Modern Plastics Handbook 2011 Part 10 pot

For cost orchemical compatibility reasons, polycarbonate may not be acceptable.That means that the next step up in volume economies for a partwhich is too large for injection molding and

cost differential can cover quite a volume before injection moldingbecomes more competitive However, as product designers create struc-tural foam designs with detail equivalent to injection-molded parts orrequire gas counterpressure molds, that advantage has diminished.When the cost of a structural foam mold is nearly that of an injectionmold, the process will be selected for its unique attributes Such features

as weight reduction, filling of hollows, and the ability to make largerparts (due to the low pressure) become the principal selection criteria.For some parts, there might be one more stop on the pecking order

before going to injection molding, that is, gas-assisted injection

mold-ing This process provides a rigid wall with a hollow channel within its

thick sections and a solid wall elsewhere Molding pressures can be aslow as 15% of injection molding pressures with the resultant increase

in the size of the part which can be molded in a given moldingmachine Gas-assisted injection molding is of particular value in theelimination of assembly operations used to fabricate box structures.There have been successful applications in the elimination of parts inautomobile door frames where the gas has been used to create the boxstructures within the frame

For most parts, the ultimate in production volume economies can bereached with injection molding At a size of about 5 ft2of surface area,however, injection molding drops out of the competition because mold-ing machines which can handle that size are very scarce and are usu-ally reserved for proprietary applications In that case, structural foammolding may be the most efficient process

Unfortunately, structural foam molding can handle a much morelimited palette of materials than injection molding can The materialsthat are readily molded with structural foam are ABS, ionomers, poly-carbonate, polyethylene, and polypropylene Most of the remaindercan only be handled with difficulty and some, like styrene and acrylic,cannot be processed at all The only high-strength and temperaturematerial in the readily processed group is polycarbonate For cost orchemical compatibility reasons, polycarbonate may not be acceptable.That means that the next step up in volume economies for a partwhich is too large for injection molding and which requires higher-temperature resistance than the structural foam materials can inex-pensively offer may be a thermoset process like compression molding

Machining. With the ready availability of most machining equipment

and little or no tooling required, machining is the ideal process for very

low-volume applications, provided the shape is one which can be ily machined Piece part cost will be highest with machining as eachpiece must be processed individually Machining also is the processwhich permits the closest tolerances for extremely critical applications

read-where the cost can be justified The high cost is due to the fact that thework must be done slowly to avoid thermal expansion or meltingwhich would result in imprecision It should be noted that some plas-tics are too flexible to be readily positioned for machining or maydeflect away from the cutting tool when pressure is applied.

Machining is also used in combination with other processes Someprocesses require that gates be removed or parts trimmed by mechan-ical means before they are ready to be used Machining may be used tocreate details the process cannot create An example of this type ofapplication would be holes in thermoformed parts Finally, machiningmay be used because the product’s production volume is too low to war-rant the additional tooling cost necessary to mold the detail into thepart An example of this type of application would be a hole in an injec-tion-molded part parallel to the parting line that would require anexpensive side action The additional amortization cost could exceedthe cost of drilling the hole for low volumes

Many plastic prototypes are machined and fabricated However, thisprocess is only useful to resolve fitment, appearance, and ergonomicissues That is because the strength, internal stresses, and environ-mental behavior characteristics of the part made in this fashion will besignificantly different from the properties of the same part made bythe ultimate production process Furthermore, many plastic parts are

in the strength and stiffness range where the characteristics of theassembly are different than those of the individual parts Therefore,testing should be performed on the completed assembly made withproduction parts Even tests performed on parts made by the proto-typing method for a given process cannot be completely reliablebecause the processing conditions would be different and plastics areprocess-sensitive

Thermoforming and pressure thermoforming. Thermoforming has the tation of being the process to use for applications requiring large partswith thin walls that have low tooling budgets Tooling for thermo-forming is relatively inexpensive because this is an “open-mold”process, meaning that a mold with only one side is used Thus, it is thefirst step up in tooling cost from simple machining Product piece partsare typically less costly for thermoformed parts than for machinedparts However, this is a three-stage process because there are twoadditional stages besides the thermoforming stage First, resin must

repu-be extruded into sheet repu-before it can repu-be thermoformed and, after moforming, the excess sheet (offal) must be trimmed and any holes oropenings cut in the parts These factors can result in fairly high piecepart costs Therefore, as the volume of a given product grows, it mayreach a point where the additional cost of injection molding tooling ismore than offset by the reduction in piece part cost

ther-Thermoforming is a stretching process which permits thinner wall

thicknesses than machining, for which walls must withstand ing forces and heat, or the molding processes that require wall thick-ness thick enough to permit melt flow For products that do not requirethicker walls for strength, this can result in lower piece part prices.There are actually several thermoforming process variations, and theycan be considered according to the thickness of the sheet used (“thingauge” or “heavy gauge” thermoforming), the manner in which it issupplied (roll or sheet), or the contour of die used (male or female).Thin-gauge (thickness less than 0.060 in) thermoforming uses mate-rial supplied in a roll It is the high-volume variety of this process and

machin-is generally associated with packaging With the exception of dmachin-ispos-ables, such as cups and plates, it is not generally used for productmanufacturing

dispos-Heavy-gauge (thickness greater than 0.060 in) material is supplied

in sheet form However, it should be noted that gauges less than 0.060

in are sometimes used in sheet thermoforming This is the variety ofthermoforming that product designers are normally concerned with Itcan produce relatively flat parts with rounded corners in its simplestform, known as “vacuum forming.” However, its “pressure forming” ver-sion can produce detail that can rival injection molding Heavy-gaugethermoforming tends to lose its competitive edge over other processeswhen the part size falls below 1 ft2 Conversely, it has a very large partcapability This allows the combination of several parts into one inmany cases, thus eliminating some assembly operations completely

Pressure thermoforming uses air pressure, often with the addition of

a plug assist, to force the material deeper into the mold cavity Withpressure, the fine detail and surface finishes associated with injectionmolding can be achieved However, injection molding usually requiresmuch higher annual volume than thermoforming to be economicallyfeasible Thus, this development increases the capabilities of the prod-uct designer by extending the range of products which can use suchdetail to those with lower annual volumes Injection molding becomes

a more serious contender for the application as its volume increases.Then the piece part cost differential can be applied to the substantial-

ly greater cost of the injection molds

The cost of the tooling for injection molding rises substantially withincreasing size, and the payoff volume, the point at which the addi-tional tooling cost is offset by piece part savings, goes up accordingly.Thermoforming can, however, make substantially larger parts thancan injection molding

Structural foam molding, gas counterpressure structural foam molding, and tion molding. The high cost of tooling is the factor which governs access

co-injec-to the injection molding process This cost can be reduced through theuse of related processes which require less molding pressure, such asstructural foam molding, gas counterpressure structural foam mold-ing, and co-injection-molding Lower pressure allows for a less sub-stantial, and therefore less costly, mold.

