12

PAUL THIENEL

ANDRÉ LÜCK ET AL.

With the birth of an optical part,it is important to focus not on-ly on the design. It is also crucial

to fill a mold with the appropriate mate-rial, using the right technology and therequired precision. The key to success isknowing how much effort must be ex-pended for a particular part geometry.The crucial factors are a profound basicknowledge and detailed understandingin various fields. To clarify the many openquestions, a joint company project,“Op-tical Technologies”, was initiated under

the leadership of the Lüdenscheid Plas-tics Institute and ISK Iserlohner Kunst-stoff-Technologie GmbH, with a total of16 participants.

The main purpose of the joint projectis to define a possible test geometry thatcan clear up all the open questions and isvery practically oriented. It includes se-

lectively influencing the light distributionby thermoplastic lens systems and reflec-tor surfaces. Using defined optical func-tions, it is possible to make a qualitativecomparison with the properties actuallydetermined. In addition, wall-thicknesssteps and intended functional and posi-tioning elements are integrated into theoptical part. Other requirements includeeconomic production by injection or in-jection compression molding techniqueswith the appropriate measurement tech-nology. The tester should have appropri-ate means for measuring the geometry.

Design for Optical Require-ments and Plastic Materials

The light guide corresponds to a circu-lar disc lens with a diameter of about

SPEC I A L

Playing with LightInjection Compression Molding. The optical properties of a part can generally be

attributed to its geometry and material. A joint company project shed light on how

the properties are influenced by the plant concept, process and processing param-

eters.

A movie about the production of the lightguide by injection-compression molding onan all-electric injection molding machine,and the mold technology used for it can befound at the following link: www.sumitomo-shi-demag.eu/solutions/optics/

On the Topic!

To assess quality, the lightguide (left) shines on adiffuse focusing screen;the resulting illuminationimage is recorded by aCMOS camera from behindthe screen (images: Sumitomo

(SHI) Demag)

© Carl Hanser Verlag, Munich Kunststoffe international 6/2010

Translated from Kunststoffe 6/2010, pp. 22–28Article as PDF-File at www.kunststoffe-international.com; Document Number: PE110412

12-17_PE110412_PE6 10.06.2010 9:14 Uhr Seite 12

13

>

54 mm and a weight of about 15 g (de-pending on the material). Using an op-tical simulation, developers have opti-mized the light guide to generate a rela-tively clearly defined 60° light cone(Fig. 1). The light source is a white light-emitting diode (LED). The brightnessdistribution and spectral composition ofthe light cone on a screen are computedwith the LightTools software (supplier:Optical Research Associates, Pasadena,USA). When simulating the total inter-nal reflection and refraction, it is impor-tant to take into account all the optical-ly relevant properties of LED light, thelight guide material and the light guide.The image on the screen determined inthis way is composed of light compo-nents that take different paths throughthe light guide. Measurement of theillumination image gives indications ofthe quality of the surfaces involved. Theregion shown in magenta is intended forreflection and would therefore act as areflector.

An illumination image is based on theassumption of particular material prop-erties. The refractive index influences themode of operation of an optical part. Thechanges produced when a different ma-terial is used can be calculated in advance.It allows a universal test mold to be builtfor different materials.

There is a contradiction between theoptical requirements and a design appro-priate to plastics material. Depending onthe function, the resulting geometry chal-lenges the production technologies tomanufacture precise and permanently re-producible parts. The effect of wall thick-ness distributions is already apparent dur-ing the filling simulation. The thick-walled regions cause a significant speed-ing up of the melt flow. Apart from theweld seam,which with some gating points

cannot be avoided, the merging of the flowfronts results in an air inclusion in the re-gion of the central lens surface (Fig. 2).Uni-form filling is produced by success adap-tation of the wall thicknesses. The simu-lation further shows that the gating pointand the thin-walled regions of the waveguide are frozen after only about 15 s(Fig. 3).At this point, the volume shrinkageof the thick-walled regions is not yet com-plete.

Mold with Compression Plungerand Sensors

The test mold combines injection and in-jection-compression molding (Fig. 4). In-jection-compression molding takes placeover the entire part surface and can be ex-ecuted via the injection molding machine(machine compression) or via hydrauli-cally operated cores (core compression).To investigate core compression, the moldis modified so that the stationary moldcore is operated via a wedge slide. Thisdoes not change the shaping mold sur-face. In machine compression, it is mod-ified into a frame mold. The melt back-flow into the sprue system during the

compression phase prevents shut-offclose to the cavity.

