Uniform Walls

Parts should be designed with a minimum wall thickness consistent with part function and mold filling considerations. The thinner the wall the faster the part cools, and the cycle times are short, resulting in the lowest possible part costs.

Also, thinner parts weight less, which results in smaller amounts of the plastic used per part which also results in lower part costs.

The wall thicknesses of an injection-molded part generally range from 2 mm to 4 mm (0.080 inch to 0.160 inch). Thin wall injection molding can produce walls as thin as 0.5 mm (0.020 inch).

The need for uniform walls

Thick sections cool slower than thin sections. The thin section first solidifies, and the thick section is still not fully solidified. As the thick section cools, it shrinks and the material for the shrinkage comes only from the unsolidified areas, which are connected, to the already solidified thin section.

This builds stresses near the boundary of the thin section to thick section. Since the thin section does not yield because it is solid, the thick section (which is still liquid) must yield. Often this leads to warping or twisting. If this is severe enough, the part could even crack.

Uniform wall thicknesses reduce/eliminate this problem.

Uniform walled parts are easier to fill in the mold cavity, since the molten plastic does not face varying restrictions as it fills.

What if you cannot have uniform walls, (due to design limitations) ?

When uniform walls are not possible, then the change in section should be as gradual as possible.

Coring can help in making the wall sections uniform, and eliminate the problems associated with non-uniform walls.

Warping problems can be reduced by building supporting features such as gussets.

The reason for draft

Drafts (or taper) in a mold, facilitates part removal from the mold. The amount of draft angle depends on the depth of the part in the mold, and its required end use function.

The draft is in the offset angle in a direction parallel to the mold opening and closing.

It is best to allow for as much draft as possible for easy release from the mold. As a nominal recommendation, it is best to allow 1 to 2 degrees of draft, with an additional 1.5° min. per 0.025 mm (0.001 inch) depth of texture. See below.

The mold parting line can be relocated to split the draft in order to minimize it. If no draft is acceptable due to design considerations, then a side-action mold (cam-actuated) may be required at a greater expense in tooling.

The reason for texture

Textures and Lettering

can be molded on the surfaces, as an aesthetic aid or for incorporating identifying information, either for end users or factory. Texturing also helps hide surface defects such as knit lines, and other surface imperfections. The depth of texture or letters is somewhat limited, and extra draft needs to be provided to allow for mold withdrawal without marring the surface.

Draft for texturing is somewhat dependant on the mold design and the specific mold texture. Guidelines are readily available from the mold texture suppliers or mold builders.

As a general guideline, 1.5° min. per 0.025mm (0.001 inch) depth of texture needs to be allowed for in addition to the normal draft. Usually for general office equipment such as lap-top computers a texture depth of 0.025 mm (0.001 inch) is used and the min. draft recommended is 1.5 °. More may be needed for heavier textures surfaces such as leather texture (with a depth of 0.125 mm/0.005 inch) that requires a min. draft of 7.5°.

Voids and Shrinkage

Shrinkage is caused by intersecting walls of non-uniform wall thickness. Examples of these are ribs, bosses, and other projections of the nominal wall. If these projections have greater wall thicknesses, they will solidify slower. The region where they are attached to the nominal wall will shrink along with the projection, resulting in a sink in the nominal wall.

Shrink can be minimized by maintaining rib thicknesses to 50 to 60% of the walls they are attached to.

Bosses located at corners can result in very thick walls causing sinks. Bosses can be isolated using the techniques illustrated.

Radius

Sharp corners greatly increase the stress concentration. This high amount of stress concentration can often lead to failure of plastic parts.

Sharp corners can come about in non-obvious places. Examples of this are a boss attached to a surface, or a strengthening rib. These corners need to be radiused just like all other corners. The stress concentration factor varies with radius, for a given thickness.

As can be seen from the above chart, the stress concentration factor is quite high for R/T values lesss than 0.5. For values of R/T over 0.5 the stress concentration factor gets lower.

The stress concentration factor is a multiplier factor, it increases the stress.

Actual Stress = Stress Concentration Factor K x Stress Calculated

This is why it is recommended that inside radiuses be a minimum of 1 x thickness.

In addition to reducing stresses, fillet radiuses provide streamlined flow paths for the molten plastic resulting in easier fills.