The part produced by structural foam molding is not solid like thepart produced by thermoforming or injection molding Within the out-

er skin, there is a cellular structure with the cell size increasingtoward the center The part must have a wall thickness of 0.187 in inorder for any significant amount of foaming to take place Thus, thestructural foam part may actually require more resin than the equiv-alent part made by either thermoforming or injection molding if thewall thickness must be increased to accommodate the foaming process

In addition, the molding cycle for structural foam molding is muchlonger than that of injection molding due to the time required for gasexpansion For these reasons, the piece part cost will be greater for apart produced by structural foam molding than for a similar part made

by injection molding The reduced molding pressure does permit

larg-er parts to be molded in a given molding machine, at least to the its of the machine platens

lim-Parts made by the structural foam molding process have a teristic swirled surface Gas counterpressure structural foam moldingand co-injection molding are variations of the process which can pro-duce a solid, nonswirled surface Depending on the application, theadditional mold cost for the gas counterpressure feature reduces thetooling savings over the injection molding alternative Coinjectionmolding permits a solid material to be used for the outer skin and afoamed material for the inner structure which can also be a less expen-sive material This process requires sophisticated equipment

charac-These three processes use closed molds and low pressures Theclosed molds are more costly than the open molds used in thermo-forming because there are two halves instead of one However, the lowpressure keeps the tool cost significantly lower than traditional injec-tion molding Unfortunately, it is that high pressure associated withinjection molding that permits its fast cycles Thus, piece part pricesare higher for these methods than they are for injection

As in thermoforming, these processes are most competitive withinjection molding for large parts The larger the part, the greater themold cost advantage over injection molding Unless, of course, onedesigns parts of such intricacy that this advantage is negated

Piece part costs are a different matter As a broad statement, partsfrom these processes will be less costly than those from thermoform-ing, but more costly than those from injection molding However, thereare other reasons for using them besides simple piece part cost

One purpose could be to change the “feel” or “heft” of a product such

as a steering wheel or a vacuum cleaner handle That permits the dle to be made in one piece instead of the older method which consist-

han-ed of two injection-moldhan-ed halves usually assemblhan-ed with screws andnuts These foam molding processes eliminate the need for assemblyand the cost of the fasteners This results in a lower product cost eventhough the injection molding process has a faster molding cycle

Injection molding. While the extrusion process is the highest-volumeprocess, injection molding is used for the greatest number of productdesign applications, largely because it provides the lowest piece partcost for volume applications Injection molding can accommodate arange of applications from huge parts which require a cycle of severalminutes to high-volume bottle caps that have been molded at the rate

of 288 parts every 4 s

Injection molding gets its name from the fact that plastic is injectedinto a mold at very high pressure That gives the process high outputcapability plus the ability to produce fine detail and tight tolerances.However, the high pressure at which the plastic is injected into themold requires sturdy, robust steel molds that are inherently expensivefor that reason alone Added to that is the fact that the process is usedfor the most precise, demanding piece part designs which also requireexpensive core and cavity details Finally, the low piece part cost pro-vided by the injection molding process is often obtained through theuse of multiple cavities Hence, the cost of each cavity is multiplied bythe number of cavities in the mold

Size is probably the major limitation to injection molding As theparts grow larger, the cost of tooling becomes prohibitive for manyapplications and the number of molding machines available that arelarge enough to make them diminishes significantly Many of the verylargest injection molding machines have been manufactured for spe-cial applications, are owned by proprietary manufacturers, and are notavailable for custom molding Gas-assisted injection molding signifi-cantly increases the size of the part which can be molded in a givensize molding machine because of its much lower pressure

5 Resin transfer molding

6 Reaction injection molding

7 Compression molding, bulk molding compound (BMC), sheet ing compound (SMC), low-pressure molding compound (LPMC),transfer molding, and thermoset injection molding

mold-Since many of the thermoset processes are best suited to largeparts, two pecking orders exist in this area as well Up to 1 ft2, thenext step after machining and casting would, depending on part con-figuration, be either compression molding, transfer molding, or injec-tion molding All of these are processes which require expensivetooling, however, the alternatives do not lend themselves to smallparts very well For large parts, the pecking order would include thefull gamut of thermoset processes with the exception of transfer orinjection molding

Machining. As with thermoplastics, thermosets can also be machined Infact, most of the thermosets are somewhat easier to machine than thethermoplastics in that the melting problem is less of a factor While local-ized heat at the machining surface can still create difficulties, the tem-peratures are much higher and charring or burning is the likely result

A considerable amount of machining is done with thermosetsbecause, with the exception of a couple of methods, the mechanicalremoval of molding flash is necessary in thermoset processes Drillingholes and cutting openings is also commonplace in some processesbecause it is difficult, bordering on impossible, to mold them in formany designs

Casting. Casting is a low-pressure closed-mold process; however, moldcosts are kept low because it is often possible to cast the mold directlyfrom the model It is difficult to create fine detail to close toleranceswith this process Casting cycles are long and, consequently, piece partprices are high

Casting is often used to enclose an object, usually an electrical ponent, in order to protect it Casting applications also include furni-ture and decorative objects where fine detail is required or simulation

com-of wood is desired in relatively low volume

Lay-up and spray-up. The largest of parts (mine sweeper hulls) can bemade by lay-up and spray-up However, the ability to create fine detail

is limited and close tolerances are not possible Machining is necessary

to create holes and trim the parts The construction is laminated withpolyester with glass reinforcement, and it is the reinforcement appli-cation method which defines the name of the process Open molds areused; therefore, one side of each part is rough and unfinished Moldcosts are low; however, they only last for a small number of parts Thusmany molds would be required for high volumes, although the patternneeds to be made only once Since these are very slow processes, piece

part costs are high However, robotics have been used to reduce laborcosts.

Cold molding or cold-press molding. This process is a step up from the mold processes previously discussed in that it is a closed-mold process.Therefore, the parts are finished on both sides However, the processdoes not quite produce the surface quality required by the transporta-tion industry Hence, cold-press–molded parts are more likely to be usedfor interior parts To a limited degree, this process can provide a partwith some inside structure such as ribs, etc However, it has poor toler-ance control and, therefore, somewhat limited application

open-Cold-press molds are plastic, which makes them comparatively pensive Relatively long cycle times and postmolding machining result

inex-in parts which are more expensive than compression-molded parts, butless costly than parts produced by the lay-up or spray-up processes

Resin transfer molding (RTM). Resin transfer molding can produce parts ofhigher finish, greater complexity, and wall thickness consistency thancold-press molding; however, the part cost can also be higher.Therefore, the processes are sometimes used together with externalparts made by the resin transfer molding process using cold-press–molded parts for internal supports Closed plastic molds areused which are sometimes plated for longer life and better finish.These molds are considerably less expensive than compression molds,thus considerable volume is required before the additional cost of tool-ing can be offset by the lower piece part cost of compression molding.Parts as large as truck hoods, small boat hulls, and car body halveshave been produced using this process

Reaction injection molding (RIM). RIM is a low-pressure process usingclosed molds However, it has a much shorter cycle than resin transfermolding, which results in lower piece part costs Unfortunately, veryfew materials are available Thermosetting polyurethane is the prin-ciple material available, with epoxy, nylon, and polyester also avail-able but to a limited extent Reinforced reaction injection molding(RRIM) is available with chopped glass used as the reinforcement.This process is most widely used in the automotive field, althoughthere have been other applications where large parts or limited vol-umes are required

Compression molding, BMC, SMC, LPMC, transfer molding, and thermoset injection molding. Compression molding is the primary thermoset process; theother processes in this group, with the exception of thermoset injection

molding, are derived from it The term compression molding also

includes BMC, SMC, and LPMC which actually describe the molding

compound used BMC and SMC refer to the type of reinforcement andthe manner in which the resin is prepared for molding.

