von mises according to the temperature

[stef_design]

Hey, guys.
I wanted to ask you the following question.
If I carry out a mechanical analysis (at t=25°c) of a component subject to a force x I find stresses max and min.
if I perform the same test at a t=100°c the max and min stresses will be equal?

In theory it is because σ = f/a0, but if I do a very fast analysis I see, for example, that if I look at von mises the results are different.
100°c analysis stress values are broadly higher than that at 25°c.
Is that normal?

Thank you.

 

[cacciatorino]

In theory it is because σ = f/a0, but if I do a very fast analysis I see, for example, that if I look at von mises the results are different.
100°c analysis stress values are broadly higher than that at 25°c.
Is that normal?

If the material you used is linear elastic, the only thing that comes to mind is that the bond conditions cause tensions due to binding reactions resulting from dilation. For example: considering z as a longitudinal axis, if your beam is bound on the flat face with a constraint of earth, the expansion of the face bound in x-y direction will be prevented, which will cause you additional stress.

How are the values of sigma away from the bond areas in the two tests?

 

[stef_design]

If the material you used is linear elastic, the only thing that comes to mind is that the bond conditions cause tensions due to binding reactions resulting from dilation. For example: considering z as a longitudinal axis, if your beam is bound on the flat face with a constraint of earth, the expansion of the face bound in x-y direction will be prevented, which will cause you additional stress.

How are the values of sigma away from the bond areas in the two tests?

Hello hunter,
instead of the beam I give you this example.
Let's say I have a connection between two tubes. inside passes a x pressure.
this pressure generates efforts on the connection.
These efforts vary if temperature varies? in theory not, but the module and material changes with temperature.

 

[Fulvio Romano]

Therefore, the increase in temperature changes the characteristics of the material. change and, y, r, etc.
However, fem analysis normally does not consider these variations, the material is always considered equal.
tensions then are not dependent on e, y, r, etc. at least in the hypothesis of small deformations.

According to me you have a hyperstatic structure, apply a deltat and for thermal expansion arise tensions. If it is as I say you should have that:

- if you remove the forces applied, tensions remain
- if you also remove the overcrowding constraints, making the isostatic structure, tensions disappear.

 

[Fulvio Romano]

Hello hunter,
instead of the beam I give you this example.
Let's say I have a connection between two tubes. inside passes a x pressure.
this pressure generates efforts on the connection.
These efforts vary if temperature varies? in theory not, but the module and material changes with temperature.

and in fact, as I said, the tension does not depend on and.
Imagine the straight bending: sigma = my/i varies to vary by e?

 

[stef_design]

and in fact, as I said, the tension does not depend on and.

Okay, but in the analysis, I have to insert some impunities. one of these are the properties of the material. the elastic module varies with the temperature and then and to 25°c will be different from and to 100°c.
It could be that vonmises also changes for this aspect. No?

 

[cacciatorino]

Okay, but in the analysis, I have to insert some impunities. one of these are the properties of the material. the elastic module varies with the temperature and then and to 25°c will be different from and to 100°c.
It could be that vonmises also changes for this aspect. No?

look that due to expansions internal tensions are also generated, even in the absence of external stresses. you see that in your case these are overflowing compared to those produced by external loads.

the classic example: a "tough" ring, that is where the difference between the inner and outer diameter is of the same order of magnitude of the diameters. do you think by applying a delta-t the solid is free to dilate freely? No, sir, some expansions are prevented by the adjoining material, so tensions are necessarily born.

the typical case is that of residual tensions, in foundry or welding.

 

[Fulvio Romano]

Okay, but in the analysis, I have to insert some impunities. one of these are the properties of the material. the elastic module varies with the temperature and then and to 25°c will be different from and to 100°c.
It could be that vonmises also changes for this aspect. No?

but why can you enter the Young module according to the temperature? What kind of software are you using? comsol?

and anyway no, if I have a beam stuck with a force at the end, the tensions remain the same both that I have steel, whether I have granite, or that I have chewing gum. It's counterintuitive for a very precise reason, if you have any doubt, I'll explain.

look that due to expansions internal tensions are also generated, even in the absence of external stresses. you see that in your case these are overflowing compared to those produced by external loads.

the classic example: a "tough" ring, that is where the difference between the inner and outer diameter is of the same order of magnitude of the diameters. do you think by applying a delta-t the solid is free to dilate freely? No, sir, some expansions are prevented by the adjoining material, so tensions are necessarily born.

the typical case is that of residual tensions, in foundry or welding.

Forgive me hunter, but I disagree.
thermal expansion is a transformation of the type:
deltav = k*v1*deltat
therefore is the equivalent of the "scale" function of autocad. if the ring tozzo is bound isostatically or even labile, there is no tension induced by expansion.

expansions prevented by the adjacent material exist in the real world (and not in the fem), but only because the temperature cannot in any case be homogeneous. the sudden cooling takes place first outside. there is a contraction of the external part resulting in plasticization (the material goes in traction)*, when it cools also the internal one, now the outside is plasticized and therefore goes in compression, while the inside goes in traction. If there were no plastic deformation there would be no residual tension.

if the cooling happened in an infinite time, there would be no residual tension. as well as if the behaviour of the material was perfectly elastic.

