Curing of Epoxy Resins for Rapid Product Development by Using Microwave Radiation

Berg K., Christodoulou P., Yariagadda P. KDV., 1998, presented at the IX. Internationales Produktionstechnisches Kolloquium PTK 98, PTK 98 in Berlin, pp. 497-503.

ABSTRACT

Rapid prototyping is a generic process for making rapidly and accurately three-dimensional physical objects of almost any shape. Existing rapid prototyping techniques rely on laser and printing techniques. The use of microwave technology in rapid prototyping has not been explored as yet. In this work, curing process of thin layers of epoxy resins using microwave radiation was investigated as an alternative technique that can be implemented to develop a new rapid prototyping technique. Curing temperature and curing time have been determined for several epoxy mixtures. The mixtures were made up of three commercially available epoxy resins, hardener, aluminium and flydust powder.

The preliminary results showed that the current achievable curing temperature and time required to cure fully the tested mixtures are ~90°C and 75 sec. The powder additions did not change significantly the hardening process. It has been confirmed that the curing process is directly correlated to dielectric properties of the material.

1. Introduction

Most rapid prototyping techniques use the same idea of manufacturing a three dimensional object by building it layer by layer. Using 3-dimensional computer aided design (3D CAD) software an electronic solid model of the object is first developed and then divided into a set of slices to be sequentially built by a relevant rapid prototyping process. In each case the slicing process is adjusted to the rapid prototyping technique to be used as the thickness of the slices and the direction of slicing affect the feature's definition and the manufacturing cost / time of the prototype. Thinner slices give better accuracy, however, the time required for building the prototype is greater and this affects the overall cost. Various Rapid Prototyping developers have developed their own proprietary software to define minimum thickness of the layer and direction of slicing. The minimum thickness of the layer depends also on the viscosity of the material, which has to be considered when mixing with a powder substance. Usually STL files are used for further processes (Jacobs 1996).

Microwave technology can be used as an alternative way for curing layers of material in the rapid prototyping process. It offers the possibility of uniform curing of the epoxy materials by generating instantly heat in the entire part independent of its shape complexity and dimensions. Materials for microwave heating need to be electrically non-conductive and should have a dipole structure. The dipoles are polarised at microwave frequency. When the frequency is high enough the dipoles cannot follow with the reorientation of the electric field and at this point microwave energy is converted into heat. Effective microwave heating is achieved, when the rate of microwave power absorption is greater than the rate of heat dissipation through convection or conduction (Metaxas and Meredith 1983).

Figure 1: Schematic diagram of the microwave system with the slotted waveguide and short circuit.

2. Experimental Set-up

Figure 1 shows diagrammatically the experimental set-up the microwave technology for the rapid prototyping process by using, which has been developed, at Queensland University of Technology. The equipment consists of the microwave control unit, Model: GMP 20K/SM which powers the water-cooled magnetron, Model: NL 10250, 2KW 54 REM at 2.45 GHz. The power output is adjustable from 0 to 1.9KW. Via the rectangular waveguide flange WR 340, the magnetron is connected to a circulator, for protection of the magnetron against reflected power coming from the single post attenuator. The second circulator takes care of the reflected power reflected by the short circuit. The WR 340 waveguide system is used as part of the resonance cavity for the dominant TE10-mode, with variable tuning achieved through the short circuit (Lance 1964).

Applying liquid onto the platform a layer of the slice is generated. The liquid is sprayed onto the platform by electrically activated spray-nozzles. The nozzles are aligned across the platform and electronically controlled, to apply the liquid as a thin film to the platform. Using information from the STL-file a contour of a layer is applied. Simultaneously while the platform moves in horizontal direction, the layer passes the slotted waveguide. The slotted waveguide is connected to a 2.45GHz microwave system generating a fringing field of electromagnetic radiation, which penetrates through the epoxy layers on the platform and accelerates the curing process. This process increases the temperature within the epoxy resin and contributes to fast curing. The vertical movement of the platform is equal to the layer thickness, so that the penetration of the fringing field at the slotted waveguide cures each layer uniformly. Depending on settings microwave radiation can propagate deep into the material making sure that the post-curing is not required.

The material of the platform is characterised by its very low dielectric constant so electromagnetic field penetrates easily the platform without generating heat. The microwave that passes the platform is contained by the water-load accommodated beneath.

The goal of this work is to determine the response of the epoxy systems to their suitability for the proposed new rapid technique. The layer has to be cured quickly to prevent any changes to its form. Nearly instant solidification of the liquid is required to change the liquid into a rubbery form. The thermosetting of the epoxy resin and the microwave heating reapplied for several times in the process of building subsequent layers will advance the curing.

3. Experimental Study

Series of mixtures of epoxy resins with varying amount of hardener (5-40 wt%) and aluminium (0-35wt%, average dielectric loss factor) or flydust (0-35wt%, low dielectric loss factor) powder additions have been tested. There are three reasons for powder additions. Firstly, ability to cure mixtures increases the functionality of prototypes (wider base for material selection). Secondly, the influence of additions with varying dielectric properties on curing process can be examined, and thirdly, the viscosity of the liquid can be controlled. Table 1 shows the compositions of the mixtures used in the present study.

Table 1: Composition of mixtures investigated.

Thin layers of the mixtures of epoxy resins were applied manually onto a glass plate. The thickness of the applied epoxy was varying depending on the composition. The average thickness for each epoxy and mixture type is shown in Table 2.