Compression molding, transfer molding (not to be confused withresin transfer molding), and thermoset injection molding compete forshort fiber–reinforced or –unreinforced applications with fine detailand close tolerances Molds are expensive, but piece part costs are low.For a product which can be manufactured by all three processes, com-pression molding will have the lowest tooling cost and highest piecepart cost Injection molding will have the lowest production cost with

a higher tooling cost, and transfer molding is somewhere in betweenthem (the use of preheated resin allows transfer molding cycles toapproach injection molding cycles) That is a gross generalizationbecause part design details will likely favor one process or the other inmost cases

In BMC, long strands of reinforcement (14to 12in) are placed in thematerial along with other additives A ball, slab, or log of this mixture

is formed and placed in the mold This method of reinforcement is lessexpensive than SMC, however, it is not as strong SMC fibers arespread into a resin paste to form a sheet Reinforcement fibers canrange from the very smallest to those of indefinite length, althoughthey usually do not exceed 3 in However, SMC is generally known as

a long-fiber process; it is used for higher strength applications such astruck tractor hoods and fenders

LPMC material is prepared like SMC but is formulated to permitmolding at a lower pressure The reduction in pressure results in lessexpensive molds, often constructed of aluminum The additional cost

of tooling requires a high production level for the piece part savings tojustify a change of process from cold-press molding or resin transfermolding to one of the compression processes However, the use ofLPMC can make compression molding competitive at relatively lowlevels

Through their large-part capability, the compression moldingprocesses permit the combination of parts with the resultant assemblysavings They are also processes which offer economies of high volume.Thus, they are the processes of choice for large-part, high-volumeapplications

Hollow part processes. Hollow products can be made either by bling parts made by most of the processes (even extrusions can becapped) or by one of the hollow part processes Which is most costeffective varies considerably according to the application and to thestate of the art of the techniques being applied The savings associ-ated with molding the part in one piece may be offset by the use of

assem-robotics in automating the assembly process coupled with a more effective molding process.

cost-For thermosets, there really is only one hollow part

process—fila-ment winding That process is very limited in shape and structure

such that, except for pressure applications, most thermoset hollowproducts are assembled

There are three hollow part processes for thermoplastics and they dohave a pecking order of sorts When proceeding with the development

of thermoplastic hollow shapes which do not lend themselves well toassembly, the following sequence should be used, starting with thelowest volume requirements:

1 Rotational molding or twin-sheet thermoforming (applicationdependent)

2 Blow molding

Processes for hollow shapes have a much simpler pecking order The

ideal shape for rotational molding is a sphere The ideal shape for

twin-sheet thermoforming is a flat panel The selection of which process to

begin with would be based on the shape of the particular part to bemanufactured, although it should be noted that rotational molding canmake parts with sections as thin as 1 in Additional factors would bethe size of the part (rotational can go larger) and the selected material,since each of these processes favors different materials

Depending on the size and shape of the part, blow molding may

become competitive as the volume grows However, there are size its to which this step can be taken as the other two processes can makelarger parts than blow molding

lim-Rotational molding and blow molding make similar hollow parts Insome cases, such as automotive ducts, the ends are removed from ahollow shape to create the final part Large containers can be molded

so they are integral with their covers which are cut off to form the twoparts Other shapes can also be molded as one hollow piece and cutapart to make multiple parts (usually two) Structural components areusually molded as double-walled parts which can be filled with foamfor greater strength and rigidity

Rotational molding, blow molding, and twin-sheet thermoformingcan also make large, double-walled parts which are relatively flat(such as pallets) Twin-sheet thermoforming is best suited to the flat-test of such parts

Rotational molding. The rotational molding process, sometimes referred to

as rotomolding, can produce parts ranging in size from small balls to

enormous containers The principal material used in rotational molding

is polyethylene, however, some nylon parts are also made For simple

parts with low appearance requirements, inexpensive sheet metalmolds can be used More costly cast molds are used for parts of a high-

er level of finish and complexity Even these molds are comparativelylow in cost as rotational molding is a low-pressure process which doesnot require heavy steel molds Piece part price can be quite high due tothe long cycle time In addition, most openings in the part must bemachined as a secondary operation For small products, such as balls, alarge number of cavities can be used to offset the low output Rotationalmolding is well suited to large parts and, for the largest of parts, it has

no real competition in thermoplastic processes

Twin-sheet thermoforming. Twin-sheet thermoforming is generally used to

produce large, flat, double-walled parts Nearly all of the tics can, at least theoretically, be thermoformed However, the bulk ofparts made by this process are usually polystyrene or ABS The high-est-temperature plastic thermoformed in any significant amount ispolycarbonate

thermoplas-Blow molding. With few exceptions, plastic bottles produced in volume are

blow molded Less well known are the other blow-molded hollow shapes

such as children’s toys, storage sheds, etc Typically, these are walled For high-volume molding of hollow shapes, blow molding is theonly option if the shape and material requirements are suitable

double-Profiles. Extrusion and pultrusion are the only profile processes,although injection molding can compete with extrusion for short piecessuch as rulers Tooling, although made of hardened steel, tends to berelatively inexpensive Neither extrusion nor pultrusion lends itselfwell to short runs due to the lengthy setup required Therefore, there

is no pecking order by volume Instead, the profile processes are cussed according to their applications

dis-Extrusion is the principal process for producing film and sheet It is

also used to create open (weather stripping) and closed (soda straws)profiles Dual extruders can be used to produce laminates, a process

called coextrusion This is often done to provide barriers to ultraviolet

light or moisture vapor Two colors can also be extruded side-by-sidefor interesting effects All thermoplastics can, theoretically, be extrud-

ed including foamed and reinforced materials However, in practice,the highest melt temperature thermoplastic extruded in any volume ispolycarbonate Thermosets can also be extruded on a limited basis asthey require special equipment and are slow to process

Pultrusion is the principal means of producing reinforced profiles It

is more expensive than extrusion and is usually reserved for ing structural applications such as light poles, wind turbine blades,and structural beams There are some consumer applications as well

demand-and they include pole vault poles, fishing rods, flag poles, demand-and skipoles The process is noted for its ability to produce a part with extra-ordinary resilience which can sustain considerable deflection andreturn to its original shape While pultrusion is not limited to ther-moset materials, its principal resin is thermoset polyester.

If there is to be any preference between extrusion and pultrusion, itwould be based on design freedom and strength Extrusion can pro-duce a wider range of shapes and has a wider availability of materials;however, pultrusion can create the stronger part Most applicationsclearly favor one process over the other for these reasons When bothprocesses can make the part, extrusion usually has the edge because

it processes at a higher rate of speed For higher-strength applications,however, that advantage can be offset by the higher cost of reinforcedthermoplastics over the reinforced polyester used in pultrusion

Ultra high strength. Filament winding is a process which stands alone

It is an ultra high-strength process which defines its own applications.Its principle competition is metal fabrication, over which it has signif-icant cost and weight advantages when employed

Filament winding is the process associated with such ultra strength requirements as rocket motor casings, rocket tubes, heli-copter blades, automobile leaf springs, and aircraft parts A variety ofwinding configurations are used to wind resin-impregnated or bath-dipped glass filaments around a mandrel, which is often retained aspart of the end product Using this process, small-diameter pipe rated

high-at 2000 lb/in2 has been created for the chemical industry Toolinginvestment for this process is low to moderate, however, piece partcosts are relatively high

Additional process selection considerations. Material limitations willaffect the process selection because none of the processes will acceptall of the plastics Therefore, the optimum process may not accept thedesired material When this problem occurs, one of the two will need

to be changed Table 8.6 lists the acceptable materials for the pal processes Bear in mind that both resin and equipment manufac-turers are continuously working to enhance their products and a listsuch as this can be made obsolete at any time If the desired processdoes not indicate that the preferred polymer can be used, it may beworthwhile to investigate further

princi-One additional determining factor in the selection of a process is thesize of the part to be manufactured Each process is limited in the size

of parts it can handle Table 8.7 gives some indication of the size rangethat each process is capable of Note that, on the high side, two sizesare indicated The column, “largest known,” lists the largest size the

TABLE 8.6 Plastics Available for Processes

Blow molding ABS, acrylic, cellulosics, nylon, polycarbonate,

polyester (thermoplastic), polyethylene, polypropylene, polystyrene, polysulfone, PVC, SAN Casting Acrylic (thermoset), alkyd, epoxy, nylon, phenolic,

polyester (thermoset), polyurethane (thermoset), silicone

Cold molding Epoxy, phenolic, polyester (thermoset),

polyurethane (thermoset) Compression molding

(including BMC and SMC) Alkyd, allyl, amino, epoxy, fluorocarbons, phenolic,

polyester (thermoset), polyurethane (thermoset), silicone

Extrusion ABS, acetal, acrylic, cellulosics, liquid crystal

polymer, nylon, polycarbonate, polyester (thermoplastic), polyethylene, polyphenylene oxide, polypropylene, polystyrene, polysulfone,

polyurethane (thermoplastic), PVC, SAN Filament winding Epoxy, polyester (thermoset)