 

[cacciatorino]

Forgive me hunter, but I disagree.

I'm not very convinced, I think about it better then I get resentful, thank you!

 

[stef_design]

but why can you enter the Young module according to the temperature? What kind of software are you using? comsol?

try to save who:http://www.cad3d.it/forum1/showthread.php?t=30888discussion #8

ciao

 

[Vmax]

and correct the static disamina of lightning.
of course if you have a hyperstatic or thermal gradient system in the material the modulemodule of young enters the account and different modules will give you different results.

 

[stef_design]

and correct the static disamina of lightning.
Obviously if you have a hyperstatic or thermal gradient system in the material the young module enters the account and different modules will give you different results.

What I'm interested in is knowing how, the young module, enters the vonmises account.
Hi.

 

[eih64]

was thegma=epsilon*e

 

[Fulvio Romano]

What I'm interested in is knowing how, the young module, enters the vonmises account.
Hi.

Forgive me stef_design, but I have the impression that you're cuffing with a fem without knowing exactly what you're doing.

von mises is not a real effort, it does not exist within the material. It is only a mathematical construct that can sum up the tension state in one number that is comparable to the tension alone. This value is valid only for steels. for plastics, rubbers, composites, mortars, etc. does not make any sense, but, so much to complicate even more the thing, for a cold rolled steel bar with considerable reduction of section, already begins to make powder. Guess why...

So, let's do a short excursus of what the fem does, not so much to answer your question, I should write at least three volumes of two hundred pages each, as to understand at least the way.

1. the fem model discretes your piece in "finished elements"
2. constraints are applied. can you know what constraints are? Are they hyperstatic? isostatic? What?
3. loads are applied, the structure is tensed and deformed
4. a temperature difference is applied, the structure expands, deforms, crushes against constraints and tensions even more.
5. the more the structure is soft (and low), the less rough against the constraints and therefore less the tensions resulting from the thermal expansion. vice versa, if the structure is "hard" (and high), the tensions are greater
6. then comes von mises that for each point takes the multi axis tension state, elevates to the square, multiplies something by three, makes a square root and pulls out a number
7. finally comes the human intelligence that, knowing intimately all the facts intercourse from 1. a 6. takes the number out and interprets it, knowing that it is a number pulled out of a model, knowing the limits of this model and trying to intuit the behavior of the real structure that will be well different from that of the multicolor model fem.

Clear?

Can you have a screenshot of what you're talking about?

 

[Vmax]

What I'm interested in is knowing how, the young module, enters the vonmises account.
Hi.

you can take the definition of vonmises, rewrite it according to the deformations, express the dependence on and how it is = alpha*e, and see how this measure depends on alpha.

 

[Vmax]

... I should write at least three volumes of two hundred pages each, ...

for the truth already exist:
the finite element method set
zienkiewicz

If you want to hurt yourself.. .

 

[Fulvio Romano]

for the truth already exist:
the finite element method set
zienkiewicz

If you want to hurt yourself.. .

Yes, it was sarcasm. .
I wouldn't even be able to deal with those arguments. . .

 

[stef_design]

Can you have a screenshot of what you're talking about?

Of course.
I made a simple model.
a bound beam and a force at the end. above this a component that generates heat.
first at 20°c and then at 50°c.
as you can see the extent of efforts and the value of vonmises are different. but it is different also the area where the max stress is present.
Hi.

 

[Fulvio Romano]

first of all in the first case you are cooling the cylinder, in the second case you are heating it. Right? Did you notice?

But it is not clear to me the concept of temperature. If I apply the temperature to one point, I need to know the time. after an infinite time I have everything at 20°c or 50°c, and therefore it is as if I had nothing. alternatively I should have imposed two temperatures, which seems so. I would say about 23°c to the disaster. Is that right?

now, you have the overlap of the effects of strength and temperature.
In the first case the tensions due to the force are greater than those due to the temperature difference, so you find a maximum at the ink.
in the second case the voltages due to the temperature difference are greater than those due to the ink, and then see "become red" the piece near the source.

nothing to do with the variation of 'is', at least, not in a visible way. if imposed and equal to the temperature range you will find that the diagrams will be very similar.

p.s.
Where is the x axis to which the central graphs refer?

p.p.s.
from the way I see the charts, you have a bad mesh!

 

[Fulvio Romano]

to understand how it works you can do the following tests:

1. remove the temperature difference. you will have the only tensions due to strength.
2. remove the force and impose the constant and constant performance compared to the temperature. you will have only forces due to thermal expansion
3. set the expansion coefficient of 0, put the force back and put the dependence on and on the temperature. comparing with 1. you will find the only effect of the change of e.

If you dare (as I understand) to do tests with all effects at once, especially then with that mesh there, you will never understand anything about what is happening.

You think?