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.


Articles from 2000 In March

By Design: Part Design 108 -- Surface finish

In this bimonthly column, Glenn Beall of Glenn Beall Plastics Ltd. (Libertyville, IL) shares his special perspective on issues important to design engineers and the molding industry.

Polishing the cores and cavities of a new tool is one of the last things that a moldmaker does. This is the finishing touch that will determine the molded part’s appearance. The degree of polish will affect the see-through ability of a transparent part, or how an opaque surface catches and reflects light. The richness of the color of a pigmented part will change with different degrees of polish. These are important considerations, as it is the surface of a product that a potential customer sees first. There must be something in the appearance of the product that encourages a person to take a second look, or to pick up the product. The feel of the product in the hand will enhance or distract from the impression created by the product.

The interior components of a product or parts destined for the industrial market have different surface finish requirements from consumer products. Approximately 25,000 new products were introduced in the U.S. in 1997, and with today’s market cycles that number has undoubtedly increased. Many of these products have similar shapes and features. The visual impression, or the feel created by a product’s surface finish, can make the difference between a successful sale and a failure. No new product is a success unless it sells and money changes hands.

The designers of plastic products are very knowledgeable regarding the marketplace advantages of specifying the ideal surface finish. Selecting the optimal surface finish is challenging, as customer preferences are continually changing. A few years ago, a nonreflecting flat black was the preferred finish for high-tech products. Today, Apple Computer is setting a new trend with a transparent matte finish that allows the customer to see the product’s interior components. This same transparent matte finish is now beginning to appear on irons, flashlights, radios, and printer housings.

Changing Appearance Requirements
The degree of polish on a part is a major contributor to the product’s appearance and marketplace acceptance. It is the product designer’s responsibility to choose and identify the degree of polish, but design engineers are not well informed on how to specify the chosen surface finish. There are many interrelated factors to be considered. Many of these factors are not well defined or described in the literature available to designers.

The surface finish of a molded plastic part is dictated by the polish on the mold’s core and cavity. It is also affected by the type of plastic and the way that material is molded.

Some design engineers have more experience with metal than plastic products. With metal, the specified surface finish is the finish that will be on the finished part. In the case of plastics, the specified finish will be on the mold, but the surface of the molded part may be different.

All of these variables have resulted in the common practice of submitting the part drawing for cost quotations with no surface finish specification. Other drawings contain notes such as "diamond polish," "Class A finish," "high polish," or "matte finish." These designations are inadequate. A moldmaker needs to know specifically what degree of polish will be required in order to prepare an accurate proposal. Hardened, dense metal can be polished to a higher level than softer metals. A high level of polish may require that the mold be built using hardened tool steel, instead of a less costly prehardened steel.

Creative designers have the ability to see things that do not exist. They can create a mental image of a product that is nicely styled with smoothly blended, free-flowing contours. The appearance surfaces of that part might have a rich orange color with a smooth finish. The appearance surface has to be smooth, but not so lustrous that it reflects light. The designer can see exactly what he wants in his mind’s eye, but it is difficult to specify that particular surface finish on a drawing.

Communications between product designers and moldmakers have always been, and continue to be, problematic. Developing a mutual understanding of the surface finish required on both the mold and the molded product is one of these recurring problems.

SPI Mold Finish Guide
In 1962, the Moldmaking Divs. of the Society of Plastics Engineers (SPE) and the Society of the Plastics Industry (SPI) cooperated in establishing mold finish standards for the injection molding industry. This standard specified six finishes, ranging from a high diamond polish to a rough, blasted surface. This standard was controlled by a set of six steel disks containing each of the six finishes.

These finish standards were very helpful in allowing moldmakers and their customers to agree on the polish that would be provided on a new mold. However, there still were differences between the finish on the steel mold and the finish on a part produced by that mold. The misunderstandings created by these differences were minimized in the mid-1970s when SPI introduced an injection molded plaque incorporating the six standard finishes. Designers could then see what the finishes looked like on a molded part. Polish could be specified according to the same numbers one through six that the moldmaker had on his six metal disks. The system worked well. It was widely promoted and well accepted by the moldmaking industry and, eventually, the product design community.

In the mid-1980s, SPE discontinued all standards-setting activities. Fortunately, SPI continues to make the mold finish standards available.

While reintroducing the standard on its own, SPI, based on the experience gained, changed the word "standard" to "guide" and expanded it to include 12 finishes (see Figure 1). The original six SPE/SPI mold finishes were encompassed within the new guide. The SPI Mold Finish Guide has become the internationally accepted standard for specifying surface finishes. Product designers would be well advised to use these guidelines while specifying surface finishes. It also would be a good idea to buy extra plaques for your suppliers.

Molding Considerations
The last "By Design" article dwelt on the importance of providing molding draft angles on part drawings. The point was made that generous draft angles reduce part cost and increase part quality.

Draft angles and mold polishing go hand in hand. Drafting the sidewalls of a deep draw, 3-D part reduces the force required to pull that part out of the cavity and push it off of the core. The benefits to be derived from draft angles will be lost, however, if those drafted surfaces contain gouges left by machining or coarse polishing. These gouges in the metal of the core and cavity form undercuts. Plastic melt flows into these undercuts and solidifies. This situation increases the force required to eject the part from the mold. This increase in ejection force results in a corresponding increase in the cooling portion of the molding cycle. The benefits to be derived from draft angles will be maximized by polishing the molding surfaces to remove these undercuts.

Product design engineers have strong opinions regarding the finish to be provided on the appearance surfaces of a 3-D part. They are less concerned about the inside nonappearance surfaces of the part. The majority of part drawings that specify a polish are referring to only the appearance surfaces of the part. It is unusual for a design engineer also to specify the degree of polish on the inside, nonappearance surfaces. With no finish specification, the moldmaker polishes these surfaces only enough to get by. This is a common but unfortunate situation that complicates the ejection process.

As a plastic material cools, it also shrinks. This shrinkage pulls the part away from the cavity to grip the core tightly. The maximum demolding force is normally that which is required to eject the part from the core. If the core that forms the inside surfaces is not smoothly polished, the ejection forces and the cycle time will increase accordingly.

There is always a last-minute rush to get a new tool finished and delivered. Polishing is one of the final functions and it is not always done as well as it could be. The cavities that form ribs or bosses and small-diameter core pins often have rough surfaces left by machining or coarse polishing. These rough surfaces increase the force required to demold the part.

In some instances, the functional requirements of a product do not allow for adequate molding draft angles. In these cases, smoothly polished molding surfaces become even more important. Polishing is normally done in a random pattern. Even a high polish leaves behind microscopic gouges or scratches in the surface of the metal. Those scratches, which are parallel to the mold’s parting line, become undercuts that increase the ejection forces.

A single scratch is not significant, but thousands of them will increase the required ejection force. Polishing the mold only in the direction that the molded part moves during ejection eliminates these undercuts. This polishing technique is called "draw polishing." Design engineers would be well advised to specify draw polishing on parts requiring minimal or no draft angles.

Material Considerations
Some plastic materials produce glossier surfaces better than others. Regardless of the finish on the mold, low-density polyethylene can never produce the shiny appearance of a polystyrene part. Heavily filled or reinforced materials cannot reliably produce shiny surfaces. Generally speaking, a hot mold will produce better surfaces than a cold mold. A poorly packed out part will not faithfully replicate the mold’s polish on all of the part’s surfaces.

With the vast majority of plastic materials, ejection forces will be minimized by providing smoothly polished surfaces on the mold. There are always exceptions. The softest polyethylenes and polyvinyl chlorides, polyurethanes, and some thermoplastic elastomers have a tendency to adhere to highly polished metal surfaces. Sticky materials of this type will normally release better from molding surfaces with a matte or lightly blasted surface finish.

Strong but somewhat flexible plastics, such as polypropylene, ABS, and impact styrene, are more tolerant of roughly polished molding surfaces than rigid, brittle materials, such as general-purpose styrene and acrylics. Draw polishing is very beneficial with these hard, brittle materials.

Design engineers must be careful to specify a surface finish that is suitable for the plastic material to be molded. Product designers, moldmakers, and polishing companies cannot be expected to know the idiosyncrasies of all of the different plastic materials. Plastic material suppliers and experienced molders can, however, be relied upon to know what mold finish is optimal for a given material.

Polishing Cost Considerations
Many parts of the moldmaking process have been mechanized or automated. Polishing is an exception. It is only the simplest of shapes that can be automatically polished. Polishers now have better equipment and materials to work with, but the process remains labor intensive.

Polishing is a step-by-step process. The polisher starts with files or coarse abrasive, 320-grit stones, which are then followed by repolishing with progressively finer materials, down to an 8000-grit diamond polishing compound. The higher the finish, the more polishing steps. Each additional polishing step represents additional cost.

There is no such thing as an average cost for polishing a mold. Some industry insiders have estimated that, on the average, polishing can represent 10 percent of a mold’s overall cost. This average will obviously vary, depending on the size and shape of the mold, the hardness of the metal, the level of polish required, and how well the mold was finished prior to polishing. Transparent molded parts require a high polish on both the core and the cavity. Optical lens molds are in another whole class by themselves.

Polishing represents a significant part of a new mold’s cost. Among the variables cited here, it is the level of polish being specified that has the greatest effect on the cost of polishing. The level of surface finishing on injection molded parts is frequently overspecified. In many instances, this overspecifying comes from the concept that "my part deserves the best finish that man can provide."

This attitude results in unnecessary added cost. Most consumers will not notice the differences between the SPI Guide’s three highest finishes on an opaque part. The two best SPI finishes should be reserved for transparent, see-through parts. In other words, why spend money on a surface finish that the consumer cannot appreciate? Overspecifying the surface finish on a molded part also results in other less obvious increases in cost. Highly polished surfaces reflect light that highlights surface imperfections. Flow marks, splay, weldlines, and minor scratches or abrasions become more obvious and result in a higher reject rate and increased part cost.

Overspecifying the surface finish not only raises a mold’s initial cost, it also increases the ongoing cost of maintaining that mold. Highly polished metal surfaces do not last forever. Molding surfaces are susceptible to corrosion, abrasion, and mechanical abuse. Highly polished surfaces will magnify these defects. Stainless steel and chrome-plated molds resist corrosion. Hardened steel molds are more tolerant of abrasion and mechanical damage. In spite of these safeguards, highly polished molds do require repolishing.

Selecting the optimal surface finish for a new injection molded part must be based on the careful consideration of three interrelated factors: (1) The marketplace requirements; (2) the efficient molding of the part; and (3) the cost of the mold and the molded parts. Between these three factors, marketplace acceptance has top priority. Cost and ease of molding will cease to be concerns if the product is unsuccessful in attracting the attention of the consumer.

The product designer is the expert on what degree of polish will charm potential customers. It is the designer’s responsibility to specify that surface finish accurately and to communicate that information thoroughly to the molder, moldmaker, and polisher. There is no substitute for a face-to-face discussion with suppliers, but that is not always possible in today’s global economy.

