Precision processing and microtopographical characterisation of tooling inserts for injection moulds.

Bliedtner, Jens ; Buerger, Wolfgang ; Rosenkranz, Sandy 等

Abstract: Ranges of plastic parts with optical surfaces can increasingly be manufactured in accordance with the users' demands at a high quality level, due to the technical development towards high-precision manufacturing in the field of injection moulding. Optical surfaces of plastic parts have to be of high quality for functional reasons. In the present case, the user of such components demands that the optically effective flat surfaces have a square value of the average surface roughness [R.sub.q] of [less than or equal to] nm.

Key words: model based systems, schedulling and planning, predicate transition petri nets, simulation, specification, verification

1. Introduction

Ranges of plastic parts with optical surfaces can increasingly be manufactured in accordance with the users' demands at a high quality level, due to the technical development towards high-precision manufacturing in the field of injection moulding. Optical surfaces of plastic parts have to be of high quality for functional reasons. In the present case, the user of such components demands that the optically effective flat surfaces have a square value of the average surface roughness [R.sub.q] of [less than or equal to] 10 nm.

In the technological process of injection moulding a process-sure fulfilment of the users' demands is absolutely necessary. Based on concrete users' demands experimental investigations were carried out in the injection moulding process of components with optical surfaces. In these investigations the roughness of the tooling inserts for the injection mould were taken into account according to their respective manufacturing processes (ultra-precision turning, polishing) as well as the roughness of the mouldings from the tooling inserts. Due to the technical roughness demands a high-quality surface testing technology has to be used in order to assure the quality of these components (Fig. 5).

2. Experimental Procedure and Results

2.1. Requirements to ultra precision processing of tooling inserts for injection moulds Tooling inserts for injection moulding of plastic optics can be manufactured in accordance with the users' demands at a high quality and precision level, due to the continuous technical development of processing techniques in the field of micro technology. In the technological process of ultra precision turning followed by injection moulding a process-sure fulfilment of the users' demands is absolutely necessary. Based on concrete users' demands experimental investigations were carried out in the ultra precision turning process of tooling inserts with optical surfaces. In these investigations the roughness of the tooling inserts for the injection mould were taken into account according to the technological parameters (rotational speed, feed rate, cutting edge radius, cutting depth) as well as the influence of different nickel-phosphorus layers. Due to the technical roughness demands a high-quality surface testing technology has to be used in order to assure the quality of these components.

The samples for the roughness measurements and the technological parameters for the experimental investigations were selected in cooperation with the company JENOPTIK Polymer Systems GmbH, Triptis (Jenoptik Group) considering the following aspects: surface hardness, type of layers, processability and quality. 41 nickel-plated (chemically and galvanically) tooling inserts with selected technological parameters (feed speed vf, rotational speed n, cutting depth ap, cutting edge radius r[??]) were processed in the test programme using the ultra precision turning machine Nanoform 200 of the company Precitech. The following roughness measurements were carried out by means of the scanning force microscope (SFM) "Dimension 3000" in the tapping mode; a cantilever scanning tip with a tip radius of < 10 nm and an angle of < 25[degrees] was used. The measurements were carried out at five different places on the sample surface. The raw surface data were evaluated by means of a roughness analysis of the selected areas, by a cross-section analysis and a 3D-image. The quality of the tooling inserts was evaluated on the one hand by the square value of the average surface roughness [R.sub.q] and on the other hand statistically by the PSD-function (Power Spectral Density function). Some results, selected from the wide range of test results, will be briefly introduced in the following paragraph.

The nickel-phosphorus layers contribute essentially to the achievable micro topography. They are used due to their good grindability; high wear resistance and corrosion protection. The layers were examined during the investigations by means of spot analysis and line scan in the cross section under a scanning electron microscope. Figure 1 shows a line scan of NiP-plated tooling insert with activating layer. The energy dispersive X-ray of the cross sections of the tooling inserts enabled a determination of existing diffusion zones and activating layers between the steel base and the NiP-layers and their chemical composition. Figure 2 shows the Vickers micro hardness of the examined nickel-phosphorus layers of different manufacturers. From this figure the following conclusions can be derived:

* The layer micro hardnesses of the companies PGE, Techmetals and Elektroform differ only slightly.

