Keyway Profile

That was my take on it. The engineers that designed it are long gone, the new ones don't have a comment.

Prevents rotation between the shaft and the impeller.

Keeps the key from falling out when mounting prop....?

The impeller sits on a shoulder and has a washer under the nut.

Is the keyway cast in or machined in post casting?

The keyway is machined to tolerances shown. I think it is for stress relief but I wanted you guys to back me up because a customer is claiming they are defective. There is nothing on Google that mentions this.

Thanks, I should have said "nothing I tried turned up anything".

U just did!

Hi, Not really an answer to the question you've posed (so maybe off-topic?) but

1. cast stainless steel is not in general brittle - there may be a grade that is, but I don't know it

2. as most engineers know, keyways are for timing, not for transmitting drive; if in a design the keyway is used for transmitting drive the components will certainly fail in service - one part or other always wears to the point of failure. I can think of very few reasons to time an impeller to a shaft (look at almost any B&S engine to see one).

3. this keyway is therefore unlikely to be required and the joint would be better without it. Should it be one of those few occasions where timing is needed then a shallower non-standard keyway (with radiused corners) would give lower stresses and a stronger job.

dieselphil,

I agree on your first point, in physical metallurgy there are always compromises where you have to sacrifice one property for another, say toughness for strength.

but your second and third points about transmitting power, I think you should step back and reexamine that and look at the picture below and explain how the power is transferred from the pulley to the shaft.

It's my opinion your looking at through a tube. so to say (figure of speech) and not the big picture.

GA! Unless the shaft and bore are tapered, the key and keyways are the only significant means of transmitting torque from shaft to wheel. Especially true when, as the OP mentioned, there is a washer on one or both sides of the wheel.

Hi chaps, I beg to differ with several of the previous comments, and as an engineer in current practice designing compression ignition engines I have as much experience as most. It is true that many engineers and most other people think that keys are commonly used to provide a torsional connection, BUT unless this is in conjunction with a frictional drive of some sort, whether radial interference, a taper giving the effect of radial interference, or a large axial clamping force together with washers etc, the inevitable "working" of the components of the joint during service will always cause joint failure. Thus keys can only be regarded as a timing or positioning device except, possibly, for short-term use.

Just think how many keyed flywheels you see on modern car engines for instance, and isn't that just the most highly loaded joint of any?

Best thingies to all,

As a counter point, one must look at the intended use and maintenance frequency of the keyed connection. Granted, many torque transfers are via an interference or locking taper - but not all... not even close.

For example, most rolling mills have a keyed connection between the shaft and the forming roll. The fit is loose between the roll and the shaft (constant roll changes to make a different diameter/wall size) and the key is designed to transmit torque, period. The entire ERW Tube and Pipe global inventory is manufactured via torque transmitted through keys.

As an engineer, do you realize your have drifted from the original post. You try to input your reasoning of no key way is required due to its only used for timing purposes and not power transference. That's not true. And instead of speculating a taper to support your reasoning, before more time is wasted, ask if there is a taper fit. ..... I'll ask for you. In the mean time, Please read or retread the original post.

And by the drawing it does not look like it it is a taper, but,.... Is there a taper fit here?

"Keyed flywheels " really is a bit over the top.

Unless we are the victim of two great countries, separated by a common language, flywheels have never been keyed. They are directly bolted to the end of the crankshaft.

I spent 8 years as the engineering manager of a wet process equipment manufacturing company.

We fabricated our own impellers from stainless steel, Titanium and Hasteloy C 276.

Motors ranged from 3/4 HP to 15HP.

ALL of our shafts were keyed. All the motor shafts were keyed. Two set screws secured each motor shaft to the impeller shafts, one on top of the key the other one 90° from those.

The shafts were machined to very close tolerance and smooth, except for the single keyway.

They were bullet proof, and easy to maintain and change impellers.

So, we have a difference of opinion.

Cheers.

This is off topic, but interesting as to the path this has taken. For for a keyless power transmission here is another example.

on a circular sawmill the hub that holds the saw blade. The hub (about 6"-8" diameter) is tapered like a cup, and the outside edge holds and drives (torques) the blade.

Dieselphil: in response with respect to answer # 2

Many swamp cooler blower pulleys have keyways and they have nothing to do with timing.

And an engineer designed the pulley that way.

Take for example : Dial Mfg. 10 " x 1 " pulley, part # 6324

Sorry, but you should reconsider what you have said.

P911 is correct in his assessment of your statements.

KEY TO KEYWAYS

Key (engineering) - Wikipedia, the free encyclopedia

In what sense are "they" ( the impeller?, the keyway?) defective? Is the impeller failing in service, and how? Or is the problem with assembly?

One thing I do note from the drawing detail, the keway width of 5.1mm seems to be excessive clearance for a standard 5mm key. This would lead to the key shearing under acceleration.

I would expect the keyway to be cast with material allowances for machining (broaching) to finish dimension on width. Depth, having looser tolerance, would probably be as cast.

I can't find any keyway broaches with that profile.

But, phoenix911's answer seem correct to me too.

As others have said, probably an effort to remove stress concentrators.

As luck would have it, I had a recent discussions with a mech eng who was presented with a gear that had the keyway machined to sharp corners. One assessment done was for rounded undercuts similar to what is shown. The FEA for this proposal was actually worse as it still concentrates the stress and also removes some of the parent casting for the spoke involved. (72" diameter gear, 18" face width, sugar mill.)

