He drew on his personal experience and on what he'd learned while at GM, using DOE-like disciplines to benchmark upper and lower control limits for a few problem processes. Immediately, he spotted cycle time variations on one insert molding job involving a four-cavity tool. It's an engine chain guide in glass-filled nylon. Curiously it ran on a relatively new Toshiba machine, a model ISG 180, with Toshiba's V-10 controller. The cycle time should have been 45 seconds, but it ranged, many times at 67 seconds, and sometimes up to 80.

He found that something as simple as standardizing the work method - naturalizing the unnatural bodily motions involved in manually loading inserts and unloading insert molded parts - removed inefficiencies while improving occupational safety. "I baby-sat that job.

I policed it. I personally trained new people on how to use the new standardized work method. I taught them how to pull up the cycle time graph on the Smart Box, so they could see for themselves what they were doing. Guess what? They got excited about improving their performance and began to exceed our expectations."

In one day alone, variation in total cycle time was reduced 26 percent. Over time, variation has been reduced 38 percent. It now takes only 15 hours to complete the run, rather than 24 hours. Cycles now run as low as 41 seconds. So morale has improved, as well as production efficiencies, press utilization, and safety.

Plans already call for equipping all 32 of Century's molding machines in Rochester, and eight machines in Tennessee (60 to 850 tons, mostly Toshibas) with Hunkar terminals by around this time next year. All will be networked into what Karpovage calls a "CIM Room" nerve center in Rochester, operated by a full-time CIM supervisor. He also hopes to eventually loop the network into automated good/bad parts separators at both plants. In this way, after-the-fact quality inspection will be eliminated. "If you do it right at the press, everything else is a given - success comes naturally," Karpovage says.

He concludes with a key word of advice for others taking a closer look at CIM: After you've identified the system you want, dedicate someone to implement its operation into your plant. "Any other priorities will be a distraction," he cautions. CIM at Century Mold is still too new for any big- picture analytical trending of process and production improvements. But the company's management has obviously liked what it's seen so far. For example, Karpovage started as a process systems manager. In only a couple of months, owing partly to his success with CIM, he was promoted to his new position, corporate quality manager.

#15 Business of Molding:


Editor's note: This series of articles on business relationships of custom molders is from consultant Bill Tobin, of WJT Assoc.

Recently I was asked by a client "Why didn't we get the job?" He had been sent a set of prints from a new customer and was asked to provide tooling and quote production. However, some of the dimensions were toleranced impossibly low: +.001 inch and as low as +.0005 inch in ABS. My client prided himself on being a high-precision small part molder and thought he would be able to get the job. However, with this kind of tolerancing, the yield rate would be low, and he anticipated a larger than normal amount of rejects.

A few weeks after the bids were due in, he got a letter from the buyer that was particularly enlightening. All the bidders were given a letter "A," "B," or "C." The buyer then constructed a chart showing the tooling and molding prices. This allowed everyone to see how he did compared to the competition. When you see one of these, with a little backwards algebra you can see if your press rates are competitive and your tooling prices are in line with the current mold building environment. My client's tooling prices were OK. However, his part prices were between 40 to 70 percent higher than the low bidder.

What was the reason? The problem was the prints. Designers are generally ignorant (in the kindest sense) of what plastic can do and about what kind of tolerances are practical. Thus, they do the tolerancing of all other mating parts and then do the stack tolerancing of the entire assembly. It is usually the plastic part that gets whatever tolerance is left over from the other parts.

Because plastic parts tend to be flexible, they usually have a self-compliant property called creep. They can be jammed, swaged, and squashed into places a diecast part would not tolerate. This is done at the peril of a product failure, but usually with a few modifications things can be adjusted so the product functions.

While you will almost have to hold a designer at gunpoint to get him to admit this, he does not care if the molded part you produce is to print. What he wants is a product that works. Thus, product function is primary, product consistency is equally important, and parts to print are nice but not necessarily important. This presents some opportunities to those who quote parts.

1. Do not insult the designer by providing a one-page quote with five pages of exceptions and a marked up print. Send your quote with a simple notation that it is based on SPI Fine tolerancing for the material or whatever tolerance classification is appropriate.

2. Partner with the designer whenever you can.

3. Remind folks of the following truths:

  • You cannot mold consistently nor can you measure +.0005 inch in molding plastics.

  • There is no such thing as zero draft - everyone puts in some draft.