* The values of the galvani[N.sub.c]iP-layers fluctuate around 550 HV 0.1.

* The chemical NiP-layers of the company Northamerican produce the highest values of 650 HV 0.1.

[FIGURE 1 OMITTED]

The roughness determination of the optical surfaces of ultra precision turned tooling inserts was carried out by means of a scanning force microscope. The square value of the average surface roughness [R.sub.q] and the power spectral density function (PSD) were determined during the investigations. Figure 3 shows the roughness analysis (left) and the 3D-image (right) of an optical tooling insert. From these investigations the following conclusions can be derived:

* The roughness values of all inserts range from 1.5 nm [less than or equal to] [R.sub.q] [less than or equal to] 7.8 nm. The users' demand of [R.sub.q] < 10 nm is fulfilled.

* The surface roughness depends exclusively on the feed f and the cutting edge radius [r.sub.[??]] of the tool.

* The influence of the cutting depth [a.sub.p] and the rotational speed n can be neglected.

[FIGURE 3 OMITTED]

Another possibility for characterizing the examined technical surfaces (tooling inserts, mouldings) is the analysis of the spectral power density distribution (PSD-function --Power Spectral Density function). The power density spectrum contains the spectrum of the spatial frequency and can be illustrated as a one- or two-dimensional PSD-function. The PSD-function allows a complete description of the surface quality characteristics and has proved to be particularly suitable for polished surfaces or applications with extreme demands. The two-dimensional power spectral density function represents the relative value of each surface roughness parameter as a function of the spatial frequency. The potential of the AFM, which was used for the measurements, enabled a characterization of the examined optical surfaces by means of a PSD-function. Figure 4 shows the two-dimensional PSD-function of an ultra precision turned surface. Two zones of the graph can be recognised. The long wave proportion of the graph until the first clear peak obviously defines the level of the form deviation. The significant amplitude at a wave length of [lambda] = 0.5 [micro]m corresponds with the feed value (f = 0.5 [micro]m) set at the ultra precision turning machine (feed speed) and the distance between the feed grooves on the surface. The short wave proportion of the graph represents the existing kinematic and geometric roughness, which depend on the technological parameters. An interrelationship between the resulting PSD-functions (average values) and the square values of the average surface roughness depending on the applied feed speeds (feed rate) can be recognised from the results of the investigations. With increasing feed speed the number of wave lengths with significant amplitudes increases too as well as the square values of the average surface roughness [R.sub.q].

[FIGURE 4 OMITTED]

2.2. Microtopographic Requirements to the injection moulding of high-precision components with optical surfaces

The roughness measurements were carried out by means of the scanning force microscope (SFM) "Dimension 3000" in the tapping mode; a cantilever scanning tip with a tip radius of <10 nm was used. A scanning area of 50 [micro]m by 50 [micro]m was chosen for the tooling inserts and the mouldings. The resolution of the scanned area is characterized by a number of 256 x 256 pixels. Prior to each measurement the sample surface was cleaned: the tooling inserts were cleaned by coating with a special cleaning lacquer and the mouldings by ionised compressed air. The measurements were carried out at three defined places on the sample surface. The primary surface data were evaluated by means of a roughness analysis of the selected areas, by a cross-section analysis and a 3D-image. The quality of the tooling inserts and the mouldings could be evaluated on the one hand by the square value of the average surface roughness [R.sub.q] and on the other hand statistically by the PSD-function (Power Spectral Density function). Statements can also be made on the standard deviations of the respective measurement results. An overview of the surface topography is given by the amplitude and height image as well as by the 3D- image (Figure 5).