Gas-assisted injection molding ABS, acetal, acrylic, cellulosics, nylon,

polycarbonate, polyester (thermoplastic), polyethylene, polyphenylene oxide, polypropylene, polystyrene, PVC, SAN

Injection molding ABS, acetal, acrylic, alkyd,* allyl,* amino,*

cellulosics, epoxy,* fluorocarbons,* liquid crystal polymer, nylon, phenolic,* polycarbonate, polyester (thermoplastic), polyester (thermoset),*

polyethylene, polyphenylene oxide, polypropylene, polystyrene, polysulfone, polyurethane

(thermoplastic), PVC, SAN Lay-up and spray-up Epoxy, polyester (thermoset)

Pultrusion Epoxy, polyester (thermoset), silicone

Reaction injection molding Nylon, polyurethane (thermoset), epoxy, polyester Resin transfer molding Epoxy, polyester (thermoset), silicone

Rotational molding Acetal, acrylic, cellulosics, fluorocarbons, nylon,

polyester (thermoplastic), polyethylene, polypropylene, polystyrene, polyurethane (thermoplastic), PVC

Structural foam molding ABS, acetal, nylon, polycarbonate, polyethylene,

polyphenylene oxide, polypropylene, polystyrene, polysulfone, SAN

Thermoforming ABS, acrylic, cellulosics, polycarbonate,

polyethylene, polypropylene, polystyrene, polysulfone, PVC, SAN

Transfer molding Alkyd, allyl, amino, fluorocarbons, phenolic,

polyester (thermoset), polyurethane (thermoset), silicone

*Special equipment required.

Blow molding 3 ⁄ 8 in deep  1 1 ⁄ 4 in long 9 1 ⁄ 2 ft long  1 ft deep 1320-gallon tank

 4 in thick,

28 in deep  44 in long Casting No limit Limited only by physical Limited only by

ability to handle molds physical ability and moldments to handle molds and moldments Coinjection molding 1 ⁄ 4  1 ⁄ 4  1 ⁄ 4 in 2  5  5 ft 2 1 ⁄ 2  4  10 ft

Cold molding 1 1

2  1 1 ⁄ 2 ft 10  10  1 1 ⁄ 2 ft 14-ft boat hull Compression molding 1 ⁄ 4  1 ⁄ 4  1 ⁄ 16 in 4  5  8 ft 1 1 ⁄ 2  4 1 ⁄ 2  14 ft

Filament winding 4 ft deep  8 in long 13 ft deep  60 ft long 10-ft high  82 1 ⁄ 2 ft deep

Injection molding 0.008  0.020  0.020 in 2 1 ⁄ 2 deep  3 ft 4 ft  4ft 6 in  7 ft

Lay-up and spray-up 1 ⁄ 4  6  6 in 150-ft minesweeper Continuous roadway

Machining No limit 10 ft wide or 15 in deep Limited by size of stock available

Reaction injection 4  12 in 3  4  10 ft 10  10 ft

molding

Resin transfer molding 1 in  3 in  2 ft 16 in  4 ft  8 ft 4 ft  8 ft  28 in

Rotational molding 1 ⁄ 2 -in -diam sphere 6 ft deep  18 ft long 12 ft deep  30 ft long

author has ever heard of That size is not realistic for general use asnot all materials can be run at the limits for a given process and theequipment that made it may not be available for custom projects.Indeed, such equipment has typically been built for some company’sproprietary needs The column, “largest commercial,” lists sizes atwhich a reasonable selection of competing bids might be obtained.However, when working near the limits of a given process, it behoovesthe savvy product designer to be certain that the equipment is avail-able for the project before proceeding with the design.

8.1.6 Tooling selection

Tooling is a critical aspect of the total plastic product design picture

because a plastic part can be no better than the tooling that created it.While it is not necessary for the product designer to be able to design

a tool, a fundamental knowledge of tooling is essential—not merely forthe design of the part, but for the process selection as well The cost oftooling for a process and the type of design detail it is capable of pro-ducing are among the principal determining factors in the selection ofthe process These two factors are determined by the amount of pres-sure required for the process and the amount of tooling that has to bebuilt For the purposes of the tooling discussion, the processes havebeen broken up according to the amount of pressure required with theopen-mold and profile processes discussed separately

Basic tooling construction. In general, the open-mold processes can beexpected to have lower tooling costs than any of the closed-moldprocesses for parts of equivalent complexity because only one side of amold has to be built Other types of molding require both a male andfemale half (core and cavity) or two female halves (split cavity for hol-low parts); however, open molds require either a male or a female sec-tion This is illustrated in Fig 8.8, which depicts a female mold which

forms the outside of the part as in Fig 8.8a and a male mold which forms the inside of the part as in Fig 8.8b.

Open-mold processes have little or no molding pressure In general,the tooling costs for the closed-mold processes increase according tothe amount of molding pressure required for the process The greaterthe pressure in the tool, the stronger (and more expensive) it must bebecause higher pressures require stronger molds which need to bemade from stronger materials that cost more and require more time tomachine However, the greater pressure permits the molding of finerdetail and results in shorter molding cycles Therefore, it is generallyaxiomatic that the piece part price decreases as the cost of toolingincreases

Closed-molds consist of either a core and a cavity as illustrated in

the injection mold in Fig 8.9a or, in the case of hollow processes (blow

or rotational molding), a split cavity as shown in Fig 8.9b Where the two (or more) parts of the mold meet is called the parting line To pre-

vent plastic from leaking from the mold, the two halves must matchperfectly When the parting line is flat and on one plane, it is relative-

ly easy to align the two halves However, some designs require the

mold to have a contoured parting line This is referred to as a broken

parting line and requires the halves of the mold to be machined with

great care Thus, a mold with a broken parting line is considerablymore expensive than one with a straight parting line

Figure 8.8 Tooling for open-mold processes: (a) female mold

and (b) male mold.

Openings in the part are created when a portion of one half of the moldcloses against the other half such that plastic cannot flow through thatarea When the hole is to be in a plane parallel to the parting line, this is

called a shut-off It is normally the core which shuts off against the ity (see Fig 8.9a) because the shrinkage causes the part to grip to the

cav-sides of the shut-off and it is preferable for that to occur on the core side

of the mold For odd-shaped or very large openings, the raised portion isusually cut in the solid core, however, round holes are most economically

made from a core pin as illustrated in Fig 8.9a.

Shutoff Undercut

Parting line Ejector pin (knock out)

Core pin

Note in this illustration that there is a hole in the side of the part.The core that makes that hole would interfere with the ejection of thepart from the mold Therefore, that core is placed in a removable sec-tion which has to be replaced after every cycle This procedure is tootime consuming for high-speed production, and mechanical devices,such as air cylinders and cams, which are activated by the relativemotion between the two mold halves, are used to achieve higher pro-

duction rates These mechanisms are referred to as slides or side

actions and they can also be used to operate the halves of a split

cavi-ty when it is used for injection molding

On the wall opposite the one with the side hole in Fig 8.9a, there is

an indentation on the core side of the wall that does not protrude all

the way through the wall of the part That indent, known as an

under-cut, would also interfere with ejection of the part from the core.

