Brew Fridge – Part V: The Manufacturing
At my last stopping point, the printer was down, and I resorted to 3D modeling a few major components in this build. My posts follow my progress each weekend, and I have no time otherwise to get much done, so this marks my fifth week into this build. On Friday evening, I decided to make some production progress over the weekend, so I finalized the tap panel in CAD and reinstalled HSMWorks to my machine. I was interested in milling some designs or logos into this panel, but I didn’t have the inspiration or time to invest, so I went with a very simple part. I’m still missing the drip tray, and Amazon is crap when it comes to accurate part dimensions, so I’m going to manually drill those holes later.
I’ve left ~15 inches of real estate below the shank holes in the panel. This area provides an adjustable range for the drip tray, a necessary component when showing this fridge to the ladies. When the tray is positioned 8 inches down, I can fit any standard size beer glass. At the bottom, the tray will provide for large growlers and everything in between. My adjustment holes will be sized for a TBD quick-release pin, and this should be a strong and elegant solution for moving the tray up and down.
To seal the freezer door, I’m using a backing MDF panel, polyurethane foam, and spray foam fill. The shanks bolt directly to the aluminum panel, with large clearance holes in the sealing stack behind it. I want this panel to be serviceable, so I’m installing six Riv-Serts into the door sized for flathead 10-24 fasteners. By doing this, I can simply open the door and unscrew the hoses from the shanks, close it, and keep the freezer cold while I remove the entire tap assembly from the front. This is useful for cleaning and swapping the taps as necessary.
The air exchange prints are running as I write this post, so it’s time to start milling the tap panel. The mill I’m using is a YCM-40 Supermax.
It’s a bullet proof machine, and simple to boot. Cooling isn’t set-up (I’ll do that manually with a spray bottle), but the main reason I’m using it is the vacuum table. This panel has a large surface area and I’m cutting slow, so there should be no issues holding it. CAM took me about 5 minutes to generate for this part: My main operations use a three flute, 0.125″ endmill. The panel is 0.125″ thick and my cooling isn’t very efficient, so in addition to a slow feed, each operation has four roughing passes and a finishing pass. I’m cleaning the perimeter before the final cut with a 0.25″ chamfer mill.
The vacuum table is a wonderful tool, but it’s not without some flaws. For those wondering, the part is not vacuumed directly to the o-ring itself. Before placing the panel, a massive sheet of double-sided nitto tape is adhered to the bottom surface. The protective sheet remains on the vacuum side (I don’t want the panel sticking to the table after-all), and this surface mates to the O-ring. Compressed, these two layers (adhesive and backing) stack up to 0.011″. This means that the reference plane for the part is 0.011″ above the table surface and tool offsets must be adjusted accordingly. For final cuts or pockets, the end mill cuts into these layers by ~0.002″ to clear all part material, but in theory, this won’t puncture the backing and kill vacuum. In reality, leaks are inevitable, and liquid cooling can be sucked into the hose and through the pump. These are the issues I have to look out for.
My chamfer is slightly smaller than I represented (My apologies ladies), but I think it looks good enough. As I expected, some coolant was sucked through the vacuum during this process, but I can tear down the pump later and do some maintenance. These things are hard to kill anyway
This panel is grained 6061 Aluminum. After I clean it, it will look really “aluminating” on this fridge. With the countersinks finished, this panel is ready to install, but before I get there, I have to make the MDF and foam backing for support and insulation.
There’s not much to these components at all. I made some rough cuts to get the right footprint, and the holes we’re cut with a 2″ hole saw. In this stack up, a foam layer (same geometry as MDF) mates to the inside surface, and the shanks have adequate clearance through this stack. The holes cut from the foam will collar around the shanks during install to fill this air gap. With this assembly nearing completion, it’s time to turn my attention back to the air exchange system. My thermal insert prints are ready to go, so now it’s time to heatsink some brass.
As I outlined before in previous posts, these air exchange housings clamshell together through the wall. The method for fastening everything hinges on thermal inserts. These are captive, barbed brass nuts that are heated and pressed into a material; plastic being a popular choice. Screwing plastic directly always leads to premature stripping, so these Yardley inserts are hot for some hole. The process is outlined below.
We wanna stay above the hole. I wanna bring everybody above the hole. We all want to live above the hole. This operation requires a drill press and thermal insert tool. The press should be unplugged from the wall, and the heat gun should be installed to the collar as normal. Turn that shit on. After the tip is sufficiently hot, slowly lower the shaft into the insert. As the insert heats up, apply a little pressure and it will slide gently into the hole. Be careful not to go too deep your you’ll get some goo stuck to the tip. After the flange has set below the surface, gently release the handle and pull out.
And fast we come to completion. Both sides have been given the my inserts, and this beauty is ready to mate with her partner. If euphemisms are wrong, I don’t wanna be right…
This concludes my first dive into the component builds for this project. I’m ready to begin cutting into the fridge and integrating these parts, but before I get there, I need to begin wiring consolidation in the fridge.