Show a renovation project for a gas distribution cabinet

Thread Content

Abstract: The valve train system is a system that can be considered simple yet also complex. Whether it is a complex system or a simple one, its reliability and reproducibility are crucial for experiments and testing; however, many researchers often overlook the importance of the quality and performance of the gas supply. Last October, we were assigned a rather challenging renovation project: the modification of the gas circuit for a membrane reaction rating system in a key laboratory at a university. This system is essentially a device similar to a gas distribution cabinet. Let’s first take a look at the condition of the original device. This is the appearance of the device: let’s take a look at its internal structure as well. It seems quite complex, but in fact it is an integrated unit that simultaneously controls 6 mass flow controllers. However, since they originally used mass flow controllers from Beijing Qixing that could only be controlled via analog signals, the wiring connections appeared quite complex; coupled with the CNC system and the piping connections, this essentially filled up the entire cabinet. Let’s talk again about the background of this equipment upgrade: This laboratory equipment has been in use for 5 or 6 years now, and several gas lines have developed faults, primarily manifested as unstable flow control and issues with the signal input and output of the numerical control unit. Although students do not have very high requirements for accuracy when conducting experiments, they do require a high level of repeatability, as repeatability is crucial for the validity of the experimental results over several days. Previously, they also tried to address the issue of poor repeatability by using the soap bubble flow meter calibration method – that is, they would calibrate the device before each experiment. They were very professional, having their own set of calibration methods and formulas; yet even with calibration before each use, it was not possible to ensure stability throughout the experiments. It’s likely that they really could take it no longer, which is why they asked us to assist in modifying the equipment. The image below shows the students calibrating using a soap bubble flow meter before each experiment. This is the formula they use for calibration. In fact, devices that have been in use for so many years generally no longer require any modifications, as modifying the gas supply system often involves more than just replacing the core equipment; the entire gas supply system needs to be rechecked and repaired. It was our first time doing this, and we had no experience; we thought that simply replacing the mass flow controller would be enough, which ended up being a costly mistake. The first proposal we made was as follows: remove all the pipelines related to the CNC unit and the pressure gauge, since the mass flow controller we selected is equipped with built-in pressure and temperature sensors, and it is possible to obtain real-time data on pressure and temperature in the pipelines through RS485 communication signals. Additionally, the 6 mass flow controllers are controlled via a host computer with six machines connected in parallel. We took into account that, during actual experiments, it is sometimes necessary to temporarily switch the gas supply; for example, if the gas source is originally oxygen but the experiment requires the use of argon, then a gas switch is needed. Therefore, when developing the software, we incorporated the functionality that allows each gas line to be switched to any other gas type as desired. When we were proud that the plan was perfect, we were completely stunned when we went to the site for testing. We removed the original circuit board and mass flow meter, replaced them with the mass flow controller we had prepared, and calibrated it using the soap bubble flow meter they had been using previously (mainly because the students said they have always used this method to calibrate gases). It was found that there is a significant discrepancy between the actual export flow rate and the set flow rate. For a while, we were unsure whether it was a problem with the soap bubble flow meter or with the mass flow controller itself. The first time we failed, the core issue was figuring out how to determine the actual value of the outlet flow rate; we really didn’t trust the readings from the soap bubble flow meter. Firstly, there will definitely be errors when using a stopwatch; secondly, could differences in temperature also have some impact? The second time, we brought a high-precision mass flow meter with integrated CNC control, connected directly to the very end of the gas circuit. As shown in the figure below, we find that the deviation amount is indeed very large. At the same time, it was observed that the pressure at the outlet end was lower than the pressure at the mass flow controller at the front end; this confirmed to us that there was a leak in the entire gas pipeline. Once we identified the cause, we analyzed the possible points where leaks could occur. Leaks can happen at any joint, and we developed a detailed testing plan to check the airtightness of each such point. Eventually, it was found that due to the old age of the equipment, many joints had rusted and were causing leaks, which is a very dangerous situation; if flammable or explosive gases leak, it can lead to serious hazards. Therefore, for greater completeness and to provide users with a better experience, we have decided to replace all the pipes and connectors; we really didn’t consider this aspect before, and it was a painful lesson. Here, I will briefly introduce the process we use for leak testing. The core device is actually the high-precision mass flow meter mentioned above. We measure that section of the gas pipeline by connecting this mass flow meter at the very end, then blocking the outlet of this flow meter. The controller is set to a relatively high flow rate, and pressure is maintained at that level (before starting this process, the pressure of the gas supply should be set around 0.4 Mpa; it should not exceed 0.6 Mpa, as we use hoses for testing purposes, and such hoses can generally withstand only a pressure of up to 7 atmospheres). If there is no leakage in the entire gas path, the pressures at both the mass flow meter and the mass flow controller will gradually increase. When the pressure reaches its critical value (the maximum pressure of the gas supply), the flow rate will decrease rapidly until it eventually becomes 0. Of course, in reality the gas circuit cannot be completely airtight with no leaks at all; it’s acceptable for there to be a drop of 0.1 Kpa after 50 or 60 seconds. This corresponds to a leakage rate of around 10^-7 pa*m3/s, which is already quite good for modifying a gas circuit. After the final modification of the gas circuit, the students were able to use it effectively after some time of adaptation. They have now abandoned the previous calibration method for soap bubble flow meters; the gas circuit we modified does not require re-calibration, as it already boasts excellent repeatability and stability! Finally, here is the final render: