Colibri Hummingbird Dynamic Sculpture - WIP

[Kim]

... My 2-56 taps aren't long enough to through tap .5" holes, so they were tapped halfway from each side.

That is fascinating to me.  If you tap from both sides, how do you make sure the threads line up?  Did you do something special to get them to align?  Or do you not need the bolts to go all the way though?

I'm really enjoying your Colibri build! This is also on my wish list.
Kim

[kvom]

Separate screws on each side.

[Kim]

Ah, OK, makes sense.
Thanks,
Kim

[kvom]

Today's project was to machine two of the large three main gears that drive the hypocycloid and cams.  Material is .25" thick, 7" wide Mic 6 cast aluminum.  After drilling the 1/8" center holes, a 3/8" 2-flute end mill pocketed out the spokes and did a roughing pass to form the teeth.  The DOC was .20" making a trochoidal slot .50" wide.  Doing this, while slow, enables room to clear chips with air blast much easier.



Next, a 1/4" 2-flute endmill does the finishing pass and cuts the bottom of teeth where the 3/8" can't reach.  This material comes with a tough plastic film on both sides.  By leaving the film on the bottom, the endmill pushes down the film but doesn't cut through.  This helps retain the gear from moving, even if there are also 4 tabs on each gear.



After removing the film, you can see that the profile was cut all the way through except for the tabs.  These are snipped in two and the stubs removed with the disk sander.



The finished gears with the smaller brass pinion.  Only other finishing needed was deburring the spokes.

[Myrickman]

Kvom, that's is certainly a neat build. I'm just starting down the cnc rabbit hole with a used Isel gantry and adding new electronics so seeing your build is inspiring to say the least.  I'll be following with great interest. Paul

[Tennessee Whiskey]

Finally had a minute to catch up. Man, this is going to be awesome. I couldn’t help but laugh and wonder: Wonder if G. Britnell is going to try this manually?  . Nice work Kirk

Cletus

[kvom]

Manually in wood with a good jigsaw and patience it's achievable.  In metal w/o CNC it's hard to imagine.  Wood with a big CNC router you could cut out all the pieces in a day.

My project for today was one of the frame supports that forms the right side of the bird.  It's only difference from the left is a hole for mounting the bearing that follows the body cam and causes the entire frame assembly to rotate up and down.

The stock is 7x17.5x.25".  Plans call for 3/8" thick, but this is what I could find.  I used a feature of the CAM program to rotate the entire drawing so as to minimize the Y extrema and thus minimize the amount the outer profile cut would eat into the aluminum vise jaw.  This left the X extrema at 16.2".  The vise jaws I'm using are 12" long on my 6" vise.  I didn't want to do much if any machining on material that overhangs the jaws, so it would be necessary to shift the work several times before completion.  Another complication is that my mill's X travel is 20", meaning that for reaching the left edge of the stock is could protrude at most 3.5" from the left edge of the jaws.

I settled on a machining origin as the center of the topmost hole which accepts a rod to attach the flower stem on the complete model.  This hole would be 1" from the top and left edges of the stock.  I used an edge finder on the fixed jaw to establish Y0, and that never changes throughout.  Each time I needed to slide the stock one way or another I merely needed to use the edge finder on the left edge at Y0 to reestablish X0.  With that accomplished, the first stage was to drill all the 1/4" mounting holes as well as the 3 leftmost pockets.



After a shift left, we have the righttmost pocket, the hole for the pivot bearing, and half of the trochoidal rouging pass for the outer profile.



After some more operations the CNC was finished, so I needed mainly to sand off the tabs and deburr.

[crueby]

quite impressive!

 

[b.lindsey]

It sure is!! You are getting a nice collection of components now Kirk.

Bill

[kvom]

The manual 1-shot oiler for my mill broke halfway through the matching B frame, so no more Colibri parts until I get that sorted.   

[kvom]

With the CNC mill out of commission for a few days, I can work on some round parts.  These spacers for the 3 large gears are brass disks 1.25" in diameter with a 1/8" hole for the axle.  An 8-32 grub screw locks it to the axle.



In the wood model a spacer would be glued to the gear.  The 1/8" steel axle would be secured by grub screws on both the hypocycloid eccentric and the spacer.  An instruction tip indicates the potential need to adjust the gear timing by disengaging it from its mates and rotating it (and the cam) by one tooth.  With the wood gear apparently it's flexible enough to bend it out of engagement.  I fear this is not the case with a metal gear.  I'm still considering the best way to accommodate this need.

[kvom]

The next pieces are similar to the spacers, but are the eccentrics that drive the hypocycloid gears.  As such they need to have a smooth sliding fit to the center of the hypocycloid gears that I had printed at Shapeways.  It turns out that there is a .005" spread in those diameters, so I turned down and parted off each after specifically fitting to a given piece.

The diameters are nominally 1.125", so I could hold them in a collet in order to drill the 1/8" hole for the axis, offset .40" from the center.





Each is secured to the shaft y a 8-32 grub screw, so it's necessary to drill from the opposite side.  I couldn't figure out the reason of the 1/16" hole in the center so I emailed Derek Hugger.  Turns out it's so the disk can pivot on a metal pin while being sanded on a belt sander.  That's how the wooden version can be turned to fit when the builder doesn't own a lathe.  But I was able to put it to use here.



The centerline for drilling is parallel to the top of the jaws of the small vise, so when mounted on the manual mill an edge finder suffices to locate the hole location.



Since the depth of the hole is longer than the teeth on the tap, I drilled the first 1/2" of the hole with an 8-32 clearance drill and the rest with an 8-32 tap drill, allowing the tap to penetrate deeper than the circumference of the disk.

After deburring the edge of the tapped hole, the part is finished.

[b.lindsey]

More good progress Kirk. I am watching closely even though I plan to use the brown stuff 

Bill

[kvom]

Minor progress today making 4 frame spacers (actually 3 as 2 are combined) in aluminum rod with a center hole.  Not worth a pic.

Rather than mount the vacuum chuck in the vise, I decided to raise the secondary table a couple of inches and mount it there (I'll need to remove the Kurt vise).  To raise it I ordered a pair of 2-4-6 blocks from MSC.

That left the afternoon open so I did a swarf cleanup of the CNC mill and  Bridgeport and floor for the first time in a while.  There was a lot of it.   

[kvom]

The oiler replacement came this afternoon, so I can get back to making non-round parts soon.  In the meantime I did a mockup of the cam action using the gear spacer to drive the motion.

<a href="https://www.youtube.com/watch?v=pSuW3rfWbao" target="_blank">http://www.youtube.com/watch?v=pSuW3rfWbao</a>

As the hypocycloid gear rotates, the 4 pins in the cam travel around its round holes.

There's a bit more friction than I expected;  however, the frame material is a lot smoother than the mockup.  I might need a bit of PTE sheet next to the frame.