However, when the undercut is shallow, its edges are adequatelyradiused and the material is sufficiently flexible, an undercut can bestripped from the mold once the cavity has been removed If the under-cut is too deep to be stripped, removable core inserts or a collapsingcore will be required Undercuts can be placed in the cavity as well;however, a split cavity, slides, or removable core inserts will berequired if the undercut is too deep to be stripped Cavity undercutscan be deeper than core undercuts by the amount of the shrinkagesince the part shrinks away from the cavity wall Removable coreinserts add a considerable amount of time to the molding cycle; splitcavities or slides add substantial cost to the mold; and collapsing coresare, generally, expensive to tool when they are feasible at all

Figure 8.9a also shows some holes which are referred to as

cool-ing/heating lines Thermoplastic processes use cooled molds for most

polymers (some require heated molds) to speed set up of the moldment.Therefore, cooling channels are drilled into the mold to allow coolwater (sometimes chilled) to run through the mold The objective is toachieve a cool mold at the precise, same temperature across its surface

so the moldment can cool uniformly and avoid distortion This ideal isdifficult to achieve, but, depending on the contours of the design, veryhigh levels have been attained Often, elaborate cooling systems areemployed to do this and that adds cost to the mold Frequently, the dif-ference in cost between sources can be attributed largely to theamount of cooling they have included in their quote

Thermosets do not have to be cooled because they cross-link duringmolding Consequently, they are run hot and the channels are used forheating media like steam or oil

Tooling for the open-mold processes. The open-mold processes arethermoforming, lay-up, and spray-up Since open molds form only one

side of the part, they permit design details only on that side Toolingfor the open-mold processes tends to be very simple Cooling systemsare employed for thermoforming and undercuts are readily createdusing break-away or removable sections which come out of the moldwith the part and which are replaced before the next cycle In high-vol-ume applications, springs or air cylinders can be used to actuate thesesections Mechanical ejection is sometimes employed using springs ormechanical devices when there is an undercut to be stripped.

In general, any openings need to be cut as secondary operationssince there is no way to close off the mold faces to create holes withinthe part For the lay-up and spray-up processes, a very large openinglike the center of a window frame could be included in the shape of themold Openings which need to be cut may call for additional tooling fix-tures, but these are relatively inexpensive Because they are sec-ondary operations which must be performed after the moldment iscreated, openings do add to the cost of the part Fine detail, whichrequires pressure to create, cannot be part of the design because themolding pressure is nearly nonexistent for the open-mold processes

There is, however, a variety of thermoforming, known as pressure

ther-moforming, that can produce fine detail using special thermoforming

equipment

Tooling for the low- to moderate-pressure processes. This category ers those processes that require closed molds which do not have towithstand high molding pressures that need hardened steel tooling.Processes that fall into this category are casting, cold-press molding,resin transfer molding (RTM), reaction injection molding (RIM), low-pressure compression molding (LPMC), plus the hollow part process-

cov-es, rotational molding, and blow molding The lower molding pressuresused for these processes permit the use of a variety of techniques formold construction, most of which involve machining a softer metal orcasting from a pattern A softer metal, such as aluminum, machinesmuch faster than hardened steel and reduces the cost of the tool.The lowest-cost tools are made by creating a pattern of the part andcasting the tool from that pattern, usually in epoxy and glass The pat-terns are often made from wood, however, they can be made from a

computer-driven method, such as stereolithography, providing the

con-figuration and size are suitable Such tools tend to wear out within alimited number of parts, in which case a new mold is cast from the pat-tern The number of parts that a cast mold can produce before it must

be replaced varies from a dozen to several thousand depending on theprocess and the complexity of the design In general, the more complexthe part design and the higher the molding pressure, the shorter thelife of the mold Complex designs require fine detail which is more

susceptible to damage and higher pressure is more likely to cause suchdamage In some cases, epoxy molds can be nickel-plated to provide abetter finish and prolong the life of the mold The most common mate-rials used for the tooling of the various plastics processes are indicat-

ed in Table 8.8

The materials used for most of the low- to moderate-pressureprocess tooling do not lend themselves well to mechanical removal ofcores Therefore, holes are machined as secondary operations or coresare manually operated, removed, and emplaced by hand This proce-dure takes more time than mechanically operated cores However,since these are not generally high-speed processes, the additional time

is not normally a problem

Tooling for the high-pressure processes. Injection and compressionmolding develop the highest molding pressures, and they require hard-ened steel molds for large-scale production However, the types of moldconstruction discussed for the low- to moderate-pressure processes areoften used for prototypes and short runs for these processes It is impor-tant to note that these methods have significant limitations when usedfor processes which develop higher pressures That is because engi-neers are accustomed to incorporating the fine detail and surface finishthese processes are capable of into their designs without taking intoconsideration that these prototype and short-run molds cannot with-stand the pressures necessary to create them for very many shots Justhow many shots a mold will withstand is design dependent.Nonetheless, it is obvious that the harder the mold material is, thelonger the tool will last without significant repair or replacement

TABLE 8.8 Mold Materials for Plastics Process

Lay-up and spray-up Epoxy (sometimes plated), aluminum, steel

(occasionally)

RIM Epoxy, glass, aluminum, steel, kirksite

Compression and transfer Steel (SMC, BMC), aluminum (LPMC)

Rotational molding Sheet steel, stainless steel, aluminum, cast aluminum,

machined metal, electroformed nickel and copper Structural foam molding Prehardened steel, aluminum with beryllium copper Thermoforming Aluminum (wood, epoxy, or polyester for prototypes) Blow molding Aluminum, beryllium copper, zinc alloy, brass,

When necessary, inserts made of a harder material, such as steel, may

be used

Molds made for prototyping and short runs often have other costsaving simplifications such as manual cores, manual ejectors, and lim-ited, or nonexistent, cooling In general, the molding cycle is largelydetermined by the length of time required to cool the part enough for

it to be rigid enough to be ejected from the mold Therefore, a significantdifference in the mold construction which affects these parameters willaffect the strength and shape of the part As a consequence, it is nec-essary to construct a single cavity with the identical cooling, coring,and ejection of the production mold if a precise prototype of the pro-duction part is required This is often done when a large, expensive,multicavity mold is to be built

Tooling for the profile processes. Extrusion and pultrusion, the profileprocesses, require steel dies because they build up back pressuresbehind the die However, the tooling costs for these processes are rela-tively low because the shapes manufactured with them are normallycomparatively simple

8.2 Design Fundamentals for Plastic Parts

The basic engineering formulas for structural design can be applied toplastic part design, within the limitations of the data available for thematerial properties, as previously discussed This information is wide-

ly available and is not particular to plastic part design Therefore, it

will not be covered in this chapter It is, however, found in Machinery’s

Handbook, 24th Edition, (Industrial Press Inc., New York) as well as

in some of the other references listed at the end of this chapter

A plastic part is like a chain in that it will fail first at its weakestlink Repair that, and it will fail at its next weakest link and so on.Therefore, the objective in plastic part design is to see to it that noweak point exists such that the part will fail below its design limit.This requires careful attention to a number of critical areas

The majority of differences between the design of plastic parts and thedesign of parts made of other materials are in some way heat or pres-sure related Heat and pressure are used to create the parts and theireffects on the part plus the effects of the subsequent cooling of the parts

to room temperature require consideration in the design phase

8.2.1 Cooling effects and the need for

uniform wall thicknesses

Most of the processes require the plastic to be heated to elevated peratures When the part cools, it shrinks away from the cavity and

tem-onto the core, if there is one Clearly, the thicker the nominal wall ofthe plastic part, the longer it will take to cool When the thickness ofthe wall varies within the part, the thinner portion of the moldmentsolidifies before the thicker segment Nonuniform cooling leads toproblems of voids, warpage, and sink, as illustrated in Fig 8.10.The extent of these problems varies considerably with the processand material (worse in higher shrinking materials) and is affected bythe use of fillers in the resin Consequently, plastic designers need tomake a concerted effort to keep the wall thickness uniform throughoutthe part In corners, this is accomplished by making the outside radius(OR) equal to the inside radius plus the wall thickness (IR  W) When

the wall thickness must vary, the transition must be gradual as

illus-trated in Fig 8.11 Every increment of wall thickness variation (T) should take place over a distance at least 3 times as great (3T) In no case should the thicker section exceed the thinner section (W) by more than 25% (1.25W) Not only will there be increased internal stresses,

but there will likely be a sink with most unfilled resins When a cal edge is required, perhaps to position the part indicated by thedashed lines on the left side of Fig 8.11, a rib can be used in place of

verti-a lverti-arge block of mverti-ateriverti-al

Excessive variation in wall thickness is not the only cause of form cooling of the part It can also result from a mold which is notadequately cooled Ideally, the mold temperature would be maintained

nonuni-at a constant tempernonuni-ature across the molding surface This ideal is tually impossible to attain because the shape of the part usually lim-its the mold designer’s freedom to place water lines where they canbest perform their function An example of this type of situation is abox configuration The core side of the box corner receives heat fromthree directions (both sides and the top), whereas the core which formsthe side of the box is heated only by the melt on that side For there to

vir-be no temperature differential vir-between those two parts of the core, thecorner must receive more coolant than the side This could require avery expensive cooling system which the budget cannot support

Figure 8.10 Voids, warpage, and sink in plastic parts (Source: Jordan I Rotheiser,

Joining of Plastics Handbook for Designers and Engineers, Hanser Publishers, Hanser/Gardner Publications, Inc., Cincinnati, 1999.)

