Keyway design - method to retain steel key 2

[awol]

We broach a keyway into an ID of a shaft and silver solder the key in place. Low temp silver solder due to the material. Although not recommended, the parts are assembled in the field using excessive force (hammers) and the solder cannot hold the key. I have investigated epoxy for more flexibility. Cannot weld, material is heat treated.
My next attempt will be using liquid nitrogen to shrink-fit the key in place. Has anyone had success in retaining a key in a keyway while a load (imact or otherwise) is applied to the key along the key slot?

 

[unclesyd]

We use a screw/screws to hold a lot of our large keys both in S/S and Alloy Steel shafts.
Getting a shft on and off shouldn't be that troublesome. I would check my tolerances.
I have never seen a key shrunk into a shaft. I take it that you are using all keyway tricks, rounded ends, ground keys (key stock, an the proper tolerances.

 

[awol]

unclesyd,

Yes. The ends are round and the key stock is ground. The keys are 1/8" square and approx. 1.5" long. Too small to use screws. These items assembly easily before they ship. But dirt/ grime coupled with impatient users who know how to use a hammer have caused another look at this design. We urge cleaning both components prior to assembly. But it doesn't always occur. I am considering a change to the broach and create a type of 'dove-tail' instead of the straight edge keyway slot. Then machine the keystock to slide into the dove-tail slot. Finish by staking the edge of the slot at each end to prevent the key from sliding.
Only the lower half of the key is machined. The upper half is per existing dimensions as to mate with existing components.
Ever seen this done?

 

[EnglishMuffin]

Another possibility might be green Loctite.

 

[unclesyd]

I seen similar setups but on larger keys. There a couple of split keys and keys with shim tabs, again on larger keys. One has to watch the amount of key left in shear and where the key is reacting. Your key is so small that I would do nothing to reduce its' cross section.
Let me do a little head scratching on this one as when keys got this small we used round keys.

Is the key pushing out the back of he shaft keyway?

Can you change to Woodruff style key in the part you are mating with the shaft?

 

[funnelguy]

The shaft has a through hole, correct? (or else how could you broach the keyway..)

The silver soldered joint at the key fails? Is there damage to the shaft as well?

Perhaps you could wire EDM the bore, eliminating the key? Very expensive and probably impractical. The key is still relatively fragile, even as an integral part.

Perhaps you could include a hardened 1095 stamping in the shape of your bore as a scraper. Include the scraper and instructions as part of your kit?

 

[bklauba]

Another possibility is to use a keyway design in the interior shaft (which might not be an option open to you) that does not exit the end of the shaft. We have made many of these for electric vibrator applications in the size range that was mentioned. We have found that a Pratt & Whitney style key & keyway design, where the keyway is cut with an end mill, rather than a Woodruff cutter, and the key has a lateral profile that is rectangular, works well.

But shaft, key and keyway fits of all sorts should be assembled with patience, willingness to examine where interference engagement is taking place, and a bit of common sense combined with skill to achieve a good running fit, which appears to be your goal.

Besides, if they are hammering these items together, not only are the damaging the shaft and key fits, the end of the shaft (which is undoubtedly mushroomed) but also the bearings and perhaps other components. Can you train or provide good instructions to the assemblers? Figuring out a way to shorten their leash should not be overlooked.

 

[EdDanzer]

You could use a spiral lock ring or two to contain the key. Smalley has some light duty rings that may work. It may require a deeper key in the hub so there is material for the ring to push the key into hub. A shorter key will also ease assembly. The other possibilities are to drive the key in after the hub is on the shaft, or use a tapered bushing. We us QD style for ease of machining.