At the very least, the surface finish must be clearly indicated on the part drawing. However, it is strongly recommended that surface finish be specified according to SPI’s Mold Finish Guide. This Guide has now become international in scope. In fact, even those in a small shop in Korea will know what an

Using technology to stay on time and under budget

Twenty years ago, three practicing doctors decided to go into the manufacturing business because they saw a need for specialized devices among ear, nose, and throat (ENT) surgeons. Today, Boston Medical Products Inc. (Westborough, MA) uses a combination of rapid prototyping, rapid tooling, and 3-D CAD to bring low-volume, injection molded silicone rubber devices to market on time and at a competitive price.

As in the early days of the company, physicians continue to play a very important role in product design. The company obtains design input from ENT surgeons, designs the final products, and turns the physicians’ concepts—devices for ENT surgery, respiratory therapy, and thoracic surgery—into reality at its manufacturing site.

The company’s product line includes tracheostomy tubes, tracheal cannulas, T-tubes, speaking valves for tracheostomy tubes, balloon catheters, and laryngeal stents. Nearly all of these devices have to be contoured to human anatomy, making for some unusual and intricate shapes. In turn, that means creating unique and complex molds to manufacture them.

Boston Medical previously designed its products using outsourced CAD services and tooling design. However, when Andrew Marsella joined the company as an engineer in 1997, its R&D department was growing, and the company gave him the go-ahead to purchase a 3-D CAD program for in-house use (Cadkey).

"We used it first for 2-D specifications," Marsella says, "and as the product changed and improved, we began using it for solid modeling and rapid prototyping." Designs can get complicated in the field of ENT surgery, and lend themselves to the benefits of solid modeling. To design a device that is contoured to the shape of the human larynx, for example, engineers can use cross-sectional data obtained from X rays and CAT scans. "These sections can be drawn and stacked in CAD, and then made into a solid using the lofting command, which sweeps a skin up and over the stack of cross sections," explains Marsella.

The typical design process starts with the solid model, to which features such as rounded edges, chamfers, and fillets can be added and removed. "When the basic part is complete, the next step is to make a solid model of the mold cavity. I create a solid block in CAD, insert the part into it, and then subtract the part from the block. That way I get a reverse of the part itself, which serves as a mold," he explains. "I add such mold features as sprues and gates." The model is then exported to a rapid prototyping machine to create the master for the sintered tool.

Rapid, Production-quality Tools
Most rapid tools are built from a prototype part, but Marsella produces his rapid prototypes of the actual mold insert on a Sanders system. "We found a toolmaker that can work with these prototypes, and returns sintered tool inserts within two weeks," he notes, "which reduces costs and lead times by at least 33 percent compared to machined molds." In addition to the time and cost savings, the sintered tooling works well for the company’s complex-shaped parts.

In the past, Boston Medical had to commit to aluminum tooling for prototype parts, but the costs and lead times can be excessive, according to Marsella. Because most of the products designed and manufactured by Boston Medical are relatively small, he says that the company can produce rapid-production-quality tool inserts in four weeks total—from creating the solid model to having a mold insert ready for the manufacturing floor. "We have not yet used these inserts in manufacturing, but they are production-ready, with tolerances of ±.005 inch," he says. "We have run the inserts in our R&D lab on an injection molding press with excellent results."

The toolmaker, NDM Inc. (Syracuse, NY), developed the process independently of any commercial rapid tooling systems. To make the inserts, heavy-duty silicone rubber is poured around the rapid prototyped mold, which creates a reverse rubber mold. Powdered metal is then poured into the reverse mold and baked in a kiln to melt the metal. This process may leave some degree of porosity in the mold, so NDM adds a filler metal that melts at a lower temperature to fill voids.

The sintered powder and metal filler form a 420 stainless steel alloy that becomes machinable when cool. The finished mold inserts can be nickel-plated to make them harder and smoother, allowing Boston Medical’s silicone rubber parts to be removed from the mold without using release agents.

Prototyping Made Easier
Marsella points out that rapid prototyping can make a model of any shape a designer can create with CAD—including those that would be a nightmare to machine by conventional methods. Rapid prototyping also helps Boston Medical save in premanufacturing costs. Before instituting this process, the company would commit to tooling, then find that surgeons would suggest last-minute changes, forcing redesign of the tools.

Marsella explains, "Surgeons may not be aware of costs and other factors involved in the tooling process, nor should they be expected to. It is up to us to devise methods that will be effective for both the doctors and the company. The quicker you can get an accurate model into the surgeon’s hand, the cheaper and more efficient it will be to settle on a design and manufacture it. Surgeons like to know the feel of a device in their hands, and they get a better concept of the part or product from a three-dimensional model they can hold than from any amount of drawings."

To meet this need, the company initially gave physicians the original SLA hard resin prototypes, but now uses the rapid tools to mold silicone rubber prototypes that are closer to the finished product and much more understandable to the doctors.

"Surgeons often want to use a prototype to test for form, fit, and function, perhaps in laboratory or even cadaver tests," Marsella says. "The products we manufacture don’t typically undergo great stress in use, but they need to fit the human anatomy." With relatively inexpensive sintered tools, Marsella can mold clinical prototypes. If changes are minor, the 420 stainless alloy tools can be machined for the resulting production runs.

Contact information
Boston Medical Products Inc.
Westborough, MA
Andy Marsella
Phone: (508) 898-9300
Fax: (508) 898-2373
E-mail: [email protected]

Cadkey Corp.
Marlborough, MA
Liz Rombek
Phone: (800) 372-3872
Fax: (508) 229-2121
E-mail: [email protected]

NDM Inc.
Syracuse, NY
Simon Bojilov
Phone: (315) 492-2933
Fax: (315) 492-3227
E-mail: [email protected]

Market Focus: Automotive Underhood

Regular readers of this section know that from time to time we like to clue you in to new phrases, buzz words, and additions to the injection molding lexicon. Notable standouts from the past include "modularization," "minimally invasive surgery," and "Ragu test." Grab a pen and paper and brace yourself for another one. The phrase is "engine beauty."

This one comes from Engelbert Meurer and Mark Witman, both of Bayer Corp. Plastics Div. Meurer's title is manager of innovative technologies, and Witman's is director of technology. At least part of their attention is focused on the automotive market in general, and underhood components in particular. They report that a recent trend in new-car buying is impacting the molding segment of automotive production.

It turns out that, aside from the usual selling points a vehicle presents (style, performance, reliability, comfort), the aesthetic appeal and organization of the engine compartment has become a factor in design and manufacturing. Dealers across the country are throwing up their hoods in the showroom to display their engines. Although designed primarily for acoustical attenuation, engine beauty covers are used to hide unsightly brackets and wires and may feature molded-in logos, all selling features intended to symbolize a well-crafted vehicle.

"Engine beauty covers with noise dampening effects and structural components with better modal [frequency] behavior will drive this market," says Meurer. "It's kind of surprising that people would be concerned about what the engine looks like, but it's a fact," adds Witman.

The genesis of this trend, like many in the automotive industry, is in Europe, where fuel efficiency, weight, and emissions standards have forced automakers to innovate differently than U.S. suppliers do. But U.S. standards are changing, prompted in part by the Partnership for a New Generation of Vehicles (PNGV), a coalition between the federal government and U.S. automakers that aims to develop lighter, more fuel-efficient cars and trucks.

Along these lines Meurer and Witman note other trends, including smaller, hotter engine compartments, increased modularization surrounding air/fuel systems (manifold, filters, air/fuel controls), and plastic/metal hybrid components.

Most notable among the latter is the grille opening reinforcement on the 2000 Ford Focus, a plastic/metal composition that won the chassis category award at the 1999 SPE Automotive Div. parts competition (see January 2000 IMM, p. 60). It's produced for Ford by Visteon using a process developed by Bayer in which a nylon 6 is molded around metal stampings inserted in the tool. The result is a part with the structural integrity of metal and the reduced weight of plastics. "We put plastic where we need it, metal where we need it, and take cost and weight out of the design," says Witman.

What does the future hold? Both Meurer and Witman agree that there are still substantial opportunities for converting existing metal parts under the hood to plastic. As materials evolve and improve, heretofore impractical plastic applications become real. This is borne out in Table 1, which predicts the increased use of plastics in passenger cars and light trucks.

Table 1

Average plastics use estimates in cars and light trucks assembled in the U.S. and Canada

  1994   1999 2004  2009  Plastics use in cars and trucks has come a long way, but how much further can it go? Jim Best of Market Search Inc. provided these estimates of average plastics use in cars and light trucks through the first decade of the century. 
 Total plastics  187 lb  257 lb  276 lb  312 lb
 Injection molded plastics  118 lb  138 lb  152 lb  167 lb
 Source: Market Search Inc., Toledo, OH

Polyimide favored for viscous fan drives

A new generation of viscous fan drives for on-highway truck and agricultural vehicles features two parts molded with Aurum thermoplastic polyimide provided by Mitsui Chemicals. The fully modulating drives control engine fan speed in direct proportion to engine cooling needs, and are manufactured by Borg-Warner Automotive Turbo Systems in Indianapolis, IN. The company switched to the injection moldable polyimide, known for its high strength and toughness and resistance to creep and wear, in an effort to improve performance and lower costs for its viscous fan drive components.

Previous models included a cast-aluminum, stamped-steel, or diecut Teflon fluid dam to redirect silicone fluid from the drive to a reservoir. In operation the dam is exposed to 350F temperatures in silicone fluid, velocities of 4500 ft/min, and maximum pressures of 100 psi. A second component called a valve system washer, which separates a bimetallic coil from the aluminum drive cover, also was previously die cut from Teflon and is now molded from the polyimide. The valve system washer is subject to a maximum temperature of 300F in an unlubricated environment.

The parts are molded by Hilco Technologies (Grand Rapids, MI), a custom injection molder of high-temperature thermoplastics. The company conducted axial and torsional vibration tests, a 10,000-hour continuous bench endurance test, and a 1500-hour accelerated engine and endurance test on the parts. The polyimide dam allows the fan driver cover casting to be used for all rotations whereas the cast metal dams required one cover casting for each rotation. The dam reportedly holds firmly in the cover without fasteners, and the strength properties have improved valve system performance. Properties of Aurum JCL 3030 include mold shrinkage of .3 percent and a melting point of 730F.

For more information:
Mitsui Chemicals America Inc.
Purchase, NY
Phone: (914) 251-4222
Fax: (914) 253-0790


One-piece bracket expands with high temperatures

A new, one-piece, self-adjusting thermoplastic bracket replaces a metal multiple-piece bracket for holding an automotive air-conditioning unit's desiccant canister in the 2000 Dodge Durango sports utility vehicle. The bracket attaches to a side wall under the hood, securing the aluminum canister and preventing rattling. The metal version required mechanical tightening.