[FIGURE 5 OMITTED]

These images provide information about height and depth relation on the surface (e.g. scratches, inhomogenities) and thus enable a qualitative evaluation of inserts and mouldings (injection moulding). The amplitude and height images show adhesions very well, which can give information about the moulding behaviour of the used plastics in the injection moulding process. Grooves on the tooling inserts, caused by the polishing process, can also be recognised very well in both images. The samples for the roughness measurements were selected in cooperation with the company JENOPTIK Polymer Systems GmbH, Triptis (JENOPTIK--Group). They were selected according to the following criteria: hardness behaviour, manufacturing costs, durability, coating behaviour and imaging qualities. The selected surfaces were optically effective surfaces. The roughness profile of flat optical surfaces is smaller than the wavelength of the visible light. The investigations were carried out with the following range of parts:

* Metallic tooling inserts (different materials and manufacturing procedures/materials: steel, nickel)

* Injection moulded plastic parts (mouldings from the tooling inserts/materials PC Lexan LS 2, COC Topas, PMMA transparent and different PC Makrolon types)

Some results, selected from the wide range of test results (e.g. in Rosenkranz), will be briefly introduced in the following paragraph. Figure 6 shows the square values of the average surface roughness [R.sub.q] of the examined tooling inserts and their mouldings with one material (PC Lexan LS 2). The following conclusions can be derived from this graph:

* The material of the tooling insert has a high influence on the roughness of the moulding. The roughness of the tooling insert is transferred to a great extent onto the surface of the PC Lexan LS 2 mouldings.

* The roughness values of all mouldings are lower than those of the tooling inserts.

* The measured roughness values prove that injection moulding with PC Lexan LS 2 is a stable technological process.

* The technical demands of [R.sub.q] [less than or equal to] 10 nm are fulfilled

Figure 7 shows the square values of the average surface roughness [R.sub.q] of the examined tooling inserts No. 2, No. 6 and No. 9 and their mouldings with different materials (PC Lexan LS 2, COC Topas and PMMA transparent).

From these results in figure 7 the following conclusions can be derived: * The applied plastic has only a small influence on the roughness of the moulding surfaces.

* The values of mouldings from steel inserts No. 2, No. 6 and No. 9 are lower than those of the tooling inserts (except material COC Topas from steel insert No. 2).

The investigations have proved that on the one hand the manufacturing process of the semi-finished products influences the roughness values of the tooling inserts (e.g. steel No. 9) and on the other hand in the moulding process from the steel inserts with certain plastics different side-processes take place which have a negative influence on the topography/surface profile of the moulded technical surfaces (e.g. the formation of grooves, Figure 8). It is difficult to find out what caused the development of these grooves as injection moulding is a complex process. The properties of the material COC can be a reason for these surface defects (grooves), especially the bleeding of the material during the injection moulding process. Another possible reason for the development of the grooves could be the adhesion behaviour of the material during the moulding process.

Figure 8 shows that in the area of 25[micro]m by 25[micro]m no grooves can be found; that is why the square value of the average surface roughness is by 46% lower than in the area of 50[micro]m by 50[micro]m.

Another possibility for characterizing the examined technical surfaces (tooling inserts, mouldings) is the analysis by means of the power spectral density distribution (PSD-function--Power Spectral Density function). The power density spectrum contains the spectrum of the spatial frequency and can be illustrated as a one- or two-dimensional PSD-function.

[FIGURE 8 OMITTED]

Basically every precision processing of technical surfaces is characterised by a specific power density distribution (PSD-Power Spectral Density function). According to Beckstette and Roth (see figure 9) three regions can be distinguished in the diagram of the power spectral density distribution as a function of the spatial frequency, which are significant for assessing the quality of the technical surface: roughness (high spatial frequency), waviness (Medium spatial frequency), form error (long spatial frequency).

The PSD-functions for different precision-processed technical surfaces could be determined in own comprehensive examinations. Figure 10 shows results of ultraprecision-turned NiP-plated tooling inserts made of steel with MKD-tools.

[FIGURE 9 OMITTED]

These examinations led to special results which are too extensive to be dealt with in detail in this article. It is for sure though that ultrprecision turning with increasing feed (all other technical parameters remain constant in this process) results in specific surface manipulations, which can be minimised through subsequent processing.

[FIGURE 10 OMITTED]

The PSD-function allows a complete description of the surface quality characteristics and has proved to be particularly suitable for polished surfaces or applications with extreme demands. The two-dimensional power spectral density function represents, according to Duparee' and Notni, the relative value of each surface roughness parameter as a function of the spatial frequency.