  • If you SPC your parts and 19 out of the 24 parts do not show any measurable difference in dimensions, your statistics are faulty. There is not enough variance to draw any conclusions. Nineteen out of 24 should show a variance, not the reverse.

  • It is smarter to build to the minus side of the wall stock tolerance. It is steel safe and more prone to inexpensive adjustment. More importantly, thinner wall stocks use less plastic and process quicker.

  • While stainless steel does not rust, its ability to conduct heat is 30 percent less than conventional tool steels. If rust is a problem, flash chrome or nickel plate the tools.

  • Make sure the RFQ has something about how the tool will be qualified and how the production parts will be inspected. Functional, dimensional, or statistical qualifications as well as inspections will vastly impact your price. If your customer is only interested in parts that work, duplicate sets of gauges that are exchanged every six months between your QC and his incoming inspection department will avoid unnecessary false rejects.

  • Quote apples-to-apples, so that everyone quotes the same type of tooling with the same cavitation. Ask for feedback from all the molders who provided tooling and part prices; it is good form.

  • Most important: Once you've got the job, make sure everyone knows the definition of a good part. This will avoid wasted production, false rejects, and miscommunication.

Quoting off the print is fine when everyone else is doing the same thing. However, you need to give the print a sanity check. Too many people look at it and quickly decide no one can or will measure some of the impossible tolerances and bid accordingly. It is these people who get the job because they will work closely with the customer to get the part modified once the product has been approved. While not cricket, that's how it is. If, however, you can teach your customer how to do it right in the first place, you'll avoid these practices.

Complex, hollow design? Consider metal core molding

Most designers associate lost core molding solely with automotive air intake manifolds. And while it is true that this process serves admirably in these applications, says CoreTech Assoc.'s Mark Battista, it has also piqued the interest of designers in other market segments where complex, hollow parts are the norm. "Although the process requires capital investment relative to part size, volumes, and the amount of automation involved, there are certain parts that cannot be made using any other process," Battista told IMM in a recent phone interview. "More designers in nonautomotive markets, especially sporting goods, general industry, and aerospace, are looking at what CoreTech trade named Metal Core Technology (MCT) to mold one-piece, internally complex hollow parts."

The GEN3 air intake manifold, found underhood on the 1997 Corvette, is produced by the MCT process using two horizontal injection molding machines. Designers from aerospace and sporting goods markets are also eyeing the process for complex, hollow one-piece parts.

CoreTech offers a range of services, from initial components development and prototyping to turnkey MCT workcells. Stand-alone core casting and melt-out equipment are available for those who wish to integrate their own production cells. Battista and partner Tom Kidd created CoreTech in 1995, following the acquisition of the division from former employer, Electrovert MDD, where they were responsible for business development and application engineering, including production of the first high-volume U.S. plastic air intake manifold for GM's Northstar engine.

For the record, Battista refutes misconceptions about this process. "MCT is both cost-effective and environmentally safe. Tin-bismuth and other metal alloys cause no harm to operator, plant, or environment. Also, cores are completely recovered after they are melted out of the part, then reused continually without degradation. The only environmental issue is handling a small percentage of heat transfer fluid created during melt-out, and that is done effectively using equipment built into the workcell."

CoreTech staffers have found that OEMs and designers ask the same questions about MCT: Can it produce quality parts at a required rate? Can prototypes be evaluated before investing in new equipment? Are the plastic, metal, and other necessary materials available and proven? In all cases, Battista and company offer an unequivocal "yes."

As director of operations at CoreTech, Battista has identified key parameters that must be satisfied before considering MCT molding. To determine if your application might benefit, take a look at this checklist:

Is the one-piece hollow configuration beneficial?
Adding value to a former metal or multiple-piece plastic part is the best way to utilize this process. MCT molding offers a competitive edge for one-piece hollow components in several cases - if they can be made only by this method, if there is a materials savings over other processes, or if significant finishing and assembly costs are eliminated.
Can the part be designed for injection molding over a low-melt alloy core?
A new design can take advantage of the fact that MCT permits undercuts without the need for core pulls or sliding cores, highly controlled internal surface finish, and tight tolerances on wall thickness. Battista cautions, however, that parts must be compatible with the process. For example, metal cores are accurate to .0015 inch per dimension for creating hollow spaces to that tolerance. Also, the metal core must survive the temperatures and pressures of the molding process, and still be melted out at a temperature and cycle time that doesn't negatively affect the plastic part.
Will the volume justify the investment?
Although each case is unique, air intake manifold volumes can serve as a guide. Tier One automotive molders are justifying systems based on annual production rates as low as 65,000, knowing that the equipment can be used for original parts and other models over a 10-year period. Most cells are constructed with standard injection molding machines already on hand - the additional investment comes from core-making systems, robotics, conveyors, core melt-out tanks, and finishing equipment.