The compression plungers are each fit-ted with a cooling system – conventionalspiral-core cooling and conformal cool-ing. The conformal cooling circuit is in-troduced by LaserCusing technology. Allthe surfaces of the compression plungerare made of the same steel. This allows themirror polish of the conformal cooling tobe compared with spiral cooling. Themirror finishes were produced manuallyor mechanically and are another criteri-on for investigation (Fig. 5). Mechanicallyspeaking, ultraprecision diamond ma-chining with monocrystalline diamondtools is used. Four compression plungersare thus available for the tests.

Compression and temperature sensorscan measure additional local process pa-rameters. To obtain detailed informationabout the mold in tests on the light guide,the following requirements are made onthe sensors:� The pressure gradient between the sen-

sors close to and remote from the gateis displayed from shot to shot.

� Thick-walled zones are monitored,with the sensor front not leaving a

Fig. 2. In the filling simulation, the influence of the wall thickness distribution can be seen: left) faster advance of the outer flow fronts forms an airinclusion, right) after the wall thicknesses have been successively adjusted, the part fills uniformly

Fig. 1. The light guidehas a distinctly de-limited light cone of60°; the light distribu-tion of the cone wastheoretically deter-mined

Kunststoffe international 6/2010

SPEC I A L

12-17_PE110412_PE6 10.06.2010 9:14 Uhr Seite 13

14

mark in the optically relevant moldregion.

� Depending on the sensor, switchoverto holding pressure is performed de-pending on the cavity pressure.

� Imaging of the contact temperaturesshould be possible.

Based on this, pressure-temperature sen-sors (type: 6189A, manufacturer: KistlerInstrumente AG, Winterthur, Switzer-land) with diameters of 2.5 mm are po-sitioned close to and remote from thegate, outside the optical functional zone.A strain measuring pin (type: 9247A, bythe same manufacturer) is used for con-tactless measurement of the cavity pres-sure in the thick-walled zones (Fig. 6).

Taking an Integrated Approach

Manufacturing products cost effectivelyin high quality requires a detailed under-standing of the individual process stepsin many areas. Besides precise manufac-ture of key components such as mold in-serts and ensuring product-related designof machine and mold technology as anintegrated unit, it is also crucial to speci-fy process-relevant parameters and theirtolerance limits. The latter is extremely

difficult and therefore determined inpractical tests with flexible recording ofthe measured values. The system technol-ogy, comprising sensors and electronics,permits a more precise interpretation ofthe individual processes taking place inthe mold. To keep the effort and costswithin reasonable limits, it also helps toconcentrate on key parameters in thequality testing of optical functions of theplastic part.

The question of which machine plat-form generates what optical precision isboth difficult to answer and product de-pendent: all-hydraulic systems generallyoffer a solid basis for thick-walled opticalparts with long holding-pressure times.Hybrid concepts are the correct platformin the thin-walled range. The oil-operat-ed hydraulic injection molding machines,by comparison,have a significantly longerholding pressure application at a high lev-el, and greater injection performance.All-electric machines permit precise move-ments for very accurate optical parts withsmall shot weights of medium wall thick-ness. The axes generally move faster by anorder of magnitude and can realize par-allel machine movements, which is par-ticularly important for simultaneous

compression via the clamping unit. Themechanical force transmission of thespindle and clamping system maintainthe selected position very stably.

Ultimately, it makes sense to performa product-based comparison of possiblemachine concepts. In the case of the lightguide, an all-electric machine (type: Int-Elect 100 Performance, manufacturer:Sumitomo (SHI) Demag Plastics Ma-chinery GmbH, Schwaig, Germany, Fig. 7)is chosen because of the large wall-thick-ness steps in the medium thickness range,which required a compression study.

The precisely designed stroke utiliza-tion of the plastication unit permits veryaccurate control of the injection velocity.The filling, compression and meteringzones have their length, flight depth, pitchand compression ratio adapted to therequirements of the materials to be pro-cessed. The moving end of the machineis additionally equipped with an integrat-

ed simultaneous compression control andadditionally with parallel core-compres-sion functions via an internal hydraulicunit. A laminar flow box (manufacturer:Max Petek Reinraumtechnik, Radolfzell,Germany) ensures consistent cleanlinessof all components in the mold area. Allthe important process data are external-ly processed via a data interface.

