Low rpm zero-backlash clutch coupling (on-off action)

[tph216]

Hi,

I'm looking for a small clutch coupling to sit at a break in a shaft, and give basically on-off action by electrical actuation. I'd specified a Warner SFC-250 initially, but now find that it doesn't quite have the torque rating I need, and is not ideal for low rpm's.

Some further details on my requirement:

* Speed is 6.5 rpm

* Design torque is 12.5 Nm

* Input and output shafts are 1/2 in (but I can alter them)

* Clutch is installed in a vertical orientation with input at the top and output at the bottom.

* Require zero backlash.

* Needs to be fairly compact, round about a 3in x 3in footprint, mounted onto the top of a product I am working on (a custom designed precision roller screw-jack box, linear output to resolution of 40 micron, output force 1000kg). I basically need to disengage electric drive from this jacking box whenever the drive motor is not driving, which will free the box up to allow manual compensations to be made.

* When the clutch re-engages there must still be zero-backlash.

With the Warner clutch unit I read that at these low rpm's the friction faces would not wear in properly, and so I would get slip until they were fully worn in (which could be years). In which case, I'd need to specify a unit with double the torque rating compared against my design torque.

Any recommendations of products or manufacturers who are likely to have what I'm looking for would be greatly appreciated.

Regards,

tp216

 

[unclesyd]

I would contact Wm Berg, Inertia, or SDP to see if they have a clutch that meets your requirements. We have used clutches and couplings from all three manufacturers for our in house machines.

http://www.wmberg.com/catalog/catpage.aspx?url=pdf/B05J007.pdf

http://www.idicb.com/electmag.htm

One of their clutches on the surplus market

http://www.surplustraders.net/specs/gb376.php

http://www.sdp-si.com/estore/CoverPg/Clutches.htm

 

[tph216]

Thanks unclesyd.

Out of interest, what kind of typical safety factor would you put on the specification of a clutch such as this.

My 'design load' of 12.5 N*m already includes a 1.5x factor above what I have calculated I need to move the rollerscrew below, which already accounts for all frictional & other losses through all the components in the load path.

I guess there is no problem with over-specifying clutch torque capacity, but there would be a problem in under-specifying it.

I'm pretty new to all this sort of stuff, and am kind of feeling about in the dark at the moment.

 

[Occupant]

Here is another example. Robatic Size 4, I think, is what you need and that would probably obviate the need for a safety factor since it is good up to 20 N*m

 

[Occupant]

I have to correct myself, Size 4 is ~4.00" Dia. Size 3 would be ~3.00" Dia., but only good up to 10 N*m - may be it's enough.

 

[BrianE22]

Wouldn't any friction type clutch have slip during turn-on and turn-off? Or maybe you don't care about slip at those points?

 

[tph216]

@Occupant: Thanks I'll take a look at those.

@BrianE22: Possibly yes, but I am only wanting to engage or disengage @ zero rpm (i.e. momentarily after the motor drive has stopped, or before it starts up again).

I've considered dog-toothed clutches, but I believe these are not guaranteed to re-engage without backlash. I think there is a slight rotation forced to align the dog teeth.

Also considered wrap spring clutches, which seemed ideal at first, but I need both CW and CCW rotation, and these seem to be either / or, not both.

 

[unclesyd]

As stated above your safety factor is OK. If you go to a higher capacity clutch you introduce other undesirable variables.

Have you looked at stepper or servo motors for your application.?

For your required precision you may also want to looks at this site.

http://www.physikinstrumente.com/en/products/index.php?ban_RD_250x300_0910-

 

[tph216]

Yeah, I initially considered steppers but had other issues with them.

For my setup, I am trying to raise and lower a 15 metre long blade over a rolling drum. There are 12 connection points on the blade, which connect to my rollerscrews each capable of exerting +/- 1000kg force to the blade. We need to be able to lift the blade in 40 micron increments, basically with all 12 connection points moving in unison, and staying within a much tighter tolerance relative to one another (+/- 10 microns or lower).

So, I have all 12 fine adjustment assemblies running from one gearmotor. The output of the gearmotor runs at about 130rpm on two output shafts, each of which goes through 6 20:1 reduction gearboxes. These outputs go through a safety coupling, then this clutch coupling I'm struggling with, and then ultimately into the fine adjustment assembly (a custom screw-jacking box with 1mm lead output).

As I recall, ensuring 100% sync between 12 steppers would be difficult. And even if I found a single stepper to substitute for my gearmotor, I still need individual on-off control at the inputs to the jacking boxes, so would still need an adequate clutch coupling.

The ultimate aim is to allow the functionality of being able to leave the motor running in whichever direction is required, and write some control software that closes the 12 clutches as required to keep the blade straight as it moves, based on feedback from 1-micron linear encoders I have installed on the blade.

But I'm getting off the point a little bit here.

I do have a backup option, which is to mount a clutch coupling on the output of the 20:1 reduction boxes, rather than the input to the fine adjustment screwjacks. There is more space there which I could use to accommodate a larger clutch like a Warner SFC-400.

It all just seems like a lot of effort to get an electrically actuated on-off function between two shafts. Is this really not a more common requirement in mechanical systems? (As I mentioned I'm pretty new to all this).