For a lower price tag, cells can be designed with less robotic and material handling equipment. Battista adds that an MCT cell can be tooled to make various short runs of different parts. For example, volumes for high-tech bicycle wheels were initially low, but the part needed the structural integrity and light weight MCT provided. After the '96 U.S. Olympic cycling team used the wheels, the OEM geared up for a full range of general, mountain, and racing wheels using the same workcell.

To make the answers a bit simpler, CoreTech offers an Application Analysis Study to provide OEMs and molders all the information needed for a decision. This includes a manufacturability assessment, cost projections for prototyping, and a guide to establishing in-house or subcontracted production. Battista also advises bringing a team together that includes the OEM, MCT process company, toolmaker, molder, and resin supplier. "Parts must be optimized for MCT to add value to the part. GM's Northstar manifold, as a positive example, can be processed off the line in less than 60 seconds, thanks to the interaction of all team players."

Servorobots help Polaroid change its modular manufacturing cells - in an instant

Polaroid Corp. may very well be the highest-volume producer of high-precision optical lenses in the world today. In the NI building at its sprawling Norwood, MA "campus-like facilities," as we called them after visiting there the first time (see August 1994 IMM, p. 32), Polaroid molds lenses with optical tolerances pressed between pluses and minuses in the millionths-of-an-inch range. And they do it in price-busting high-cavitation tooling - up to 16 and 24 cavities in many cases - making many millions of lenses each year.

Polaroid has used the same kind of process engineering savvy that makes this kind of world-class production possible to create truly flexible manufacturing cells. These manufacturing cells allow Polaroid to make even better quality lenses today, with better labor efficiencies.

Polaroid has made all of its secondary stations into mobile self-contained modules that can be plugged in and out of manufacturing cells and travel with molds from machine to machine.

Theodore A. Parker is a principal engineer in Polaroid Camera Div.'s Optics Group: "We have a large number of active products, well over 100. And we run them in 24 molding machines. But we don't believe in large inventories, so we do lots of mold changes. That's why flexibility is a key operating principle around here."

Manual . . .

Not too long ago, flexibility involved higher labor costs and lower lens quality. Parts were removed, degated, inspected, and packaged manually, with an operator opening and closing a safety gate, interrupting the cycle, and potentially contaminating the high-precision lenses simply by handling them. Parker says his group decided early on to automate parts removal and handling. But they wanted to do it with true, closed loop, all-electric servodriven robots. "We had a couple of pneumatic robots with electric traverse strokes. They did the job, but they were not very easy to set up. Setup involved a lot of tweaking and trial and error, all manual, and they offered little programming flexibility." He had learned from prior experience that ease of programmability in robots is extremely important. "We did not want anything with ladder logic. We wanted any of our engineers to be able to program the robots."

. . . or Automatic?

Smarter robots were needed because Polaroid had gotten smarter. It planned to fully automate lens manufacturing. Robots would hand lenses off to Polaroid's own servoautomated beside-the-press product-handling stations. There, automatically, the lenses would be ultrasonically degated, oriented, and transferred to PVC packaging tubes or vacuum-formed trays or palletized in fixtures for special optical coating in high-vacuum coaters. So the robots had to be able to interface and communicate fluently with the molding machines and with the secondary automated handling stations, as well as with the operator. Better control over the quality of the process would eliminate the need to do 100 percent inspection for product quality. One person could handle the output of two or three workcells at a time.

But remember, Parker says Polaroid does lots of mold changes. Production efficiencies could be improved if the secondary stations were mobile self-contained modules that could be plugged in and out of manufacturing cells, like a camera's film cartridges. Such cells could travel with molds from machine to machine as business warranted. And new cells could easily be plugged in with existing injection machines and robots running new projects in new molds. Nothing would be hard-wired.

There would be no single host supercomputer running everything. Control would be decentralized among the discrete controllers of the units in the cell, requiring only a "handshake" communications protocol to share safety, timing, and sequencing routines between themselves. And individual cell units could be serviced off-line. Manufacturing would truly be flexible.

