Centrifuge gear design

[Mooselake]

I was asked to help with a hand cranked centrifuge.  The initial design used mod 1 gears, 72 and 16 tooth.  While this is way beyond my knowledge level, it seems that the teeth are too small for something that will be hand cranked.

Any suggestions?  It doesn't need to be a pair of gears, it just needs to spin the centrifuge, and hold up.

Or, more simply, Help!!

Kirk

[ArtF]

Kirk:

  So it sounds like they need a 4.5:1 ratio on whatever they intend to use. The gears you specify are
mod 1, so a 72 and 16 tooth are 74 and 18 in diameter with 44mm between centers. Pretty small space
to get that 4.5:1 in..

  A centifuge usually spinds pretty fast to get gravity as high as it can, so its hard to say why those gears were
chosen. Were it me, Id be tempted to simply use a belt and 2 pulleys since syncronicity usually isnt important
in a centrifuge. Perhaps it goes too fast for a belt?  As  I understand the problems with centrifuges is the high speed is very hard on the mechanics, bearings, and any gears used. No way around it, high speed is aways hard...

Art

[Mooselake]

I don't know why they were picked.  I'll need to get some more info since iirc  it needs 2,000 rpm, and they expect the hand crank to be turned at 45 rpm, so the numbers don't work out - I get 45:1, not 4.5:1.  I'll also try to find out how big a space they have to fit in.

Kirk

[JustinO]

The ones I've seen have worm gears driven backwards.

[Mooselake]

I got some better information - the goal is 3000 rpm with a cranking speed of about 30 to 40 rpm.

Kirk

[John S]

You need Popeye ;D

[Mooselake]

It's just me, but I'll eat my spinach!

My first thoughts, details to follow and still very rough, are to use three parallel shafts.  #1 will be connected to the crank, #2 is the appropriate distance away, #3 is concentric and larger than #1, and connected to the spinning gadget.  Some magic bearing substance (maybe delrin tubing? I have some delrin rod sitting here) will be between #1 and #3, and for the hubs of the gears that spin on their shafts.

I'll design it with GT after I get back from todays physical therapy for my knee.  Not sure how to put concentric shafts in GT, but now's the time to learn.

Kirk

[JustinO]

D?j? vu. Parts of an old bike and a bicycle dynamo -- I think I must have made a blood analysis rig when I was a kid.

What diameter and weight capacity does the armature need to have?

[JustinO]

Meta subject:
Whoa. I spelled deja vu correctly with the diacritics in the text entry window, and it turned into dej? vu in the forum listing.

[JustinO]

http://www.ebay.com/itm/VINTAGE-BENCH-H ... 3a8452f371

You need to buy this to get a feel for the engineering task.

I find it hard to believe that the mechanical engineers of a hundred years ago didn't figure out everything that could be figured out in this genre.

[Mooselake]

I find it hard to believe that the mechanical engineers of a hundred years ago didn't figure out everything that could be figured out in this genre.

Me, either, although I'm guessing it's a worm gear drive.  I'm trying to get a preliminary design before going OOT next week, so I'll postpone a purchase (although it would fit nicely into my shop).

I'm muddling through with a 6 gear train, 3 x 4.5, and finding some things in GT that aren't working like I'd expect.

I'm unable to get a snapshot using the snapshot button under media on the project page.  Here's one from shift-ctl-printscr.  I'm unable to make a video file, and the select codec checkbox won't check.  Most likely I'm missing something.

There's three pairs of gears, colored white, green, and red using colored plastic material types.  The material type isn't restored when I save and reload the project file.

I can't seem to find a way to connect the gears together as I'd like using linkages. &nbsp;For example, I'd like the small white gear to be connected to and drive the big green one, and the small green gear to drive the big red one. &nbsp;Whoops, the small red gear doesn't drive anything, looks like it needs to be an even number of pairs <smacks head>. &nbsp;Right now, when I hit simulate, half the gears disappear. &nbsp;I exit the program and reload the project file to recover, loosing the colors in the process. &nbsp;I've tried removing linkages, but can't seem to get a list of available unlinked gears when I hit Add Linkage. &nbsp;I'll keep trying and see if I missed something.&nbsp; Zipped copy of the project file attached.

I don't think there's a way to put a bigger hollow shaft over a smaller one, or am I missing something there too?

Kirk

[Mooselake]

Since I'm not a metric guy I don't have any real intuitive feel for metric dimensions or terms.&nbsp; I'd understood incorrectly that a mod 1 gear had teeth that were one mm per tooth, so a (say) 36 tooth gear would have a 36mm circumference and about a 11.5mm diameter.&nbsp; Looks like I'm wrong.&nbsp; Doing the research I should have done a mod 1 tooth is 1 x pi, or 3.14mm per tooth, about 1/8".&nbsp; Not the 1/25.4" I'd assumed.&nbsp; This metric stuff isn't as easy as "they" claim.

So, forget what I said about the original mod 1 gears teeth being too small.&nbsp; However, it still needed to be redone, just for different reasons.&nbsp; It also means that a 4 gear solution should fit, although that may not be the best solution.

I accept my public humiliation <sigh>.

Kirk

[Mooselake]

How do you calculate the load an individual tooth can handle, and therefor determine the best gear thickness and/or module, given the appropriate properties of the material used?&nbsp; I'm about 45 years past my properties of materials class, the book is in storage somewhere, and there wasn't a lot of need for it while I was a datacomm programmer.

Is there a reasonable way to (gu)estimate tooth and hub wear (if the hub spins on the shaft, and the shaft is a lot harder), and what properties of the material need to be known?

With a speed changer is it correct that the load on the teeth increases by the ratio of each change, plus a bit for friction on the shaft bearings and at the gear teeth?&nbsp; For example, when using 10:1, a one unit force at the output will require a 10+ unit force at the output?

It's been interesting having a real project, and is the way I learn best.&nbsp; It's really nice, after spending the last almost 4 months recovering from knee surgery and being stuck in bed for a lot of it.&nbsp; The meds and pain made it pretty much impossible to do anything useful for a good part of that time.&nbsp; If no pain/no gain is true then I'm looking at a lot of gain :)

Thanks!

Kirk

[ArtF]

Kirk:

&nbsp; Ill have to let someone much more knowledgeable than I handle that one... Im just a mere
mortal who doesnt get that deep into&nbsp; the materials engineering side. :)

Art

[Mooselake]

Me either, but I'll get blamed if the teeth strip off :)

Kirk