The measuring software of the SFM, which was used for the measurements, enabled a characterization of the examined optical surfaces by means of a PSD-function. Figure 11 shows the two-dimensional PSD-function of the PMMA moulding from a steel insert. For the comparison of measured values the envelope for each measuring point can be established and the different measuring points can be illustrated in a diagram by means of a spreadsheet. The existing primary data can also be compared by means of a PSD-function (Figure 11). Figure 12 shows an example for measuring the average value of the PSD-function for the steel inserts No. 2 and No. 9. The graphs can be analysed statistically by suitable statistical methods (e.g. regression analysis).

The evaluation of the determined power spectral density distributions (PSD-functions) of selected surfaces of the tooling inserts and their plastic mouldings aims at increasing the statistical certainty of the quality evaluation of these technical surfaces on the one hand and detecting the differences between the technical surfaces

On the other hand. The evaluation of these specific technical surfaces can be qualitatively increased by the power spectral density distribution.

[FIGURE 11 OMITTED]

[FIGURE 12 OMITTED]

The problem with the evaluation of the determined power spectral density distributions in relation to wavelength and frequency is to define appropriate criteria (parameters) for the evaluation of the differences. Basically, the following methods are suitable in order to define such parameters:

* Characterising the increase of the functions PSD = f (wavelength) or PSD = f (frequency). A problem here though is the separation of the flaw from the roughness proportion with the wavelength and the spatial frequency for the examined specific surfaces (x-coordinate). There are no experiences from experimental investigations for the injection moulding of technical surfaces with optical surfaces.

* Characterising the distances between the functions PSD = f (wavelength) and PSD = f (frequency) in the direction of the y-coordinate. Definition of a parameter for the evaluation of systematic differences in relation to the roughness of the different tooling inserts and their mouldings.

The regression analysis was used for the evaluation of the systematic differences of the determined functions. The test variable for the comparison of the regression coefficients [[??].sub.1] and [[??].sub.2] can be calculated, according to Rosenkranz as follows:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

[??] = test variable

[??] = regression coefficient

[s.sup.2.sub.y x x] = square standard error of estimator

n = amount of spot samples

[Q.sub.x] = auxiliary variable [right arrow] [Q.sub.x] = (n-1) x [s.sup.2] with

s = standard deviation of the delogarithmised x-values

The PSD-functions of the average values of steel inserts No. 2 and No. 9 are statistically compared (Figure 6). The logarithmical values of the PSD-functions were used for the statistical evaluation. The steel inserts are compared by means of the increase characterisation. The regression analysis resulted in the following: Statistical evaluation of the PSD-function: comparison of the increase of the functions by null hypothesis [[??].sub.1] = [[??].sub.2] (bilateral comparison).

Comparison of the regression coefficients [[??].sub.1] and [[??].sub.2]

Test decision: [[??].sub.1] [not equal to] [[??].sub.2] with a statistical certainty of P = 99%

Test variable: [??] = [absolute value of -12,1571]

Quantile of the t-distribution: [t.sub.252,0,995] = 2,5852

[[??].sub.1] [not equal to] [[??].sub.2] as [absolute value of -12,1571] > 2,5852

From the investigations can be concluded that, with a statistical certainty of P = 99%, there is a significant difference in the roughnesses of steel inserts No. 2 and No. 9 (steel insert No. 2: [R.sub.q] = 2.94 nm, steel insert No. 9: [R.sub.q] = 9.74 nm). The roughnesses of PMMA mouldings of these two steel inserts show analogue results (statistically significant difference).

3. Conclusions

The extensive experimental investigations have proved valuable for the objectification of production issues in the injection moulding process of components with optical surfaces, for further optimisation of the manufacturing process and for the quality assurance of such components. Selected results of the investigations are talked over e.g. in publications.

4. References

Frohlich, M. (2004). Charakterisierung ausgewahlter Eigenschaften optischer Funktionsflachen an ultraprazisionsgedrehten Werkzeugeinsatzen in Abhangigkeit von technologischen Parametern. Diplomarbeit Fachhochschule Jena

Duparre', A.; Notni, G. (2002). Charakterisierung nanorauher Oberflachen. In: DAKOM 2002: Charakterisierung von optischen und technischen Oberflachen. Darmstadt, 27. Februar 2002

