Brew Fridge – Part IV: The Modeling
It’s been a week since the keezer teardown, and I’m close to procuring the last of the necessary components for this system. My 120mm PC fans have arrived, and I’ve picked up some paint, connectors, and dryer vents from ACE Hardware. I’ve been running designs through my head, but haven’t settled on the final layout and look for this fridge. Since the fermentation side maintains higher temperatures than a standard fridge, I’m considering a clear panel to that can display (and perhaps illuminate) the beautiful SS Brewtech Chronical. Glass is sexy albeit expensive, but compared to Polyurethane foam, it’s thermal conductivity is a lot higher (0.96 W/mK compared to 0.03 W/mK). This should be within reason for this application, but Acrylic is only an order of magnitude worse than foam (0.3 W/mK) and I can get a custom cut panel from Tap Plastic for a reasonable price. I haven’t seen this done in a side by side yet, so I’m interested in pursuing it.
The clear path forward is to model the physical system. This will guide the layout of components as well as the design of the 3D printed parts in various applications. The critical sections are the controller door mounts, the tap panel, the fridge cap, and the air exchange system. I’ve begun with a basic model of the fridge.
I’m not planning to route many cables in this model since SolidWorks can be akin to drawing with a brick in this department, but TBD. My first foray into this design will be the layout of the air exchange system. This is a pretty simple architecture. There are two exchangers: a push into the top of the fermentation chamber, and a pull from the bottom. Each of these sections feature a fan, dryer vent, and 3D printed enclosure that clamshell mounts through the wall. The fan I’m using is an AC Infinity AI-120SCXD. I’ve heard these can be a bit loud, but they have a speed controller and a 120V plug on the end. I’m keeping it simple. It’s time to start modeling.
The dryer vent has very simple geometry. After 15 minutes of free-form measuring with my Mitutoyos, a vent model is born…
The fan model will be simpler. I’ll forgo the cables and controller for now and focus on the housing itself. The interface between the fan body and vent is the critical parameter here.
I’m going to cheat a little bit and design the 3D printed shroud as a multi-body part containing these two components. When I export my STLs later, I can suppress these models before I enter the Stratsys-phere. I’ve adopted this ugly habit as of late, but reference geometry is so much easier to work with in models like this. I’ve constrained the vent and fan in the shroud environment. Now it’s time to add a cross-section of the fridge wall and start playing with the interface.
The shroud will be made with ABS filament using FDM, and the machine capable of this task (capable in the sense that the stars are aligned properly) is a Stratasys Uprint SE Plus. My mock-up for the top shroud is below.
To make this a little more clear, I’ve included an exploded view below. Fastener models can be downloaded from McMaster-Carr, and since this is printed ABS, I’ll be using Yardley thermal inserts for any threaded holes. Since this plastic is less dense than machined ABS, these holes should be around 5 thousandths undersized from spec.
The vent side clamshell houses all of the inserts. 10-24 fasteners secure the vent on the fermentation side, and the fan fasteners pull double duty on the freezer side: securing the fan and fastening the two housings together. A tortuous channel passes through both halves for cable routing. I will fill any gaps with seam sealer or soft foam to prevent excess heat transfer through this wall. It’s time to print these pieces and test fit them.
I remember a distinct conversation with Stratasys a long time ago. Print quality of the support material was continually degrading during the first few months with the SE Plus. The culprit appeared to be moisture causing chemical changes to the soluble support. Having replaced these spools multiple times during my troubleshooting phase, my reaction was, “Just so you guys are clear, I’m not printing in a rain forest!” After months of frustration, they finally sent a new print head and the problem was solved. The thing is, 3D printers are just like regular printers, only much worse. The consumables are astronomically expensive, multiple expensive components require periodic maintenance or replacement, and believing the control logic will behave a certain way (hello material spool encoder!) is just as ignorant as letting your cat do your taxes. 3D printers are just as bad as inkjets, only much more proprietary.
Why am I ranting about this now? Well after printing one half of the clamshell, the printer decided to give up and ignore the model material in the last 10 layers of the second half. There goes 10 hours and $50. That famous scene in Office Space is funny in or out of context, but the white hot hatred towards unruly hardware is only felt by the unfortunate souls who have to engage with these miserable plastic and steel hell-incarnate boxes on a regular basis. I’m cooling down though, and might as well make the best of it. The fan and wiring fits great inside the freezer half, but I have to wait until the next post to test this assembly. These parts are only one area of prints I’m working on. The Inkbird controllers also need a swap-friendly housing that I can mount in the doors. Time to whip out the calipers again.
My ITC-308 Model should be a fairly accurate reference for the door housings, though the curves forming the case footprint may be suspect. My strategy should be to minimize any geometry that needs to interface with these surfaces and focus on the orthogonal areas. With the controller mounts underway, it’s time to shift to one of the more important features of this build; the keg side. I have an 1/8th inch grained 6061 Aluminum panel in mind for the faucet and drip tray assembly, but this model won’t be complete without some representations of these beautiful stainless steel Perlick faucets. These Mitutoyos are getting quite the workout!
Kudos to Perlick. I can try and try, but accurately modeling the beautiful design language defining these faucets is a futile endeavor. I’ll give myself some credit and agree my lofts could be a little sexier if I spent more than ten minutes defining profiles and guide curves, but this will be more than suitable for the layout of the panel. Lastly, I want to make sure I can fit the SS Brewtech Chronical into the fermentation side. This will be the last model for today, but it’s critical for understanding additional volumetric constraints in the chamber. This will be a quick and dirty mock-up.
I’m at a good stopping point for this post, but there is much left to do. I have to model the Inkbird housings, define the assembly of the top cap, and designate all of the holes and cuts for these components. In addition, I actually have to manufacture these parts, but that’s reserved for another day. I’ll leave this post with a first pass at the overall concept so we can see a little progress and promise.
The original is of course above in white. The concept is below in charcoal, and it’s beginning to assume a new identity with these new components. I’m excited to see how much the final product changes down the road.