MAS.863 How To Make (Almost) Anything – Fall 2014

Richard Li

Final Project Ideas

Week 0

Many ideas are floating around in my head and I haven’t quite decided on one yet, especially given the time constraints of the semester (or rather the last week in prototyping the idea). I like to build things that are tools that can be put to immediate use. Ideally, if I can build things for this class that can also help my everyday labwork or safety, then that’s a big win! So here is one idea:

Idea 1: Vacuum Controller/Pulsator

For my graduate research on nano-engineered composite materials, I fabricate high performance composite specimens for mechanical and multifunctional testing. One of the composite fabrication techniques we employ quite commonly is Vacuum Assisted Resin Infusion (VARI) which entails first laying up your dry fabric (e.g., carbon fiber, fiber glass etc.), vacuum bagging it inside a mold, and flowing epoxy resin through it until it full wets before stopping the flow and curing the resin in place. Once the piece is cured, the part can be peeled out of the vacuum bag, and machined down to the section you want – thus your desired composite is formed. A diagram is shown below:

However, this is a process that is very much manual and intensive, and requires constant monitoring. Depending on the mold geometry, the infusion process can last from 45 minutes to over 8 hours. Additionally, vacuum conditions are absolutely critical to ensure low void content in the specimen – a high content of which results in large stress concentrations and drastically impacts the mechanical properties of the specimen. To really help remove entrapped air bubbles inside the resin, it’s helpful to pulsate the the vacuum level so that the bubble sizes change, and likely loosen from any fabric surface they temporarily linger onto. This would help these voids to advect down towards the vacuum and off of the sample. In the past, this was done very manually, by either kinking the downstream tubing (so the vacuum decreases as the flow progresses) and then unkinking it to bring it back to max vacuum levels. This was usually done in intervals of 5 to 10 minutes for the full duration of the infusion. One can imagine how crazy this gets if we do it for 8 hours. Thus, it is of interest to automate this system! Here’s the vision:

A device with 3 tubing ports – one to the vacuum pump, one to the air (venting), and the last one connected to the downstream line of our sample for which we’re trying to infuse. This device will have a pressure sensor (analog vacuum gauge?), and a valve that opens and closes to the vacuum pump or air. When connected to the computer, this device will allow a user to specify a vacuum waveform to track. For instance, if we want to pulse between 29 and 15 inHg in periods of 16 minutes, then the device will switch open the the valve to the vacuum pump until 29 inHg is reached within the first 8 minutes, and then open the ambient valve slowly so it vents to 15 inHg by the 16 minute mark and etc. On the interface side, this would be great to control remotely as well – say using wifi.

While on one extreme, I might thinking of building the valves from scratch, these are a highly specialized parts that must reliably cycle through long durations of labwork. I think what makes the most sense to implement this would be to integrate commercial off the shelf valves and sensors into one glorified vacuum control box/panel that can communicate to a computer. Before I CAD anything up more sensibly, I’ll have to really dig around the internet to find the size of these valves, and then design my control box around these devices. Perhaps the closes item to this concept is shown here: http://www.vacuubrand.com/us/page800.html.

My idea is to have something more DIY, integrated, and made of cheap commercial off the shelf solenoid valves:

To visualize this better, here is CAD in solidworks of the parts minus the wiring. The red is the pressure transducer, the dark grey are the solenoid valves, the blue is the power supply, and the copper are the circuit boards to be designed which may need a DAQ mount. The front of the box has a cutout for USB mount, and the rear has a cutout for the IEC power connector.