Munich-Furthermore, the cooling lines must avoid the ejection system as well.The mold designer must often choose between an ejector and a coolingline Fewer ejectors creates more pressure per ejector which can alsoresult in distorted parts or an extended molding cycle.

The shape of the part will also dictate gate location Ideally, thegates should be located where they permit the cavity to be filled uni-formly If one part of the cavity fills before the rest, it will begin to coolimmediately, resulting in differential cooling through the part Thisphenomenon can also result in premature gate freeze-off which pre-vents adequate packing of the part Improper gate location or inade-quate number of gates are other ways to cause nonuniform cooling andpart distortion

8.2.2 Shrinkage and the use of draft in

plastic parts

The shrink rate is the amount the part will reduce in size as it cools

from processing temperature to room temperature It is described interms as the amount of shrinkage per inch of part size Thus, a piecepart 1 in long made of a material with a shrinkage rate of 0.005 in/inwill require a mold 1.005 in long if the final part is to wind up a true

1 in in length Mold dimensions for other sizes would be determined byusing 1.005 as a multiplier Thus, the mold for a 1.500-in dimensionwould be determined by multiplying 1.500 by 1.005 (1.508 in).Normally, the mold designer performs these computations and this isnot a concern for the part designer However, the part designer canhelp reduce mold costs by specifying part dimensions which take theshrinkage rate into consideration Thus, a part specification of a 0.995-

in diameter, instead of a 1.000-in diameter, would permit the use of astandard 1.000-in-diameter core pin which would require significantlyless work for the moldmaker than the larger diameter The savingsmight not be great for one cavity, but could be considerable for a mul-ticavity mold

Figure 8.11 Nominal wall thicknesses in plastic parts (Source:

Jordan I Rotheiser, Joining of Plastics Handbook for Designers and

Engineers, Hanser Publishers, Munich-Hanser/Gardner

Publi-cations, Inc., Cincinnati, 1999.)

The shrink rate for plastics as found on the specifications supplied

by the resin manufacturer is often indicated as a range (0.004 to 0.006in/in) That is because the rate of shrinkage will vary according to thethickness of the wall, with thicker walls shrinking at a higher ratethan thinner walls Depending on the material, this shrinkage canrange from very slight (0.001 to 0.002 in/in) to as much as 0.050 in/in.The shrinkage rate for some resins is uniform both in the direction offlow from the gate and in the transverse direction (isotropic shrink-age) The shrinkage for other resins varies according to the direction

of flow of the resin in the cavity (anisotropic shrinkage) In most

cas-es, it is greater in the transverse direction

The use of fillers can significantly alter both the amount of age and its direction For example, glass fibers align in the flow direc-tion which causes lower shrinkage in that direction However, there islittle or no reduction in shrinkage in the transverse direction Theresult is in an increase in differential shrinkage and a greater tenden-

shrink-cy to warp

In the injection molding process, the shrinkage will be affected by thedensity of the plastic in the part which can be controlled by “packing”more resin into the mold The processor is the person most familiar withthe behavior of a given resin in the equipment and it is the processorwho determines what shrink factors will be employed in the construc-tion of the mold, often after consulting the resin manufacturer

Compression molds, which are not cooled, operate at elevated peratures which cause the molds to expand If this expansion isgreater than the shrinkage of the material, the mold may be actuallybuilt smaller than the intended piece part to compensate for thisexpansion

tem-When the part shrinks onto a core, it grips the core very tightly andcannot be removed without considerable force That force will tend todistort the part; therefore, the cycle must be extended long enough forthe part to be removed from the mold without damage The force can

be reduced by placing a slight angle on the walls of the part dicular to the parting line, as indicated in Fig 8.12 This angle is

perpen-referred to as draft and is common to all molded plastic parts.

In the past, we followed the general rule that 1° per side of draft isideal and 12° per side is a bare minimum However, economic pres-sures require ever shorter cycles, and improvements in equipment,materials, and tooling allow this demand to be met However, in order

to permit parts to be removed from the mold in less time, they must beejected at higher temperatures when they are softer Therefore,greater draft is required Thus, the use of 2° and 3° per side draftshave become commonplace For deep parts (beyond 6 in), a 3° per sidedraft is definitely called for and more would be better

Note that there are two ribs in Fig 8.12 When the part shrinks,multiple ribs create a locking effect which causes the part to grip thecore with greater force Such ribs need to be well drafted to avoid thisphenomenon When multiple ribs are external to the part, they can

cause the part to adhere to the cavity—an occurrence known as a

cav-ity hang-up Since most molds are built with their ejection

mecha-nisms in the core side, a cavity hang-up leaves the molder with noconvenient way to remove the part from the mold The usual remedyfor this problem is to add tiny undercuts to the core side until the cav-ity hang-up is eliminated However, this changes the design and addsadditional stresses to the part

When texturing is added to the surface, the draft must be increasedbecause, microscopically, the texture is composed of thousands of tinycrevices which must clear each other on removal from the mold toavoid scrape marks The amount of draft required varies a bit depend-ing on the texture, however, the general rule for textures is 112° perside plus 112° per side for each 0.001 in of texture depth In tight situ-ations, this can be reduced to 1° per side plus 1° per side for each 0.001

in of texture depth for some applications

8.2.3 Stress concentrations in inside corners

Plastics are particularly vulnerable to the concentration of stresses oninside corners as most of them tend to be notch sensitive The graph in

Parting line Drafted walls

Figure 8.12 Draft in plastic molds.

Fig 8.13 illustrates the manner in which stress concentrates in inside

corners by plotting the ratio of the inside radius (R) to the wall ness (WT) against the stress concentration factor At ratios below 0.5,

thick-the stress concentration factor rises dramatically At thick-the ratio of 0.25,

it has reached 2.25 That should be regarded as an absolute minimum.Below 0.25, the stress concentration factor reaches toward astronomi-cal levels Even a little radius is better than none at all In addition tothe reduction in stress concentration factor, radiused corners alsoimprove the flow of plastic in the mold resulting in a more uniformmelt and a shorter molding cycle

8.2.4 Rib and post design

The elimination of sharp inside radii is important to the design of ribs

as well For free-standing ribs, that raises a paradox in that it becomesdifficult to provide generous radii at the base of the rib and still main-tain the nominal wall rule of not increasing the wall thickness by more

than 25% The effect of changing the inside radii from 0.25W to 0.5W

is demonstrated in Fig 8.14 The 25% increase in wall thickness is

represented by the circle which is 1.25W In Fig 8.14a, the radii at the base of the rib are 0.25W, and they are 0.5W in Fig 8.14b Note how much thicker the rib is in Fig 8.14a (Y1) than it is in Fig 8.14b (Y2).

There is a danger that the tip of the rib will become too thin whendraft is taken into consideration; 0.040 in should be regarded as a

Figure 8.13 Relationship of inside corner radius to the stress concentration factor.

(Source: Jordan I Rotheiser, Joining of Plastics Handbook for Designers and Engineers,

Hanser Publishers, Munich-Hanser/Gardner Publications, Inc., Cincinnati, 1999.)

minimum If the rib is used for locating another part or is lightly

loaded, the use of inside corner radii in the 0.25W range is perfectly

adequate However, for load-bearing ribs, the greater the inside

radius, the safer the design The point of diminishing return is 0.8W.

