Making a vacuum chamber for an electric propulsion experiment

Paschen’s law has made some enormous contributions to the study of electrical discharge in various gases and their properties in the field of electric propulsion. This law illustrates a relationship between the applied voltage and the product of the pressure and the distance between the electrodes. To experiment and conduct further research into this field, an electrical breakdown facility has to be developed. This facility is required to withstand up to 1 millitorr of pressure for a suitable experimental environment, with a transparent vacuum chamber and also a movable electrode configuration. This vacuum chamber was to be completed in 10 weeks.

Design

It was critical to completely understand the basic requirements of the project before starting any work. A transparent vacuum chamber with a movable electrode configuration that is able to withstand 0.001 Torr of pressure was required for usage as a teaching aid. These requirements indicated that the usage of Pyrex glass or acrylic would be effective as they are transparent materials with the required strength. Various auxiliary needs arose as a result of the main requirements. As this chamber would be used for demonstrating plasma discharge, it was also important to have an insulated system apart from the two electrodes to ensure the isolation of the discharge only between the two electrode plates. This meant that the electrode configuration had to be insulted. Understanding the requirements and their consequences clearly allowed for the conceptualization of design solutions.


Conceptualization

Solutions to engineering challenges often require multiple solutions that have to be comparatively analyzed and tested. Three final designs were discussed and conceptualized in CATIA. It was decided that the standardized existing components in the market would be used to ensure the simple design and lesser time delay for manufacturing new components from scratch. This proved to be less time consuming and effective as these parts also followed universal standards which allowed for flexibility in design. Computer-aided design (CAD) models of parts for forming sub-assemblies were downloaded from a widely recognized website focusing on vacuum solutions. These parts were put together into sub-assemblies and assemblies to result in a conceptual model of the vacuum chamber design. Here are some models which were developed in the process.


Manufacturing & Machining

Based on the above design with some additional modifications, parts, and raw materials were ordered. Once the parts and material stock arrived, they were machined according to the design for assembling them into a chamber. It was noted that a lot of material and stock were used in machining processes such as facing and cutting, which needed more stock than required for the design. For example, a cylinder of Teflon with a diameter of 6.5 inches was purchased for a required disc of diameter 6 inches. This ensured that enough stock was left for manufacturing after the machining processes. Many times, the parts were damaged and rendered unfit for use while machining which required us to have excessive inventory to prevent a backlog from ordering the parts repeatedly.


Assembly

Upon completion of machining various parts and components, assembling was the final step for the manufacturing process of the chamber. It was seen that the tolerancing on various parts was immensely essential for good fits and fixed between different parts. For instance, an ISO-160 half nipple with an outside diameter of 6 inches would not fit axially to an interior of an acrylic tube with an inside diameter of 6 inches. But understanding their tolerance enabled a good fit because both the parts had some leeway as they were +/- 0.0005 inches off from the specified measurements. Since the majority of parts ordered online from the same vendor, it was extremely simple to join two components as they were standardized fits. At the conclusion of the assembly of the vacuum chamber, a table setup was developed to rest it on, so that it can be easily seen by students in the classroom. The complete model is shown in the image below.


Testing

At the conclusion of the design assembly, testing was a critical step to ensure that all the requirements were fulfilled. The testing of the vacuum chamber was a multiple step-based process. The chamber had to be tested in terms of the following aspects: chamber pressure levels, electrode insulation, and protection, assembly portability, and maintenance. Here are some of the plasma discharge images from the testing phase.

Vacuum chamber pressure levels

It was observed that the acrylic walls of the chamber and the ISO-160 half nipple had to be well sealed to attain the required pressure readings. This arrangement was a challenge because acrylic tube and the metallic half nipple had to be mated well with each other. Due to this, a viton o-ring was used to ensure proper sealing under vacuum. This led to a good configuration which brought the chamber to around 70 millitorr but wasn’t efficient in getting the pressure lower. After close inspection, it was noted that his was a result of the unpolished ends of the acrylic cylinder which had minute radial cleavages which provide an easy channel to the air molecules to get in the chamber. This led us to use an o-ring – acrylic ribbon sandwich setup which minimizes this problem.

Electrode Insulation and protection

While testing for the plasma discharge in the initial cycles of running the chamber, it was noticed that the plasma was discharged with the backside of the plate due to its null potential acting as a ground rendering it to be an easy path for the discharge to occur. This was prevented by using a ceramic tube inside the steel tubing to insulate it. This design proved to be an effective one. Another addition to refraining any arcing of the front electrode face with its back was to add a Teflon housing to the entire electrode disc setup to insulate it from further arcing. This also proved to be working well during the test cycles.

Assembly portability and maintenance

After the chamber was assembled, it was essential that it was portable and easily set up anywhere. So, a u-plate setup was used to mount the half nipple on it to have both the ends of the chamber stand without rolling. Further these, plates were mounted on two 80/20 parallel rods to ensure sliding for setup and to have the half nipples stand independently on their own for experimentation setup and maintenance. This proved to be a very useful and efficient arrangement.


Conclusion

There were numerous learnings obtained from this design challenge which will help further researchers and lab associates to improvise the vacuum chamber. There are a few flaws which need to be rectified but a functional vacuum chamber is still present to conduct experiments with. There can always be improvements as, in an engineering challenge there can be numerous iterations with design, testing, improvising and experimentation which leads to an indefinite potential to improvise the current design. Keeping that in mind, the current design of the vacuum chamber is made to be extremely flexible and modifiable which allows future changes to be made with ease.


Acknowledgments

I would like to thank the SURF program for giving me this wonderful opportunity. I would like to convey my gratitude to the Artisan and Fabrication Lab for supporting the manufacturing of the project. I would like to give a special thanks to Omar, Xingxing, Brandt, and Glynn for their never-ending assistance and help. Last but not the least, I would like to thank Professor Shashurin and Animesh Sharma who has introduced me to this great field of research and guided me throughout the way.

Further details can be found via associated documentation related to the project in the files below.


Description

I get selected to do a research project in an electric propulsion lab. 

Professor: I want a vacuum chamber that can withstand 1 millitorr of pressure, preferably portable and easy to work with.

Me: No idea where to start. Somehow managed to make this, and learned a lot along the way...