ÂÌñÏ×ÆÞ

In a previous article we reviewed rapid forming of thermoplastic composites (TPC). In the second part of this three-part series, we’ll explore the thermal processes and how these can be managed, particularly during rapid forming.

Major advantages of TPC are that they can be formed quickly and can be reformed multiple times because melting and solidification are physical processes and do not involve a chemical reaction. However, the thermal processes of heat-up, melt processing and cooldown must be managed to ensure the polymer is in the desired state, particularly during final cooling.

The key process steps are different between thermosets and TPC. For thermoset composites, the critical stage is heat-up, because that is when viscosity reduction, resin flow and gelation occur. For thermoplastics, though, the cooling phase is the more important.

As we previously discussed, the thermal cycle for TPC can be very well managed and reproducible but it is necessary to understand the steps and the aspects that are critical in each step. A typical stamp forming thermal cycle is shown in Fig. 1, illustrating thermal traces using embedded thermocouples for a carbon fiber reinforced polyetherketoneketone (PEKK) part with four key process steps: rapid heating, melting, very fast cooling on contact with the tool and then equilibrating at the tool temperature. We’ll refer to this example throughout the article.

Polymer thermal properties

Materials suppliers can provide polymer thermal properties such as:

• T_g: Glass-rubber transition temperature (polymer property)
• T_m: Crystalline melt temperature (polymer property)
• T_p: Process temperature
• T_c: Crystallization temperature.

These properties are readily available, but it is still valuable for processors to assess these characteristics and differential scanning calorimetry (DSC) is widely used to measure these thermal properties. A typical heat-hold-cool DSC trace for a carbon fiber/PEKK composite is shown in Fig. 2. The inflection at T_g is small because of the reinforcement and because the relaxation at T_g is due to the amorphous portion of the polymer only. The polymer crystallizes on cooling, and it is notable that the T_c is lower than the T_m due to “super cooling” effects in polymers.

Single values for T_m and T_c are reported in datasheets, but in practice the processes take place over a range of temperatures. While T_g and T_m are polymer properties, T_c depends on the process conditions as we’ll see later.

Amorphous vs. semi-crystalline polymers

TPC polymer selection is based on application requirements and cost. End-use requirements will dictate whether an amorphous or semi-crystalline polymer is appropriate, and this will have an impact on the processing. The main characteristics of amorphous and semi-crystalline polymers are shown in Table 1.

Morphologies of the major polymer families are shown in Table 2 and typical polymer stiffness versus temperature is shown below in Fig. 3. Amorphous polymers do not have any ordered structure and start to flow above the T_g. The polymer viscosity decreases with increasing temperature, resulting in a wide process window. In semi-crystalline polymers, a portion of the polymer is in the ordered, crystalline state, typically 20-40%, with the remainder of the polymer being amorphous. This amorphous fraction relaxes at T_g, but the polymer will not flow because the crystalline portion constrains mobility. Above T_g, a semi-crystalline polymer is in a solid, rubbery state and retains stiffness and strength. The polymer will only start to flow when the crystalline melt temperature (T_m) is reached and therefore the process temperature (T_p) must be above T_m. Many of the polymers of interest for TPC are semi-crystalline due to their combination of properties, including high mechanical properties, toughness and resistance to high temperatures and chemicals.

Several of the semi-crystalline polymers in Table 2 have been used for decades, and managing processing is well understood, although some concepts may be new to the composites community.

Heat-up and melting

TPC can be heated at any rate to T_p. The major difference is that amorphous polymers will change from a viscoelastic solid to viscous fluid at T_g and flow will increase with temperature. Semi-crystalline polymers have a distinct melting point where the polymer chains in the crystallites disassociate and start to flow past each other, and therefore T_p will always be higher than T_m. Material suppliers provide T_p ranges in their technical datasheets, although some experimentation may be needed to determine the conditions for a particular application.

