Drive Shaft Disconnect Design Help

[oompaloompa]

I'm making a conveyor type of machine to move a product in a linear path about 36" long. I have a drive shaft and two other shafts running beneath the conveyor belt. The shafts are at different heights (Y-axis let's say) relative to the floor.

Currently, I'm using a hex socket to drive the conveyor shafts. They are spring loaded, allowing them to be pulled back to a locking position without the use of hand tools (tools are not allowed).

Part of the requirement of the system is that the entire conveyor assembly must pivot up in order to allow cleaning beneath. This pivoting occurs 1-7 times per week. In addition, once a month the entire assembly must be removed and brought to a clean room to be washed down. Again, no tools allowed for this, and all materials must be stainless steel.

The problem with the hex drive I am having, is that it is extremely difficult to align the conveyor shafts (axis) with the drive shafts once the conveyor returns from the pivoted position. Sometimes, the couplings won't even re-engage, due to shaft misalignment or rotational position of the hex.

I'd like to solve the problem with a different type of coupling if possible, rather than redesigning the pivots, etc. Gears are ok, some type of U-joint, sliding shaft etc.

Your thoughts are greatly appreciated.

 

[oompaloompa]

Here are some screenshots of the setup:

[

http://i280.photobucket.com/albums/kk180/dynaspeed/shaft_side.jpg

]

[

http://i280.photobucket.com/albums/kk180/dynaspeed/overhead_view.jpg

]

[

http://i280.photobucket.com/albums/kk180/dynaspeed/3qtr_view.jpg

]

 

[DesignerGuy16]

Have you considered your run of the mill spider coupling?

James Spisich
Design Engineer, CSWP

 

[MikeHalloran]

Looks like a natural application for a crown gear coupling.
( Look at the throttle linkage inside an old outboard motor handle hinge. )
I'm not sure anyone makes them commercially, or if they can carry the load.

An alternative is a bevel gear on the end of each shaft, mating with an idler bevel aligned with the hinge axis.



Mike Halloran
Pembroke Pines, FL, USA

 

[aafuni]

You could just use a splined clutch plate. By this I mean a flat disk with grooves that mate with an oposing plate. How much torque is transmitted?

 

[oompaloompa]

Jspisich - Spider coupling may have a similar problem to what I'm experiencing now with the axial alignment. Also, let's say one shaft turns 15 degrees. The teeth won't mesh. Depending on which shaft it is, it may not be able to be turned by hand to align it either. And the insert may get lost.

Mike - I will take a look at the crown gear coupling. The bevel gear we've discussed here at work. I'm not sure how easy aligning those teeth would be either after the conveyor is lowered from a tilt. I like the concept though.

aafuni - that sounds interesting. that might work. The drive shaft has the most torque. I don't know the rating off hand but it's a 1/6 hp motor.

A friend also suggested a magnetic coupling device. I haven't found too many compact ones, or ones with decent torque ratings. They were also pricey.

Thanks to all that have responded so far.

 

[hydtools]

Look through the Lovejoy catalog

http://www.lovejoy-inc.com/uploadedFiles/Catalogs/g.pdf

Ted

 

[ChrisDuncan]

I'm imagining a 45deg chamfer about 1/2" deep leading into the hex socket, to cause them to align axialy, and then the hex sockets being spring loaded once they are brought from the lock back position. Then the belt is rolled manually until the hex sockets line up radialy and spring into position.

 

[ChrisDuncan]

You could also have a 12 point hex socket for possible easier alignment, but is has less cross section area for strength. You would probably also want some kind of indicator/mark to confirm full engagement.

 

[MikeHalloran]

The advantage of the crown gear coupling is that the teeth are not disengaged in a tilt operation. The disadvantage is that they can't have optimal tooth geometry or great strength, precisely because mating teeth remain engaged while their axes rotate around the contact patch.



Mike Halloran
Pembroke Pines, FL, USA

 

[dvd]

Probably doesn't have too high of a shaft speed. You could make a special coupling that had 3 or more Destaco clamps around the periphery that are un-toggled to disconnect.

 

[EngineerTex]

If your conveyor only runs in one direction and you don't need it to reverse, then you might consider a sprag (overrunning clutch) incorporated into most any part of the rotating assembly and it can go on either side, or it can become part of your slider hex.

Lots of manufacturers produce these parts in a drawn-cup type of overrunning clutch with needle bearings incorporated. And they're cheap. I found one on the INA/F.A.G. site (model number HFL3530) that is good for up to 90 ft-lbf of torque.

http://medias.schaeffler.de/medias/hp.ec.br.pr/HFL*HFL3530

When I previewed this post, it appeared that the full url above doesn't present itself correctly. Copy and paste so that you include the "*HFL3530" on the end of the link.

This may allow you to use what you already have, so that you can index the shaft backwards by hand so that you can align the two parts.

Engineering is not the science behind building. It is the science behind not building.

 

[ChrisDuncan]

OOmpaloompa,

If you can please post how you solved this problem.

 

[dicer]

If a clutch can be used,and its a one direction deal, what about a roller or sprag type clutch?

 

[unclesyd]

How about an automotive universal like the one used with the high rise truck kits.

Even at very acute installation angels and high loads they will provide a reasonable life and relatively are inexpensive.

 

[unclesyd]

Another approach for a very compact and inexpensive quick connect couping would be to use a universal impact socket. There are numerous configurations available, square drive, spline drive, pin type, pinless, held with safety pin, ball detent or elastomeric ring,

In the same vane there is the Wobble socket. It too has several locking possibilities.

http://stegmantoolcompany.thomasnet.com/category/sockets-and-wrenches-sockets?

http://www.lancetools.com/wobblesockets.html

 

[oompaloompa]

Sorry guys, my computer went down last week and I've been pretty tied up.

Based on some of the input here, I purchased a lovejoy coupling to experiment with. It has a split spider with coupling nut that can be removed by hand and is also stainless steel, one of the requirements here. The cost was around $200 for each. Sorry I don't have a part number on me, but they are pretty standard. Mcmaster carries them too.

I also sketched up another idea for a coupling, using some ideas here as inspiration. I'd machine it in-house. Has a 45 degree taper on each side. Three press fit pins would drive it. Your thoughts are welcome on that as well.

http://i280.photobucket.com/albums/kk180/dynaspeed/new_couplings.jpg

The sprag clutch I'm not sold on, only because I'm not sure if it helps with the disconnecting or not.

The wobble sockets are pretty sweet. I'm currently using an impact socket (6 point only) to drive this now, so if I can press that into my hand screw, that might work as well.

I have a little time before I have to fabricate the next unit so I'll play with these ideas some more.

Thanks again everyone.

 

[EspElement]

I see you put slots in the counterpart that can allow for misalignment. Im susupecting this isnt a very fast system?

 

[oompaloompa]

EspElement - no. The drive shaft and the one it is coupled to move at 6 feet per second max. The third (middle) shaft might run at 3600 rpm, but carries very little load. It just turns an eccentric shaft used to agitate the belt.

 

[EspElement]

It doesnt run in reverse ever?