Shaft key calculation 1

[Ledombre]

Dear all,

My question is the following:

I have to check the shaft-key connection of a 3000kW - 1000RPM asynchronous motor which is intend to drive a pump.
When I compare the shaftend diameter with other motors (similar torque) the shaft-end diameter is smaller. To be sure I want to check it.

I have made some calculations as described in the DIN 6892.
I understand the way of calculating, but have the following questions:

Is there a rule (and stated in a standard?) for selecting the maximum torque? (f.e.: 2x Teq???)
Is there a minimum safety factor?
(or is this already included in the Ka factor)

Are there other standards which describes this calculation?

Thanks in advance.

 

[DVJones]

In my somewhat limited experience, there is no factor of safety in the Ka factor. The FOS is a seperate factor in the calculation of the shaft diameter. Do you know if FOS is standard in your department/company concerning similar designs? I hate xerox engineering, but baselining on succesful designs is always a good thing. I would not use a FOS any lower than past successful designs. FOS can be a bit of a nebulous thing , imho, and is influenced by quality of material data, past testing (or succesful designs), analytical model quality, your company history, among other things.

http://www.roymech.co.uk/Useful_Tables/ARM/Safety_Factors.html

seems to have a reasonable guideline that is in good agreement with one of my favorite refernces ('Machine Design: An Integrated Approach' by Norton). Of course, company policy may override these guidelines.

One question - are the designs you baselined the same material? If so, that would indicate to me that more work is needed fleshing out the proper FOS. If they are different materials and your current material is 'better' for the application, then I wouldn't be quite so concerned.

Unfortunately I dont have direct experience with the DIN spec, or a good rule of thumb to give you concerning torque.

Hope my ramblings can be of some service to you, and sorry if this is all very obvious to you.

 

[Cockroach]

This is actually a nice calculation, but I need to know the size of shaft in question in order to get a solution. Unless you want theoretical, you need to state the size of motor driveshaft.

I figure it must be big, 3000 kW at 1000 RPM is 4023 ft lbf of torque! Any idea material specification? Must be something on the order of AISI 4140/4130 HRc 28-32, minimum?

Regards,
Cockroach

 

[KllrWolf]

Cockroach - shouldn't that 4023 be the HP, with 21121 ft lbf of Torque, or am I having another one of those confused days?

 

[Cockroach]

Oops, correct, that is 4023 HP and 21121 ft lbf torque. Material specification by the DIN 6892 would be EN-GJL-250, C45K or 1C60N.

My confusion KllrWolf, before the first morning coffee.

Regards,
Cockroach

 

[KllrWolf]

No problem, just hope the OP lets us know if the information helped him.

 

[Ledombre]

Thanks for all response, more details from my side:

Nominal torque: 28650Nm
Application factor (Ka)1,5
Equivalent torque: 42975Nm (=Ka x Tnom)
Maximum torque: 85950Nm (= breakdown torque of AC motor (=2) x Ka)
Shaft end diameter 180mm
Key width: 45mm
Key height: 25mm
Key length: 290mm
Functional key length: 267,5mm (= length - (0,5x width))
Shaft material S355J2G3 (St52.3 N)

When I calculate with this data I reach a safety factor of 1,08.
This is according the DIN OK.

My concern is when I compare this motor with a motor which we have delivered in the past, the shaft diameter was 210mm (with the same key lenghth, shaft material and torque (3000kW/ 1000RPM))

Must be noted we have also delivered a 5000kW/ 1000RPM motor with 210 mm shaft diameter en same functional key length.

Best regards.
Ledombre

 

[KllrWolf]

I did some quick calcs and do not see a problem with the shaft diameter or keyway.



I used the breakdown (highest) torque for the calculations.

I got the material strength from a quick search online.

My calculations are not according to a specific standard, but they are what I use for our motors and have not let me down yet.

 

[Cockroach]

Looking at your input specifications, I do not find an issue with your system. I get 1.33 as the lowest expected factor of safety found at the motor startup where torque is the highest.

I have attached the computational model for your reference. Note that I also found a material specification for this S355J2G3 shaft which listed a proof strength of 285 MPa. I used this in the calculation since it would be reasonable to expect your material to have similar behavior performances.

So highest torques, lowest material specifications given your key and shaft geometry, FOS=1.33 (worst case). I think you're fine.

Regards,
Cockroach

 

[kingnero]

Cockroach, You've used the yield strength of the shaft for your key, because "it would be reasonable to expect your material to have similar behavior performance".

I've already seen keys in 303 stainless, having a yield strength of ~15-20% less, in CK45, at about 200% of the yield strength, and in all sorts of steel in between.

I'd advice Ledombre to double-check this value before assuming your statement, especially if the material is softer than the shaft.

I do applaud your step-by-step calcs as tehy're very informative for an anonymous forum...

 

[Cockroach]

I use Key material yield strength because, in the worst case, the onset of plasticity I would consider the beginning of failure. I don't want the Key to be permanently deformed at any point in the application.

My FOS would be very conservative, say compared to using ultimate strength, which frequent the literature. But this would be a matter of designer preference, my suspicion is that LeDombre's calculation used this, hence the much lower FOS.

...and thanks for your kind comments.

Regards,
Cockroach

 

[Ledombre]

Dear Cockroach,

Thanks for your calculations. It is similar to the DIN 6892 method

The DIN 6892 calculates in 2 ways; first a calculation with the equivalent torque and second with the maximum torque.

Next to this the DIN takes more external factors in consideration, for example the frequency of changing the rotation direction of the shaft (not applicable for a pump) and the frequency of load peaks.

My final conclusion is, when the factor of safety in both calculations is above 1, the connection is sufficient.