interference fit shaft spline 3

[brian823]

I am trying to find information on designing an interference fit involute spline. Should it be a side fit spline with interference where the teeth contact each other? Or should it be a major diameter fit with the interference between the major diameters? Do the standards address the amount of interference, or is that up to the engineer to analyze? I would prefer to use a spline according to ANSI B92.1, but am open to other types.

 

[MikeHalloran]

There are other options.

I don't know if they still do it, but back at Brand F, we broached the u-joint splines straight, and rolled the pinion splines with just a hair of twist, so that a little, ahem, affection, was required to assemble them.



Mike Halloran
Pembroke Pines, FL, USA

 

[tbuelna]

brian823,

I don't believe any of the spline types or tolerance classes described in ANSI B92.1 involve an actual interference fit. The major diameter fit splines are intended to create a close tolerance sliding fit between the tips of the external teeth and the roots of the internal tooth spaces, to ensure accurate radial positioning. With side fit splines, there is always some amount of clearance intended between the mating flanks.

As MikeHalloran notes, splines are often machined with lead correction to compensate for torsional wind-up in splines with high L/D aspect ratios. While this "pre-twist" may result in some interference along the flanks when the spline is unloaded, as soon as the spline is subject to operating loads, the shape of the spline teeth should normalize.

 

[gearcutter]

The DIN 5480 involute spline standard would be the best for you to use, see image attached.
The ANSI standard, as far as I'm aware, does not include specifications for interference or transition fits.
I don't understand why so many designers are hanging on to the ANSI standard. From a designer's point of view, I find it to be somewhat inflexible.


DIN 5480 by ronvol, on Flickr

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[brian823]

gearcutter,

Would the tolerances in the chart you provided be applied to the space width and the tooth thickness?

 

[mfgenggear]

Gearcutter

I will ask the question.
what application would required
an interference fit?

the whole Idea behind ANSI is no special tools
are required for assembly.
Interference fit would require a press, & or thermal expansion.
what are the press interference fits

Thanks

 

[mfgenggear]

Brian

My question to you is why do you need an interference fit?
is it alignment or or is it ?

 

[brian823]

mfgenggear,

We are sending out a replacement shaft in an application where the shaft hangs vertically from a hub. On previous design, hub was threaded onto shaft and 4 holes were drilled and tapped parallel to shaft axis at the hub/shaft interface, with set screws inserted to act as keys. Set screws have not been sufficient to transmit torque so project engineer suggests using an interference spline with the set screws. I'm pretty sure the interference is to hold the shaft in the hub and the set screws are more or less a positive backup for this purpose. I asked him if the set screws with a clearance fit spline was sufficient, but he still wants the interference spline.

Brian

 

[mfgenggear]

Brian

with the minimum info

My recommendation is to make sure the assembly stays together is use a spline with a bolt or bolts, or thread & nut application.
get rid of the set screws. but I an sure there is lower cost methods. even with a press fit application secure it with bolts or
threaded stud & nut. depending on the amount of press & size of the splines it can separate.

 

[gearcutter]

We regularly manufacture parts where an interference fit is required.
To help reduce the problems with assembly; we supply the customer with a 'taper lock' bush which has been splined to suit the mating shaft, usually with a class 6 tolerance (most shafts are machined to class 5).

If you have the space for a pair of taper lock hubs, then I would recommend you use something the same or similar.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[gearcutter]

Would the tolerances in the chart you provided be applied to the space width and the tooth thickness?

Yes, that is correct.
For 'diameter location' type fits, separate tolerances are given for the diameters.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[gearcutter]

I will ask the question.
what application would required
an interference fit?

I've seen quite a few through the years.
The last job that comes to mind was an application within a furnace environment.
The spline interface needed to be tight at room temperature.
That application specified a shrink fit, so the hub was warmed up to a specific temperature before assembly.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[MikeHalloran]

I will ask the question.
what application would required
an interference fit?

I think the 'root' answer, if you will, whether flank interference, twist, Loctite, pinch bolts, or shrink fit, is to reduce the incidence of fretting wear in splines that are not periodically (re)lubricated.


Mike Halloran
Pembroke Pines, FL, USA

 

[tbuelna]

Due to the combined tolerances in the mating parts of a spline connection (lead/profile deviations, index errors, runout, etc), it is impractical for even very accurate splines to ensure uniform load distribution with an interference fit. Even a small amount of angular shaft displacement will create large stress concentrations in the spline contacts. Thus it is generally best to design the spline connections with some small amount of clearance to permit sliding.

With regards to fretting of the spline contacts due to relative motion, it is basically impossible to prevent this from occurring, even with an interference fit at the spline flanks. So you must either design the spline to accommodate the fretting (as is typically done), or you must use some other form of shaft connection that will not experience any relative motion at the connected interface (like bolted flanges, curvics, shrink fits, etc).