Typically, at corners, the inside radius is 0.5 x material thickness and the outside radius is 1.5 x material thickness. A bigger radius should be used if part design will allow it.

The use of ribs

Ribs increase the bending stiffness of a part. Without ribs, the thickness has to be increased to increase the bending stiffness. Adding ribs increases the moment of inertia, which increases the bending stiffness. Bending stiffness = E (Young's Modulus) x I (Moment of Inertia)

The rib thickness should be less than the wall thickness-to keep sinking to a minimum. The thickness ranges from 40 to 60 % of the material thickness. In addition, the rib should be attached to the base with generous radiusing at the corners.

At rib intersections, the resulting thickness will be more than the thickness of each individual rib. Coring or some other means of removing material should be used to thin down the walls to avoid excessive sinking on the opposite side.

The height of the rib should be limited to less than 3 x thickness. It is better to have multiple ribs to increase the bending stiffness than one high rib.

The rib orientation is based on providing maximum bending stiffness. Depending on orientation of the bending load, with respect to the part geometry, ribs oriented one way increase stiffness. If oriented the wrong way there is no increase in stiffness.

Draft angles for ribs should be minimum of 0.25 to 0.5 degree of draft per side.

If the surface is textured, additional 1.0 degree draft per 0.025 mm (0.001 inch) depth of texture should be provided.

Boss Design

Bosses are used for the purpose of registration of mating parts or for attaching fasteners such as screws or accepting threaded inserts (molded-in, press-fitted, ultrasonically or thermally inserted).

The wall thicknesses should be less than 60 % of nominal wall to minimize sinking. However, if the boss is not in a visible area, then the wall thickness can be increased to allow for increased stresses imposed by self-tapping screws.

The base radius should be a minimum of 0.25 x thickness

The boss can be strengthened by gussets at the base, and by attaching it to nearby walls with connecting ribs.

Hoop stresses are imposed on the boss walls by press fitting or otherwise inserting inserts.

The maximum insertion (or withdrawl) force Fmaxand the maximum hoop stress, ocurring at the inner diameter of the boss, max is given by

Failures of a boss are usually attributable to:

High hoop stresses caused because of too much interference of the internal diameter with the insert (or screw).

Knit lines -these are cold lines of flow meeting at the boss from opposite sides, causing weak bonds. These can split easily when stress is applied.

Knit lines should be relocated away from the boss, if possible. If not possible, then a supporting gusset should be added near the knit line.

Counter Bore/Sink

Counter-sinking is often done to accommodate heads of flat head screws. However as can be seen from the figure, there is a sideways component of the thrust which could split the countersink due to the generated hoop stresses

Counter-boring is done to accomodate pan-head, fillister-head or round-head screws or other screws with flat-bottomed undersides.

Counter-bored screws exert only force in the axial direction, thus operate mostly under compression, with no sideward component to the applied force vector. Such design is inherently more robust than counter-sinking.

Inserts

• Inserts are used in plastic parts, to allow the use of fasteners such as machine screws. The advantage of this is that since these inserts are made out of metal, they are robust. Further, machine threads also allow great many cycles of assembly and disassembly

• Inserts are installed using one of the following methods:- 

Ultrasonic insertion. The insert is vibrated using an ultrasonic transducer, called the "horn" mounted in an ultrasonic machine.

The horn has to be specially designed for each application for optimum performance. The ultrasonic energy is converted to thermal energy due to the vibrating action, which allows the insert to be melted inside the hole.

This type of insertion can be done rapidly, with short cycle times, low residual stresses. Good melt flow characteristics for the plastic are necessary for the process to be successful.

The ultrasonic equipment is relatively expensive, and also needs a custom horn for optimal production rates.

- Thermal Insertion. The inserts are heated by placing them over the hole and pressing them in with a heated tool. The tool first heats the insert, then the insert is pressed in.

The advantage of this method is that the special tooling necessary is relatively simple, usually a cylindrical tool, which can be easily, fabricated in the machine shop. The cycle times are usually short.

However, care has to be taken, not to overheat the insert or the plastic, or it will lead to local plastic degradation.

- Press Fitting.The inserts are designed with barbs (straight knurls that are interrupted) and can be press fitted inside the hole in the boss.

This process is fast and requires no special tooling. However, high hoop stresses are generated, so the boss design has to be robust. Also, the retention is strictly based on press fitting and the small amount of material that flows inside the barbs. Thus retention is not very high.

