static-structural analysis on wb: how to bind this piece?

[ezekiele89]

Hello everyone,
They assigned me the filter attached to the figure, I simply have to calculate the deformations that you have at an internal pressure of 8 bar.
experimentally it has been seen that the feet open to a mò of banana, sliding on the ground.My problem is that I can't reproduce this situation, as I don't know how to bind my feet to the ground. I thought of "displacement", but I have to assign a move. . .
How do I do that? ?

to rest on the ground are the parts highlighted in red.

 

[Onda]

tie a foot to the ground completely and others with the compression only bond.

 

[meccanicamg]

as first approximation if the feet are attached to cemete castings in the ground with prisoners/barre threaded I would say that the feet are stuck (as bond). then the "hard body" system is isstatic and you can proceed to the load of forces/pressions.

If you're just leaning, try to put the vertical reaction vink. if it is not enough and the simulator does not part you have to put virtual springs that stabilize the system (automatically you will be provided k and x virtual springs to stabilize the system after simulation) and you will get the deformations of your tank.

Try a little and then update us.

 

[Arturo79]

if the feet are swallowed, or blocked somehow, fix them blocking all degrees of freedom.
s are only supported (but it seems strange) you can build a virtual base (a simple parallelepiped) with a material with very high rigidity and imposed as a bond, a contact with friction. as friction coefficient you can use a value from 0.15 to 0.3.
Obviously the base of the parallelepiped must be completely blocked.

p.s., but why don't you write how you solved it once you find the solution? ?

 

[ezekiele89]

Excuse me if I only make myself feel right now, but I could do very little simulations because of a move.
I have not yet come to a solution, but I have understood the bond to use for the feet, it is "compression only" (I tested it on a table with tilted feet, designed as a single pieceIt worked.

the problem is that on this filter instead the postprocessor not from the expected results: I believe that the guilt is of the many parts of the design and the constraints that I grant to this last.
to alleger the calculations I am working only on the base, I attach the screen of the connections I used to connect the bottom with the feet. are all welded connections, and the entire filter is structural steel.

did I incorrectly set up some input data or did I completely wrong the procedure?
thanks in advance!! ^

 

[ezekiele89]

Sorry for the miniature image, I am a noob :tongue:
I attach the link that is perhaps best. .

 

[Arturo79]

Excuse me if I only make myself feel right now, but I could do very little simulations because of a move.
I have not yet come to a solution, but I have understood the bond to use for the feet, it is "compression only" (I tested it on a table with tilted feet, designed as a single pieceIt worked.

the problem is that on this filter instead the postprocessor not from the expected results: I believe that the guilt is of the many parts of the design and the constraints that I grant to this last.
to alleger the calculations I am working only on the base, I attach the screen of the connections I used to connect the bottom with the feet. are all welded connections, and the entire filter is structural steel.

did I incorrectly set up some input data or did I completely wrong the procedure?
thanks in advance!! ^

Hi, it's not fair to understand where you possibly commit the mistake. I think it's more useful if you show the contour plot of stress and deformation. In this way I think it is easier to understand how you have bound the component.

 

[Arturo79]

Then there's that mistake... but does the solutor leave or not? You probably make some mistakes. but are you considering rigid bodies? If that's not how I don't understand why you use joints. You could use simple connections by setting the nature of contact regions.
I would have done this way: foot to the ground: only compression bond (or wanting to be perfectionists, more realistically as I told you, with friction);
feet-bottom: completely bound contact (as the components are welded). if you have to calculate welds, the programam provides you with the forces exchanged in the contact area and use for the sizing as defined in the cnr-uni 10011 standard.
I don't understand your perplexity. It's a pretty simple problem. . .

 

[ezekiele89]

Then there's that mistake... but does the solutor leave or not? You probably make some mistakes. but are you considering rigid bodies? If that's not how I don't understand why you use joints. You could use simple connections by setting the nature of contact regions.
I would have done this way: foot to the ground: only compression bond (or wanting to be perfectionists, more realistically as I told you, with friction);
feet-bottom: completely bound contact (as the components are welded). if you have to calculate welds, the programam provides you with the forces exchanged in the contact area and use for the sizing as defined in the cnr-uni 10011 standard.
I don't understand your perplexity. It's a pretty simple problem. . .

thank you for the help, I start to start simulations following your advice and in the day place the results and I let you know how it goes to finish!

