Some folks have asked for more design details on the post-crushers we use in our pedal grinder. I’ve put a link to a PDF drawing under “Cider Equipment Plans” over on the right. If anybody wants actual CAD files, email me at firstinitiallastname at gmail.
The concept is basically that you want two drums with shallow ‘teeth’ that are deep enough to engage the apple pulp and draw it through, but not so deep as to get clogged up. We were constrained in designing the tooth shape by the cutters that were available at local stores (in the end we used a router bit that we found at Sears), but we played around a bit until we found something that made a decent interlocking shallow tooth. It’s not perfect, it’s not an involute, and you have to be careful that the shapes have clearance when the tooth is lined up directly with the slot on the mating drum, but also that they pass clearly into and out of that point – the ‘gearing mate’ feature in solidworks was handy if I remember correctly.
As to the fabrication process, IIRC it went something like this: UHMW Polyethylene round stock; it routinely comes larger than advertised, so 5″ should be fine – at least from McMaster. We bought a 12″ chunk and cut it into 2 pieces. Chucked it in the lathe with a 3-jaw chuck, and drilled it through at 3/4″ – may have reamed it too, if we were feeling anal. Then we broached it to 3/4″ hex. That hex broach will cost real money new, but Holly has managed to find decent used ones on Ebay for about $75. We then pressed in some pre-made shafts of 3/4″ stainless hex bar with the ends turned down to 5/8″ round. We’ve gotten lucky 2 out of 3 times with the hex bar coming in a little over dimension, so it makes a nice press fit. But it occurs to me that if you heated the PE rounds to above 50c or so in a water bath before broaching, it would probably broach like butter and shrink down to make a nice press fit even if the hex bar was at nominal dimension.
It could be argued that we gild the lily a bit by using stainless hex bar for our shafts; our philosophy has been that we’re building stuff that’s going to last the rest of our lives, and we want the core of it to be rock-solid. And the time we’ve invested making it is substantial enough that spending a few hundred dollars on good materials seems warranted – hell, some people spend way more than that on bicycles or HiFi equipement. Anyway, once the PE was pressed on the hex shafts, we put one end in a 5/8 collet and the other end in a live center, and turned the OD to dimension (5″) – I bet we cleaned up the ends and took them down to length as well.
The fussiest part of the whole thing was getting the setup on the mill to cut the teeth on the drums. We used a manual rotary table, propped upright on the mill table so its axis of rotation was horizontal and along the long dimension of the mill table. The rotary table came with a tailstock, IIRC, and we had to fuss around with the test indicator for a while to get the shaft parallel to the X axis and to the mill table. Then we cut the grooves; I think we did a roughing pass all around, then once again with a finishing pass. The surface finish was acceptable, but not fantastic, as it was a cheap carbide router bit running at less than 2000 RPM. A little roughness is OK, but you want to try for decent finish, cuz UHMW is hard to sand smooth.
I recall we fiddled with the groove dimensions in Solidworks once we had them rough cut on the machine; it’s likely that the woodworking bits aren’t super precise to the dimensions that they seem to be made to (it’s rare that a hardware store router bit comes with an actual dimensioned drawing; for instance, is a core box bit really actually hemispheric? Anyway, the dimensions shown will work, and other dimensions could be made to work, so long as you lay it out in CAD ahead of time and make sure it gives a nice uniform crush everywhere on the tooth profile. Here’s a screenshot of the profile we designed:
There are a couple more details to the setup that are important. The (cheap, lubed-for-short-life) pillow block bearings we use are from McM; they mount to T slot extrusion on either side so it’s easy to adjust distances etc to get clearances and chain tension right. To set up the post crushers, we basically fix one drum and then slide the other drum up until it touches, then tap it the minimum distance back to get some clearance – well under 1mm is OK; apple seeds, stems, and other hard stuff seem to feed through by elastically deforming the polyethylene or springing the shafts a bit – you really don’t want anything hard that’s much bigger than an apple stem to feed through there though, or it will put a big gouge in the drums (or worse).
The chain drive is not shown in the images, but we drive the primary cutter drum and the left-hand post-crusher at the same speed from the same piece of bike chain. As mentioned above, the left crusher drum won’t drive the right one passively; it will make a horrible chunking noise if you try. So we use a separate piece of light 25 pitch chain to keep them in time – there are photos on the blog that show how this looks. On one of the drums the sprockets are keyed to the shaft, but on the other it’s just set screws, since you need to be able to manually set the phase of the one drum to the other and then lock it down. It works fine until the set screws vibrate loose; better would be a keyed flange bushing on the driven crusher’s shaft, with a circular slot arrangement that would allow greater precision and strength for locking in the phase.
Alternatively, if someone comes up with a more sophisticated (but still fabricatable) crusher shape that does self-drive, that would be brilliant. We designed the shape we use basically on intuition and what cutters we could find, and we pretty much just got lucky – I’m not sure how many more times we would have tried, since it was a couple late nights in the shop to make the first set. It’s possible that a proper involute with substantially deeper teeth would still work and still fling the pulp out of the grooves, so long as you were spinning fast enough. Sadly, few enough people have access to even a bridgeport; I don’t expect many folks can do their own proper gear hobbing.