There is a point of diminishing returns for the height of the rib as

well—3 times the wall thickness (3W) For corner gussets, that height

is twice the wall thickness (2W), as shown in Fig 8.15 The corner

angle is effective at 30° from the vertical If necessary, corner gussets

and ribs can be as close to each other as twice the wall thickness (2W),

however, that may be excessive for many applications considering theadditional gripping force to the core that results External ribs andgussets require a generous draft to prevent them from causing thepart to stick in the cavity when the mold opens

Posts are designed with the same cross section as a rib and use thesame design parameters However, free-standing posts are difficult to

Figure 8.14 Standing rib design: (a) R  0.25W and (b) R  0.5W (Source: Jordan I.

Rotheiser, Joining of Plastics Handbook for Designers and Engineers, Hanser Publishers, Munich-Hanser/Gardner Publications, Inc., Cincinnati, 1999.)

Figure 8.15 Design of corner

gussets (Source: Jordan I.

Rotheiser, Joining of Plastics Handbook for Designers and Engineers, Hanser Publishers, Munich-Hanser/Gardner Publications, Inc., Cincinnati, 1999.)

cool uniformly and have no support Therefore, they have a strong dency to mold distorted Gussets will improve this condition A bettersolution, however, is to increase the diameter, hollow out the center,and create a small boss That configuration is stronger, more control-lable, and easier to design without sinks and molded-in stress Bossdesign details are discussed in Sec 8.2.7.

ten-Certain processes, such as blow molding, lay-up, rotational molding,spray-up, and thermoforming, cannot create standing ribs For theseprocesses, the box rib illustrated in Fig 8.16 can be used For most

applications, the width of the rib (a) can be 4 times the nominal wall thickness (4W), and the height of the rib (b) can be 4 to 5 times the

nominal wall thickness Inside corner design criteria (c, d) are the same

as for standing ribs Some large parts are difficult to adequately force with integral ribbing which will not create sinks even when pro-

rein-duced by a process which can create them Box ribs (known as hatbox ribs) can be manufactured independently and attached to the large

part, usually with adhesives

8.2.5 Gas traps

The interior of a closed mold is not empty before it is filled with plastic.The space is occupied by a gas, air to be precise The plastic displaces thegas, chasing the gas before it as it fills the mold Certain design configu-rations, such as the free-standing rib shown in Fig 8.17, create a trapwith nowhere for the gas to escape Such a design detail is referred to as

a gas trap and the compression of the gas into it causes the gas to burn.

The moldmaker can place very small vents (on the order of 0.0005 indeep by 0.25 to 0.50 in wide) at the parting line for the gas to escape themold These are often added after the first molding trials, which revealthe gas traps A better design is also illustrated in Fig 8.17 where the rib

is extended until it attaches to the outer wall of the part where the gas

is able to escape Plastic part designers must be aware of the possibility

of creating gas traps and avoid such designs Another solution is

illus-trated in Fig 8.17b, that is, the use of an ejector pin pad on a rib The

a

b

c d

w

Figure 8.16 Box rib design.

vent is then ground in the side of the ejector pin The constant action ofthe ejector pin keeps the vent clean.

The problem of gas traps also exists in posts and bosses In the case

of a post which is trapping gas, the usual solution is to place an tor pin on the post similar to the one used for the ejector pin pad onthe rib Bosses use a core pin that is stationary to create the hole.While the core pin can be modified to create a vent, that vent will bestationary and can plug with debris in time That can require a moreexpensive solution: the use of an ejector sleeve around the core pinwhich then has to be fixed to the bottom plate of the mold base

ejec-8.2.6 Knit or weld lines

Knit lines, also known as weld lines, occur in parts made of the

processes in which the plastic fills out a mold (cold press, compression,

Attached rib

Gas trap on free standing rib

injection, reaction injection, resin transfer, structural foam, and fer molding), except when the shape is very simple and there is anunobstructed flow path Cores, variations in wall thickness, depres-sions, changes in flow direction, etc., cause the melt to divide (oftenrepeatedly) as it moves through the mold Where the melt flows rejoin,knit lines are created Knit lines are formed because the leading edge

trans-of each flow cools somewhat as it moves through the mold and forms apartially solidified skin When the edges rejoin, the pressure of themelt bursts through the skin and intermingles with the melt from thejoining flow to create a weld Unfortunately, the weld does not extendthrough the total wall thickness at full strength, even if there is no vis-ible knit line

In some cases, there is no weld at all—a condition known as an open

knit line and which is uniformly regarded as an unacceptable part.

Knit lines which are visible, but not open, can vary considerably instrength from 10% to approximately 75% of that of the surroundingmaterial; the harder to see, the better the knit line In the author’sexperience, the maximum knit line strength was 85% of that of thesurrounding material, with 50 to 65% being typical Since they are theweakest link in the wall of a part, knit lines are one of the most com-mon causes of plastic part failure

Adjustments to molding conditions can improve the quality of theknit line However, since it is difficult to guarantee the strength of aknit line, it is obviously undesirable to have one occur in a highlystressed section of the part Depending on the process, relocation of thegate or the charge can alter the location of the knit line Ribs, whosesole purpose is to provide a channel for molten material to speed to thedesired area (path of least resistance) or flow interrupting depressionscan also be used to relocate a knit line The use of multiple gates willchange the flow within the mold as well, however, an additional knitline is added for each new gate Computer simulations are availablewhich can forecast the locations of knit lines with reasonably goodaccuracy

Openings in the part that are not to be cut into it require the use ofcores When the melt encounters a core, it divides and passes around

it Therefore, there is always a knit line around a hole or boss on theside opposite from the direction of flow For relatively simple parts,that would be the direction of the gate It is also difficult for fillers toreach these same areas and they may contain little or no filler—a con-

dition known as resin rich If knit lines or lack of filler result in a loss

of strength beyond what the application can withstand, it may beadvisable to cut the holes instead, even though the cost is higher Withthe proper equipment, machining the holes can also be more precisethan molding them

8.2.7 Holes and Bosses

For those processes in which the plastic fills out a mold (cold press,compression, injection, reaction injection, resin transfer, structuralfoam, and transfer molding), openings in the part can be molded intothe part through the use of cores Holes in plastic parts are normallycut with a router (or drilled if round) for the remaining processes(although large openings can be molded into parts to be rotational orblow molded in some cases) When holes are created with cores, therewill always be a knit line on one side If the hole is for a fasteningdevice, such as a screw, it is important to locate the knit line awayfrom the side of the hole that the force is applied to

When a boss is used, it will be formed by the melt running both intothe cavity and around the sides simultaneously until it fills the cavitycompletely Figure 8.18 depicts a partially formed boss in which the

melt has not completely closed around the core (known as a short shot).

Clearly, there will be a knit line along one side of the boss Every bosswill have such a knit line and it is rare that a boss fails other than atits knit line

There are two criteria for the design of bosses: the appearance ria and the strength criteria The appearance criteria is illustrated in

crite-Fig 8.19a In this situation the 25% maximum increase in wall ness rule has been applied (D  1.25W) in order to reduce the molded-

thick-in stress and avoid sthick-ink When the draft is taken thick-into account, the wallthickness at the open end of the boss can become quite thin It is gen-erally unwise to permit it to become thinner than 0.040 in Bosses ofthis type are usually used for location or as stand-offs with very little

load, therefore, the inside radii can be 0.25W, which will result in a boss wall thickness of approximately 0.5W when the rule is applied The boss in Fig 8.19a will not supply the strength necessary for

structural applications like threaded metal inserts and self-tappingscrews Structural applications require much stronger boss walls like

those (f ) in Fig 8.19b As the circle in the lower left corner illustrates,

the 25% maximum wall increase rule is exceeded by a considerable

Figure 8.18 The formation of a

molded boss (Source: Jordan I.