- Molded-in Inserts. The inserts are placed in the mold prior to the injection of plastic. The injection of the plastic completely encases the insert on the outer diameter and provides very good retention. In fact retention of such inserts is the best compared to other process (barb/knurl design being the same).

This process slows down the operation of the mold, since the inserts have to be manually placed inside the mold. Inserts can be automatically placed in a mold, but this greatly increases the complexity and cost of the mold. This can only be justified if the volume of production is very high to offset the cost savings in shorter cycle times.

Self-Tapping Screws

Self-tapping screws are often used to fasten plastic parts together. Self-tapping screws are available both in thread-cutting type as well as thread-forming type.  

Thread Forming

Pros Cons

High tensile valuesHigh hoop stress, boss has to be designed to withstand this stress

High amount of torque to

strip threads

Large number of cycles of assembly and disassembly possible

Stress relaxation possible

Suitable for plastics that are non-brittle

Not suitable for brittle plastics

 

Thread Cutting

Pros Cons

Lower Hoop Stress Lower Tensile Pullout

Suitable for brittle plastics

Lower amount of torque to pull the threads

Snap Latches

• Snaps allow an easy method of assembly and disassembly of plastic parts. Snaps consist of a cantilever beam with a bump that deflects and snaps into a groove or a slot in the mating part.

• Snaps can have a uniform cross-section or a tapered cross section (with decreasing section height). The tapered cross-section results in a smaller strain compared to the uniform cross-section. Here we consider the general case of a beam tapering in both directions.

When Rh=1 and Rb=1 , the above formula does not apply, L'Hospital's rule applies and the formula is simplified to the following:

• Disassembly force. The disassembly force is a function of the coefficient of friction, which ranges from 0.3 to 0.6 for most plastics. The coefficient of friction also varies with the surface roughness. The rougher the surface, the higher the coefficient of friction.

• There is an angle at which the mating parts cannot be pulled apart. This is known as the self-locking angle. If the angle of the snap is less than this angle, then the assembly can be disassembled by a certain force given by the above formula.

The self-locking angle = tan-1(1/µ)

where µ is the coefficient of friction which ranges from 0.3 to 0.6 for most plastics.

This computes to angles ranging from 73° for low coefficient of friction plastics to 59° for high coefficient of friction plastics.

If this angle is exceeded then the snaps will not pull apart unless the snap beam is deflected by some other means such as a release tool.

This property can be used to advantage depending on the objective of using the snaps. If the snaps are to be used in the factory for assembly only (never to be disassembled by the end user), then the ramp angle the self-locking angle should be exceeded. If the user is expected to disassemble (to change batteries in a toy for example), then the angle should not be exceeded.

• Tooling for snaps is often expensive and long lead time due to

- The iterations required achieving the proper fit in terms of over travel. The amount of over travel is a design issue. This will control how easy it is to assemble, and how much the mated parts can rattle in assembly. This rattle can be minimized by reducing the over travel or designing in a preload to use the plastic's elastic properties. However, plastics tend to creep under load, so preloading is to be avoided unless there is no other option.

 

- Often, side action tooling (cam actuated) is required. This increases the mold costs and lead times.Cam actuated tooling can be avoided if bypass coring can be used that results in an opening in the part to allow the coring to form the step.

 

• Some common problems of using snaps: 

- Too high a deflection causing plastic deformation (set) of the latch (the moving member). Care has to be taken that the latch does not take a set. Otherwise, the amount of latch engagement could reduce, reducing the force to disassemble. If the set is bad enough the engagement might even fail.

- The moving arm could break at the pivot point due to too high a bending stress. This can be avoided by adhering to the design principles and not exceed the yield strength of the material-in fact it should be kept well below the yield strength depending on the safety factor used.

- Too much over travel leads to a sloppy fit between mating parts resulting in loose assemblies that can rattle.

• Good snap design practices 

- Design the latch taking into account the maximum strain encountered at maximum deflection.

- In general, long latches are more forgiving of design errors than short latches for the same amount of deflection, because of the reduced bending strain.

-  Build mold tooling with "tool safe condition". By this we mean that the deflection or over travel, or length of engagement can be changed easily by machining away mold

tooling, rather than add material to mold tooling, which is more expensive and not good mold practice. This "safe" condition allows for a couple of tooling iterations of the latch, until the snap action is considered acceptable.

