Table of Contents ECNDT '98 Session: Materials characterization

On-line Process Monitoring of Thermosets by Ultrasonic Methods Joachim Döring, Wolfgang Stark Bundesanstalt für Materialforschung und -prüfung, Berlin 12205 Berlin, Unter den Eichen 87 Gerhard Splitt Krautkrämer GmbH & Co. 50354 Hürth (Efferen), Robert-Bosch-Straße 3 Corresponding Author Contact: Joachim Döring Email: Joachim.Doering@bam.de , URL: http://www.bam.de

Introduction

The demands on industry in relation to productivity, quality and price , means an ever increasing need to improve the quality system. At the heart of this system in the manufacture of plastic components by compression or injection moulding, lies process control. Of which, the most important element is the measuring device or sensor which measures process variables such as mould temperature and pressure. The data yielded by the sensor is processed by the computer (integrated), often with incorporated quality control features such as SPC and possibly through a computer network to a controlling computer which monitors several machines simultaneously (bi-directional system). In order to optimise process control sufficient information is required about the whole process (e.g. mould temperature, pressure etc.)which often varies depending on the material used. [1].
Up till now it has been difficult to obtain data about the current mechanical state of the material being moulded. This is an important factor in processing control particularly for thermosetting polymers which undergo a chemical cross-linking reaction (curing) with the application of heat e.g. during processing. The degree and rate of curing, which depends on processing parameters such as mould temperature, curing time, injection speed etc. have a substantial effect on the quality and mechanical properties of the final product. This paper describes a method using ultrasound technology which provides a means for the in process cure monitoring of thermosetting materials.

The measuring equipment used monitors the evolution of mechanical properties in the thermoset during curing. The through-transmission (through the material) ultrasonic measurements are made with ultrasonic sensors fixed within opposite sides of the mould walls. The variation of the mechanical properties, for instance viscosity or mechanical modulus, is measured as a variation of ultrasonic attenuation and ultrasonic velocity. In this measurement the ultrasonic transducers register the relaxation process of molecular segments. The idea of curing control is not new [2], however most methods were not viable for practical application in industry. Thus this new technique has been developed for use in industry, in order to improve both quality and reliability of thermosetting products.

Determination of the mechanical moduli

The following steps illustrate the thermosetting moulding process:
• Softening (70°C...90°C),
• Mould filling (140°C...180°C),
• Moulding under pressure (200bar...600bar) and net work building (curing) and
• Ejection of the product.
In order to determine the mechanical state an ultrasonic pulse signal is transmitted through the moulding compound. This pulse excites molecule segments which start oscillating. This relaxation process attenuates the ultrasonic waves [3].
Therefore the attenuation depends on the actual molecular structure and consequently on the degree of curing of the moulded compound. The sound velocity is also dependent on the material structure and thus on the degree of curing (the velocity of sound increases with the increasing degree of cure). If the attenuation coefficient and the sound velocity are used as a base, then the longitudinal module L* is [4]

L* = L' + iL''
L' = vlong2 * ,	with vlong - longitudinal wave sound velocity, - density,
L'' = 2vlong3* long / ,	with long - longitudinal wave attenuation coefficient,
	- angular frequency.

The transversal waves yield the shear modulus G*:

G' = vtrans2 * ,	with vtrans -shear wave sound velocity, - density and
G'' = 2vtrans3* trans / ,	with trans - shear wave attenuation coefficient,
	- angular frequency.

Both moduli are necessary, in order to determine the complete set of mechanical modules:

K* = L* - 4/3 G* ,

with K* - compression modulus.

The more important wave type is the longitudinal wave. It can be detected in thermosets even when they have a low viscosity.

Measuring equipment

Fig. 1:Measurement equipment

The moulding compound is pressed and heated in a compression mould. Mounted within opposite mould walls are two ultrasonic transducers as sensors [5]. They were specifically developed for use under industrial conditions (high temperatures and pressures). One transducer acts as an ultrasonic transmitter and emits short ultrasonic pulses which are received during curing by an ultrasound transducer at the other side of the moulding.
The detected signals are analysed, in order to determine the amplitude of the ultrasonic pulse and its delay time. For this purpose an ultrasonic receiver is used (Fig.1). Simultaneously variations in sample thickness are measured (which are required for sound velocity calculations). The final data processing is made by a PC.
A typical measurement, showing amplitude response is illustrated in Fig.2.

Fig. 2: Typical amplitude response during the thermoset moulding process Breakdown of the curing curve (s. Fig.2):

• On closing the compression mould the ultrasound signal reaches the receiver. The curve shows a first maximum.
• On heating, the thermoset material softens and the amplitude decreases, i.e. the material goes through the glass-rubber-transition.
• Softening is following by curing.
• Amplitude stops increasing as the curing process comes to an end.
• The moulding process ends with the ejection of the product (mould opening).

