Cylinder filled with water deflection study

[roldy]

I am having some problems understanding how to apply a distributed load to a cylinder due to it being filled with water. Attached is an image of the cylinder (placed horizontally). The cylinder is symmetric about the y axis. Water is to be filled from the left through the 1" hole. The cylinder is part of a closed-looped system which means that the cylinder will fill all the voids except the ones on the ends of the cylinder. I modeled the water using the combine/subtract feature so that I could find what the volume of the geometry is. To keep the model simple, I left out the holes in the plates. I am using SolidWorks 2012 to do this.

The cylinder is fixed at both ends. The loading is due to the weight of the water and the weight of the cylinder. I applied a stainless steel 304 material to the cylinder. The external loads are gravity and a remote load for the water.

Procedure:
Here is where I'm not sure I did things right. For a remote load I selected the inside surface of the cylinder (the inside wall) in the 4 regions. The load type is direct transfer. My y direction force component is 1496.85 N which is the weight of the water. Note: The dimensions of the cylinder is in English units. I made a spreadsheet to convert the units to SI so that I could figure out the volume and thus the mass which I used to calculate the force.

I meshed the cylinder and ran the simulation.

I would like to know if my procedure for the applying the water load is correct.

 

[Nidum]

(1) Is this a homework/study problem or a real world problem that needs to be solved ?

(2) How fast is tube rotating ? At dead slow rotational speeds this is just a simple beam problem but since your tube is long relative to it's diameter instability effects certainly need to be looked at for higher rotational speeds

(3) On the face of it this is not a problem that requires Finite Element modelling methods . In any case until you have a clearer picture
in your mind of what the loads are and what is likely to happen in terms of deflection I suggest that you do simple manual calculations only .

 

[Nidum]

Why not do a simple preliminary calculation just to get the feel of this problem ?

Treat as a simple static beam with distributed loads based on :

(1) weight per unit length of tube .
(2) weight per unit length of water .

Obtain a value for mid length deflection .

 

[roldy]

(1) Is this a real world problem or a real world problem that needs to be solved ?

(2) How fast is tube rotating ? At dead slow rotational speeds this is just a simple beam problem but since your tube is long relative to it's diameter instability effects certainly need to be looked at for higher rotational speeds

(3) On the face of it this is not a problem that requires Finite Element modelling methods . In any case until you have a clearer picture
in your mind of what the loads are and what is likely to happen in terms of deflection I suggest that you do simple manual calculations only .

1) This is a real world problem that I am curious about.

2) The tube is rotating very slowly. I am considering the tube to be static.

3) I have a spreadsheet where I was doing hand calculations. I wanted to compare the results.

 

[roldy]

Why not do a simple preliminary calculation just to get the feel of this problem ?

Treat as a simple static beam with distributed loads based on :

(1) weight per unit length of tube .
(2) weight per unit length of water .

Obtain a value for mid length deflection .

Here is where it is a little foggy. Correct me if I am wrong but is it as simple as finding the weight of the water or tube and then dividing said weight by the correct length of the tube? So for instance, q for the water would be the weight of the water divided by 232" (the length of the water "tube"). Once I find both q's do I add them together to use in the deflection equation with L being 264" (shaft end-to-end measurement)?

 

[Nidum]

Here is where it is a little foggy. Correct me if I am wrong but is it as simple as finding the weight of the water or tube and then dividing said weight by the correct length of the tube? So for instance, q for the water would be the weight of the water divided by 232" (the length of the water "tube"). Once I find both q's do I add them togethe to use in the deflection equation?

That's it .

 

[roldy]

Ok, thank you. I would still like to know whether or not I am using the right procedure in SolidWorks for doing such a problem.

 

[roldy]

I ran some numbers in my spreadsheet and I am getting very large numbers. For the sake of simplifying, I left out the weight contribution of the end plates and the supporting plates that are inside the cylinder. Attached is the spreadsheet I made for the calculations. I can't find where I went wrong.

 

[SteamKing]

I ran some numbers in my spreadsheet and I am getting very large numbers. For the sake of simplifying, I left out the weight contribution of the end plates and the supporting plates that are inside the cylinder. Attached is the spreadsheet I made for the calculations. I can't find where I went wrong.

You're using the wrong modulus of elasticity in your deflection calculations.

According to the attached data sheet, 304 Stainless Steel has an E value of 193 GPa. Your spreadsheet is off by a factor of 1000.

 

[roldy]

You're using the wrong modulus of elasticity in your deflection calculations.

According to the attached data sheet, 304 Stainless Steel has an E value of 193 GPa. Your spreadsheet is off by a factor of 1000.

Wow, thank you. I can't believe I missed that. Was that the only error in my spreadsheet?

 

[SteamKing]

Wow, thank you. I can't believe I missed that. Was that the only error in my spreadsheet?

It was the only one I found.

 

[lingesh]

from ur calculations,i observed you are considering shaft as hollow,whereas actual model has 1" hole only for small length.So actual wt of this shaft should be more than what you have taken.

 

[roldy]

from ur calculations,i observed you are considering shaft as hollow,whereas actual model has 1" hole only for small length.So actual wt of this shaft should be more than what you have taken.

Thanks for pointing this out. I removed the 1" hole in the model and re-ran the simulation. I get a deflection of 0.002339 in which is wrong. Another individual ran the same model using ANSYS and they got a deflection of 0.118 in. I believe my problem is in SolidWorks as I do not fully understand how to set up a remote mass on multiple surfaces. I'm not sure if when I select all the surfaces for the remote mass, if it divides up the remote mass to the surfaces or it applies the remote mass to each surface.

 

[lingesh]

Deflection at centre is =WL^3/48EI
With your values as its in spreadsheet, i got 16mm..What value you are expecting?

 

[roldy]

Deflection at centre is =WL^3/48EI
With your values as its in spreadsheet, i got 16mm..What value you are expecting?

Deflection = 5qL^4/(384EI) for both ends fixed. I should be getting close to 0.118 in. I've included the updated Excel spreadsheet. I am going to try and do this simulation in Abaqus this weekend to see if I can get 0.118 in. I have no idea what I am doing wrong in SolidWorks or in the spreadsheet for that matter.

 

[lingesh]

Deflection = 5qL^4/(384EI) for both ends fixed. I should be getting close to 0.118 in. I've included the updated Excel spreadsheet. I am going to try and do this simulation in Abaqus this weekend to see if I can get 0.118 in. I have no idea what I am doing wrong in SolidWorks or in the spreadsheet for that matter.

I thought its SSB.

Deflection = 5qL^4/(384EI) for both ends fixed. I should be getting close to 0.118 in. I've included the updated Excel spreadsheet. I am going to try and do this simulation in Abaqus this weekend to see if I can get 0.118 in. I have no idea what I am doing wrong in SolidWorks or in the spreadsheet for that matter.

5wl^3/384EI

THAT gives 128.5mm. Whats new value from simulation? There will be 5 to 10% error between FEA and theoretical value..

 

[lingesh]

can you share the 3d model? i will try to simulate..

 

[roldy]

can you share the 3d model? i will try to simulate..

I am at work right now but will upload the model when I get home.

 

[roldy]

can you share the 3d model? i will try to simulate..

I attached the SolidWorks file.

 

[roldy]

I am at work right now but will upload the model when I get home.

I was wondering if you had a chance to run your simulation? I attached the file in a previous post.

 

[lingesh]

I was wondering if you had a chance to run your simulation? I attached the file in a previous post.

i am really sorry..i got busy on some other project..so i can't do it now...