Bibliografie

Avallone, E.A., Baumeister, T. (eds.) (1997), Marks' Standard Handbook for Mechanical Engineers, 11th ed., McGraw-Hill, Inc. (New York). Banister, R.A. (1987), "Designing Hinges That Live", Machine Design (July 23rd). Bauccio, M. (ed.) (1994), ASM Engineering Materials Reference Book, 2nd ed., ASM Intl. (Materials Park, OH). Beer, F.P., Johnston, Jr., E.R. (1981), Mechanics of Materials, McGraw-Hill, Inc. (New York). Chow, W. (1977), "How to Design for Snap Fit Assembly", Plastic Design Forum (Advanstar Communications) (March/April). Domminghaus, H. (1993), Plastics for Engineers: Materials, Properties, Applications, Hanser Publishers (Munich). Dym, J.B. (1983), Product Design with Plastics: A Practical Manual, Industrial Press (New York). Franciose, R.F., Khouri, R.N., Gomez, C.J. (officers) (1983), Dimensioning and Tolerancing ANSI Y14.5-1982, American Society of Mechanical Engineers(New York). Giesecke, F.E., Mitchell, A., et al. (1985), Technical Drawing, 7th ed., Macmillan Publishing (New York). Hanna, R.D. (1961), "Molded-in Hinge Polypropylene Components", ANTEC (SPE) Conference. Holman, J.P. (1986), Heat Transfer, 6th ed., McGraw-Hill, Inc. (New York). Kutz, M. (ed.) (1986), Mechanical Engineers' Handbook, John Wiley and Sons (New York). Lincoln, B., Gomes, K.J., Braden, J.F. (1984), Mechanical Fastening of Plastics - An Engineering Handbook, Marcel Dekker (New York). Lindeburg, M. (1997), ME PE Review Manual, 10th ed., Professional Publications (Belmont, CA). Norton, P.B., Esposito, J.J. (officers) (1994), Encyclopaedia Britannica, 15th ed., Encyclopaedia Britannica, Inc. (Chicago). Oberg, E., Jones, F.D., Horton, H.L., Ryffel, H.H. (1996), Machinery's Handbook, 25th ed., Industrial Press (New York). Popov, E.P. (1968), Introduction to Mechanics of Solids, Prentice Hall, (Englewood Cliffs, NJ). Pruitt, G. (1986), "Geometric Dimensioning and Tolerancing", Manufacturing Engineering (July). Rao, N.S. (1991), Design Formulas for Plastics Engineers, Hanser Publications (New York). Roark, R.J. (1975), Formulas for Stress and Strain, 5th ed., McGraw-Hill, Inc. (New York). Shigley, J.E. (1972), Mechanical Engineering Design, 2nd ed., McGraw-Hill, Inc. (New York). Simonds, H.R., Weith, A.J., Bigelow, M.H. (1977), Handbook of Plastics, Van Nostrand Reinhold (New York). Timoshenko, S., MacCullough, G.H. (1957), Elements of Strength of Materials, 3rd ed., Van Nostrand Co. (New York). Tres, P.A. (1998), Designing Plastic Parts for Assembly, 3rd ed., Hanser Publishers (Munich).

Joining Plastic

There are several techniques for joining plastic parts. Equipment cost and labor for each method vary considerably. Most techniques have limits on the sizes and types of plastic that can be joined.

Mechanical fastening: The simplest way to join plastic parts is to design a fastening element (hinge, latch, detent) into the parts. Only stronger plastics are suitable for this method since the joint must survive the strain of assembly, service load, and possible repeated use. This form of fastening is suitable only for lightly loaded, nonrigid assemblies where precision is not critical.

Mechanical fasteners (screws, rivets, pins, sheet-metal nuts) are the most common joining method. They require a plastic that can withstand the strain of fastener insertion and subsequent high stress around the fastener. Conventional machine screws are rarely used except with extremely strong plastic.

There are a number of fasteners designed for use with plastics. Threaded fasteners work best on thick sections. Thread-forming screws are preferred for softer materials, while thread-cutting screws work best on harder plastics. Push-on locknuts and clips may be better for thinner sections. If a fastener has to be removed a number of times, metal inserts are recommended. They may be molded in place, forced, glued or expanded into holes, or inserted ultrasonically.