Power Reserves and Problem Zones

In injection-compression molding, thevolume of the cavity can be very finely ad-justed. It allows the properties of the op-tical parts to be directly influenced. Com-pared to injection molding, the dimen-sional stability and warpage can also beimproved, along with lower melt stress-ing as a result of reduced material shear-ing, less orientation and sink marks, andeasier deaeration.

SPEC I A L

© Carl Hanser Verlag, Munich Kunststoffe international 6/2010

Fig. 3. Gating and thin-walled regions are already frozen after 15 s, while the thick-walled regionscontinue to shrink significantly

Fig. 4. With the test mold, the parts can beproduced by injection molding and injection-compression molding; injection-compressioncan be performed either via the machine or viahydraulically operated cores

Fig. 5. To determine how the mirror finish influenced the quality of the test sample, two inserts – one polished manually (left) and one by machine (right) – were produced and tested with a stripreflection

12-17_PE110412_PE6 10.06.2010 9:14 Uhr Seite 14

15

>

The pressing of optical plastic parts viathe clamping unit of an injection mold-ing machine provides a substantial pow-er reserve. Machine compression is par-ticularly advantageous for asymmetricalfilling processes,complex geometries, longflow paths, high pressures and rapid com-pression movements. Mold compressionversions with hydraulic plungers permit avery good pressure transmission over acomparatively long time. The compres-sion plunger is a particularly suitable sys-tem for parts with small problem zones,such as very thin-walled chip regions ofcredit cards (localized compression) andrelatively large part thicknesses, such asmagnifying glasses. For the light guide, itis very difficult to predict which effect willbe produced by which version of pressuretransmission. More information can beascertained from a comparison of the twocompression techniques: what pressurelevel is transmitted over what time, andwhat shrinkage compensation – an im-portant quality criterion – the injectionmolded part experiences.

Even in the important injec-tion phase, the easily opened cav-ity proves to be advantageousduring injection-compressionmolding. The melt flow-front ve-locity can be influenced via theadjustable part thickness. The ad-

ditional volume counteracts thematerial shrinkage. The shrinkage

compensation takes place much earlier,and not solely via the plastic core (Fig. 8).While, in injection molding, this compen-sation is generated hydrodynamically viathe melt by an applied holding pressure,the compression movement presses thematerial virtually hydrostatically andevenly (Fig. 9).

Appropriate Material Selection

A specially designed questionnaire ori-ented to the project participants ought toclarify which materials can be preferredfor which light guide systems. The aim isto produce an independent and statisti-cal relationship between the material, testgeometry, part size and manufacturingtolerance. Based on the quoted materialsand the effects on the light guide, fifteenmaterials were shortlisted. Besides thetypical polymers polycarbonate (PC) andpolymethylmethacrylate (PMMA), poly-

methacrylmethylimide (PMMI), cyclo-polyolefin copolymers (COC), styreneacrylonitrile (SAN) and polyamide (PA)are also tested.

To investigate the shaping fidelity, asimple plate geometry (wall thickness2 mm) is chosen as test geometry. Themirror finish surface of the test mold isprovided with pyramidal notches of dif-ferent depths (DIN hardness test, HV 10,20 and 30). The defined position of thenotches permits comparative measure-ments between the material forming andmold insert.

The test samples are measured with achromatic white light sensor, which re-produces the plastic surface three dimen-sionally, and the roughness values Ra and

Rq are determined. The results can be dis-played in a 5-axis network diagram as per-centages of the roughness values and theheights of the pyramids (Fig. 10). The de-gree of reproduction is a measure of thereproduction accuracy of the materialsinvestigated.

Two PC grades and one PMMA andone PMMI grade each show the best re-sults for reproduction accuracy. They areselected for further studies. High-impactPC, scratchproof PMMA and high-tem-perature resistant PMMI have differentoptical properties (Table 1).

Three Test Series, Two Evaluation Strategies

The test series can be classified into threeareas: how the flow-front velocity is in-fluenced by the stepped velocity profileand compression gap, comparison of in-jection molding and injection-compres-sion molding processes, and how the pol-ishing and cooling concepts affect the sur-face of the light guide.

