Pump Shaft and Impeller

Take my advice, weld it! There is no way to secure a working impeller to a shaft with a "bolt", such as you suggest.

Or, put a sleeve inside the impeller, drill and tap it and use set screws to secure the impeller to the shaft.

Nearly every impeller is "bolted" to the shaft.

Shaft has either a male or a female thread and impeller is "bolted" "nutted" onto it

Welding is NOT the solution, nor is the use of set screws, except maybe as a positive lock for the impeller nut or bolt.

So many questions about your project......like how long do you expect the pump to run for?

Are you going to us a L/Hand or R/Hand thread? and which direction is your pump rotating......?

Many pumps are built with keyed shafts - cutting a key shaft will coast the same as a therad and bolt but run or last longer...preform more cycles before failure?

You could go to the local pump shop and just buy a pump....

By welding you are adding stess to to the shaft?

What is you budget for this project?

and are you planning to get more work from your customer?

Vanuta.....please feel free to e-mail me if you want anymore help

info@alnoproductservices.com.au or http//:www.alnoproductservices.com.au

Hi Vanuta,

This is not an uncommon problem. You want to avoid this thing falling off, to put it quite simply.

(1)There needs to be a taper section on the last part of the shaft and a key way cut to accommodate at least a 'woodruf' key, the retainig bolt assembly should have a thread cut in the opposite sense to the rotation i.e the turning will tend to tighten the assembly. The assembly process is in two stages. firstly put it together and torque the bolt to a desired level; then remove bolt and reassemble using friction grip washers under the head (some also add thread grip solution). This will provide a reliable fit add still afford the ability to remove the impeller when the need arises. Do not try welding anything apart from dissimilar materials the ensuing distortion/embrittlement and lack of being able to remove the impeller at all renders this a poor option.

Clearly this answer calls for some fairly clever machining and if this is not available I offer the next version

(2) The ''agricultural option''. Forget the taper. Locate impeller on shaft,tighten bolt as before then cross drill right thru' the impeller boss and shaft and drive in an interference fit mild steel pin which is flushed off at either side. Crude but effective, I've done it and it works!! Final tip to avoid the risk of the pin ever coming out treat the hole with a conc solution of ammonium chloride and make sure there is no oil or grease around. The pin will be rusted solid within 24hrs and you will need to drill it out. I know this might sound belt and braces but the combination will work.

Have fun!!

Thanks everyone,

I was also wondering what are the forces, i should consider when designing the nut for this application?

Hello Massey,

I really liked your answers. I think I used the 'c' shaped lock shaft, on my own machines and they were in a greasy area and never came loose. In the water they will rust and lock even better?

I agree that the screw in the end on the spindle is not a great answer to this problem. Not enough locking force, and it is likely to unscrew as the shaft rotates. The OP could have the screw go in against the rotation of course. But chances are the screw will fall out under load and vibration.

I suggested a square shaft, though this may not work with the pump end ?

I believe you called them an "interference" fit? I never knew what the correct name was for them. Just something I seemed to always have in my workshop and used where appropriate

GA to you Sir ...........

Hi babybear,

Good to hear from you. Yep, like you my workshop experience started young but I had an excellent teacher who did not like having to revisit problems especially when they were in highly inaccessible areas such as down boreholes!

Things working loose like pulleys and other drive devices were often given a new lease of life by counter boring and fitting a sleeve which was typically a ''few thou'' under size to ensure that it had to be mildly flogged on. Tired key ways were always an issue especially if the drive was direct from a diesel which when idling could set up a vicious torsional vibration!

The square on the end of the shaft is another option though I suppose the technical purists might decide that the cuts into the circular cross-section could be considered to be potential weaknesses as stress raisers! Keeping things as simple as possible seems very attractive no matter which side of the argument you are approaching from.

All the best.

Hello Massey,

I thank you for your kind remarks.

With regard to the thread, and answers, it has to be a learned thing and you have to see potential stress points, just as you have pointed out with the square shaft.

My reasoning was if the OP had not got anything yet and was in 'the design stage............. They could use an impeller with a square drive?

I know exactly what you mean when you mention about, "having to revisit problems especially when they were in highly inaccessible areas such as down bore-holes"!

My machines were very fast and there was substantial wear and tear on very heavy and awkward machine stripping, so I preferred "getting it correct the first time", rather than wasting time having to do it again. That is also a reason for 'working 'clean'. It takes just one bit of grit to ruin any work done! This is why you always see the floors of F1 racing Car Manufacturers is 'spotless'!

Take care my friend.

Square shaft is just asking for a failure - too many uncontrollable stress raisers. Also makes the machining of the impeller unnecessarily difficult. Round with key slot is so much easier.

Simple key and centre bolt or nut with positive locking in case of reverse rotation is all that is used on nearly every pump in production today.

Taper fit of impeller is totally unnecessary complication.