Burger, W.; Bliedtner, J.; Rosenkranz, S. & Muller, W. (2004). Roughness measurement at injection moulded plastic parts with optical surfaces. In: Annals of DAAAM for 2000 PROCEEDINGS & of the 15th INTERNATIONAL DAAAM SYMPOSIUM "Intelligent Manufacturing & Automation: Globalisation--Technology-Men--Nature", 3-6th November 2004, Vienna, Austria, 2004, pp. 063-064, ISSN 1726-9679

Bliedtner, J.; Burger, W.; Muller, W. & Roeder, J. (2005). Mikrotopografische Anforderungen beim Spritzgie en: Optische Oberflachen im Blick.,Kunststoffe, vol. 54, no. 3, pp. 48-55, Carl Hanser Verlag, Munchen

Bliedtner, J.; Burger, W.; Roeder, J. & Muller, W. (2003). Rauheitsmessungen an spritzgegossenen Kunststoffteilen mit optischen Funktionsflachen. In: VDI-Berichte Nr. 1806 "Oberflachenmesstechnik", pp. 239-348, VDI Verlag GmbH, Dusseldorf

Rosenkranz, S. (2004). Charakterisierung ausgewahlter Eigenschaften verschiedener spritzgegossener Bauteilsortimente mit optischen Funktionsflachen. Diplomarbeit Fachhochschule Jena

Weckenmann, A.; Ernst, R. (2003). Anforderungen und Randbedingungen fur den Einsatz von Me systemen in der Mikro- und Nanotechnik. In: VDI-Berichte Nr. 1530 "Sensoren und Me systeme 2000", pp. 297-307, VDI Verlag GmbH, Dusseldorf

Authors' data: Prof. Bliedtner, J.[ens], Buerger, W.[olfgang], Rosenkranz, S.[andy], Mueller W., Froehlich, M[aik], University of Applied Sciences Jena, Department SciTec, Carl-Zeiss-Promenade 2, D-07745, Jena, Germany, Fresnel Optics GmbH, Flurstedter Marktweg 13, D-099510 Apolda, Germany, JENOPTIK Polymer Systems GmbH, Am Sandberg 2, D-07819 Triptis, Germany, Jens.Bliedtner@fh-jena.de ,Wolfgang.Buerger@fh-jena.de, sandy.rosenkranz@fresnel-optics.de, maik.froehlich@jenoptik-ps.de

This Publication has to be referred as: Bliedtner, J.; Buerger, W.; Rosenkranz, S.; Mueller, W. & Froehlich, M. (2006). Precision Processing and Microtopographical Characterisation of Tooling Inserts for Injection Moulds, Chapter 05 in DAAAM International Scientific Book 2006, B. Katalinic (Ed.), Published by DAAAM International, ISBN 3-901509-47-X, ISSN 1726-9687, Vienna, Austria

DOI: 10.2507/daaam.scibook.2006.05

Fig. 2. Vickers micro hardness analysis of different Nip-layers
of different manufacturers

PGE
Probe 14 559
Probe 16 541

Northamerican
Probe 22 650
Probe 36 643

Techmetals
Probe 27 548
Probe 41 554

Elektroform
Probe 25 551
Probe 38 561

Note: Table made from bar graph.

Fig. 6. Square values of the average surface roughness [R.sub.q]
of the examined tooling inserts and their PC Lexan LS 2 mouldings

 tooling insert moulding

PC Lexan LS 2

steel no 1 4,04 3,52
steel no 2 2,94 2,33
steel no 3 5,10 3,28
steel no 4 3,26 2,62
steel no 5 2,86 2,46
steel no 6 3,26 1,79
steel no 9 5,74 5,25

Note: Table made from bar graph.

Fig. 7. Square values of the average surface roughness [R.sub.q] of the
tooling inserts steel No. 2, No. 6 and No. 9 and their mouldings
(tooling inserts No. 2 and No. 9--optically polished, tooling
insert No. 9--ultra-precisely turned)

 tooling insert moulding

steel no 2
 2,94
PC Lexan LS 2 2,33
COC Topas 3,07
PMMA transparent 2,57

steel no 6
 3,26
PC Lexan LS 2 1,79
COC Topas 2,56
PMMA transparent 1,28

steel no 9
 5,74
PC Lexan LS 2 5,25
COC Topas 5,70
PMMA transparent 5,43

Note: Table made from bar graph.