Fusion bonding: Plastic parts too complex or large to be fabricated on available molding equipment are sometimes made as subcomponents and welded together by fusion bonding. Holding fixtures ensure accurate mating and alignment of the parts to be joined.

To plasticize the part edges, the fixtures press the parts against a heating platen. As the platen melts the part surface, the plastic material is displaced while the fixture maintains part pressure against the platen. This initial fusing produces a smooth surface by removing surface imperfections, warps, or small sinks. Melted material continues to be displaced until "melt stops" on the platen contact the holding fixtures. Material is then no longer displaced and the parts are held against the platen until each part edge is plasticized to a predetermined depth. Melt depth is regulated by contact time, which is usually from 3 to 6 sec.

After part edges are plasticized, the fixtures open and the platen is withdrawn. The fixtures then reclose, forcing the parts together until "seal stops" on the fixtures come into contact. The parts are held together under pressure as the melted material cools, bonding them together.

Usually, fusion bonding joins thermoplastics such as polyethylene, polypropylene, and ABS. However, with slight changes, the process can be used to join both filled and unfilled nylon.

To fuse unfilled nylon, a thin, nickel-chrome blade, rather than a platen, is heated to about 1,100°F. The heated blade is brought to within 0.0005 in. of the material in a brief heating cycle. The blade is then removed and the part conventionally fixtured. Glass-filled nylon can be softened using direct heat, if care is taken to clean the blade or platen with a steel brush between each cycle.

Cooling time normally is the same as melt time -- from 3 to 6 sec. When cooling is complete, the gripping mechanism in one of the holding fixtures releases the part, the fixtures open and the finished part is removed manually or automatically. Total machine time from start to finish is generally about 15 sec, well within the range of the injection-molding systems often used in conjunction with fusion bonding.

Hot-gas welding: This is a low-speed process for fabricating large structural parts from sheet stock. A thermoplastic rod is heated with the parts to be joined until they soften and can be pushed together. The heat source is usually an inert gas. Top speed on long, straight welds is about 40 in./min. Intricate parts require more time. Operator skill is critical for both weld strength and appearance.

Vibration welding: Vibration welding produces pressure-tight joints in circular, rectangular, or irregularly shaped parts made from almost any thermoplastic material -- even in dissimilar materials having a melt-temperature spread as great as 100°F. The process is particularly suited for hollow, container-type components having the weld joint in a single plane.

The parts are frictionally heated by pressing them together and vibrating one of the parts at 120 to 240 Hz, in the plane of the joint.

After 2 to 3 sec, vibration is stopped at the exact required relative position of the two pieces. Pressure is maintained briefly while the softened plastic cools. Joint strength is very near that of the parent material. Cycle time, including manual loading and unloading, ranges between 5 and 8 sec for most parts. The process is adaptable to fully automated systems.

Vibration welding accommodates large parts that are impossible or impractical to weld by other methods. Parts can be rectangular or irregular, as long as the weld joint is in a single plane and a small amount of motion (at least 0.12 in.) in that plane is possible. Components with weld surfaces as long as 20 in. have been successfully joined.

Solvent bonding: Plastics are softened by coating them with a solvent, then clamped or pressed together. The plastic molecules mix together, and the parts bond when the solvent evaporates. This process is limited to thermoplastics. Fusion time is a function of the solvent's evaporation rate and may be shortened by heating.

Pure solvents provide the simplest, lowest-cost bond. Doped solvents, which cost more, contain solutions of the plastic being bonded to fill gaps in imperfectly fitting parts. Next in complexity and cost come monomer and polymerizing solvents. These materials contain catalysts and promoters added to doped solvents to produce polymerization at room temperature or a temperature below the softening point of the thermoplastic.

Solvent-bonded parts must be pressed together for 10 to 30 sec before the joined parts can be handled. Pressure is critical, as too much pressure causes parts to distort. A day or more at room temperature or several hours at elevated temperature may be needed to cure the bond.

Ultrasonic welding: Pulses are transmitted to the part by a resonant vibrating tool called a horn, causing two plastic materials to vibrate against each other. Vibration heats and fuses the parts together. Plastic products including blends or alloys of different resin families can be joined by ultrasonic welding. Such dissimilar parts should be designed carefully, and both the resin and equipment suppliers should become involved early to ensure that ultrasonic techniques can produce a suitable bond.