Molded of BASF's Ultradur B thermoplastic PBT polyester, the bracket can reportedly be installed with greater ease than the metal design, which uses tension bands. The plastic bracket has two raised hollow ribs molded into it, which allow it to expand slightly when the canister is inserted. The expansion ribs allow the PBT bracket to increase in diameter when high temperatures cause the aluminum canister to expand.

Flambeau Corp. molds the bracket at its Baraboo, WI plant using Ultradur B-4300 G2, a glass-fiber-reinforced PBT polyester. Reported benefits of this material include creep resistance, high flexural modulus, and retention of strength in high temperatures. The material reportedly has a continuous use temperature of 285F, and is resistant to chemicals and oils that are typically present under the hood.

Flambeau injection molds the bracket for Hutchinson FTS Inc., which supplies the assemblies from its Reading, MI plant to Chrysler Corp.

For more information:
BASF Plastic Materials
Mount Olive, NJ
Phone: (973) 426-3910
Fax: (973) 426-3912


Assembly converted to plastic

Engineering thermoplastic is now used to make the integrated transfer case/automatic shifter assembly on Chrysler's Jeep Grand Cherokee. The assembly incorporates an automatic gear shifter and transfer case shifter for the four-wheel drive, all molded with Vydyne nylon, a reportedly strong and rigid material with heat and chemical resistance. Dow Automotive, manufacturer of the resin, reports that the material is suited to powertrain and chassis applications.

The shifter is molded from Vydyne nylon 6/6 in a multicavity tool by N-K Manufacturing Technologies in Grand Rapids, MI. Engineers for the company say it was chosen for its design flexibility and overall rigidity. Of the 18 components that make up the shifter, half are produced by N-K, which also does a portion of the subassembly. Final assembly is done by Dura Automotive Systems Inc. in Rochester Hills, MI.

One challenge presented by this part was meeting dimensional requirements; this is the first time a plastic has been used for this assembly. The nylon reportedly has good structural integrity and long-term creep resistance properties.

For more information:
The Dow Chemical Co.
Midland, MI
Phone: (248) 391-6300
Fax: (248) 391-6417


Nylon balances strength, cost

Designers at Nagle Industries, Clawson, MI, wanted to streamline the design of a metal cable actuator assembly, and selected Verton RF, a long-glass-fiber-reinforced nylon 6/6 from LNP Engineering Plastics, for the job. Engineers report that the nylon has strength characteristics similar to that of diecast metal, and costs less.

The assembly is located in the steering column and actuates a cable that releases the tilt mechanism, allowing the wheel to move up and down. The unit has an actuator arm and housing that is snapped over a metal bracket without the need for fasteners. This helps to minimize space requirements and save costs.

Designers of the actuator report that Verton RF produced little deflection or twisting when under load and can be molded in thick sections with minimal shrinkage. The actuator is molded by M&N Plastics in Sterling Heights, MI.

For more information:
LNP Engineering Plastics Inc.
Exton, PA
Phone: (610) 363-4500
Fax: (610) 363-4749


Nylon resin fuel filters replace collapsible tinplate

Traditional oil and fuel filters using tinplate can collapse during cold starts of diesel engines, during which fluid pressure can shoot up to 100 psi. Apparently solving the problem is Zytel nylon from DuPont.

The nylon canisters hold filter elements of tightly wound paper, up to four stacked in housings. Open-end canisters (pictured at right) are made from unpigmented resin and are used for filtering diesel fuel. Black-pigmented, closed-end canisters are used for oil filtration.

The filter canisters, manufactured by Filter Technologies Pty Ltd. in Rocklea, Queensland, Australia, were developed with assistance from DuPont. The parts are injection molded by Tooltech Pty Ltd. in Carole Park, Queensland. The mold was designed and built in Australia by Industrial Tool Service, Underwood, Queensland.

Tooltech reports that molding the parts presents a challenge in that each canister has zero draft on the inside wall, and the outside wall has only .5 mm difference in thickness from top to bottom.

As a result, these virtually straight-sided canisters can be difficult to eject from the mold. The mold itself is designed with a stripper sleeve that travels the full length of the part. A foot in the base of the closed-end canister increases the stripper contact area.

The canisters measure 6.9 inches high and are 4.3 inches in diameter. The resin grade used is Zytel 70G33HS1L, a heat-stabilized nylon with 33 percent glass reinforcement. The parts also look sturdier than tinplate, which reportedly is a marketing advantage.

For more information:
DuPont Engineering Polymers
Wilmington, DE
Phone: (800) 441-0575
Fax: (843) 856-8277


Aluminum manifold scrapped for nylon

The aluminum air intake manifold in Renault's four-cylinder, 16-valve engine used in its Lagune and Espace models has been replaced with a nylon molded version. The manifold allows reuse of the oil mist from the decanting system of the rocker cover underneath. The design includes two injection molded shells, assembled by vibration welding, that reduce the number of parts to be assembled. There also is a weight savings, as the cable guides and depression sensors have been integrated into the air intake module.

The material used is Technyl A 218 V35, a nylon 6/6 reinforced with 35 percent glass fiber. Produced by Rhodia Engineering Plastics, the material is reportedly resistant to chemicals, greases, brake fluid, and oils under the hood while holding its dimensions in high temperatures. Development teams say that the rigidity of Technyl has ensured a strong seal between the air intake and the cylinder heads, while keeping the part's walls thin. Good static and dynamic behaviors create resistance against engine backfire and stress from internal pressure variations.

Rhodia helped Renault develop the manifold and was heavily involved in part and mold design, prototype development, moldfill simulation, and final product testing. Production parts were checked for leakage, burst pressure, and unwanted resonance. They've been subjected to vibration endurance and acoustical behavior optimization. Mark IV Systemes Moteurs in Orbey, France molds the part and reports that Technyl A 218 V35 is an easily weldable material, allowing the company to use thin welding lines that do not alter the high burst-pressure resistance of the part.

Renault already uses plastic air intake manifolds on its Twingo, Clio, and Kangoo models, although the new manifold is the first to be produced in two welded halves.

For more information:
Rhodia Engineering Plastics Inc.
Mississauga, ON
Phone: (905) 270-5534
Fax: (905) 270-4737

IMM's Plant Tour: Tour de force in micromolding

Going the extra mile-or in this case, the extra micron-is all in a day's work at Micro Precision Plastics (MPP). Built from the ground up for micromolding and small parts production, this Canadian custom molder pays attention to even the tiniest of details in a major way. What else elevates MPP to the "tour de force," or exceptional achievement, level? During IMM's recent tour, many of the answers could be found both on the shop floor and under the microscope-amazingly complex and demanding micro parts produced to zero defect levels for OEM customers that include heavyweights such as Xerox.

Make no mistake, micromolding can strain the limits of even the most exacting molder. Gary Lane, president, explains that MPP differs from others in the field because it is geared entirely to this niche. "When my wife, Chris, and I started the business in 1994," he says, "we saw an opportunity for molding small-sized, high-precision parts on small machines. Other molders were running small or micro parts on large machines, and were unable to control dimension and sizes. Often they would have to put so many cavities in to balance the shot size that they'd end up fighting with dimensions. Instead, we invested in mainly 25- and 35-ton machines that are sized for small part molding."

Apparently, the Lanes' hunch was correct. MPP has doubled sales every year for five years. Initially, the plant supplied parts to other molders. These early customers needed someone to mold parts for assemblies-parts that were too small for their machines. Although MPP still does a fair amount of this type of work, most of the business is now geared to OEM end users in business machine, electrical connector, and automotive markets.

Creative Design
Roughly one year ago, the Lanes moved operations to the current building. From the time MPP designed its new plant to the time construction began, business had grown so rapidly that it had already outgrown the space, even though it had yet to move in. The Lanes hit on the idea of creating a second floor mezzanine when they looked at the 24-ft ceilings still under construction. "Most of the space was empty, so putting another floor there nearly doubled our manufacturing space," he explains. The plant now has two floors, each with 12-ft ceilings.

Small advantages proved important here, too. The machines are not as large or heavy as typical 300- to 1000-ton presses. As a result, the mezzanine didn't need much reinforcement-just enough to aid in dampening vibrations. This second floor is actually a free-standing structure floating on its own support beams within the building, and there are 26 presses on it. With larger machines, this creative use of space would have been impossible.

The entire plant is fully air-conditioned to control the environment year-round. For health and safety reasons, the system also includes an air-to-air heat exchanger so that there is no fume buildup in the air (MPP runs a good deal of acetal for gears). "We've designed the plant for uniform part quality, and eliminating variations in environment is important to that," Lane says. In addition, the plant includes a central chiller instead of a cooling tower to control water temperatures. Air, water, and power utility lines are mounted on the ceiling of the first level, so they are routed up either to the mezzanine or down to the first floor machines.

The second floor consists of two rows of Battenfeld Plus machines ranging from 25 to 35 tons. Each press is equipped with part-removal robotics and runs 24 hr/day during the week. All 25 of these Plus machines have been acquired over the last five years. Lane says he carefully researched the small-machine market and chose the Plus after reviewing the data and consulting with other molders who had used the machine and its competitors. Their biggest upside, Lane adds, is a user-friendly control system that simplifies setups and offers consistent accuracy.

In addition, on the first level, there are six Milacron Roboshots devoted to running connectors, four Nissei HM7 machines, and six Bantam 9-tonners. On weekends, there are eight to 10 machines running.

Five years ago, when MPP opened its doors, there was a problem getting auxiliary equipment that met the needs of the small-parts molder. That scenario is changing, though. Today, the plant includes special mold temperature controllers (Conair MicroTemp, the smallest on the market) and a dryer system (Dri-Air) that is made to handle 10 lb/hr of material.

"The majority of dryers are in a central location," he says, "but we are slowly moving them to individual machines because we are now scheduling machines to run only one material. For example, some machines will run only nylon, so those will be equipped with individual dryers. For some of the smaller machines, though, it's impossible to mount a dryer on them because of their size. In these cases, we'll continue to use a central dryer arrangement."

For the 26 machines on the mezzanine, it takes only one setup person, two operators, and one inspection specialist to keep production running. The nature of the business and the demands of just-in-time requirements dictate many short runs. MPP often does 20 or more mold changes per day.

When it comes to waste, there is a paradox involved in micromolding. The runners tend to be larger than the parts themselves. To cope, MPP has installed a closed loop regrind system on many of the Battenfeld presses. "We can store weeks worth of material in a Rubbermaid trash can," Lane quips. The runners are sent through a granulator (Rapid); a proportional feeder picks up the regrind and mixes it into an autofeeder. MPP regrinds and uses its production scrap in percentages agreed to by clients. Regrind can be as much as 30 percent but is most commonly around 20 percent.

Tooling Tactics
Tooling for small and micromolded parts is an area in which customers often need to be educated. "We've pushed customers to go to fewer cavities. Instead of running an eight-cavity mold with half of the cavities blocked off, we go with two four-cavity molds on smaller machines to control dimensions better. I always feel it's better to have two molds than one for a part that is running all the time, because if the mold goes down, you at least have the other one going. On one project, we ship 35,000 parts per day from two four-cavity molds. Having two molds also gives us time for maintenance."