Table 2: Average thickness of epoxy resin layer applied on microscope - glasslide.

Note: Applying the epoxy resin to the glasslide was done manually.

The specimens were cured in a commercial microwave oven, Menumaster, Model 3100 at 1.5KW power. Microwave ovens usually have very complex wave propagation of electromagnetic fields in their cavity. To ensure uniform radiation the specimens were placed in the centre of the power concentration, which has been determined experimentally.

The temperature of the specimen as a function of exposure time to microwave field, and the time taken to cure the specimens were determined by placing an epoxy sample in the microwave oven. The measurements were taken in 60-second intervals. Every 60 seconds the power was turned off and the temperature of the specimen was measured using type K-thermocouple. At the same time the specimen was checked if it is cured by touching with a sharp needle. The specimen was tested for curing through putting light pressure on the surface of the layer. The results are shown in Table 1 and are summarised graphically in Figure 2 (1-6).

The dielectric constants define the interaction mechanisms of electromagnetic fields with the materials. These interactions depend on the frequency of electromagnetic field and temperature (Sutton et al.1994). Since the dielectric constant determines the ability to absorb microwave energy (von Hippel 1995), dielectric constant for each of the epoxies tested has been determined by using a Network analyser, HP 8510C, connected to a high temperature dielectric probe HP 85070B (Hewlett Packard 1991).

The measurements were performed at different temperatures. The epoxy resin samples were heated using a temperature controlled auxiliary heater. The results of measurement of dielectric constants are given in Table 3 and 4.

Table 3. Dielectric constant e' and dielectric loss e" measured at 2.45 GHz.

Dielectric constant (e') characterises the penetration of microwaves into the material. Dielectric loss (e") indicates the material's ability to dissipate energy. 4. Results and Discussion

he results of the measurements have been summarised graphically. The curing temperature and curing time are presented as a function of chemical composition of the mixture (see Figure 2). Generally, the curing time and the curing temperature are decreasing with the increase of hardener and powder additions (Figure 2). It appears that using LC 3600 resin lower curing temperature and shorter curing time can be achieved. Low curing temperature helps to avoid distortion in the final product. Short curing time is necessary to avoid deformation of the liquid layer, which can flow on the surface. This also depends on viscosity of the mixture.

It is apparent that in all cases the curing temperature and curing time depend almost exclusively on the concentration of the hardener. This is especially evident in the case of LC 3600 resin with one exception where Al additions change significantly the FGI R180 epoxy behaviour. However, not to the desirable extent. The powder additions only marginally change these characteristics.

The measurements of the dielectric constants have been conducted at the frequencies from 1.0GHz to 20.0GHz (broadband) to determine the relaxation frequency as shown in Figure 3. Following the broadband investigation (measurement is established over a wide range of frequencies), there were also dielectric test performed at a narrow-band, frequencies from at 2.2GHz to 2.55 GHz as shown in Figure 4.

The broad band measurement from at 1.0GHz indicates a slight decrease of the dielectric constant (e') and the dielectric loss (e"). Over a wide range of frequencies, the dielectric property does not change significantly until the relaxation frequency at approximately 17GHz.

At this frequency the microwave treatment would be the most effective. However, the operational frequency of the equipment available is 2.45GHz, where the dielectric loss (e") has a value of about 0.9 for epoxy resin Ciba Geigy, LC 3600 at 230C, as shown in Table 3, and also in Figure 4.

As it can be seen the dielectric constant and dielectric loss factor do not vary in the range of 2.2 to 2.55 GHz frequency. Table 3 and 4 show that LC 3600 resin has larger e' and e" constants. Especially, the dielectric loss factor is significantly larger. Considering the curing temperature and curing time with respect to these two constants it appears that a correlation exist. The curing temperature is proportional to e" and the curing time reciprocally proportional (see Table 5, which is formed by extraction of the data from Tables 1, 3 and 4). The good performance of the LC 3600 resin can be attributed to its dielectric properties.

5. Conclusions

Mainly hardener content controls the curing time and curing temperature. The powder additions influence them only marginally. The best results are achieved for the LC3600 epoxy resin. The curing temperature is about 90°C, and curing time is about 75 seconds. Our aim now is to achieve a rubbery state. This will be a part of further investigations. A direct correlation between the dielectric properties and epoxy performance has been confirmed. Based on these results a new set of experiments will be designed to determine relationships between dielectric constants, mixture composition, and temperature.

Figure 2: Curing temperature and time as a function of the composition of the mixtures.

Table 5: Correlation between parameters tested and dielectric properties of the resins.

6. References

Hewlett Packard, 1991, "High Frequency Dielectric Materials Measurements", pp. 62-74.

Jacobs P., 1996, "Recent Advances in Rapid Tooling from Stereolithography", 3D Systems, P/N 70270/10-15-96, pp. 95-118.

Lance A.L., 1964, "Introduction to Microwave Theory and Measurements", McGraw-Hill, New York.

Metaxas A.C., Meredith R.J., 1983, "Industrial Microwave Heating", London: Peter Peregrinus Ltd.

Sutton W., Brooks M., Chabinsky I., 1994, "Microwave Processing of Materials: An Emerging Industrial Technology", National Research Council, Publication NMAB-473, National Academy Press, Washington, DC 1994, ISBN 0-309-05027-8, pp. 61. von Hippel A., 1995, "Dielectric Materials and Applications-Reprint", John Wiley & Sons, Inc., New York.