Filtern
Erscheinungsjahr
Dokumenttyp
- Zeitschriftenartikel (9)
- Beitrag zu einem Sammelband (6)
- Beitrag zu einem Tagungsband (6)
- Vortrag (5)
- Posterpräsentation (5)
- Dissertation (1)
Schlagworte
- Aushärtung (2)
- DMA (2)
- Epoxy (2)
- Ultraschall (2)
- Ultrasound (2)
- Activation Energy (1)
- Compositbauteil (1)
- Composite material (1)
- Composites (1)
- Cross-linking (1)
Eingeladener Vortrag
- nein (5)
Epoxy composites used for high-end structural applications are typically cured under the influence of temperature and pressure causing a number of complex chemical and physical transformations. On heating a mould, temperature gradients will occur through the component which will depend largely on the thermophysical properties of the specific composite. The crosslinking reaction is exothermic leading to additional heat release, thus complicating heat transportation models. If such effects are not accounted for, it can lead to variations in resin flow, poor fibre wetting causing voiding and inhomogeneous cure, leading to shrinkage and unfavourable variations in moulded part geometry. Limited information is available for thermal models used in the manufacture of reinforced thermosets. Autoclave [1, 2] and laser or infra-red curing processes [3, 4] typically use constant values determined on fully cured parts. In this work, the variation in thermal conductivity (K) (W/(m K)), thermal diffusivity (a) (m(2)/s) and specific heat capacity at constant pressure (c(p)) (J/(g K) is determined for a carbon fibre prepreg during cure. It is the intention to improve understanding of how these parameters are related to chemical or physical transformations occurring during cure, and where estimates or shortcuts may be used for heat transfer models
Epoxy carbon-fibre prepreg, Hexcel Type 6376 HTS, was investigated using Dynamic Mechanical Analysis (DMA). The DMA characteristic parameters are storage modulus E', loss modulus E' and loss factor tanδ. These parameters are ideally suited to observe the vitrification, referred to as glass transition, resulting from the cross-linking reaction. Detection of the cure state may also be achieved by determining the momentary glass transition temperature of partially cured samples. The consequent use of a multi-frequency measuring regime was used to derive the apparent activation energy for the glass transition process. Different temperature programs were also applied to monitor the curing process directly, as well as to investigate the different states of incomplete cure reached in preceding curing steps. The intention was to provide better understanding of the consequences of an interrupted autoclave curing process and to use DMA to detect the cure state achieved. With DMA, the continuation of an incomplete curing process also can be monitored. DMA measurements up to 300 °C showed, furthermore, that the final glass transition temperature was reduced by thermal degradation at high temperatures.
The technique normally used to measure Cp during isothermal cure is Temperature Modulated - Dynamic Scanning Calorimetry TM-DSC. It is however not standardised, experimentally complicated and quite time intensive. As will be shown, Cp may also be estimated during isothermal cure just from using dynamic heating experiments on a fully cured sample. Such values are often sufficient for isothermal heat transfer models that otherwise employ a constant Cp value obtained from the fully cured epoxy. Secondly, the results from dynamic heating experiments provide a quick means, in comparison to isothermal TMDSC measurements, of estimating Cp variation during cure as well as providing a good estimate value for Cp towards the end of isothermal cure. As will be shown, such values obtained from a standardised measurement procedure are very helpful in setting up TMDSC experiments that are more sensitive to experimental error influenced by factors such as sample weight and geometry.
The DSC results illustrate that the measured heat capacity Cp for a fully cured epoxy over a temperature range are very similar to values for samples partially cured at corresponding isothermal temperatures, under the prerequisite that vitrification takes place. In such cases the primary influence on Cp is specific measurement temperature and not degree of cure. For isothermal cure temperatures investigated between 150 and 200 °C, the total change of Cp during cure is nearly constant and correlates well with values published by authors on other epoxy based systems. Taking Cp variation as constant, it is possible from just dynamic heating experiments on the cured epoxy to estimate Cp for the uncured epoxy system at specific cure temperatures. The next step would be to estimate the full Cp profile during isothermal cure, however, in such cases, the time to vitrification would also be needed as additional information.
The objective of this work is to demonstrate the practical application and sensitivity of ultrasound as a high frequency Dynamic Mechanical Analysis DMA technique for the characterisation of polymers. Conventional DMA techniques are used to determine thermo mechanical behaviour of polymers by typically employing dynamic shear or tensile loading modes at defined frequencies between 0.1 and 50 Hz. Sound waves may also be employed for DMA applications and depending on type of wave propagated, shear G´, G´´ and longitudinal L´, L´´ storage or loss modulus and tan (δ) may be determined from the measured acoustic parameters sound velocity and amplitude. The primary advantage of ultrasound DMA is that due to the compact sensor size it can easily be integrated into most manufacturing processes. To demonstrate the sensitivity of ultrasound to variations in the viscoelastic properties of polymers, the acoustic properties of a cured epoxy with an observed glass transition temperature of 86 °C (tan(δ) peak, 1Hz) were monitored in a temperature range from 20 to 200 °C and compared to conventional DMA results. The influence of measurement frequency, dispersion, hysteresis, reflections at material boundaries, and changes in material density on the measured sound velocity and amplitude were taken into account. To support conclusions a wide range of experimental data was evaluated using sensors operating in the frequency ranges 400 to 800 kHz and 3 to 6 MHz. The ultrasound results are compared to the tensile moduli E´, E´´ and tan(δ) measured using a conventional DMA technique operating at 0.1 to 33 Hz. Using different evaluation strategies such as the Williams Landel Ferry WLF equation it was possible to study the sensitivity of wave propagation to variations in the viscoelastic behaviour of a polymer. Taking advantage of this background knowledge, further experimental results are presented with the aim of demonstrating the sensitivity of this technique for cure monitoring applications and to the material transformations: gelation and vitrification. For this purpose an epoxy resin was cured at a range of constant temperatures whereby the curing reaction and the corresponding change in viscoelastic properties were monitored. Analysis techniques employed included ultrasound at 3 to 6 MHz, Differential Scanning Calorimeter DSC and Rheometry at 1 Hz. All results were summarised and presented graphically. Additionally an Arrhenius relationship was employed enabling direct comparison of results obtained from analysis techniques based on different working principles. Using this information, it was possible to demonstrate the practical application and the sensitivity of this technique to even small changes in viscoelastic properties of polymers.