Dynamic Shaft Analysis 2

[softmech]

Hi ,

Am trying to work on step shaft analysis( Dynamic).
Details.

Shaft Rotation ----2900 Rpm.
Radial load---- 60 Kg
Axial Load ---165 Kg
Impeller weight 6 Kgs.
Bearing distance -- can be assumed.
Dia 1 of shaft -- 35mm, Length -- 120 mm
Dia 2 of shaft --30mm, Lenght -- 90 mm
Assume any suatble material

I am trying to determine the shear stress induced at the step in the shaft during shaft rotation at 2900 rpm. Plz tell as to how to approach the problem and kindly furnish me with details scuh as :

Package to be used in ansys(structural or DYNA or suitable).
Element type to be used.
Steps involved in ansys to get the solution.
Any other important details.

PLLLLLLzz help me

 

[brad]

Unless I'm missing something, this problem sounds like it could be more readily handled using basic elastic calculations (as you have stated that material doesn't matter, this appears to devolve to a simple Solid Mechanics problem).
Isn't FEA overkill?
Brad

 

[machineryguy]

Softmech,

I would agree with Bradh in that this does not seem like a problem warranting FEA. This is providing you do not have some unusual geometry in your rotating shaft that you can't find tabulated stress concentration values for. Definately start with a hand calculation as opposed to FEA.

That aside, in my opinion, it might be a GREAT idea to do BOTH. It sounds as if you are relatively new to FEA, so solving a model that you can easily verify with hand calculations will most likely be insightful for you. (Just dont stop when you see colorful results, and try to get a feel for what is going on with what you see or don't see).

At the risk of sounding like an *******, I am going to suggest you peruse the ANSYS help manuals. You should find a problem/solution similar to what you are describing above, that you can use as a benchmark/guideline.

Take care

 

[softmech]

Thank you very much for response guys. I am new to Ansys. I have solved this problem manually and through static analysis.

The reason as to why am doing this in ansys is we have a key way at the begining of the step shaft. We have received a few sheared shafts coinciding with the location near the keyway . So i would like to do dynamic analysis .

1) Is it possible to simulate a rotating shaft with all the stresss being displayed.

2)Are there any other means of doing this real time analysis so as to see how the stress is varying in the shaft


key way -- 3m radius
key way length -- 6mm

bye

 

[GregLocock]

1) I've never seen it done, but I bet you aren't the only one trying to do it. Applying rotating loading to a 'normal' FE model sounds tricky, but given linear superposition I suppose it is possible.

2) Yes, I think so. Export your FE solution to a mechanisms analysis program and let it model a rotating shaft for you as a time based simulation. Solve time might be extravagant! You may have to reimport the shaft deflections into FE to get stresses.

Cheers

Greg Locock

 

[corus]

I don't see why this is a dynamic analysis. Your loads are those defined plus the torque in the shaft. If your loads are not axisymmetric you can use a 2D asymmetric loaded axisymmetric model of the shaft to obtain your nominal stresses which you can then factor for the keyway. Use a fatigue assessment based upon rotating shafts to see if the stresses at the keyway are within allowable values given the surface quality, material, and shaft thickness.

 

[brad]

corus makes a great point with his 2d suggestion.
Even if you feel you need to do a 3d analysis, the only "dynamic" thing about this is the radial load as the shaft rotates.

You have a changing load due to your oscillation (2900 rpm). This is @50 Hz. As long as your first bending frequency is greater than 150 Hz, this load can be treated as quasi-static.

In this case, even if you want to model the full 3d model, you could still treat this as a static analysis with several applications of the radial load.

If you know how to use linear superposition, this devolves into three applied loads (one axial, and two orthogonally-applied radial loads). These three loads when properly combined should be able to describe the stress state of the shaft as it revolves (presuming that the quasi-static assumption does indeed hold).

Brad

 

[desertfox]

Hi softmech

I maybe over simplfying your problem however at any point in time during the rotation of the shaft, the resultant stress would be given by the combined load of the torsional stress and the tensile stress.Therefore the stress your looking for would be a principle stress, which would be a maximum at a line acting at 45 degrees to the centre line of the shaft. The principle stresses can be calculated using the Mohr's circle, which I assume you are familiar with. In addition having constructed the Mohr's circle this would give you not only your maximum principle stress but also your minimum principle stress from which you could do a fatigue analysis having now established your stress range.
Also you mentioned a shear failure near the keyway if you look at the two fracture surfaces do they run at roughly 45 degree's to the centerline of the shaft because if they do your failure is due to the principle stresses, if however the fracture surfaces run at 90 degree's to the shaft centre line then they have failed purely by shear.
I also have some stress concentration factors for keyways here which I detail below so when you calculate the stresses
you can factor them.

Kt= 1 + (h/r)^0.5

where Kt=stress concentraction factor
h=keyway depth
r=radius at the base of the groove or keyway

for a semi-circular groove Kt=2

for a square groove with an corner radius of about (1/10)of the keyway depth Kt is approximately 3.

Finally according to my mechanics book (Mechanics of Materials by EJ Hearn (Pergamon)ISBN 0 08 031151 2)you can reduce the stress concentration factor at the keyway by reducing the corner radi in the keyway or by drilling several tiny holes adjacent to the keyway on each side at an angle of 45 degrees.

I hope this is of some help