Using the measured ultrasonic amplitude and the thickness response the variation of the attenuation coefficient can be calculated:

= ln(A/Ao ) / d,	Ao - amplitude in the saturation part of the curve,
	A - current amplitude and
	d - specimen thickness.

The attenuation coefficient equals plus a constant. In this way the change of the attenuation coefficient can be determined as a material parameter. Fig.3 shows a typical curve. This curve is similar to that shown in Fig.2, but it is independent from the specimen thickness.

Fig. 3: Change of attenuation coefficient

Experimental results

The experiments were carried out using a compression moulding machine. This machine was used in order to imitate industrial conditions. One motive was to verify the practical use of this method. Fig.4 and Fig.5 show such examples. Although it is well-known, that the curing time depends on the temperature, this can now be quantitatively determined. This data could be used as a basis for the precise control of thermoset processing machines.

The curing curves of Fig.5 illustrates the behaviour of different thermosets. Typical materials are urea - UF, melamine-phenolic - MP, melamine - MF, phenolic - PF, epoxy - EP and unsaturated polyester. Each material has a specific amplitude response, like a finger print. Also these characteristics could be used to set machine parameters. For that purpose they are calibrated by a correlation with current empirical methods (e.g. boiling test).
In this way it is possible to determine the point of mould opening for an optimal cure.

Fig. 4: Amplitude responses at different temperatures

Fig. 5: Amplitude responses of different thermosets

A more complete view of the mechanical material properties during the curing process is obtained using longitudinal and transversal wave sensors. The change of attenuation coefficient and of the sound velocity is shown in Fig.6 and Fig.7, which were measured by both kinds of sensors. A thermoset was investigated at first with longitudinal wave sensors and then with transversal wave sensors (Fig.6). In this way the whole set of mechanical moduli can be determined. The same procedure is used for a cast resin (Fig.7). Note, that the signal of transversal wave sensors cannot be detected before curing has begun. During softening the attenuation is very high.

Fig. 6: Change of attenuation coefficient and sound velocity of melamine-formaldehyde (MF)

Fig. 7: Change of attenuation coefficient and sound velocity of a cast resin.

The investigations of curing control for injection moulding machines have direct practical importance. The amplitude response, obtained by sensors built into an injection moulding machine, is shown in Fig.8. The behaviour of the material in injection moulding machines differs from that in compression moulding machines, because it is heated before it comes in the mould, therefore the softening part is very short.

Fig. 8: Cure monitoring in an injection moulding machine.

Fig. 9: Amplitude response and delay time for polypropylene

This technique was tested for thermoplastic process control too. Thermoplastics are injected in molten state into a cooled mould, where they solidify. In order to find the optimal point for mould opening, the point of solidification must be known.
Fig.9 shows the curve of amplitude response and the curve of delay time for polypropylene. The transition point is clearly seen in transition time (increase of sound velocity) and attenuation.

Summary

The usual methods of the determination of curing degree are not suited for the modern quality requirements. Therefore a sensor equipment was developed for cure monitoring in compression and injection moulding machines. The investigations have shown, that the measurement equipment characterised the curing process well. This measurement equipment is a solid base for curing process control.

Acknowledgements

The authors wish to thank the following colleagues for close co-operation and friendly support:
Dr. W. Mielke, Prof. H. Wüstenberg, Dr. J. Kelm, Dr. G. Kalinka P. Fengler,
V. Lehmann, J. Roltz, M. Schneider, (BAM),
Prof. P. Thienel (Engineering College Iserlohn) and
M. Stahl (Bakelite Letmathe).
Special thanks to M. Rath, G. Vollmer, J. Mc Hugh and B. Hantschke for the experimental work performed.

References

1. K. Schröder: "Qualitätsüberwachungssysteme", Spritzgießen von Duroplasten, Hrsg. K. Niemann, K. Schröder (Praxis Kunststoff-Verarbeitung; 7), Dr. Alfred Hüthig Verlag, Heidelberg, 1994
2. G.A. Sofer; G.H. Dietz; E.H. Hauser: "Cure of Phenol-Formaldehyde Resin", Industrial and Engineering Chemistry 12(1953)2743-2748
3. H. Kuttruff: "Physik und Technik des Ultraschalls", S. Hirzel Verlag Stuttgart, 1988
4. Alig I. Alig; K. Nancke; G.P. Johari"Relaxations in Thermosets.XXVI. Ultrasonic Studies of the Temperature Dependence of Curing Kinetics of Diglycidyl Ether of Bisphenol-A with Catalyst", Journal of Polymer Science B 0(1994)1465-1474
5. J. Döring, W. Stark "Aushärtung von Duroplasten mit Ultraschall",Materialprüfung, 7-8 (1997)308-311