Also, with small machines there is no need to go to more expensive and difficult multicavity molds to gain production economies. MPP can

be competitive with single-cavity molds, even on parts in which production is as much as 100,000 or more per year. In fact, one single-cavity mold makes a million parts per year. Though the company has worked with hot runner molds in certain situations, with one or two cavities it is just as easy to design the mold with very small runners and spare the expense of hot runners.

Interacting with its toolmaker is no problem for MPP, because its supplier, C&L Molds, is located in the same building. In an unusual arrangement, C&L Molds, owned by Dave Coles and Steve Lee, owns the building and land jointly with MPP, yet each operates as a separate company. C&L consists of seven toolmakers and builds molds exclusively for MPP, which buys only about 10 percent of its molds elsewhere, and usually then only to meet workload or specialized technology needs.

With more than 800 part numbers, either as full molds or inserts, MPP builds 80 to 100 new molds per year. Work frequently comes in sets of between 10 and 20 molds when a major customer releases a program for a new product. MPP uses standard mold components and inserts frequently to lower costs and to cut startup time on a new project. In case of a customer emergency, a pulley or gear mold can be turned around in a couple of days. Usually, though, lead times are four to six weeks. "For a good customer, we are happy to make whatever effort it takes," says Lane.

Although they have used standard mold bases, MPP and C&L have designed their own line of bases for micromolding. For gear molds, the toolroom has a supply of blank inserts already cut that need only to have gear teeth and drops put in. "In this way, we can improve our response time to the customer," adds Lane. When a client had a crisis with an acetal gear, for example, C&L was able to cut a cavity and core for a prebuilt P-20 tool and Lane had parts to the customer the next day. This is highly unusual, but indicative of the lengths to which MPP will go.

In addition, some small parts have required brushes in the mold to sweep out parts that are sticking. This requires specific expertise, but when the parts are too light to be handled by air, this technology is needed to ensure consistently good part removal. "Because they are so small and light, static will make them stick to the mold. We use compressed air from the materials handling system [filtered and dried] to remove most parts, but at times, we have to combine this with brush removal. The brush is a toothbrush fitted into an air cylinder to impart a light action," he says.

Quality Details
With a zero-ppm defect rate for customer Xerox, it's clear that MPP pays close attention to quality. SPC and

P-charts are used extensively in concert with ISO 9002 procedures. On most parts, checking weight is almost impossible because parts don't have the mass required to give good resolution on a scale. As a result, inspection equipment includes two video measurement systems, a Microview 250x magnifier, and a newly added computerized gear checker.

Each molding machine has its own sorting table, located away from the press, where operators pack and sort parts. The table also doubles as the station for part information. Every job has a Part Information Book (PIB) that contains work orders, shipping labels, quality plans, SPC inspection charts, drawings, and change sheets. The PIB is central to communications.

For a new job, after mold trials are performed, the engineer produces a mold setup sheet that must be approved by the engineering department. Any changes to an approved setup must also be approved. SPC checks are done every hour, and at startup, every half hour. At shipping time, the QA inspector produces a ship log audit, and is the only person who is allowed to close a box of parts and certify them for shipment. No one can break that rule.

According to Lane, quality is paramount at MPP. "And cleanliness is just as important," he adds. Rather than employing air blowers, MPP vacuums the floors every shift, using a central vacuum system installed on each floor. It helps reduce the dust level in the plant, which directly assists quality. The environment would meet Class 100,000 cleanroom standards if it were audited. "We could easily switch to cleanroom molding if that becomes necessary," he says.

As the company was getting started, an automated material feeding system and robotic pickers were considered too costly. That opinion quickly changed as Lane realized how their benefits override the cost. He explains, "If there was one thing I would change in the company's history, it is that we would have automated more and faster."

Lane buys only new equipment; most of the current equipment is less than three years old. He also prefers that everything on an injection molding machine come from that machine's manufacturer, including robotics and auxiliaries. "You buy it as a complete system, and if there are any problems, you only have to make one phone call."

A Micro Future
Lane sees many growth possibilities both with current customers and new ones. Awareness is increasing among designers about the possibilities for small and micro parts in plastic. In addition, many products already on the market are being miniaturized, and emerging electronic marvels are aimed at being small, personal, portable, palm-sized, and wristwatch-sized. The parts inside of these products, which include everything from mobile phones to laptops, must be small and precise. And as the world miniaturizes, MPP is cheering it on.

Lane has no plans to move to larger machines. "There is plenty of room to grow in the small parts market, and there are some very significant market advantages for a manufacturer like MPP," he explains. "When you can ship a three-month supply of a part to a customer in a package that is less than one cubic foot, the customer can be almost anywhere in the world, or at least anywhere UPS, FedEx, or DHL will go."

Contact information
Bowmanville, ON
Gary Lane
Phone: (905) 697-1466
Fax: (905) 697-1759
E-mail: [email protected]

Predicting damage for stripped IM parts

At the MoldMaking 1999 Technical Conference, Clinton A. Haynes, vp, and David A. Tekamp, principal, of Stress Engineering Services, presented their ideas on how to analyze part designs that include stripped undercuts. The goal of their method, of course, is to avoid damage and maintain finished dimensions on stripped parts.

Stripping undercuts on injection molded plastic parts remains one of the more experiential aspects of plastic product and tooling design. Rules of thumb available to part designers are by no means precise. Instead, they lack quantitative insights regarding the expected final dimensions or quality of the molded part. Consequently, designers often find long and costly tooling rework loops, focused either on simply getting the part out of the tool, or modifying the part design to enable the stripper to get the part ejected without damage.

In recent years, nonlinear finite-element analysis (FEA) has evolved. It is now possible to predict reliably the interaction between tooling design, part design, and end-of-cycle processing conditions. Using this technique, issues of final part dimensions and damage can be simulated before steel is ever cut. Here is a significant opportunity for OEM designers, toolbuilders, and molders to reduce dramatically the time and cost associated with the development of plastic components and products. Thoughtful application of nonlinear finite-element techniques can provide a detailed, physics-based window into the tool/process/part interaction while in the design phase of the project.

FEA to the Rescue
Hampered by poor technical results produced by the "we can do it all" software packages that came and went in the 1980s, the acceptance of nonlinear FEA has been rather slow, and for good reason: The problems are very difficult. Unlike metals or other more-or-less linear elastic materials, under most conditions, plastics exhibit nonlinear behavior.

For example, designers of metal components like fasteners do not get concerned about creep behavior until the structure reaches 650F or higher. However, the designer of a polyethylene component must deal with this complex material behavior even if the product is used in freezing conditions, much less in elevated temperature situations.

Unlike linear FEA applications in which the solution to structural problems is relatively simple and direct, nonlinear applications are solved iteratively. During each numerical iteration the computer solves thousands of simultaneous equations as the loads are gradually applied to the structure. For this reason, nonlinear analyses require much more time to complete than linear finite-element simulations.

Three types of nonlinearities must be accounted for when modeling the stripping of a plastic part:

  • Material.
  • Boundary.
  • Geometry.

Of these three, material and boundary nonlinearities dominate when using FEA to predict the effects of stripping on plastic parts.

Material Matters
Accurate material data is critical to making quality predictions of the part/stripper interaction. Modeling the stripping of an undercut for an injection molded part using FEA techniques requires a true stress vs. true strain curve for the particular material at the ejection temperature. In addition, it is also important to have this data at a strain rate characteristic of the stripping operation being modeled.

A true stress vs. true strain curve differs from an engineering stress vs. engineering strain curve because it accounts for the reduction in cross-sectional area associated with the elongation of the gauge length. These types of curves show the dramatic differences in nonlinear response for a broad class of polymers.

Polypropylene, for example, exhibits strong nonlinearity, characterized by a more-or-less continuous curvature of the stress/strain curve at room temperature. For this material it is not possible even to speculate about a single value of a yield stress since the material accumulates inelastic strains even at very low stress levels.

Conversely, polyester exhibits more metal-like behavior at room temperature with a region of approximately linear correspondence between the applied stress and the strain. The material also has a fairly well-defined point at which gross yielding will occur.

At the time of ejection from the mold, in most applications, the part will be warm, typically between 110F and 140F. At these temperatures, the mechanical properties of most polymers will be much different than at room temperature. For polypropylene at 140F, there is a significant reduction in strength. The stress that is created in the part as a result of the stripper action will produce higher strains and dimensional variation in the final part at this elevated temperature.

One additional material complication that should be considered when modeling the ejection of a part from an injection mold is strain rate dependence of the material. Typically, as the strain rate increases (the speed at which the deformation takes place on a defined cross section), the stress vs. strain curve will change.

Stress vs. strain curves are generated by conducting a tensile test of the same material at crosshead speeds, proportional to strain rates, ranging from .004 in/sec to about 4 in/sec, or four decades in strain rate. Although the change is not dramatic

(a change of four orders of magnitude in strain rate produces a change in ultimate strength and tensile modulus of only about 30 percent), this variation will impact the quality of the FEA simulation results. To maximize the accuracy of the predictions, it is useful to use the material curve that best represents the state of the material during stripping.

Boundary Nonlinearities
Contact between the plastic part and components of the mold—core, cavity and stripper—causes boundary nonlinearities in FEA. The primary boundary nonlinearity of importance in the simulation of part stripping is the boundary between the core and the part. This must always be modeled.

If undercuts are not present and the focus of the analysis is to determine if the stripper causes damage to the part during ejection, the friction between the part and core must also be modeled. The friction establishes the resistance the stripper will experience during the ejection. This in turn defines the stress level that will be developed, which will be tracked by the material model during the FEA solution.

Generally, the contact between the cavity and the part can be neglected because the shrinkage of the part onto the core during cooling will typically minimize the effect of any contact with the cavity.

If a smooth undercut is present, as in the polypropylene part in Figure 1, the stretching of the plastic over the core pin will create stresses that are proportional to the diameter over which the plastic must pass. Figure 2 illustrates a series of stress contour plots during part ejection.

As the undercut region of the plastic part stretches over the corresponding protrusion in the core pin, the stress level increases. If the stresses reach a sufficiently high level, the plastic will be permanently deformed. This phenomenon is illustrated by plotting the equivalent plastic strain in the ejected part. The red region shown in Figure 3 illustrates the area where the plastic is actually deformed or wiped away because the material could not support the level of stresses encountered there.

Another useful output of the FEA simulation is the calculation of the force required to eject or strip the part from the core of the mold. This knowledge aids in sizing mold components, particularly for high-cavitation tools.

If the undercut is sharp, as in a snapfit feature, the role of friction is diminished and the stripping force is dominated by the material properties of the plastic. In this situation having representative material data near the ejection temperature is important. The stiffness and strength of the material governs the ability of the part to be removed from the mold with minimum damage.

Dosing Cup Mold—A Case Study
Stress Engineering conducted FEA for an injection molded dosing cup that has a continuous snap bead around the opening. It measures a prescribed volume of liquid soap on a large economy-size container. The clarified polypropylene cup is also a protective lid, and is attached to a blowmolded neck finish via intermittent snap beads on the bottle.