Rotheiser, Joining of Plastics Handbook for Designers and Engineers, Hanser Publishers, Munich-Hanser/Gardner Publi- cations, Inc., Cincinnati, 1999.)

amount This will result in a sink (e), indicated by the dashed line,

beneath each wall at the base of the boss The injection moldingprocess will result in the greatest amount of sink, however fillers can

be used to reduce it Foaming will also result in a reduction in sink,but there is a significant drop in physical properties when foam is usedand the surface appearance is affected Surface textures, labels, andsome hot stamping patterns will also tend to hide sink The other fill-ing processes demonstrate sink to a lesser extent

Bosses for self-tapping screws. The inside diameter of the boss for tapping screws is dependent on the size and type of the screw Threadforming screws are generally used for plastics with a flexural modulusbelow 400,000 lb/in2, however they can be used for stiffer plastics insome cases They are preferable to thread cutting screws because they

self-do not break the skin of the moldment However, they can creategreater stress in the boss which may be a problem for some applica-tions Thread cutting screws can be used for plastics whose flexuralmodulus exceeds 200,000 lb/in2 The inside diameter for a thread form-ing screw boss should allow for engagement of 70 to 90% (with consid-eration for draft) of the thread depth The inside diameter for a threadcutting screw boss should permit engagement of 50 to 70% of thethread depth The generally recommended screws for plastics are type

Figure 8.19 Molded boss configurations: (a) appearance criteria and (b) strength criteria (Source: Jordan I Rotheiser, Joining of Plastics Handbook for Designers and Engineers,

Hanser Publishers, Munich-Hanser/Gardner Publications, Inc., Cincinnati, 1999.)

AB for thread forming screws, type BF or BT/25 for thread cuttingscrews These screws provide for more plastic between each metalthread than most of the other types Specially designed screws forplastics are also available.

Once the inside diameter of the boss is established, the remainingdimensions can be readily determined The length of thread engage-ment should be 2 to 3 times the screw diameter and the outside diam-

eter (a) of the boss should be 2 to 3 times the inside diameter (The outside draft is added to that.) The lead diameter (b) should just clear the outside diameter of the screw and the depth (c) should be enough

to permit one full turn of thread For screws 0.250 in in depth, the side gussets should be used The gusset design can follow the same cri-teria indicated in Sec 8.2.3

out-The two most common modes of failure for self-tapping screw bossesare boss cracking and low stripping torque Poor weld lines are an obvi-ous cause of boss cracking; however, too small a hole diameter, too large

a screw diameter, too great a flexural modulus, and too long a threadengagement are also causes Unfortunately, the remedy for one mode offailure may be the cause of the other as low stripping torque can be theresult of too large a hole size, too small a screw diameter, too low a flex-ural modulus, and a length of engagement which is too short

Bosses for threaded inserts. Bosses for threaded inserts are very lar to those for self-tapping screws The inside diameter is determined

simi-by the diameter of the insert; the exact size for heat, press-in, or sonic postmolded inserts are those recommended by the insert manu-facturer for a given plastic who will also provide the lead diameter.(For self-tapping inserts, follow the same inside diameter criteria asfor self-tapping screws.) In this case, however, the lead is an 8° includ-

ultra-ed angle instead of a step The outside diameter of the boss is 2 to 3times the diameter of the insert and the minimum depth of the hole isthe insert length plus 0.030 in

8.2.8 Design for multiple part assemblies

Thus far, the design discussion has centered around the design of vidual piece parts However, most products require multiple partassemblies, often consisting of parts made of different materials Thefirst step is to ensure that the parts fit together properly—not merely

indi-at room temperindi-ature, but indi-at the temperindi-ature extremes of whindi-at theproduct may be expected to encounter (For example, a force fitmentthat works perfectly at room temperature may loosen at elevated tem-peratures or fracture at low temperatures.) That involves the deter-mination of the fitments after the relationship of the parts to each

other has changed due to differences in the coefficient of linear mal expansion Thus, the establishment of acceptable dimensional

ther-limits, generally known as tolerances, for the fitment dimensions is of

critical importance

Dimensional control. The next major consideration is the fact thatplastic part quality is “process sensitive.” That means that the part’ssize and shape can vary according to variations in process parameters.The thermoplastic processes generally operate with a cool mold, withthe moldment remaining in the tool until the part is rigid enough towithstand the forces of ejection If the part is ejected while it is too hot,

it can be distorted and dimensional control lost Furthermore, a moldcore can act as a shrink fixture if the part is left in the tool beyond thatpoint Also, the temperature of the melt has an effect as a hotter melt

is less viscous and can be injected into a mold at a higher speed.However, a hotter melt can lead to greater shrinkage, more distortion,and take longer to cool

Plastic molding cost is also process sensitive because machine cost

is a major expense component The faster the mold is run, the lowerthe part cost However, since running the mold fast requires highertemperatures and pressures which result in poorer dimensional con-trol, it costs more to mold tight tolerances—often a great deal more.For this reason, molding quotes are typically tied to a drawing whichhas tolerances The tighter the tolerances, the higher the part cost.Piece part quotes are not final until there is agreement that the partcan be produced to the desired tolerance with the available tooling Ifthe molds are not capable of producing the parts within the requiredtolerances at the quoted prices, the molds will need to be upgraded

or the piece part prices increased accordingly For these reasons, it issometimes said that “the money is in the tolerances” in plastics molding

In the event of a dispute, the tolerances indicated on the drawingwill determine legal liability Therefore, it is important for the engi-neer to understand the establishment of plastics tolerances complete-

ly, for they are more complicated than for other materials Plasticstolerances are divided between the toolmaker (one-third) and theprocessor (two-thirds) Therefore, a dimensional tolerance of ±0.006 inwould require the toolmaker to build the tool to within ±0.002 in If it

is further presumed that the plastic has a shrinkage rate of 0.005in/in, it is clear that a 1.000 ±0.005-in dimension requires a precisionmold and very careful molding

A shrinkage rate of 0.005 in/in is not really very high for plasticmaterials as some can reach as high as 0.050 in/in What might beregarded as a “commercial” tolerance for one resin could be considered

“fine” for another and “impossible” for a third As a consequence, theSociety of the Plastics Industry has prepared a series of charts for theprincipal plastics indicating what can be regarded as the “fine” and

“commercial” tolerance ranges for each resin These are availablefrom:

The Society of the Plastics Industry, Inc.

Literature Sales Department

1801 K Street NW

Washington, DC 20006

Processes also vary in their ability to produce fine tolerances.Injection and transfer molding can produce the finest tolerances (withthe proper resin) Processes like rotational molding and thermoform-ing require tolerances on the order of 0.010 in/in and can produce tol-eranced dimensions only on the mold side Lay-up and spray-upcannot produce high-toleranced parts

Casual requests for ultra-tight tolerances are widespread Whenprocessors and toolmakers observe such tolerances, they will askwhich are “critical tolerances,” that is, those which really must beheld This practice undermines the validity of the entire tolerancingsystem since all tolerances should be held and no unnecessary onesshould be specified Since the drawing is part of the purchase contract,what is written on it is what will count the most

Deviations from drawing tolerances can be approved if they turn out

to be excessively tight when the actual parts are available When partsare accepted with deviations from the contract drawing, a writtenrecord should be retained and the drawing should be altered accord-ingly to reflect the newly approved tolerance

Regardless of how it is specified, the objective remains the same,namely, that the parts must fit together readily and stay togetherwithin acceptable parameters

Establishing tolerances. A drawing or CAD (computer-aided design)file without tolerances is like a time bomb waiting to explode anddestroy the project Prior to the development of CAD, the engineerand a checker would examine the drawing before releasing it for toolconstruction and would accept responsibility for its accuracy Thissystem of checks and balances served industry well for many years.The purchase order referred to the drawing and the toolmaker andprocessor were responsible for meeting the stated tolerances if theyaccepted the purchase order In the event of a dispute, the issue wasclear-cut

When a CAD file is produced without dimensions or tolerances, it

is very time consuming and tedious to check Hence, this step is