Thanks again!

 

[ezekiele89]

Then there's that mistake... but does the solutor leave or not? You probably make some mistakes. but are you considering rigid bodies? If that's not how I don't understand why you use joints. You could use simple connections by setting the nature of contact regions.
I would have done this way: foot to the ground: only compression bond (or wanting to be perfectionists, more realistically as I told you, with friction);
feet-bottom: completely bound contact (as the components are welded). if you have to calculate welds, the programam provides you with the forces exchanged in the contact area and use for the sizing as defined in the cnr-uni 10011 standard.
I don't understand your perplexity. It's a pretty simple problem. . .

So things went wrong.. (ahimè, with wb are still at first arms)
I tried to work only on the base with the compression bonds only to the feet, and between feet and base I made him generate automatically (also because I did not know which to put exactly), this is the result:
(All: 21cgjex.jpg)

for what I have to do, it is acceptable. .

repeating the same procedure for the whole filter comes out a disgusting one:
(all: 2dhe803.jpg)

another thing: why if in addition to gravity, I also insert internal pressure, the entire filter moves 5-6 meters? I am desperate. . .

 

[Arturo79]

So things went wrong.. (ahimè, with wb are still at first arms)
I tried to work only on the base with the compression bonds only to the feet, and between feet and base I made him generate automatically (also because I did not know which to put exactly), this is the result: http://i54.tinypic.com/21cgjex.jpg for what I have to do, it is acceptable. .

repeating the same procedure for the whole filter comes out a disgusting one: http://i54.tinypic.com/2dhe803.jpganother thing: why if in addition to gravity, I also insert internal pressure, the entire filter moves 5-6 meters? I am desperate. . .

hi from what I can interpret, the first file you sent can go well. You should see deformation in real scale and not in that way as it can deceive you.

on the fact that it moves 5-6 meters, there can be many reasons. For example, you have bound with only compression your feet to the ground and therefore the bond is unable to transmit horizontal forces and this should not give problems as the weight force is a vertical load and the internal pressure should be a balanced load. the only reason why your filter (moves the whole filter rigidly, doesn't it?) moves is that the internal pressure is not uniform and therefore the result also has a horizontal component (not balanced for the reasons I wrote).
must be attentive to the internal geometry of the "recipient": when applying pressure are you sure that you select all internal faces of the filter?? That could be the reason. Let me know. Bye.

 

[ezekiele89]

hi from what I can interpret, the first file you sent can go well. You should see deformation in real scale and not in that way as it can deceive you.

on the fact that it moves 5-6 meters, there can be many reasons. For example, you have bound with only compression your feet to the ground and therefore the bond is unable to transmit horizontal forces and this should not give problems as the weight force is a vertical load and the internal pressure should be a balanced load. the only reason why your filter (moves the whole filter rigidly, doesn't it?) moves is that the internal pressure is not uniform and therefore the result also has a horizontal component (not balanced for the reasons I wrote).
must be attentive to the internal geometry of the "recipient": when applying pressure are you sure that you select all internal faces of the filter?? That could be the reason. Let me know. Bye.

I put the deformed scale only to show the type of deformation that is expected in extreme conditions.
Yes, the container moves completely. . Thinking about it, you will certainly be right about uneven pressure: There are 4 "boccaporti" of which one open to allow the unloading. I just need to build the base and create contact with friction. .

but regarding the filter body, how do I connect the individual parts?
if I generate automatic connections it gives me errors.... at the end is this one that shoots me in front, because the problem of the base at the end is almost solved

 

[Arturo79]

I put the deformed scale only to show the type of deformation that is expected in extreme conditions.
Yes, the container moves completely. . Thinking about it, you will certainly be right about uneven pressure: There are 4 "boccaporti" of which one open to allow the unloading. I just need to build the base and create contact with friction. .

but regarding the filter body, how do I connect the individual parts?
if I generate automatic connections it gives me errors.... at the end is this one that shoots me in front, because the problem of the base at the end is almost solved

It is not said that with friction solve the problem, it depends if the friction force, given by the weight precisely for the friction coefficient, is greater than the horizontal component of the resulting pressure. I don't think in reality the filter is simply grounded. Are you sure it's not fixed at least with simple bolts?
If so, it blocks the translations along the three axes and good night.

for the other problem, I should see at least what errors gives you. All the contact bodies in the cad model, by default they are considered as glued by wb, practically like a single body ("internal disaster"). then manually you can change them according to the real model.
I don't know, you should send me something more detailed.