The dosing cup unit tool was designed using conventional methods and finite-element simulation was not used prior to fabricating the mold. Upon startup, it was discovered that it was difficult to strip the cup from the core without damaging the snap bead. After three weeks of trial-and-error tool modifications, it was decided to use a more systematic approach and resolve the problem via FEA. The goal was to use the FEA to evaluate redesign alternatives and select the most promising candidate, rather than commit more time to an open-ended trial-and-error method on the unit tool.

A finite-element model was constructed that focused on duplicating the geometry in the current tool that was experiencing the problem. The purpose in modeling this baseline situation was to validate the finite-element approach being used and learn as much as possible about the problem. Results of the baseline simulation are illustrated in Figure 4, opposite. Comparison of the simulation results with the actual stripped part indicated good agreement with the physical damage produced by the stripping action.

Based on this result, approximately 13 redesign concepts were defined and subsequently evaluated using finite-element methods. This allowed two to three designs to be evaluated per day to save time and money. Several of the more promising candidate designs failed FE simulation. For example, Design #5 tried to incorporate a tall foot below the snap bead for flexibility, but simulation results predicted that the foot would buckle during ejection. In Design #12, adding a short foot and increasing the nose radius showed severe crushing of the foot.

Design #6 incorporated three design modifications—an increase in the snap bead nose radius, an increase in the lower ramp angle, and an added small foot at the base of the part. The results showed only small amounts of permanent deformation on the inside foot corner.

Based on the analysis, Design #6 was selected as the leading candidate and the unit tool was subsequently modified to reflect that snap bead geometry. The molding trial proved the new design to be satisfactory, producing good-quality parts for a broad range of processing conditions.

Tips for completing an analysis

Simulating the stripping of an injection molded component using finite-element techniques can be divided into six steps:

  • Construction of part geometry.
  • Definition of boundary conditions and contact surfaces.
  • Fitting material data to material model.
  • Establishing temperature distribution in part.
  • Conducting simulation.
  • Postprocessing results.

Construction of part geometry. Several solid modeling software packages such as I-deas, Pro/Engineer, SolidWorks, MSC/Patran, and Unigraphics have significantly reduced the tedium formerly associated with the development of geometrically complex parts. Any one of these software packages can be used to create the geometry of the molded part and tooling components. Although all of these software tools are capable of generating the geometry needed to complete the analysis, the ease with which they interface with a nonlinear FEA software is highly variable. It is recommended that the analyst research the compatibility of the software packages and select programs with a high level of compatibility.

Definition of boundary conditions and contact surfaces. The method and requirements for defining the boundary conditions and contact surfaces depend on the FEA package used. However, it is recommended that an analysis package be selected that offers generalized contact capabilities and rigid surfaces. Contact and rigid surface algorithms enable the analyst to reduce the overall size of the finite-element model and avoid some unique problems that can occur between the plastic part and mold components during the simulation.

Generalized contact eliminates the need to define local contact rigorously before conducting the analysis. This is often a difficult task since the deformation of the plastic part is not known before the analysis is completed. Also, rigid surfaces enable the analyst to take advantage of the fact that stiffness of the mold is at least an order of magnitude greater than that of the plastic. Therefore, for all practical purposes, all the deformation takes place in the plastic and the steel can be assumed as rigid.

Fitting material data to material model. Commercial FEA packages generally provide the user with a feature to fit their stress vs. strain data to a range of material models. Some of the packages provide the capability to implement their own user material. Currently, there doesn’t appear to be one particular material model that has been universally adopted by all commercial software packages. Therefore, we recommend consulting the software vendor about how to go about fitting specific material data to models in its software. Also, it is recommended that test cases be conducted to verify the range over which the material model is valid.

Establishing temperature distribution in part. Depending on the objectives of the analysis, two approaches can be used to establish the temperature of the part at the time of ejection. The simplest approach is to estimate the temperature at ejection and use stress vs. strain data at that temperature for the analysis. This is a reasonable approach and will generally provide sufficiently accurate results.

If, however, the part is sufficiently complex, or the thickness varies in a way that makes it difficult to estimate the temperature, a transient heat transfer analysis can be completed using the same FEA model and mesh. For the purpose of modeling the stripping, this analysis can be completed by defining the temperature of the part to be the melt temperature of the polymer at time-zero. The steel temperatures in the core and cavity can be assumed to be somewhat warmer than the cooling water temperature. For those unfamiliar with these temperatures, the mold operator can provide this estimate.

With the initial temperatures defined, a transient analysis can be completed for a period of time equal to the cooling portion of the molding cycle. The temperature distribution in the part at that time can be used to select the appropriate material curve. Although this approach will generally predict ejection temperatures that are cooler than they really are, it is a reasonable and quick way to estimate the end-of-cycle conditions.

Conducting FEA simulation. The information and data described above provide all the needed material and process conditions for the simulation. The specific details regarding load steps, translations of the rigid surfaces, and stability of the solution will depend on the software package used. Regardless of the software package used, it is not uncommon that numerical instabilities will arise that cause the software to stop and require attention. However, provided that the software is capable of solving the problem, with time and experience, the process of successfully modeling this problem becomes more straightforward.

Postprocessing results. Once the finite-element simulation has been successfully completed, the engineering analysis can begin. The relevant stress, strain, and deformed shape plots should be examined and compared with hand calculations, as well as actual parts when they are available.

Contact information
Stress Engineering Services
Mason, OH
Clinton Haynes
Phone: (513) 336-6701
Fax: (513) 336-6817
E-mail: [email protected]

The Materials Analyst, Part 31: When trouble is obvious, keep testing

This series of articles is designed to help molders understand how a few analytical tools can help diagnose a part failure problem. Michael Sepe is our analyst and author. He is the technical director at Dickten & Masch Mfg., a molder of thermoset and thermoplastic materials in Nashotah, WI. Mike has provided analytical services to material suppliers, molders, and end users for 15-plus years.

Often when clients come to us with a problem, they have a preconceived notion of its origin. Such insights are never to be taken lightly, since practical information is often the key to a good interpretation of the results. At the same time, however, these glimpses into the thought processes of the customer cannot be taken as gospel, particularly when the objective evidence does not support the theory. In these cases it is important that the analyst change direction quickly using common sense and experience to adjust the problem-solving strategy. This is one of those cases.

A client came to us with a brittle ABS part. The part was being made using natural resin and a range of color concentrates that were blended at the press. The parts that we received had been molded from a mixture of 70 percent virgin plus a beige color concentrate and 30 percent regrind of the same color generated from a previous run. The base resin was a medium-impact ABS, hardly the toughest compound in the product line.

However, even a casual review of the parts revealed obvious problems. The parts contained cracks from routine handling; they had not even been in the field yet but it was clear that if they ever got that far there would be immediate problems.

It was a simple matter to stress the cracks by hand and propagate them rapidly through the part. There was no sign of the stress whitening that usually appears when ABS is placed under a severe load. Most importantly, the cracks showed severe signs of delamination. The prevailing theory from the customer was that either the concentrate had been mixed too heavily and was causing the product failure or the material was simply degraded due to abusive processing.

They asked specifically that a test for pigment loading be conducted along with the melt flow test to compare virgin material to the molded part. Unfortunately, no samples of the color concentrate were available, so for the moment at least we had to assume that it was of the correct composition and not based on a "universal" carrier.

Start by Testing Color
We began the assessment of the pigment loading by analyzing the part using thermogravimetric analysis (TGA). For those of you who have been reading this column for a while, you already may recognize this as a weight-loss technique that is performed to gain information about composition.

In the good old days, when most pigment systems were based on heavy metals, this was the best way to determine the relative loading of a pigment system since most of the colorants were stable beyond the point at which the polymer degraded. They could be collected as residues in the same way as a filler or reinforcement.

With the advent of more environmentally friendly systems, though, this has become much less of a sure thing. Some pigments, like titanium dioxide, the most common ingredient in white colorants, can still be directly measured as residue. But other systems that have changed to organic pigments or dyes may decompose along with the polymer. Since we did not have access to the pure concentrate, we decided to run the test and compare the output to historical results for ABS of similar colors.

Figure 1 shows the weight loss for the molded part. Most of the weight loss, more than 94 percent, occurs when the polymer burns off between 400C and 525C. When the atmosphere in the TGA sample chamber is changed over to air, an additional 3.65 percent of the material is lost. A very small amount of residue, just less than 1.4 percent of the original mass, is left at the end of the test. Its color was close to that of the molded part, an indication that this ash was related, at least in part, to the pigment system.

In checking older test results for ABS materials of similar color, we found that this TGA result was almost a perfect match. The pigment loading did not appear to be the issue. The melt flow test was also a blow to our customer’s expectations. The data sheet specified the melt flow rate for the virgin material at 3.5g/10 min. The parts measured 4.12g/ 10 min. This represented a 17.7 percent increase, well within the allowable range for good processing.

Two More Tests
At this point we had checked the product for the items that the customer believed had gone wrong and had come up empty. But the existence of a serious problem was undeniable. As we began to take a fresh look at the parts, it appeared that the problem had more to do with contamination by another material.

Usually when this type of problem surfaces in a styrenic, the culprit is a polyolefin like polypropylene or polyethylene. The incompatibility of styrenics with olefins is well documented, and we frequently find polyethylene or polypropylene in molded parts because these materials are the basis for many purging compounds. Failure to purge the barrel thoroughly during startups can cause molded parts to be produced that contain relatively large amounts of foreign material. Since both polyethylene and polypropylene are semicrystalline materials with strong melting points, it is fairly easy to detect this type of contamination in an amorphous material like ABS where no strong melting event exists.

Moving in this direction, we ran a sample of the brittle part by DSC, another thermal analysis technique. This one is designed to detect transition temperatures such as the melting point. Figure 2 shows the result of the DSC test for this brittle product. Much to our surprise, there was nothing out of the ordinary here. The main event is the glass transition temperature for the ABS at 109C. This is followed by a very weak event at 125C and an even weaker one at 162C.

While these two temperatures correspond to the melting points for some grades of polyethylene and polypropylene, there is a problem. First, many ABS materials incorporate additives that melt in the region of 120 to 130C, so this first point is not conclusive. The second transition is uncharacteristically broad for polypropylene. A part this brittle should have contained a much higher level of contamination and the melting point should have been at least as large as the ABS glass transition. While the event might be real and might even be a foreign material, there was no way that we could advise our client on the nature of the problem based on these results.

While it is relatively easy to find contamination of an amorphous material by a semicrystalline material using DSC, finding a second amorphous material is another matter. Often the offending material has a similar glass transition temperature and simply cannot be detected. This would be the case if, for example, an acrylic were mixed with the ABS. And even in cases in which the glass transition temperature is different, as with polycarbonate or PVC, the level of contamination is often too small to register a clear sign of the problem. In these cases, we often turn to infrared spectroscopy.