The evaluation is based on two differ-ent strategies. First, the relevant processparameters are directly recorded and con-trolled. On the other hand, a downstream

SPEC I A L

Kunststoffe international 6/2010

Fig. 6. The combinedpressure-temperaturesensor is positioned close toand remote from the gate, acontactless strain measuring pin in the thick-walled zones measures the cavity pressure (below)

Fig. 7. The test machine is an all-electricinjection molding machine with integratedsimultaneous compression control

Injection molding(Filling volume ≤ 100 %)

Injection-compression molding(Filling volume ≥ 100 %)

Gate

Compressiongap

Fig. 8. The material shrinkage is expressed differently in injection molding than in injection-compression molding, consequently the final part volumes are different

© Kunststoffe

Material Refractive index Transmission[%]

Abbe number Critical angle[°]

PC 1.58 89 30 39.3

PMMI 1.54 91 53 40.5

PMMA 1.49 92 59 42.2

Table 1. Comparison of optical properties: Polycarbonate (PC) has a high refractive index forplastics, this results in low transmission; in the case of polymethylmethacrylate (PMMA), theopposite is the case, and polymethacrylmethylimide (PMMI) is roughly in the middle

12-17_PE110412_PE6 10.06.2010 9:14 Uhr Seite 15

16

optical measuring bank determines thequality parameters and their dependen-cies. The optical part radiates onto a dif-fuse focusing screen (Title photo). A CMOS(complementary metal oxide semicon-ductor) camera records the illuminationimage on its reverse side. To evaluate theillumination distribution, characteristiclines are defined, converted into a grayvalue distribution curve and comparedwith a reference curve.

Pressure Transmission Affects Quality

The flawless introduction of the materi-als requires a sensitive injection profile.Inaccurate positioning during change-over to holding pressure leads to fillingproblems of the small foot. The compres-sion gap favors mold filling when it reach-es about 5 % of the average wall thicknessof the lens. The thick-walled regions ofthe reflector ring and foot (Fig. 11, item 2)

are located behind the cold runner gate(item 1). They are directly followed by the

thin-walled edge (item 3) of the inner,very thick main reflection surface(item 4). The inner lens region (item 5)is also thin walled. The injection unit ofthe machine must be capable of achiev-ing the necessary velocity changes at theindividual points rapidly, precisely andreproducibly, and so eliminating surfacedefects such as flow lines.

The accurate contour reproduction ofthe optical surfaces is a prerequisite forthe functioning of the light guide. Dur-ing injection molding, in all test series, theholding pressure transmission falls offrapidly, independently of the set pressurelevel. After about 20 s, all the cavity pres-sure signals fall below 200 bar (Fig. 12 left).

The failure in the holding pressure in-evitably causes sink marks,which serious-ly alter the desired illumination distribu-tion. Compression via the hydraulicplunger also doesn’t provide adequatecavity pressure values.

Machine compression, on the otherhand, provides significantly better re-sults. With an “ideal” setting, the inter-nal pressure level is still sufficiently high,at about 500 bar after about 90 s (Fig. 12,right). For the optical quality, it is impor-tant for the internal pressure to settle ata constant high level after about 30 s dur-ing machine compression. Moreover, thestarting point of compression is of cru-cial importance. If the compressionstroke starts too early, too much materi-al is forced back out of the cavity. If itstarts too late, very high pressures aregenerated in the mold, since the thin-walled regions are already frozen. Thepart has significantly greater internal

SPEC I A L

© Carl Hanser Verlag, Munich Kunststoffe international 6/2010

700

500

300

200

500–700

500–700

500–700

Injection molding Injection-compressionmolding

Fig. 9. Theoreticalcavity pressurebehavior in injectionmolding, the holdingpressure is hydro-dynamically (200–700 bar) applied viathe melt, in compres-sion, via a plungerover the surface andapproximately hydro-static (500–700 bar)

© Kunststoffe

Ra

Rq

HV 10HV 20

HV 3020

0

40

60

80

100

Mold insert PMMA

Fig. 10. The notchesadditionally intro-duced into the mold,and roughness valuescould be used to de-termine the demold-ing degree in percentof polymethylmeth-acrylate (PMMA)

© Kunststoffe

Time

1,800

1,500

1,200

900

600

300

0

60

50

40

30

20

10

0

bar

0

Spec

. pre

ssur

e

Volu

me/

velo

city

0.5 1.0 1.5 s 2.0

cm3 und cm3

s5

43

2

1

36.0

30cm3/s

3cm33.6

Volume

Velo

city

1

2

3

4

5

5

43 1

5

Spec. pressureScrew volumeInjection velocity

Fig. 11. To avoid stress cracking, gramophone record effects or air inclusions, a sensitive injectionprofile is established; the actual curve on the machine with polycarbonate as material

© Kunststoffe

12-17_PE110412_PE6 10.06.2010 9:14 Uhr Seite 16

17

stresses and in the worst case is damagedduring demolding.

