calculation of motor shaft 2t

[Brambilla]

Hello, I'm doing a reverse project for university on a chainsaw for home use along with 3 other students.
the motor shaft consists of 2 semi-albers connected by force; we imported into ansys workbench the assembly of the two semi-albers in .stp; Informing us around it was recommended (to obtain more realistic values) to insert identical nominal diameters (12 mm) and then set a "connection with friction" and enter a interference value ("offset for the connection surface").
the radial interference between the button and the hole in the sleeve is 0.035 mm. inserting these parameters in ansys, however, the only interference generates a maximum voltage around 1400 mpa: this level is much above the breaking limit for our material (965 mpa - reclaimed steel).
What's wrong with us?
we can not understand if it is a problem of setting in ansys, if the value of interference is realistic or exaggerated (we have tables of the manufacturer, but we have a date for a semi tree and a table revised a few years later for the other tree, if we miss something for the type of internal efforts to the material (traction-compression).
Thank you.

 

[cacciatorino]

we can not understand if it is a problem of setting in ansys, if the value of interference is realistic or exaggerated (we have tables of the manufacturer, but we have a date for a semi tree and a table revised a few years later for the other tree, if we miss something for the type of internal efforts to the material (traction-compression).
Thank you.

I don't know ansys, but I had made a simulation with another fem and I had modeled the parts with real geometry, not as you did at the nominal quota by imposing an offest.
the results were quite consistent with those obtained by analytical way.

However studying interference is quite complex, since it is necessary to introduce non-linearity and mesh must be very dense (in fact, these analyses are done with a symmetrical 2d axial model, not on the real 3d).

I suggest you work first on a simplified model: make a full pin d20 mm and put on to interference a thin ring with thickness 0.5 mm. In this way you can easily calculate with the theory of elasticity the tensional state of the ring (you will assume that all the deformation you have to recover the interference will be loaded with the ring).
Then you'll study a similar fem model and you'll do all the "tuning" you need for numerical results to be similar to those of theoretical analysis. Once you learn all the tricks to make ansys work as you need, you will return to your engine tree.

 

[Brambilla]

thanks for the sudden response, however, the mesh used is quite dense... the geometric interference values that provides us ansys, that is due to the internetration of the two mesh is around 1e-9 mm, I would say it is a good approximation considering the level of precision that interests us.
Now we're going to check on a simple model to verify that the results match between theory and fem, actually we didn't think about it.
but about the figures you got, can you tell us something? on a shaft of these dimensions 0.035 mm radial interference is a plausible value?
the inner pin is subject to compression, the inner surface of the hole is instead subject to traction right?

 

[cacciatorino]

the inner pin is subject to compression, the inner surface of the hole is instead subject to traction right?

for the pin yes. For the hole, you'll have a tangential sigma that obviously is traction, and a radial compression sigma. The tangential sigma should be considerably higher than the radial one if the ring is thin. If the ring is nice, then things will change a little.

Bye.

 

[cacciatorino]

but about the figures you got, can you tell us something? on a shaft of these dimensions 0.035 mm radial interference is a plausible value?

I don't know, I remember when we were a kid, we changed the scooters once a season, and we did it with a very small press!

Actually, I don't know how to answer your question. I should also do the analysis.

 

[stevie]

the problem you are facing is a classic axial-symmetrical problem and can be solved analytically with formulas of lamè.
once you notice the contact pressure between pin and hole (linked to the imposed interference) you can evaluate radial and circumferential stress in the two parts.
the value of the interference set is linked to the torque that must be transmitted, that is, once you notice the torque to be transmitted you will find the least interference that thanks to the friction is able to transmit that torque.
I have spoken of minimal interference because this kind of forced juggling induces static stresses in the organs that are added to normal exercise demands, so it is convenient to reduce them to the minimum indispensable.

 

[Brambilla]

Yes, if we talk about rotating trees we do this, and that's exactly what we can't explain; This type of force does not need to connect 2 concentric trees (it is not a "normal" shaft-wire link), we talk about tree to elbows and the forcing is near the crank pin on one of the 2 sides. the passage of couple therefore should not be tied to the force but to the structure of the manovellism... or are we wrong?

 

[Er Presidente]

I would not want it to be a chainsaw the engine is not considered "to lose" and not intended for heavy replacements.
in this case the manufacturer may have decided for a fitting with definitive deformation.

I agree that friction should be low in this case, even absent.

 

[sampom]

Hello,.... chainsaw.... .
Thank you.

Other:biggrin:

..uhmm.,
chainsaw, Tuscan, student, ca..
by chance one of the 3 "soci" is pippo82?

in mechanical design on several occasions it has been widely discussed of the thing..:biggrin:

greetings
Marco:smile:

 

[sampom]

the passage of couple therefore should not be tied to the force but to the structure of the manovellism... or are we wrong?

No, you should be right:
because here the transmission of the bike is assured, even if the crank pin was free to turn.

the stable coupling is to guarantee the "linearity/rigidity" of the tree, practically as if at the end it was "monolitic". It is necessary to avoid that the two flyers "svergolino" and the pin is infected (but also that it twists) under the resistant torque of the shaft given by the transmission (the chain that cuts the wood could nail itself).

very generally speaking as they are completely fasting of ansys and fem.

greetings
Marco:smile:

 

[Brambilla]

@er president: This observation is very interesting, especially for the purposes of drawing up the specific plausible of the machine, given the user to which it is intended. the question is, why would they do that? Where is the convenience compared to a more blatant force?

in mechanical design on several occasions it has been widely discussed of the thing..:biggrin:

greetings
Marco:smile:

I took a look and I don't think anything like this has ever been treated. ..cmq I can easily imagine how ever pippo82 and its interventions may have remained in the heart

As for your answer I don't know, the clutch is centrifugal, in case of locking the blade I think it would be this to resent (fast deterioration) and not the tree... but sincerely I did not make calculations on the friction force to regimen. Axial shifts remain, but for these are enough friction forces far below those that come with such interference.