Along with DSC and TGA, an infrared spectrum is another way of fingerprinting the material. It can also hone in on smaller sample areas in order to magnify regions where a suspected problem might be. This is particularly important for products in which delamination occurs because the contaminant can be trapped in thin layers that are sandwiched between sections of good material. DSC and TGA often rely on samples that are too large to isolate the problem product.

Figure 3 shows the infrared spectrum for the brittle ABS above a spectrum for the good product. While there are regions in the brittle ABS where radiation is absorbed in a manner consistent with standard ABS, there are other areas where the sample is absorbing when ABS does not. Three of these abnormal absorptions are indicated in the figure.

Identifying the Contaminant
At this point it is a great temptation to turn a computerized library loose on this result and let it find a material that matches. The problem is that if this spectrum is caused by a mixture of two or more polymers, the computer will make a mistake because it will interpret the spectrum as belonging to a single material. It will do its best to come up with a material from its archives that best matches all of the peaks that it sees in the sample; and it will undoubtedly be wrong. Instead, we used the computer to subtract the ABS spectrum from the spectrum of the brittle part.

Figure 4 shows the result. Subtracted spectra are always a little risky and we have talked previously about the mistakes that can be made if the technique is misused. A good example of a potential problem can be seen in the highlighted area near 2240 cm-1. This is the region where the nitrile bond in acrylonitrile absorbs. It is very sharp and distinct. If we subtract a pure ABS from a diluted one, this peak actually becomes negative and will go down rather than up from the baseline. This type of response can cause some problems, but if the contamination is significant it will still come through as a distinct set of features.

This was the case here. Several strong absorption peaks left little doubt that there was a substantial amount of a PPO/polystyrene blend present in the brittle molded parts. Finally we had a result that explained the problem. PPO/PS and ABS do not mix at all and when blended they will produce the results that we were seeing. A conversation with the client revealed that they did indeed run parts for the same product line using a PPO/PS of the same color. This material had found its way into the ABS in the form of the regrind.

Notice also another possible hazard of spectral subtractions. Both the ABS and the contaminating PPO/PS blend contain styrene. When the subtraction is performed, the two peaks that analysts often look to for a confirmation of styrene have been seriously reduced. Here again, a skilled analyst helps make sense of the result in a way that blind use of the computer can never hope to match.

This case is a good example of how complex the task of picking the correct test can be. Helpful hints on where to look for the problem may shorten the job, but they can also create a preconception that can delay the discovery of the final solution. It might seem reasonable to have stopped looking for problems after three tests proved negative. But when common sense dictates that there is a problem, you have to keep looking until an answer that makes sense is found.

Contact information
Dickten & Masch Mfg. Co.
Nashotah, WI
Mike Sepe
Phone: (262) 369-5555, ext. 572
Fax: (262) 367-2331
E-mail: [email protected]

Precise two-shot tools make the difference

Two-shot molding is high on the list of manufacturing trends today, especially in consumer markets. Making the tools to produce these parts, however, is another story. It's a relatively demanding job, one that requires high precision and often cavity interchangeability. One company that specializes in these molds, Atlas Precision, offers a glimpse into the process.

A prominent custom molder in Arden, NC, Atlas began in 1979 as a one-man tool shop and branched out into molding in 1987. Today, its tooling division and molding operations consist of 225 employees. IMM recently spoke with three key members-Johann Hofschuster, ceo and president; John Fore, gm, Tooling Div.; and Robert Bulla, sales engineer.

Think Two-shot First
According to Fore, two-shot tools are still a relatively new experience in the U.S. "We started moving into it five years ago, and are one of the few shops in the Southeast that can build two-shot tools," he adds. About 80 percent of the tools built at Atlas go to OEM customers, while the rest are built for its own molding operation.

All three Atlas executives agree that two-shot technology does not reside in a two-shot press, which is essentially a molding machine with two barrels and two sets of controls. The technology comes from the tooling and the design behind it. So one of the first steps Atlas takes is to evaluate designs for two-shot molding.

"Many parts are molded in several components that could be molded in the same tool," says Fore. "OEMs often aren't aware of the advantages of two-shot in reducing cycle time and cost of the part. It reduces assembly work, assembly equipment, and handling equipment. And you're replacing two molds and two presses with one tool and one press."

Opportunities are not limited to soft-touch TPEs on a rigid plastic. Applications can include two rigid plastic components molded together. "Handle-grip toothbrushes are the first things you think of," says Fore, "but we have successfully molded two-shot parts without any TPEs involved."

Other applications ripe for two-shot are parts that have to have rubber gaskets applied. These can be overmolded with the gasket in place. "Two-shot has moved to the point where it can go nearly anywhere your imagination can," says Hofschuster. "As a prime example, headlight lenses for cars these days contain two to four different materials. This is a typical project for us. We work as a team from the prototyping stage with our customer, the press manufacturer, and automation suppliers. At times, we also help with part design."

Atlas has been successful with one project in which an OEM customer now manufactures its products 40 to 50 percent faster as a result of the switch to two-shot. "Especially for OEMs that have a captive molding operation," says Hofschuster, "we are helping them to cut out nonvalue-added steps. Part quality also improves as a result of eliminating opportunities for defects."

Evaluating costs
Two-shot molding doesn't generally make economic sense for low volumes (less than 100,000 parts/year). Tools are more expensive and the machine rate is higher. However, for midrange volumes-from 100,000 to one million parts annually-two-shot often makes sense. "And at one million parts and above," says Fore, "it is definitely a no-brainer to combine two or more parts in a two-shot design."

As for complexity, parts that are extremely simple don't lend themselves to two-shot economics. The more complex they become, the easier it is to justify two-shot tooling. Atlas uses a case-by-case analysis with its customers to help them determine current equipment costs and the payback for making the switch.

"Typically in the past, soft-touch materials were glued on," says Hofschuster. "Cycle time savings with two-shot include the elimination of this assembly time. For molding, total cycle time is determined by the material with the longer cycle."

Cost savings come not only from combining molding steps, but Atlas also integrates two-shot tools with other manufacturing processes such as insert molding to eliminate other assembly lines and part travel. Several colors, materials, and inserts are integrated into one line. "We have used this in several markets-automotive, consumer, personal care, and electrical," Hofschuster notes.

Keys to Success
Before building a two-shot tool, moldmakers need to be aware of several demanding requirements. Prototype tooling, for example, should normally be cut from production tool steel. "For two-shot, we try to mimic the production tooling as closely as possible because we are looking for shrinkage and any potential filling or cosmetic problems," Fore explains.

Another key element of building these tools is that they must be extremely accurate-both for repeatability and to offer the molder a wide processing window. Accuracies of ±.002 inch are common.

Fore also recommends checking critical dimensions more closely than usual. "We compare the wireframe CAD file to the cavity steel," he says, "to ensure that tolerances are maintained throughout the tool."

Atlas achieved its ability to build high-tolerance interchangeable tooling by building tools for a large connector company when it first opened its doors, according to Hofschuster. "All of those parts had to be interchangeable, so we applied it to all our customers and their close-tolerance tooling."

The tooling division has a catalog in which a part number is assigned to the components of every tool manufactured. In this way, new inserts can be ordered as needed. "In toolmaking, every insert is typically numbered because it has been hand-fitted in the frame," says Fore. "Our tooling inserts are never numbered because they are built to such high accuracy. We also build all of our mold frames in-house. Inserts can be taken out of different positions in the mold and interchanged as a result."

The company's training program is important to maintaining interchangeable tools, according to Hofschuster. An apprentice toolmaker program has been in place for 18 years, and is approved by state agencies within North Carolina. An apprentice engineering program, now 11 years old, is also essential. In fact, Atlas hired the first graduate of North Carolina State University's engineering apprentice program.

Of course, precision equipment also makes a difference. "In EDM sinker and wire, we have the highest-accuracy Agie and Charmilles equipment available," says Fore. "In our CNC department, we can make electrodes with an accuracy of ±.002 inch. Every electrode is thoroughly inspected. Basically, we learned how to use the equipment beyond traditional limitations."

When investigating new forms of cutting equipment, Fore conducts experiments and tests before purchasing. He sends test pieces out to vendors, then evaluates how well the equipment cut the piece. "Last year, we had a company claim they could do something on one piece of equipment rather than two, but the surface finish didn't pass the test," he says.

Molding and Assembly
One of the reasons that the tooling division works so well with OEM customers is that the molding and assembly areas of Atlas give the company's toolmakers detailed knowledge of the customers' challenges. "Our molding facility takes the same attitude as the tooling division," explains Bulla, "aiming for high quality and high precision."

With 33 presses ranging from 38 to 720 tons, the shop floor also features air conditioning and tile floors to keep contamination down and temperatures constant. Materials are stored in the basement, then fed up through pneumatic handling systems to the presses. "Our number one goal is repeatability," he adds.

Assembly areas are specially designed to produce the most assemblies per day using the least amount of floor space and minimum movement of parts between workers. All of the employees are cross-trained, so that one assembly line worker can go to another assembly line easily, and temporary workers can be brought up to speed quickly. Assembly line people can also move into a molding operator position if necessary. Atlas Precision's quality record relates to this attention to assembly. The company now has completed two years with no late deliveries and zero defects on components, many of which have 50 to 100 parts on one unit.

Contact information
Atlas Precision
Arden, NC
Robert Bulla
Phone: (828) 687-9900
Fax: (828) 687-2700
E-mail: [email protected]

Blueprint for the 21st Century: Design challenges

To design, create, and envision. To model, invent, and contemplate. These words are at the roots of mankind’s greatest achievements. They conjure up images of Leonardo da Vinci or Thomas Edison hard at work. Yet, while today’s designers still thrive on these tasks, they are also being asked to shoulder more practical concerns such as cost reduction and time-to-market issues. And increasingly, engineers and designers must create and invent in a team environment to address the complexities of manufacturing in the 21st Century.

To capture the challenges facing those who design for injection molding today, IMM talked with OEMs, molders, and toolmakers involved in all aspects—industrial design, part and product design, and tool design. Our panel included Eric Chan, principal, Ecco Design Inc. (New York); Steve Spanoudis, engineering manager, Motorola (Plantation, FL); Julio Castañeda, gm, Product Development Technologies (Coral Springs, FL); Robert Braido, health care market development manager, GW Plastics (Bethel, VT); Malcolm Smith, vp, Palo Alto Products International (Palo Alto, CA); Dave Kelly, engineering manager, Pitney Bowes (Danbury, CT); and Mike Pyle, senior project engineer, Hamilton Beach (Glen Allen, VA).

To Market, To Market
Overall, our panel agreed that getting products to market on time continues to strain the resources of a majority of designers. On the extreme end of this dilemma lie consumer electronics such as PDAs and handheld Internet browsers, many with a six-month development window and a mere 18-month life cycle. However, even consumer appliances with a five-year shelf life are feeling the pressure.