The lower bound of T_p is based on reaching a sufficiently low viscosity for the polymer to flow and form the part. The upper bound is based on avoiding undue polymer degradation, which depends on a combination of time and temperature (Fig. 1). Many TPC fabrication methods take place in a normal atmosphere and polymer degradation is faster in the presence of oxygen. For example, as a rule of thumb, the maximum T_p for polyaryletherketones (PAEK) is 400°C to avoid undue degradation. The polymer should spend at least a short time in the T_p range — e.g., 2-5 minutes — to ensure that polymer chains are truly disassociated and can flow past each other. Part consolidation or forming must take place while the polymer is in the melt state and before it cools to T_c for semi-crystalline polymers, or T_g for amorphous polymers.

Cooling and solidification

TPC with amorphous matrix polymers must be cooled below T_g to be form-stable. Parts fabricated with amorphous polymers can be cooled as fast as possible because the cooling rate does not affect final performance, and room temperature tools may be used in stamp forming, which enables very fast cycle times.

For composites with semi-crystalline polymers, the cooling conditions may affect the degree of crystallinity and thus the properties of the finished part. Crystallization occurs over a specific temperature range and is dependent on cooling conditions. The cooling rate can be controlled in some processes such as compression molding, oven consolidation, etc., and it is important that the polymer spends a sufficient time in the crystallization temperature zone.

The effect of cooling rate on crystallization is shown in the DSC charts in Fig. 4, where the T_c is not a fixed value and depends on cooling rate. With faster cooling rates, the crystallization temperature decreases and the crystallization temperature window widens. All samples in Fig. 4 develop full crystallinity even though it occurs at different temperatures.

In some processes — such as stamp forming and dual-press forming — the molten blank is transferred to a cooler tool, held at a constant temperature and the crystallization takes place under isothermal conditions (Fig 1.). The rate of crystallization varies considerably with the isothermal temperature as shown in Fig. 5, which highlights the crystallization rate on a log scale.

The data to develop this type of chart must be measured using different methods as described in the paper “Crystallization and Melt Kinetics for Process Modelling of PEEK Matrix Composites” by Gordnian, Vaziri and Poursartip from the 2017 SAMPE Conference in Seattle. The interpolation of the data shows that the rapid crystallization range is 210-250°C, with the fastest rate of crystallization at ~230°C. The tool temperature should be in this range to ensure rapid and full crystallization at a constant temperature. The crystallization behavior for any thermal history can be modeled using predictive software such as RAVEN from .

Dimensional change

The polymer volume change for a semi-crystalline matrix composite during cooling from the melt is shown in Fig. 6. At T_p the polymer will flow in the melt state, and as the polymer cools the viscosity increases until T_c is reached. This is the “stress-free temperature” for TPC. The polymer undergoes significant volume change when it crystallizes and solidifies. As the part continues to cool the polymer stiffness increases and the volume decreases, creating internal stresses. Below T_g there is continued decrease in volume although at a lower rate due to the change in the amorphous regime below T_g. As we discussed in Part 1 of this article, residual stresses and part deformations can be calculated using available modeling tools, which can be used to design thermally compensated tools, to ensure that the parts meet the required dimensional tolerances.

Repeated thermal cycles

TPC can be thermally processed multiple times, as happens during each step in this typical process chain: fiber placement of plies, laminate consolidation, part forming and welded assembly. In each process, the polymer is heated above T_m, although the thermal cycle will be different. For example, fiber placement uses a high temperature for a very short time, whereas consolidation is typically at a lower process temperature but for a longer time. A representative cumulative thermal history with multiple process steps is shown in Fig. 7.

Thermoplastics are very robust but there will be a small amount of degradation in each cycle which will typically cause a small increase in viscosity, and for semi-crystalline polymers, a decrease in the rate and potentially the level of crystallinity. The combined cycles can be assessed by running a sample of material through a DSC cycle which represents the actual process, or performing DSC on parts that have been through the complete cycle. In general, thermoplastics used for high-performance composites have been shown to be very robust to repeated processing so long as conditions are controlled in the normal ranges.

Extensive information on processing of TPC is available from materials suppliers, and modeling methods are well developed to simulate process cycles. Fabrication of TPC is well established, with millions of parts manufactured every year using a range of methods. With an understanding of the materials and process methods, thermal cycles can be established, and parts can be processed successfully and reproducibly.