Robots would have to be smart enough, not only to remember detailed choreographed parts removal and handling routines for molds, but also to remember parts-handling programs for the parts-handling modules. In addition, robots would need a fourth closed loop servodriven axis, a wrist rotation, to accurately place lenses removed from vertical multicavity mold faces in differing pattern layouts into different horizontal holding fixtures in different secondary stations. Some pattern layouts are radial and some are rectangular. The latter requires the robot end-of-arm tool to deposit, say, eight parts in a row in a pallet, then flip around and deposit the other eight.

Polaroid's ideas were ahead of its time. Servorobots like the ones they wanted weren't around when they started their project earlier in the '90s. After a few outright refusals to quote, false starts, and time-consuming setbacks, the company found a servorobot supplier it could work with that was a 3-hour drive away - Wittmann Robot & Automation Systems (Torrington, CT).

Polaroid is used to working with its capital equipment suppliers, and working on their equipment. It modifies almost everything that comes in the door. For example, the majority of its 26 hydraulic molding machines are from Shinwa Seiki (50 to 300 tons). Polaroid used its process engineering savvy to come up with a number of improvements to the machines for its own purposes in-house, mostly improvements in machine control software. The OEM has incorporated Polaroid's suggestions into its latest series of closed loop machines.

Polaroid worked with Wittmann to customize the robots to the "plug-and-play" flexible manufacturing cells it uses today, particularly in adapting its controller to accept its redundant safeties, its standardized interface cable connectors, and its software protocols. These are shared by the robot, the molding machine, and the product-handling peripheral.

There are 128 I/Os in the robot controllers for auxiliaries. Wittmann Robot CAN-BUS control-area-network architecture also came in handy. If necessary, robot controllers have the intelligence to control the entire product-handling workcell. In palletizing workcells, the robot controller coordinates pallet filling. The number and arrangement of slots in a pallet are not consistent with the number of parts in a shot, so the robot tells the pallet when to index when the time is right.

Polaroid builds its own end-of-arm tooling and grippers. It has even developed its own positive-contact and digital part sensors for different types of grippers. Of critical importance, Parker and his associates find programming the robots to be intuitive, thanks to their easy-to-use interface. "I don't think we've yet found everything a Wittmann robot can do," Parker admits.

Wittmann provided training for all three shifts. The robots are clean enough to work in Polaroid's Class 10,000 cleanroom molding area. Polaroid has eight Wittmanns now. In about two years of operation only one cable had to be replaced on one robot. Other service problems were equally minor. There's only one pneumatic robot left in the shop.

So far, the capital investment in machines, robots, and secondaries in the project has been between $2 million and $3 million. Cycles are rock steady and stabilized, and manual handling has been all but eliminated, so lens quality has improved. There is still a way to go to fully automate the entire plant, but the hard part is over. "We're fortunate that the company over the years has been willing to invest in new technology. To compete in the marketplace, we have to continuously improve our production efficiencies," Parker concludes. But it took savvy to make production capabilities flexible enough to adapt to change. For more information on servorobots and automation solutions from Wittmann, circle 237. - Carl Kirkland

At its facilities in Norwood, MA, Polaroid molds lenses with optical tolerances pressed between pluses and minuses in the millionths-of-an-inch range.


A Flexible Polaroid Cell In Action

Here's how it works with a radial-pattern multicavity mold: A parts-handling module safety enclosed in clear plastic is rolled up and docked into place on a docking plate beside the press after a mold is changed. Since the module is enclosed, there's no need for permanent, floorspace-wasting guards around the back of the molding machine's clamp. Saved setup programs are loaded into the controllers. As programmed, the Wittmann robot automatically discards a predetermined number of parts after a cold start-up. Gripping the runner, the robot removes the lenses, moves up and over the safety gate, then precisely places them into a holding fixture on the parts-handling module. Suspect parts detected by the Shinwa Seiki's process controller are automatically discarded.

Empty PVC packaging tubes are indexed into position around the circumference of the holding fixture under guide chutes. Like a Scara robot, a servodriven ultrasonic welder rotates into position over the fixture and degates the parts. Lenses are transferred through guide chutes into the PVC tubes, separated by cavity, while sprues and runners are automatically discharged into waste containers.

After the proper number of lenses has filled a tube, the tube is automatically packaged, and is automatically replaced with an empty tube. This automatic degating/ packing system is a three-axis servodriven machine in and of itself, providing closed loop control over the positioning of the welder, the shot, and the tubes. Magnetic safety locks on access doorways on the modules are automatically engaged while the robot is in motion.