Coping with the push to reduce product development cycles has, in itself, created more challenges. "To bring products to market faster, designers and product engineers must first have a clear definition of the product," says Hamilton Beach’s Mike Pyle, "and often, key areas are not defined. These include factory cost, competition, packaging requirements, quantities, forecasts, suppliers, manufacturing facilities, industrial design, and knowledge of the end user. So although the ship date never changes, these areas must first be defined by the engineer."

Steve Spanoudis, who manages a combined prototyping and tooling center at each of two Motorola plants, agrees that part of the task lies in looking at a design from the perspective of manufacturing, both upstream and downstream. "Not only must designers employ some of the traditional DFx [design for variable x] methods such as design for manufacturability and assembly, they must now also look at issues such as component supply and postcustomization."

Product development specialist Rob Braido sees the demands of high-tech product design and time-to-market as the catalysts for new relationships. "OEM customers, molders, toolbuilders, designers, and product developers have to build a true alliance these days," he adds. "Only by integrating all of these functions together as a virtual enterprise will companies be able to bring high-quality products to market in a profitable way."

Virtual enterprises are not common today. Instead, Braido sees an intermediate step—virtual companies that seamlessly blend all resources and technologies required for new product development. He cites an independent study that found these types of companies took half the time to develop products and recoup their investment vs. traditional ones. "The benefits are astounding," he says, "but creating such an environment means a change in corporate culture—from individuals to teams, from leaders to mentors, and from managers to coaches."

Team Engineering Hazards
Concurrent engineering has been touted as the fix for time-to-market and cost-reduction pressures, but even here, the prognosis isn’t always a rosy one. "Products in the telecommunications market are no longer evolving—they’re mutating," says Spanoudis. "One way to cope is to engineer these products simultaneously so that software, electronics, tooling, and housings are all being developed in parallel. This becomes a complex process and can often lead to bumps in the road as unforeseen snags crop up. Good communication is essential."

To obtain the best part design possible, a plastics engineer needs a working knowledge of tooling, processing, resins, and decorating, according to Pyle. Again, concurrent, or team, engineering is assumed to be the answer here. Unfortunately, working together as a team brings its own set of hurdles.

"We are utilizing concurrent engineering, and it’s working, but not to the level that it could," Pyle says. Hamilton Beach recently held a Plastics Roundtable to identify shortfalls in the current system. "This is still a learning process. For example, in cross-functional teams, we assign a representative for each group. Those individuals make decisions for their whole group. One issue is how to address the thoughts of individuals within each group."

According to Dave Kelly at Pitney Bowes, concurrent engineering is vital to internal product development and external customers. It reduces lead time, aids in material selection, and helps his team design parts for the lowest total cost. "This type of interaction fosters relationships for our custom molding business," he says.

On the downside, one pitfall for the team approach rests with its potential for abuse. For example, a customer recently approached Pitney Bowes with a concept. Kelly’s designers stepped in, created a part design, and produced an SLA prototype. After the work was done, the customer began sending the job out for bids and settled on another molding operation. "There is no way to guarantee that if you do all of this work, you’ll get the business," he adds.

Kelly also finds that OEMs lack a working knowledge of IM tooling and its costs. "OEMs don’t typically understand tooling prices. When you start a project to produce five different molded parts, and quote one of the molds at $50,000, customers often respond that their projected tooling budget would only cover the cost of that one mold."

Tooling and Prototypes
Most of the panel agrees that one of the lengthiest and costliest portions of product development for IM parts remains the tooling. While lead times of 24 weeks were typical several years ago, today’s pace more commonly requires six to 12 weeks. Add prototype tooling to the mix, and you have the proverbial bottleneck. Designers trying to escape this time crunch often resort to releasing tool designs in parts so that toolmakers can begin working on one area while part designers are finalizing others. This practice can succeed in saving time, but there is an equal chance that it can add days or weeks.

Julio Castañeda of Product Development Technologies believes that the bottleneck can be broken if designers are willing to bypass prototype tooling in favor of designing production tools more accurately. "I’m not suggesting the elimination of prototypes for high-risk design items," he confirms. "But prototypes used for validation testing and analysis don’t supply results with any confidence unless they are produced on a production tool."

Rather than waiting to test the first shots from a prototype tool, Castañeda offers an alternative scenario in which concepts go through rigorous mechanical design and analysis, followed by rapid prototyping. The SLA or SLS part is then used to evaluate fit, form, and function, with results incorporated into a production tool design.

"The challenge here is to do a thorough job in the design and analysis stages of product development, making accurate conclusions so that only real design issues are addressed," he says. "With prototype tools, these issues are prolonged."

Prototype tools, production tools, and analysis (or virtual prototyping) are beginning to converge, according to Spanoudis. "There is a battle going on between simulation and prototyping. Can I analytically decide faster than I can make a prototype? What will give me the best mix of speed and confidence? The time frame to do a good analytical study and a good prototype tool are getting closer and closer."

Changing Consumer Tastes
Variables exist at all locations on the design spectrum, but the impact of consumer demands is most keenly felt during concept and early product development stages. Industrial designer Eric Chan feels that personalizing the current and future wave of consumer electronics is an issue facing many designers today.

"To stand out in the personal electronics marketplace, OEMs need to offer a product that not only has function, but also fashion," he says. "Tactile qualities and decoration are coming to the forefront."

As industrial designer Malcolm Smith walked through the recent Consumer Electronics Show, he found it amazing how far OEMs have taken decoration. "The selected use of these techniques will be valuable," he says, "and designers need to be well versed in these methods. They also need to know how the decoration affects the supply chain."

Likewise, a flood of new products is challenging designers from concept to production stages. Handheld Internet browsing units, for example, have competing requirements. "They need to be small and lightweight," Smith says, "yet the battery must have sufficient runtime. While hardware and software are being designed, so are the IM plastic components. And somehow, this all has to fit together."

Digital design quality missing the mark

According to Prescient Technologies’ CEO Gavin Finn, the issue of quality in engineering data is reaching critical levels. "Inaccurate or incomplete CAD models affect the product development process significantly because these models are now used by toolmakers, other suppliers, and manufacturers. Errors in design data add rework time and cost downstream, while also wreaking havoc on data exchange."

To help determine the scope of this problem, Prescient conducted a series of engineering quality audits using its Audit QA software customized to look at engineering standards that were the greatest concern for each specific company. More than 3000 models were analyzed from various-sized companies spanning six different industries, including aerospace, automotive, and electronics. Greater than 90 percent of these models failed to meet standards set by the companies themselves. Seventy percent of the models failed standards that companies categorized as "critical."

"These results aren’t limited to company size or market," says Finn. "It’s clear that this problem is a consequence of the increased pressure to use digital data throughout product development to improve time to market." Three main design problems were the root of failures to meet standards: interoperability, iteration errors and omissions, and manufacturability. Across all industries, the big winner was iteration errors and omissions, which accounted for 67 percent of failed models. Specifically, this category includes geometry-based problems such as unattached features, identical elements, and errors in communicating design intent.

Pricing your work: The machine rate enigma

What press rates do molders get for a 150-ton press?"

"Do you have a list of the standard press rates molders charge for various sized presses?"

Those are just two commonly asked questions that inquiring molders want to know when it comes to setting hourly machine rates. Yet it should come as no surprise to most molders that there are no "standard" rates.

And there shouldn’t be, says Sid Rains of IMM Performance Products, a consulting firm in Medina, OH. "There’s nothing common about the service provided," Rains notes. "We forget our business is really being a job shop and by their nature [job shops] provide a huge variety of services to their customers."

Surveys have shown that molders’ machine-hour rates don’t even fall into any rational pattern. They are all over the map from extremely low to extremely high for the same tonnage press. For example, an IMM survey of 47 molders revealed that hourly rates for 25- to 100-ton presses ranged from $9.07 on the low end to a high of $35.

Several molders charge the same hourly rate regardless of press size. One molder charges $36.64. Another charges $44/hour for all presses.

Molders generally attempt to charge various hourly rates depending on the size, realizing that costs to operate different sized presses vary. That would seem to make sense—costs do vary. Yet, there appear to be no standards there, either.

One molder, for example, charges $35/hour for presses in the 25- to 100-ton range and up to $55/hour for presses in the 601- to 700-ton range. Yet another charges $16.30 for presses in the 101- to 200-ton range and only $25.40 for those in the 501- to 600-ton range.

The problem is that this information doesn’t support a single conclusion.

"It’s obvious that there are a lot of strange rates that hurt the industry," says one molder when asked to decipher this survey of hourly machine rates. "The bottom line is that few people know what their costs really are."

Know Your Costs
Knowing costs to manufacture is critical to establishing meaningful machine rates, yet for too many molders, costs to manufacture are ambiguous.

David Brentz, vp of marketing and sales for Plastech Corp. in Forest Lake, MN, agrees. "Most molders, no matter what size they are or their accounting structure, are very aware of what a molding machine costs, their labor rates, and some even know their maintenance and electricity costs," he says.

"Where things fall down is in the burden or overhead rate, and that has a significant impact on how they establish rates," he adds. "How they spread their burden is critical."

For example, maintenance must be allocated, explains Brentz. Molders without a good cost accounting system might not assign an allocation of maintenance costs to individual jobs, inflating profitability on some jobs. Other molders attempt to spread their burden evenly across all press sizes, which results in hourly rates that might appear high at the lower-tonnage ranges and low at the higher-tonnage range, like the molder cited above that charges $44/hour regardless of press size.

Jeff Mengel, a manager for the CPA and consulting firm Plante & Moran in Southfield, MI, says, "If you spread [your burden] like peanut butter across all your presses, you’re distorting the costs of what it takes to run a big press."

Many molders do charge higher machine rates for those presses that incur the greatest overhead, and less for those at the lower end of the tonnage scale. Brentz says that’s why many molders appear to be making money on every job—because they have a handle on their capital costs, labor, and material. However, in actuality they’re losing money because they’re not covering their burden.

The whole issue of establishing pricing that is competitive, yet profitable, is much more complex for the custom molding industry. "You would think the market would dictate pricing structure," notes Brentz. "It does and it doesn’t, at least not as much as in other industries where pricing can be much more clearly dictated, such as in a proprietary product atmosphere."

Know Your Jobs
Another key to establishing machine hourly rates is knowing the types of jobs that run in the various presses. Because molding is a job shop business, not every job runs equally even in the same press. Rates need to be established for jobs depending on the costs to run those jobs.

There are both fixed costs and variable costs in establishing machine rates. Fixed costs, such as the monthly machine payment, electricity required, and maintenance, are level. Variable costs such as labor, robotics, the mold, and material can fluctuate with the job.

Plastech uses three levels of labor rates depending on the job’s use of automation, or the labor intensity, or whether it’s a combination of labor and automation, depending on the type of job.

Steve Larson, marketing and sales manager for Southwest Mold Inc. in Tempe, AZ, says that company knows the jobs and the gross margins of each job. "We know that a 440-ton machine will generate more revenue than a 40-ton machine, but it also consumes more of everything per hour," explains Larson. "And when we’re carrying debt, the 440-ton has a higher debt load. It might be generating $100/hour of revenue, whereas the 40-ton might be generating $40/hour."