 

[awol]

englishmuffin.
The 290 Loctite may or may not work in this application. But from the Loctite webpage I found some epoxies that list high impact resistance and have high strength. I will get a sample to try. Thanks.

unclesyd.
The solder is letting go of the key as the outer assembly (with the key) is forced onto the shaft. Once assembled, the key cannot go anywhere and functions fine. These parts are then disassembled. That is when the key pops out loose.
This whole assmebly/ disassembly works great with the key. We solder the key in for convenience to the customer. But when it comes loose they call and think the machine is broke. Current (corporate) response is "the machine will function. It only requires the loose key to be stored and then replaced in the keyway prior to asssembly". I am getting tired of this answer and want to permanently solve this issue.
I may not have the room to try a woodruff key. I will check the layout.

funnelguy.
Mostly they knock the key out. Shaft is nitrided 4340. If we get the key back for repair, it is usually flared on one end from the force. We can tell what they did. Getting someone to admit to it is the hard part. We end up fixing most just to save face on the customer level. Got to keep customers happy. The scraper would be easy to create and supply. I will investigate.

bklauba.
Patience during assembly? Training and instructions? You are at the root cause! I don't think there is a design change I can make to be bulletproof for this application. I am hoping, at best, to find a method of securing the key to resist "some" of the abuse this assembly will experience at the hand of the customer. Once the key starts to 'mushroom' , nothing short of a resizing tool could prolong the life.

ALL GREAT COMMENTS!!
-awol

 

[BOBV]

Since the screw idea works in some applications but doesn't work for your smaller keys, could you consider small split/roll spring pins? It seems that you could use some kind of an expansion mandrel to hold the key up, in the keyway, and then tap in two small pins and then remove the mandrel.

 

[KevinJohnson]

How was this resolved?

 

[Barry1961]

If this is a hard duty application you might consider using a keyless bushing. It would cost more but would save some installation problems and be a lot stronger. There are several people that make them, B-Loc, Peters, Ringfeder are a few.

Barry, Titular Despot Emeritus

 

[rimbaldi86]

There is a polymer product called 'Belzona'. Best thing ever invented. Very strong, can be filed, resists heat to 200 degrees. An industrial grade recovery compound. Often used to repair engine blocks that have had a whole blown in the side from a thrown con-rod. Good stuff!

 

[MDAPSEY]

Have you looked at Grooved Keys

Check them out here

http://www.driv-lok.com/grv_machine_key.htm

 

[metman]

Have you ever seen a tapered key (tapered radially that is)? It has a tang on one end so that the brutes that bang it into place cannot overdrive it. Naturally they can mash over the heel of the tang if they so desire because as you say nothing is idiot proof.

 

[griffengm]

Metman is on one good track, I think. The tang provides a larger surface to hammer.
BUT, is there a reason that the key cannot be installed after the hubbed component is in place? If features are so abused that the key cannot be installed at this point in assembly then any warranty should be voided. This usually gets the attention of the bean counters and the problem is solved at the customers facility.
You are adding cost on a feature that is counter to standard maintenance practice. You are either pricing yourself out of the market or eating into your margins.

Griffy

 

[pennpoint]

awol
Are you saying, you are trying to attach the key to the shaft independant of any hub or pully. Useually a set screw in a hub or pully is located over the key and locks the key into both the shaft and hub. Maybe I missed something or was reading too much into the AWOL's post of Dec 17. It's been bugging me about his statment about the screws being "too small". So I'am picturing small screws counter-sunk/bored into the key and anchoring it into the shaft before the hub/pully or thingy gets put on the shaft.

Please correct me if I'am in error on this.

Thanks

Regards

pennpoint

 

[metman]

awol,

Haven't gone awol on us 'ave ya?

I was so intent on adding my two cents that did not give full attention to MDAPSEY's post.

From everything I gather from your needs, this seems like an ideal answer and therefore have repeated that link here.

http://www.driv-lok.com/grv_machine_key.htm

 

[awol]

I have been reading all these great suggestions and I intent to try a couple. There are actually two shafts and two keys. The first component onto the shaft must slide completelly over the first key and come to rest on the second key. Then a second component is installed onto the first key. The keys must be in place and in line during assembly. These keys are .250" and 3/16" square and makes mechanical means (set screws) difficult to install to hold the key. The belzona idea has an added feature of 'filling in ' the keyway behind the key and acts in a compressive capacity to prevent the key from moving. This fix will be messy with the two-part mixture. I had overlooked the drive-loc idea, believing what was meant was a one-way, round driven pin. The drive-locs have "teeth" that grab the keyway during installation. I will contact the company for a test run of the product. Thank to all who have kept this issue alive. Great solutions are just a forum away.... awol

 

[rhodie]

Redesign the part(s) to have splines, not keys.