tree to elbows

[salvatore87]

Hello everyone I am a mechanical engineering student,and I am using ansys,for the analysis of a tree to elbows.for the binding,I would have two possibilities.one is to bind the shaft to the bearings,and the wrench,the other is to consider a strong couple equal to that generated by the forces on the pistons.in the second case how do you apply a torque on ansys?
In the first case how should I bind the quarry of the wrench?

 

[Arturo79]

the couple applies as they apply all other forces.
If the axis of your tree is the x axis you must apply a mx moment.
the optimal solution would be to use contact elements. in this way you would get results close to reality but contact problems are non-linear.
another hypothesis can be to rigidly connect the compression side of the key to the seat. the other side let him free, in the sense not restrain him.

p.s. I personally would not even use contact elements in a dream.

 

[salvatore87]

for the bearing seats I had thought to make an isostatic assembly, so rigidly lock the first, and leave free the slide along the axis on the second,can go?

 

[salvatore87]

to connect rigidly you mean to frame the compression side of the quarry?

 

[Arturo79]

Aspè, we clarify ke with "hardly lock" I mean "glue" the two bodies in contact.
in practice that contact behaves like an incaster ke transfers forces and moments.
in 3d models, simulating the bond of a bearing is not trivial. You'd need to see how you shaped the tree. .
However, you cannot "glue" all the knots of the first bearing as you say.
If you think about what happens in reality you immediately realize that this solution is not good (then it depends on the analysis you want to do.
but do you use ansys or ansys wb?
As for the key (or better tongue, I'm sure it's a tab) you must always see how you modeled: you can select the nodes on the side of the key that transfers the moment and paste them to the corresponding nodes of the seat.

 

[Arturo79]

I happened to see simulations (not made by me of course) where the inflection of the end of a tree was calculated using a 3d model and setting all the knots of the bearing seat.
I immediately noticed the error but I was not heard. When the inflection on the royal tree was determined, the result was that I was right!

In reality, the bearings are built in such a way as to allow rotation (inflation) of the shaft, so they are not perfectly cylindrical.

 

[salvatore87]

Given that I am conducting a static analysis, how could I bind bearing areas?

 

[Arturo79]

risk a hypothesis: in both bearings, impose on the knots on the "contact" cylindrical surface with the bearing seat (pins) a transverse shift equal to zero.
then, regarding the mixed bearing (radial+axial) blocks also the lungitudinal shift.
should work, of course it is a solution that goes well in the first approximation.
do not go down to the details of the deformed you will get.
If you have to determine the stresses in the tree, it's more ke well.

p.s. you see ke studies engineering. . Not even on August 12th find peace:

 

[salvatore87]

thank you so much,however it was the same solution that I had adopted, it is clear that I referred to the locking of contact surfaces, not all the knots

 

[salvatore87]

I'm sorry, but I'm doubtful, but are the values of the forces in ansys expressed in newton and speed in turn per minute?

 

[Arturo79]

thank you so much,however it was the same solution that I had adopted, it is clear that I referred to the locking of contact surfaces, not all the knots

ok xò nn block anke rotation around the axis of the couple otherwise create an ink

 

[Arturo79]

I'm sorry, but I'm doubtful, but are the values of the forces in ansys expressed in newton and speed in turn per minute?

if you set the units in mpa (/units,mpa) you have the strength in newton and the angle speed in rad/s.
However, you must set the following values for the material of the tree:
en=210000 (or 206000)
prxy=0.3
dens=7.85e-9

of course you will get displacements in mm, radiant rotations and mpa stresses
see also in the help

 

[salvatore87]

to tell the truth I did not set anything at first, where do you see?

 

[Arturo79]

to tell the truth I did not set anything at first, where do you see?

If you have set nothing you should have all units of measures in the international system.
cmq test on a classic overhanging beam and remove any doubt.
in the help type /units

Hello