I suspect that your customer might be correct in challenging this design.

what is the advantage of using keys with round bottom?

Woodruff keys are retained and held in place without a fastener

I see what they are trying to do, but feel they have misapplied the principal in two ways:

First, the under cut profile should be on the shaft since that is where the key fully seats. There is clearance between the top of the key and the bottom of the keyway in the bore that negates much of the need for undercutting.

Second, the above not withstanding, the concept is applied incorrectly. Rather than an undercut, the corner of the keyway should simply have a radius tangental to the sides and bottom of the keyway, and the key itself chamfered. You will still incur the same loss in regards to percentage of engagement, but the stress-riser mitigation will be far superior.

I've come across this in my past Machine Design days woking for a company that had a long and storied history. I was instructed to always look at maximizing material at point of interface and the solution with the least amount of profile changes is generally the correct one.

As a secondary thought - material and load type play a huge role in what type of stress riser mitigation can/should be applied and what the fit should be... the standard called for in the Machinery's Handbook is not a one-size fits all solution.

Its a multiple use impeller, I saw those years ago when I was a pool & spa service technician. It could be used in a wet end to either push or draw water. I was told that the curves were there to lock the impeller to the shaft if the rotation was clockwise or counter clockwise.

There seems to be a difference of opinion on the proper application of a keyway on a rotating shaft.

From Wikipedia:

In mechanical engineering, a key is a machine element used to connect a rotating machine element to a shaft. The key prevents relative rotation between the two parts and may enable torque transmission. For a key to function, the shaft and rotating machine element must have a keyway and a keyseat, which is a slot and pocket in which the key fits. The whole system is called a keyed joint.^[1][2] A keyed joint may allow relative axial movement between the parts.

• Every time I used a KEYWAY, it was to prevent the free rotation of an element on a shaft.
• It is also used to prevent damage to the more expensive parts of a machine by the keyway shearing off when the rotating parts stopped too fast or too much torque was applied or something seized on the machinery!
• It is an inexpensive way to join a shaft and other element.
• As for the particular profile of Unredundant's picture, I have not seen one like that. Several of the other posters have mentioned that it is for stress relief and I truly believe they are correct, plus the manufacturer makes a bit more money by selling the end user a custom made key for a non-standard keyway.
• It looks like you could use a standard square key in a pinch, but that may lead to warranty issues if it fails and damages the shaft or impeller?

Good luck.

Guys, I am back (Had to work), The impeller, key and shaft are 316 (Swedish SIS 2343) the drive is 13hp.

These pumps are intended for operation in ph range 2-10 (Temperature dependent)

The key is square,

The undercut may be to alleviate stress due to corrosion.

We have never seen one crack at the keyway.

Thanks to everyone for the input.

usually a key will roll before it'll crack...... well, I have seen one roll (probably due to too loose of tolerance, wrong size key or a shade tree mechanic's work), but I have never seen one crack.

Supposedly, standard 316 stainless steel (not 316L) does have an issue with chloride-induced stress corrosion cracking. Probably since this part of the impeller is wetted, the desire to reduce stress points must be related to some failures seen early on in the generations of this part.

I know it to be a hard fact that 316 will fail under conditions with significant heat, high chloride residual concentration in the water, and some considered stress inherent within the design of the metal (a coil for example). A flat plate (annealed) not nearly as much problem, but could still fail, depending on how this is mounted.

Sometimes stainless steels will fail under stress corrosion cracking mechanism when there is a high concentration of organic acids (acetic, citric, etc.) Heat also exacerbates the issue.

This may be a bit off topic but back when Lockheed was developing the SR-71, some of the machinists would use regular black ink marking pens (like a Sharpie) to lay out patterns on some of the titanium and special stainless sheet that was being formed or machined. One machinist drew a straight line across a 3' square sheet at the end of the day, the next morning he came back to the sheet, tossed it on a bench and the sheet cracked and separated perfectly on the ink line he drew the previous day. He took his findings to his superior and they discovered that the ink had a chloride component and the ink had caused a chemically induced stress fracture. They immediately stopped using the pens and had a new pen and ink developed strictly for the specific material. You can see how a simple thing like ink could cause a failure of the material

If my eyes are not fooling with me (again), I am seeing a cut-away of the same pump where the key-way exists, also showing the key stock in place?

Nice drawing!

I see two ladies dancing.

The cut-away is of a smaller model, The vanes on the one in the drawing overhang and depend on the upper diffuser to form a portion of the channel. This requires adjustment with shims behind the impeller to get minimum clearance, one reason for the simple parallel key. An artist by the name of Krueger does cut-away's for a lot of companies in Europe.

I have overhauled fire pumps ranging from 8 HP to pumps generating over 2,000 GPM at over 200 PSI. I have never seen an impeller that did not have a key locating it to the shaft. Some were tapered, some not. But all had key-ways and keys.

If not designed to transmit torque, what was the purpose of the keys on older axles? What about today's splined axle shafts? is that not just multiple keys machined into the axle?

Yep.

Maybe the Brits do it differently.

I used to car pool with a guy who drove a Triumph TR6.

His passenger door NEVER closed properly. The plastic door fascia was fragile and I was afraid I'd break it if I really slammed the door.

I had a '58 Triumph Motorcycle that I really liked. It was a "Marlon Brando, Wild One" model.