Massey,

I was just wondering why it has to be two stages? Why do you have to torque the bolt to a desired level then remove the bolt and reassemble using friction grip washers. Why cant we directly use the grip washers?

the biggest load is the weight of the water you are lifting. calculate the weight of the water, thats the torque resistance you will need on a nut holding the prop to the shaft.

Hello vanuta,

Hope you are well?

I was wondering if a square or square tapered shaft would work with your pump? And the impeller of course would have to be square internally.

Other than that the "Interference fit or "push fit" lock shaft is a very good suggestion by Massey 726. I used these all the time and even though they were in a greasy area they never came loose. The ones I used were almost a complete circle with about 2 mm to allow for the smaller hole they fitted. looking from the end, had a 'C' shape and were sprung and fitted very tightly indeed. They needed a hammer and punch to remove them!

Keep in touch and if possible send a picture?

Good luck...........

Could the impeller be heat-shrunk onto the shaft, in a similar manner to the way that railway locomotive tyres are heat-shrunk onto their wheelsets?

Hello PWS,

Nice idea. Not sure is something so small could be heat-shrunk onto the spindle? Probably just not enough metal there?

Take care............

Why would you possibly want to do this, it makes it difficult to remove and as impellers are nearly always cast with as little metal as possible in the eye for hydraulic reasons you would just be asking to crack the poor thing in tension.

Bollt it on the way every pump manufacturer does.

I regularly work with impellers in the 6" to 16" diameter range used on fire pumps. The industry is very strong on over kill reliability. The standard way impellers are installed is with a shoulder on the shaft to seat on. A keyway in the impeller and shaft. A square key, and a threaded shaft along with a castellated nut and cotter pin.

I hope this picture explains it.

Hello bob c,

Hope you are well?

Appreciate your description and schematic also. On thinking about this I completely forgot about the most common way (a Key-way) to prevent an item like an impeller from just spinning round on the shaft!

Thank you once again.

Hi Bob C

At last some sanity. The only comment I have is that in the petro-chem industry we usually use the design as in post 12 (13) in order to prevent the shaft or the thread from coming into contact with the process fluid. The cap nut has an elastomer seal between nut and impeller. Otherwise the principle is as you describe it which is the same as in post 12 picture. Positive locking of the impeller nut is by direction of rotation (left or right hand thread) and a positive locking system, either grub screw or lock washer or even a tack weld.

The other way is to have a female thread in the shaft and use a bolt - same principle.

Not another word about sanity please, I'm planning to plead temporary insanity to get out of a traffic ticket.

I have a truck that is 92,000 pounds GVW. There is a power divider between the engine and the automatic transmission. There is a shaft in there that is roughly 1.5" in diameter. It is inserted into a gear about 2 " long and drives the truck. This shaft to gear assembly has no keyway, splines, or any other method of limiting rotation between the two parts. The gear is heated with a heating torch until the shaft will drop into it. It is then allowed to cool. There was .003, or .005 interference fit. I thought the design was junk. But if the clearances are correct, there is no way for the gear to shaft to slip. If there was a keyway, or splines, it would just be a starting point for a crack. We know of no one that has had the interference fit gear come loose. The manufacturer of the power divider is Cushman. Go figure.

Hello bob c,

Sounds good and there is a whole lot of torque going through that?

Reference the keyway fit, I suppose it is a little like drilling holes in a rafter to feed wires through.................... it weakens the rafter and can cut the size down from 6" X 2" (~150 mm X 50 mm) down to say 4" X 4" (~100 mm X 100 mm)? But I would imagine either the steel 'type' is harder if a keyway is used, or the overall shaft dia' is lager? Having said that, the few gearboxes I have worked on as far as I can recall all used key-ways. But at that time engine 'revs' and the torque on the boxes I saw was much less.

These days it seems the idea is to make things as small as possible, using 'better' quality and harder steel. It keeps the weight down, and if a car goes, as is often the case, several hundred thousand miles ..................(you can guess at the equivalent metric), they must be doing something right? ;=)= That last part of the 'face is supposed to be a tongue btw!

Take care my friend.

Nothing wrong with keyless interference fits, just not the norm for pump impellers (per OP's question) as you pointed out. Not uncommon for the couplings of larger engineered pumps or sometimes for the impellers of multistage pumps (though these are often keyed and light interference).

Many large compressor, steam turbine and gas turbine couplings are keyless hydraulic taper fits (transmitting many many MW). Many large compressor wheels (impellers) are keyless shrink fits.

BOB C,

Do you have a bigger page or could you please send a link to where i could look at the picture. Thank you.

The picture that I copied is from Waterous Fire Pump and Equipment Co. They can be reached at www.waterousco.com . Or you can call them at 651-450-5287. The picture was drawing #"SPL 82198". The model of the pump was "CR". They should be able to fax you bigger drawings. Good luck.

Hi all, recently read an article on: Acme Keylocking Studs.

This product discribes about locking the studs. Our requirement here is locking the nut on the male thread (at the shaft end), right?

I think with proper modification the concept of Acme Keylocking Studs can be used for this purpose as well.