 

[ezekiele89]

It is not said that with friction solve the problem, it depends if the friction force, given by the weight precisely for the friction coefficient, is greater than the horizontal component of the resulting pressure. I don't think in reality the filter is simply grounded. Are you sure it's not fixed at least with simple bolts?
If so, it blocks the translations along the three axes and good night.

for the other problem, I should see at least what errors gives you. All the contact bodies in the cad model, by default they are considered as glued by wb, practically like a single body ("internal disaster"). then manually you can change them according to the real model.
I don't know, you should send me something more detailed.

Basically I'm doing the internship at a company where they understand less than me in fem (and you want)
they sell this product: http://www.irritec.it/template.asp?...&idproduct=91&dsc=filtri_a_sabbia_di_quarzite (as you can see from the image, the feet are simply resting on the ground)
mò tried to create the same filter, a bit larger and with 4 feet, but it opened as described. . .
through me, they would like to know what happens by changing the type of steel.

boh, now I launch the umpteenth simulation and then I'll take you. . .

 

[Arturo79]

Basically I'm doing the internship at a company where they understand less than me in fem (and you want)
they sell this product: http://www.irritec.it/template.asp?...&idproduct=91&dsc=filtri_a_sabbia_di_quarzite (as you can see from the image, the feet are simply resting on the ground)
mò tried to create the same filter, a bit larger and with 4 feet, but it opened as described. . .
through me, they would like to know what happens by changing the type of steel.

boh, now I launch the umpteenth simulation and then I'll take you. . .

Is it open?? In what sense, did you see the broken filter or at least some photos?

 

[Arturo79]

I put the deformed scale only to show the type of deformation that is expected in extreme conditions.
Yes, the container moves completely. . Thinking about it, you will certainly be right about uneven pressure: There are 4 "boccaporti" of which one open to allow the unloading. I just need to build the base and create contact with friction. .

but regarding the filter body, how do I connect the individual parts?
if I generate automatic connections it gives me errors.... at the end is this one that shoots me in front, because the problem of the base at the end is almost solved

then, I think if you want to check the filter, you need to know well what the load conditions were to break (see that it is a real case).
to not have horizontal shifts you must select all internal faces of the filter when applying the operating pressure.
Now I know well the operation but if the discharge happens with a nice pressure difference between inside and outside, the filter can move horizontally or even turn over if it is not well sized. and this also happens in the real case not only in simulation.
If you leave the exhaust phase and consider the filter "isolated" from the outside, you can assign the pressure on all internal faces and the simulation is simple (like the one you are doing). If you want to take into account the exhaust phase, you should set a fluiddynamic analysis as, I don't know if you have studied hydraulic or fluid dynamics, there are effects related to the variation of the amount of fluid motion itself and therefore, in your case, a horizontal thrust. to give you an example, consider a ball to play football. If you give him a hole, in addition to swelling, you see that it starts in the opposite direction to the hole. this is due to the variation of the amount of motion. Now I don't want to put strange things in your head, but what you have to do is know under what conditions the filter broke. only so you can set a correct analysis. Let me know.

 

[ezekiele89]

fluid dynamic analysis would have set it up later, then correct it all (I saw several guides where this is feasible).
I finally talked to the technical office manager explaining all the issues I met, and for the purposes of drawing up the relazioncina (for the unit) I performed a simple analysis with feet bolted. .
I can finally delete the files of this damn filter that gave me so many sleepless nights, while the company will study the original case to people more competent than me.
Thank you all for your help, especially arturo79, which has always been very helpful and very prepared.

soon, maybe for my graduation thesis
ezekiel