Illumination Image as a QualityCharacteristic

The type of polishing is not found to influ-ence the optical quality, for the same testset up, either in the illumination image orin the light distribution.In the cooling con-cepts, it is found that the use of spiral coreand LaserCusing also generates the same il-lumination images and light distributions.Quality is only affected by the mold tem-perature: Too high feed temperatures, as aresult of the greater thermal contraction,lead to insufficient reproduction fidelity, ir-respective of the raw material used.

The photographed illumination imagesillustrate the quality differences very accu-rately. The theoretical illumination imageis schematically composed of two regions:a circular area in the center, superimposedwith a circular ring somewhat further out(Fig. 13, left). In all injection molded opticalparts, the circular ring has illuminationmaxima at the inner and outer edges (Fig. 13,center). One cause of these are sink marksin the thick-walled regions, which divert

the light rays towards the inside and out-side.The uniform pressure transfer in ma-chine compression and the improvedmold filling, with a slightly enlarged cavi-ty, lead to an illumination image (Fig. 13,right) that approximately corresponds tothe simulation of the optical design.

Outlook

Provided that all influences are adequate-ly understood, complex optical parts canalready be manufactured in good quali-ty. Modern processing and mold technol-ogy offer some interesting approaches. In-jection-compression via the clampingunit of the machine, combined with pre-cise mould technology, increases the re-production accuracy of the parts, andtherefore the illumination function.Proper evaluation of the influencing pa-rameters requires ascertaining a mean-ingful relationship between the producedquality and optical function.

There is still plenty to do in the future,and therefore a new joint company proj-ect “Optical Technologies 2” was startedin February 2010. The main issues are theintegration of diffractive structures into

optics, the possibilities for state-depend-ent demolding and the influence of var-iotherm temperature-control technolo-gy. Furthermore, the possibilities of LSRprocessing will be investigated for theirsuitability for manufacturing highlytransparent optical parts. This project hasalready started with thirteen participants;others can still join at a later point.�

THE AUTHORS

PROF. DR.-ING. PAUL THIENEL, born in 1944, headsthe plastics processing lab of the Fachhochschule Süd-westfalen in Iserlohn, Germany, and is scientific advis-er to ISK Iserlohner Kunststoff-Technologie GmbH,Iserlohn, and the Lüdenscheid Plastics Institute.

DIPL.-ING. (FH) ANDRÉ LÜCK, born in 1973, worksin technology development at Sumitomo (SHI) DemagPlastics Machinery GmbH, Schwaig.

DIPL.-ING. (FH) MATTHIAS MILITSCH, born in1980, is responsible for process optimization at ISK.

DIPL.-ING. (FH) MICHAEL TALHOF, born in 1978, ishead of process engineering at the Lüdenscheid Plas-tics Institute.

DR.-ING. THOMAS ABEL, born in 1955, is the pro-prietor of the engineering consultancy of the samename in Lüdenscheid, Germany.

DOMINIK CORDES, born in 1980, is applicationsspecialist for Plastics at Kistler Instrumente GmbH.

Kunststoffe international 6/2010

Time

1,600

bar

1,200

1,000

800

600

400

200

0

160

°C

120

100

80

60

40

20

00

Pres

sure

Tem

pera

ture

10 20 30 40 50 60 70 80 s 100

Time

1,600

bar

1,200

1,000

800

600

400

200

0

160

°C

120

100

80

60

40

20

00

Pres

sure

Tem

pera

ture

10 20 30 40 50 60 70 80 s 100

Strain sensor Pressure GC Pressure GR Contact temperature GRContact temperature GC

Injection molding Injection-compression molding

Fig. 12. In injection molding, the built-up pressure (left) falls within the first 30 s to below 200 bar; in injection compression (right), the pressure remainsat about 500 bar both close to (GC) and remote from (GR) the gate

© Kunststoffe

Fig. 13. The theoretical illumination (left) is schematically composed of two regions; the injection molded part has an illumination maximum (center) atthe inner and outer edge; the uniform pressure distribution during compression reduces the illumination maxima (right)

SPEC I A L

12-17_PE110412_PE6 10.06.2010 9:14 Uhr Seite 17