Southwest visits the issue of machine hourly rates each year when management puts together its budget. And that’s good, says Rains. "If you’re not changing your machine hour rate every year, you probably don’t know what your real costs are," comments Rains. "They might go up. They most certainly shouldn’t go down. They should never stay the same. They will vary depending on your customer mix, how well you schedule, how many mold changes you’ll have, how many colors you’ll run—all these logistics factors. Without those you can’t calculate a meaningful machine hour rate."

That’s where many of the differences in hourly rates for the same tonnage press come into play, say molders. Every molder’s overhead is different.

Machine Hour
Rate Basics
 All machine hour costs contain all items from the Machine Cost column. Other molding operations would require optional items from the Optional Costs column. Special molding requirements like a medical cleanroom would require items from the Special Costs column. These variations may change machine hour rate comparisons by 50 to 100 percent.
Building cost
Average hourly rate plus burden
Material cost plus scrap
Secondary Services
Tooling Cost
Robot or picker
Temperature control units
Inspection equipment
Packaging equipment
Special presses:
1. Two-component
2. Coinjection
3. High speed
4. High pressure
Crane for mold changes
Quick mold change equipment
Special screw or nonreturn valve
Special color equipment:
1. Liquid color
2. Dry color
3. Concentrate feed
Inspection or quality control
Engineering support
Tooling support
Material testing
Special equipment for fillers
Conveyors for:
1. Packaging
2. Labeling
Mold storage
Mold maintenance
Seven day/week schedule

Other Factors

Automation plays a big role in many of today’s custom molding plants. However, there may be a misunderstanding of the costs associated with robotics. How extensively robots are used and what functions they perform determine whether costs will be higher or lower.

For example, Brentz explains that you might reduce your direct labor costs with the elimination of a human operator at the press. At the same time, your higher-value technical resources (technicians, engineers, and maintenance people) must be dedicated to a highly automated operation. “Inexperienced molders can underestimate the financial commitment necessary to staff and maintain a highly automated operation,” adds Brentz.

Plante & Moran’s Mengel agrees. “You have less direct labor but more indirect labor costs,” he says.

“Robotics cost money so you can’t use the same rate for a press without robots as with robots,” he explains. “Pick-and-place is one thing, but if you have robots performing complex functions, you need to have a separate hourly rate for the robots.”

Robots increase predictability in manufacturing, but often increase costs as well. Mengel encourages molders to segregate their costs to obtain a more accurate hourly rate.

Given all the variables that exist, not only from job to job, but also from company to company, of what value would identifying industry-wide average machine rates be? None, says Rains.

“Nobody should be concerned with what anyone else is charging for machine hour rates,” Rains says emphatically. “No two operations are identical so no two can have the same machine hour rates.”

Additionally, publishing machine hour rate averages merely simplifies a very complex issue, he says. “If you treat this as a simplistic issue,” adds Rains, “you’ll create problems for a guy who can’t make money with published machine hour rates.”

Still, many say they would find it beneficial to have access to a rate chart showing average hourly machine rates, primarily to see how they stack up against the industry as a whole.

Southwest’s Larson says that there is a perception in the industry that “we’re not in line with the norm.” Molders think “they’re missing the boat by not charging enough or charging too much.”

Larson believes one thing that’s hurting the custom injection molding industry as a whole is molders that give machine time away just to keep the presses busy. “In s

Transitioning a tool, Part 2

Editor's note: Consultant Bill Tobin of WJT Assoc. spends his time helping molders diagnose molding problems. Here is Part 2 in a series of articles about the ins and outs of moving molds.

With the pressure to reduce costs, more and more clients think they can simply tell one molder to "pack up" their mold and ship it to someone else. Molders are doing the same thing to thin the herd -- moving less profitable jobs to subcontractors to open up room for new clients. However, the physical moving must be well thought out and properly executed or the entire experience will be a guaranteed disaster. Moving molds is a lot like square dancing. The moves are simple, the variations are easy, but if a step is missed or someone goes in the wrong direction, everything goes haywire. Here are the steps to take if you're an OEM moving a mold.

Before the Move

  • Do a complete first article inspection on a typical molded part. Sometimes you'll be surprised that the part that works is not to print. Since it does work, change the print.

  • Make sure you have enough inventory to make the transition seamless.

  • Make sure the current molder's accounts have all been paid. Have future orders been canceled, raw material returned or transferred, work in progress completed? If you don't settle the accounts the molder might not ship anything.

  • Prepare the new vendor. Wouldn't it be nice if the new molder had all the materials on hand, had reviewed the prints, and knew what an acceptable part was before the molds showed up?

    All too often the incompetent lead the ignorant in these transfers. The "mold move team" often consists of well meaning folks who buy the parts. However, they usually lack the expertise to say anything more than "the molds look beat up." If this is the best opinion that can come from the people buying parts, hire a consultant who is familiar with these projects, take his advice, and do what he recommends.

    Also there are usually too many bosses in this type of project. Make sure there is only one person who acts legally as the company agent from both the molder and the client. These people make the final decisions. All decisions should be written down before anything happens.

    In Preparation for the Move

  • Make a shipping shopping list. Make sure you know what tools, jigs, fixtures, drawings, gauges, and miscellaneous gadgets are related to this molded part. Be certain that all of these items travel with the mold. Keep in mind that the current molder may assemble using drill presses, sonic welders, and some hand tools that are not part of your tooling program. Unless you can show clear legal title to something like a sonic welder, your new molder better have one or you'll be providing one.

    Sometimes, if you are lucky, you can make arrangements to engage one of the old molder's employees to help in the startup of the project at the new molder. While usually this is wishful thinking, asking isn't stealing.

  • Inspect the mold for damage or undue wear. This is something you must attend to and not inflict on your new molder.

  • Look at the finished part. Are there trim marks? This means someone has been removing flash. The mold needs to be fixed or the molded part price will need readjustment.

  • Get drawings and procedures for everything in a format you can read (hard copy or an electronic media program your folks can read and understand).

  • If possible, get the old tooling contract and maintenance history. The original moldbuilder is usually the best for doing repairs and maintenance as well as telling you what he built under the old purchase order and therefore what you own.

  • Make sure you, not the potential new molder, fully understand how to mold and make the product. Don't expect the old molder to train the new one. You've taken business away and therefore taken money. Enthusiastic cooperation is not going to be particularly high on the previous molder's list of priorities. Get the process on videotape, in digital photos, written down, tape-recorded, or in some other useful format. You'll usually find videotape is easier than written procedures that are rarely followed to the letter.

    Preparing the Mold
    Molds are usually made of steel. Steel plus water and oxygen equals rust (Mother Nature's rule of chemistry).

  • Make sure the waterlines are blown out with compressed air and are completely dry.

  • Bring the mold up to room temperature (if it's from a cold warehouse), open it up, and spray the inside and outside with a good mold preservative.

  • Wrap the mold in heavy-gauge plastic sheeting.

  • Tape the sheeting closed. You now have a waterproof package. Cut holes for the handling holes and install the safety hoist rings.

    Chock, Skid, and Box
    Conventional skids are not robust enough to hold a mold unless it weighs less than 300 lb. If you are shipping small molds, the wrapped molds can be strapped to the skid and shipped by conventional freight carriers.

    If your molds are heavy, a custom-made skid or box is required (see Figure 1). The bottom of the skid is made of the conventional 1-by-4-inch hardwood. The rails are made of 4-by-4s spaced to allow access with a forklift and notched on the sides to afford side lifting. The top of the skid is made of CDX exterior plywood. The final skid should be at least 4 inches on all sides larger than the mold box. Ideally all molds need to be boxed.

    The mold is put on the skid, then strapped to the skid using steel straps anchored by drilling through the bottom and the 4-by-4-inch rails.

    The sides of the box are made by cutting the exterior plywood to size, nailing or screwing 1-by-4 inch plywood around the perimeter of the plywood, and then nailing or screwing the panels together. The top is made the same way, and then nailed or screwed to all four sides. When you build a box of this type, you might consider overbuilding the skid to allow for other boxes to be attached that will hold typical molded parts and spare mold components.

    Chocks are made of 2-by-4s to keep the mold from slipping around on the skid.

    Attaching the Mold to the Skid
    The following first three steps are good enough for short moves without a box. However, long moves or extended storage should use all the steps below.

  • Put the mold in the center of the skid with the safety rings on the top.

  • Nail or screw the chocks to the plywood bottom of the skid around the entire perimeter of the mold.

  • Strap the mold to the skid with steel bands by going around the chocks through drilled holes in the skid, under the bottom, and back to the top.

  • Nail or screw the four sides of the box to the chocks and to themselves.

  • Put the sample parts or whatever else is necessary into the box before installing the top.

  • Nail or screw the top in place.

  • Label with a stencil or dry marker the contents of the box, its owner, destination, and weight.

    Labeling and Storage
    Each skid/mold/component box must be adequately labeled to identify easily what is in it without having to open everything up. Molds that have been properly sprayed, wrapped, and boxed can be stored in a warehouse almost indefinitely. For those wishing an extra measure of security, once the box is closed several good coats of waterproof paint can be added to protect the mold from the elements, or you can wrap it.

    Unless you so specify, there is no guarantee that your molds will travel in a covered truck. A mold that is strapped without protection to a skid on an uncovered truck in a sleet storm in Illinois will certainly show up in Kansas City rusty.

    There is a neat solution to this called "ponchoing." You take the entire skid and staple a heavy-gauge plastic sheet over the top of it. This poncho gives it an extra level of moisture protection. Don't think stretch wrap will do it. Stretch wrap allows capillary action to wick and hold moisture in.

    Things to Keep in Mind
    Molds are a means to an end; they make the parts that make the money. This makes a mold a valuable asset that should be preserved and maintained with the highest care. The method described to preserve and ship may seem expensive, but it is worth it compared to the cost of a rusty mold.

    Finally, talk to your freight carrier and buy insurance for the trip. If the truck tips and the molds end up in a creek, you'll be out many thousands of dollars in lost sales in addition to the replacement cost of the mold.

    When changing molders you will almost always suffer some episodes of "the good old days" syndrome. This happens when the molded parts come out a slightly different color, with slightly different dimensions, or some other subtle variation that happens simply because someone else is manufacturing the part. Unless you can ensure that everything is supplied from the same company, you will see some subtle changes.

    When done properly, moving molds is easy. When not done properly there will be a very testy telephone call from the new molder saying the mold is rusted shut, the cavities have corroded, the parting line is ragged and this-is-not-what-I-quoted. These are the telephone calls that require steady nerves and medication. The proper prescription is to do it right in the first place. I hope this article helps. If not, take two antacid pills and call me in the morning.

    Transitioning a tool, Part 1

    Contact information
    WJT Assoc.
    Louisville, CO
    Bill Tobin
    Phone: (303) 499-3350
    Fax: (303) 499-4116
    E-mail: [email protected]