Ultrasonic assembly is often done at 20 kHz to achieve the vibrational amplitude and power needed to melt thermoplastics. However, higher frequencies that produce less vibration can also join thermoplastics, especially engineering thermoplastics such as reinforced polymers. For some applications, use of 40 kHz means less material degradation. Tooling used for 40-kHz welding is smaller than that used for 20 kHz; therefore, the welds produced at 40 kHz are generally smaller.

Pressure is critical. Too much causes the parts to vibrate as an integral structure with no heating. Too little does not provide enough contact friction or heating.

Ultrasonic welding is fast. Assembly rates of more than 25 parts/min are possible with a single station. There are no secondary operations, such as coating, inserting, or cleaning. The process requires fairly rigid materials. Dissimilar-material sonic welds can be made, but the melting temperatures of both materials must be quite close, otherwise only the lower-melting material will soften and a bond will not form.

Ultrasonics can also be used to insert metal components into plastic, stake metal and plastic parts together, or spot-weld plastic sections.

Induction welding: Induction welding can be done by pressing two pieces of plastic material together around a metal insert. When passed through a magnetic field, the encased metal is heated, and the compression produces a fusion weld. The metal remains sealed inside the part.

Thermoplastic bonding agents filled with either electromagnetic or ferrite materials may also be used in induction welding. The material, in the form of a preformed ring or strip, or as a hot melt, for instance, is inserted between the mating parts before induction heating. If metallic particles are used, the alternating magnetic field induces current flow within the particles, generating heat. When ferrite is used, no current is produced. Instead, heat is produced by molecular friction as the particles try to retain their magnetic charge when the fields are reversed.

To eliminate the need to add metal to the joint, metal powder can be added to the original plastic molding, but a much higher frequency is needed.

Induction welding is a high-cost technique and is suitable for difficult-to-weld plastics such as polypropylene, and for shapes that cannot be fitted into an ultrasonic welding machine. The process is best suited for bonding most polypropylene, polyethylene, styrene, ABS, polyester and nylon in high-volume, highly automated joining operations.

Heat-resistant polypropylenes that cannot be joined with adhesives or other welding techniques may be successfully joined using induction welding. Bonding agents heated inductively reach temperatures of 300°F in 0.1 sec to fuse with the heat-resistant substances.

Dielectric welding: Dielectric welding is used on films and thin sheets up to about 60 mil, primarily in packaging. The technique uses the breakdown of plastic under high voltages and frequencies (13 to 120 MHz) to produce dielectric heating and fuse the plastic. Welding speed is a function of dielectric-loss factor, material thickness, and the area subjected to the voltage. Dielectric welding is ideal for PVC materials.

Living Hinge

Living hinges are thin sections of plastic that connect two segments of a part to keep them together and allow the part to be opened and closed. Typically these are used in containers that are used in high volume applications such as toolboxes, fish tackle boxes, CD boxes etc.

The materials used to make a living hinge are usually a very flexible plastic such

as polypropylene and polyethylene. These can flex more than a million cycles without failure.

Besides meeting the design guidelines, the hinges have to be processed properly. The molecules have to be oriented along the hinge line for the hinge to have acceptable life.

As molded the fibers of the plastic are somewhat random in orientation. In order to orient the fibers to aid in prolonging the hinge life, some or all of the following practices should be followed:

- The gate location should be such as to allow the plastic to flow across the hinge for maximum strength.

- As the part comes out of the mold, it needs to be flexed a minimum of 2 times while it is still hot, for optimum strength

-  Coining is often done to give the hinge, enhanced properties. The coining process compresses the hinge to a pre-determined thickness. The strain induced is greater than the yield stress of the plastic. This will plastically deform the hinge (i.e. place it outside the elastic range into the plastic range). The amount of coining (compression) should be less than the ultimate stress, to keep the hinge from fracturing.

- The finished thickness after coining should be from 0.25 to 0.5 mm (0.010 to 0.020 inch). This keeps the stress in the outer fibers from exceeding the yield strength when being flexed.

This process can also be done by heating the hinge or the coining tool to a temperature below the glass transition tempertature of the plastic. This allows

for easier coining and somewhat enhanced properties, as the plastic "flow" easier when being heated.