Hoop stress

[sna1]

Hi,
Can anybody help me. I have got stress in X and Y direction with SAMCEF software (I use cartesian coordinate systems) and i would like to deternime the hoop stress. Can anybody tell me how to get hoop stress from principal stress (X and Y).

 

[sna1]

hoop stress

 

[SnTMan]

You have a cylinder, right?

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[rb1957]

is one of the principal stresses aligned to the hoop direction ?

if you have a cylinder (or a curved surface) and a pressure, then you can calc hoop stress ('cause you'll have R, t, and p)

another day in paradise, or is paradise one day closer ?

 

[sna1]

My model is a cylinder under pressure and i would like to determine hoop stress from x and y stress

 

[TGS4]

Without knowing the location, you cannot perform the coordinate transformation between a Cartesian coordinate system and a cylindrical coordinate system. The X and Y stresses (in your Cartesian coordinate system) need to be transformed, but the transformation is location-dependent.

Can't you generate a cylindrical coordinate system?

 

[rb1957]

if you have a cylinder, you can calculate hoop stress, and you'll probably get a number different to the FEM ('cause the FEM is responding to the constraints and geometry changes. looking at you model there should be a obvious "hoop" stress from you elements, probably not one of the principal FEM stresses, but the stress aligned to the tangential direction of the elements (as opposed to the longitudinal direction).

another day in paradise, or is paradise one day closer ?

 

[sna1]

I would like to calcul hoop stress and strain in one noeud so i know the location of the point but how can i determine it?

 

[TGS4]

Transform the Cartesian stresses into cylindrical stresses by performing a coordinate transform. Something that you should have learned in your undergraduate engineering education.

 

[pdiculous963]

While transforming to a cylindrical system is the best answer, another option would be to pull the stresses in the X direction at node in the Y-Z plane(assuming the axis of rotation is the Z axis). These would be the hoop stresses.

 

[rb1957]

do you have stresses in the element axes, or global axes (99% of FEA would be in element axes) ?

so you should have one element aligned to the tangential direction (and so telling you hoop stress).

or you can calc hoop stress by hand. (and compare to your model results)

another day in paradise, or is paradise one day closer ?

 

[sonia518]

This is my model. I have found that (in one element) X stress = 67.7 MPa
Y stress = 334 MPa
Z stress = 11.1 MPa

How can i calcul hoop stress ?

 

[TGS4]

Transform the Cartesian stresses into cylindrical stresses by performing a coordinate transform. It depends on the location.

 

[sna1]

I would like to calcul stress in element num 218 as shown in thee figure.
[cos theta -sin theta 0 ]
[sin theta cos theta 0]
[0 0 1]
To obtain stress in cylindric cordinate system should i multiply this matrix * the vector [ stress (x), stress , stress(z)] to obtein the stress in cylindric system?

 

[dhengr]

Sna1:
Given your latest sketch of the cylinder (your 30NOV15, 16:58 post), why don’t you consider the following. At any given Z plane (latitude), and in (or very near to) the X plane, won’t the Y stress component be the hoop stress? At the same Z plane, and in (or very near to) the Y plane, the X stress component should be the hoop stress, and it should be essentially equal to the Y hoop stress you found previously. At any other location on the circle you have to do some adjusting and summing of the stress components (X & Y) to transform them to be pointing in a radial and circumferential direction. The Z stress component should be the same all the way around. The above, assumes of course, an infinitely long cylinder in the Z direction, without any anomalies which cause significant stress changes. You shot a major bull when you modeled this cylinder by not taking advantage of symmetry, by not placing the Z axis at the center of rotation of the cylinder, and by not using polar (or cylindrical) coordinates to model it. Remember, a computer program does not a thinking engineer make.

 

[rb1957]

what FE code gives results in global axes ?

and doesn't provide a local option ??

with such lousy results, pick an element where the global axes align with the hoop direction. previous posts about using vector summing also apply.

another day in paradise, or is paradise one day closer ?

 

[TGS4]

sna1 - yes, that is the vector transform.