Boron nitride nanotubes (BNNTs) are a new, man-made molecular fiber with revolutionary potential. They are an electrically insulating and high-temperature analog to carbon nanotubes (CNTs), with the highest strength of any man-made material. When produced with a small number of walls and a very high aspect ratio, they have very unusual properties. For example, a BNNT has a thermal conductivity eight times that of copper but with a high insulating ability. Among other virtues, these nanotubes have the potential to impart strength, temperature and oxidation resistance, thermal conductivity, and radiation protection to composite materials.
Until now, scarcity has hindered the adoption of this material. Although the material was discovered in 1995, synthesis poses many challenges, and the material has not been available commercially until recently. BNNT, LLC is the first company to manufacture and sell high-quality BNNTs on a commercial basis. Our high temperature/high pressure (HTP) method requires working with extremely high temperatures at moderate pressures, which provides interesting challenges when managing the synthesis operation. At BNNT, we use simulation products obtained through the ANSYS Startup Program to help us explore new avenues in synthesis and to optimize our process.
Specifically, we use ANSYS Fluent to study many effects of gas flow in our synthesis chamber. For example, in their natural, unprocessed, and uncollected state, BNNTs have a fluffy, cotton-like appearance, and tend to float around. They like to stick to anything they encounter. Using Fluent, we were able to design the process gas entry to our chamber to help keep a critical viewport clean. The figure below shows our original configuration, which was simply an inlet port that we placed near the window:
Actually, while this configuration wasn’t bad, we thought we could do better. The flow is low, but we don’t want to increase it too much. We added an air knife arrangement to try to keep contaminants off the window. If you look at the image below, the problem looks as though it has been solved, but a section through the middle shows that the recirculation is just as bad, and contamination from the chamber can still be entrained into the flow near the window.
Next, we added a baffle to try to reduce the recirculation. The cross-section shown below left looks very good, but now on the right, the coverage over the window is not very good:
We thought we could perhaps improve the coverage by adding vanes and changing the baffle design. Fluid dynamics is tricky, and often things do exactly the opposite of what you think they should do. Adding vanes and an additional baffle made everything worse:
On close examination of the flow patterns at different flow rates and in different regions, the reasons for this became obvious, and finally, after several iterations of changing the vane geometry, we found a solution that produced good results and was relatively intolerant to changes in flow rate, as shown below:
As we accumulate knowledge from running our process and become more comfortable with the advanced multiphysics capabilities of ANSYS products, we will continue to study more complicated parts of our operation. We expect to see major increases in efficiency, quality, and production rate as we begin to understand and control the conditions more completely.