API 650 Tank model

[Anonymous]

Hello, I am a student in electromechanical engineering. It's my first time posting in this forum.
This is my first time working on such a project and with Robot structural analysis. I am asking for help to check my model is there any default because the calculation with only the Dead load is off and there is a separate structure somewhere. I am sorry if it seems incompetent on my part to ask such a thing.
Also, for some reason, the columns seem not to be calculated.
I will delete this if it goes beyond the forum rules.

https://www.mediafire.com/file/966gpafig87b9xv/newshellwithsupport.rtd/file
I couldn't include the file directly in the forum.

[drenzoche]

Hi, you have to define supports in the columns and to avoid the separate structure you have to define again the roof structure because it is not connected to the walls and beams.

[Anonymous]

Thank you for your insights. I would love it if you can tell me a method of defining the roof.

[Stephane.kapetanovic · Accepted solution]

hi @Anonymous 

First, let me ask you if you are sure you want to design your tank with this sheet metal format as you have a lot of panels and welds. You chose to offset the vertical welds by 10 inches but you could space them further apart by staggering them at middle of the lower panels. You can model the bottom or an anular part of it rather than opting for the simple embedding in the base. After checking the general stability and checking your thicknesses, you will be able to subsequently model the support devices more precisely.
Typically there is a problem with the assembly of the panels like the example in the screenshot.

You need to resolve the connections between the edges of the panels and change the mesh options for better shape evenness.

For the roofing sheet, you must connect all the points of the upper circular panels one by one and join the tip of the cone by sector.
For the roof stiffeness, it's up to you to find the right principle but look to the following principle. (sizing and number of sector to be define)

Best regards

[Anonymous]

Dear sir,
I defined the roof as you advised, but I got in another error even after rotating the plates to the half ( incoherent mesh on edges ). I used different mesh options as I switched off Regular Mesh but it didn't work. 

What do u think is the best sheet metal format for this model. I used arc and lines to fill in the panels. 
https://www.mediafire.com/file/9077kl2qumhypd4/modeladviced.rtd/file
Thank you for your insights.
Thank you.

[Stephane.kapetanovic]

hi @Anonymous 

Solving the mesh problem simply consists in reducing the size of the elements of the upper sheets (EP7 for example delaunay size 0.5m). However, your design has not changed for the roof and bed. It does not seem to me possible that a simple 5mm metal sheet can carry wind loading over 18m without beams for stiffening. So I think that efforts to resolve the incoherent meshing will be in vain at the next stage (sizing). 
Also, you have modeled a circular crown at the jonction with the rafters which will be used to distribute the efforts of the roof, but this one not being designed you cannot finish its cutting. To do that you must draw the rafter bolted connection to see how this part could be done.

Best regards

[Stephane.kapetanovic · Accepted solution]

hi @Anonymous 

I saw you online so allow me to add the following point.

There is a small difference on the junction points of the arches between roof and walls. To see this, increase the number of decimals of the unit of length and check for example panel 377 with panels 468. This gives the following result:

377

468

that give a difference of

Do not correct these errors right away. Before choose a stiffening mode for the roof with a junction with the compression ring in accordance with the connection details given in API 650.

physically, the sheet metal formats of the roof will certainly not be cone sectors but rather simple rectangular formats placed on the framework and welded on site.

best regards.