DSM Engineering Plastics

Controlling temperature is key for additive manufacturing with semi-crystalline engineering plastics

  • Patrick Duis · Innovation manager

  • 5 February 2018

DSM, a global leader in engineering plastics and 3D printing with Somos® SLA (stereolithography) materials, shares more than 25 years of knowledge in the automotive industry. Today, the groups are shifting gears by joining forces in application and material knowledge for additive manufacturing to address the market needs for more materials that are suitable for this industry. With a deep understanding for the needs of the automotive market, DSM has developed high performance polymers for Filament Fused Fabrication (FFF) and Selective Laser Sintering (SLS) and materials for SLA. This paper focuses on the development and challenges of bringing semi crystalline polymers for FFF to the market. By using DSM’s extensive knowledge of application, process and material properties, developments can be sped up by predicting processing and material compositions by modeling technologies including FFF, SLA, SLS and Multi Jet Fusion (MJF). 

Additive manufacturing with plastics is moving into demanding applications with a continuous requirement for increasing complexity, placing additional demands on material suppliers. When considering Fused Filament Fabrication (FFF) or Fused Deposition Modeling (FDM), two polymers currently dominate the market — the biopolymer polylactic acid (PLA) and the more traditional polymer acrylonitrile butadiene styrene (ABS). These two materials are popular for several reasons, they have good printing behavior and they’re not too expensive and, in the case of PLA, it is also more “eco-friendly.”

The downside of these materials is that the mechanical and thermal properties, especially for PLA, are underwhelming. In addition, they do not have excellent resistance to high temperatures or weathering. So, it’s clear that the FFF 3D plastics printing market needs materials that have a better combination of printability and in-use properties; meaning producers of engineering thermoplastics need to develop grades specifically for FFF. However, producers of FFF 3D printers and their customers need to be aware that engineering thermoplastics may demand special attention to processing conditions, which may require some hardware engineering and software tweaking.

DSM has developed a small family of semi-crystalline thermoplastics for FFF 3D printing, including grades of Novamid® ID polyamide 6 and 6/66, Arnite® ID PETP and Arnitel® ID thermoplastic bio-based co-polyester elastomer. As a major developer and producer of engineering plastics for many processes and applications, DSM understands that a deep insight into the 3D printing process is key to understanding material behavior and this insight is leading the company to develop products quicker and more accurately. 

With the current state of FFF 3D printing, the mechanical capabilities of the technology are exceeding their thermodynamic capabilities. Printers can reach speeds of up to around 300 mm/s, however in many cases when engineered thermoplastics are used, the mechanical properties of the finished parts are inadequate because of poor inter-laminar strength caused by poor thermal bonding. The principal problem is that while the printing equipment is highly capable of depositing the filament very accurately at a high speed, the equipment’s capacity does not match this ability to melt the polymer sufficiently. Next to this, a controlled print chamber helps to print with lower warpage of the part. 

To raise the temperature to this point, heat is added from the surface of the nozzle with an external heater block. As already mentioned, the thermal conductivity of a polymer is low (0.3 w/mk). Therefore, it will take some time to reach a homogeneous temperature distribution at the nozzle exit. Using numerical simulations of the heat balance in a frequently used hot-end, a “v6” from UK 3D printer component supplier E3D, it is possible to demonstrate how much the homogeneity of the temperature distribution decreases with increasing printing speed.  


Figure 1. Typical FFF nozzle temperature distribution with increased printing speed.

The calculated temperature contours in the hot-section of the E3D extruder as shown in Figure 1 are achieved with a nozzle temperature of 240°C and a filament feed temperature of 25°C. The flow rate used is that for printing widths of 0.5 mm in 0.2 mm layers, using the indicated printing speed. Also, a longer nozzle can enhance the material heating (ex. volcano nozzle by E3D). Material parameters are for Arnitel® ID 2045, a bio-based grade of co-polyester TPC, which was developed by DSM for FFF printers.

Next to this effect, controlling the print environment is increasingly important when changing from amorphous to semi-crystalline materials. First, we need to ensure that the material is melted in a controlled method. The trend to semi-crystalline materials is driven by the engineering performance of these materials, including better mechanical properties, long term heat stability, chemical resistance, etc. The main challenge is the increased volume change when cooling down from melt to solidification, typically known as shrinkage due to the crystal structure. This shrinkage results in warpage and prints that fail. The typical FFF build plate is made of glass and from this glass substrate, the part will need to be built using adhesives to keep the polymer bonded to the glass. If the ‘warpage’ forces become too large, the part-glass interface fails and the part is torn off the substrate.   

A way to reduce this is to “control “the printing environment and print with higher environment temperatures (above the glass transition temperature - Tg), which minimizes the chance of a bad print. DSM has developed an analytical model to predict the crystallinity and thermal behavior during the FFF process, which creates better insight in the way layers start to crystallize in the process. By predicting the processing settings, DSM can tune the process settings to reduce the effect on warpage. An anther advantage is that also putting in material parameters and modeling a material that can print well, while keeping the warpage under control.  

Figure 2. Shrinkage effect between layers.

Figure 2 includes two examples on the temperature control and distribution in the nozzle and how the crystallization behavior can be influenced by the process. A simple geometry is modelled with all the parameter settings of the process (print speed, temperature of the melt, bed, and chamber and print speeds). DSM combines this data with all the material parameters to model the heat distribution and the crystallization behavior throughout the print and now can look at the printed substrate at a certain position. In Figure 3, the print is at 2.5 and 8mm during printing. All parameters can vary to control the crystallization speed.  

Figure 3. Model of a printed sample for modeling the process.

Figure 4 shows the thermal history of the part during the print job. The ambient temperature is 25°C and the Temperature of the heated glass substrate is 125°C. The first layers are close to 110°C, however, at a higher build geometry the temperature is close to 80°C. The effect of the heated bed fades away when the build is evolving.

Figure 4. Thermal history and temperature behavior at Tamb 25°C and Tbed 120°C.

When the temperature history is known, we can link the model to crystallization and volumetric shrinkage behavior. In this way, we can “control” the crystallization behavior during the build of the part. In Figures 5, 6, 7 and 8, the chart on the left shows the temperature history that is plotted at a certain condition and the plot on the right is the crystallization conversion. When the material is fully crystallized, the value is 1. When we apply different settings, you can observe that the layers in crystallization behavior can be controlled. The dotted line in the right chart shows the difference between standard PA6 and optimized PA6 (Novamid® ID 1070) for additive manufacturing. The difference in crystallization speed is clearly demonstrated. Slower crystallization helps to fuse the layers together, hence a stronger layer-to-layer adhesion.

Figure 5. This figure demonstrates that the first layers crystallizes first.

Figure 6. When all temperatures are low, crystallization is hindered.

Figure 7. When the ambient temperature is high, the last layers crystallize first.

As shown in the figures 5,6, 7 and 8, we can control the crystallization of a printed FFF part by controlling the environment and bed temperatures. In figure 8, we clearly see that when we apply both temperatures around 100 degrees, we can print Novamid® ID1070 (PA6) with limited warpage effects. 

We can “steer” the warpage and create parts that can be used for functional prototyping, as well as end use parts. Next to the crystallinity, we can also model the effect on layer thickness and printing speed on the fusion strength between layers. 

Currently, DSM is conducting trials on newly developed machines with heated chambers to obtain the right settings for the printer material combination. By fully understanding the process and material parameters, we can build a strong part with low warpage to meet the demands from the automotive industry.