Theoretical to CAESAR model spanning study.

Posted by: Mattcaroni

Theoretical to CAESAR model spanning study. - 05/06/19 07:15 PM

Good day.

I am currently new on the software and would appreciate any comments and suggestion. As a new stress engineer I've read some subjects of the following:
-Roark's formula for stress and strain
-Stress Design of piping systems Mw-Kellog
-Pipe Stress engineering by peng
-Piping introduction to pipe stress kannappan


I am aware that based on my theoretical calculations from simply supported at both ends and uniformly loaded beam with both ends fixed does not provide any idea on how the orientation of the pipe should be. I am trying to use my computed length of supports around equipment with this formulas. However, when I try to model it through CAESAR II, the shown displacements under SUS only are way too conservative.


Any insights, ideas would be appreciated as I am aware I lack or missing certain understanding which leads to my confusion. Thank you.



Posted by: anubis512

Re: Theoretical to CAESAR model spanning study. - 05/07/19 05:43 AM

I'm having trouble understanding your issue. You did manual calculations for a pipe segment (presumably including only two supports) and you say the equivalent C2 model is overly conservative (ie, displacements exceed your expected values)?

What kinds of supports? Temperature/pressure? What piping code?
Posted by: Michael_Fletcher

Re: Theoretical to CAESAR model spanning study. - 05/07/19 02:13 PM

I recommend you provide calculations so that we may better comment.

Please understand there are safety factors built into the software and the code that take account for real world effects, such as mill tolerance, uneven corrosion, etc.
Posted by: Mattcaroni

Re: Theoretical to CAESAR model spanning study. - 05/13/19 05:48 PM

I've been trying to reply to this thread with my calculations, however it keeps on saying. Encountered a problem. The reason reported was
Database error only visible to forum administrators
Please click back to return to the previous page. Does anybody know how to resolve this so I can post my calculations? THank you
Posted by: Richard Ay

Re: Theoretical to CAESAR model spanning study. - 05/13/19 10:11 PM

Are you using the full "Reply" option instead of the "Quick Reply"? The full "Reply" has a File Manager option that should let you upload files.
Posted by: Michael_Fletcher

Re: Theoretical to CAESAR model spanning study. - 05/14/19 07:33 AM



Posted by: Mattcaroni

Re: Theoretical to CAESAR model spanning study. - 05/17/19 07:42 PM

My apologies if took quite some time for me to reply. As you can see I have a figure 1 and figure 2 model. I am not sure if which of the two is correct when modeling and looking for the deflection, however as you can see from load case SUS only, I only get less than 3 displacement with a 9 meter span, but in my calculation (theoretically) it is way far from 12.5 mm sag that I used. I think I'm missing some multipliers (safety factors) which were used by CAESAR II. Because I am having difficulties explaining my computations since it is too far from CAESAR II results. Any info will be greatly appreciated especially safety factors that I wasn't able to include. Kindly, include the reference or the title of the book if there are any for me to consider. Thank you.
Posted by: Vannella

Re: Theoretical to CAESAR model spanning study. - 05/20/19 08:29 AM

Mattcaroni,
normally the approach is considering piping "supported somewhere between simple support and fixed support".
These formula are valid for straight pipes. If you have bends, some reduction factors or additional equivalent lenghts have to be considered.
Please check the linear weight (looks a little too low) and do not forget mill tolerance, corrosion, insulation and valve weights if any.
By the way, yes: the theoretical calculated results are "often" overconservative if compared with CAESAR II calculated spans.


To investigate more, some good reference books are:
- Pipe Stress Engineering (Peng)
- Design of piping system (Kellogg)
- Introduction to Pipe stress analysis (Kannappan)

Corrado Vannella
www.vannella.com
Posted by: Rdlm

Re: Theoretical to CAESAR model spanning study. - 11/19/19 02:25 AM

Originally Posted By: Vannella

By the way, yes: the theoretical calculated results are "often" overconservative if compared with CAESAR II calculated spans.


Any reason?
Posted by: Michael_Fletcher

Re: Theoretical to CAESAR model spanning study. - 11/19/19 09:52 AM

If you're given a pencil and paper and are told to design something and make sure nobody gets hurt, your natural inclination would be to add conservatism in to account for human error.

When you make an application as a collaborative effort in order to make it foolproof, the natural inclination will be to improve accuracy with the intent to limit cost.
Posted by: Rdlm

Re: Theoretical to CAESAR model spanning study. - 11/21/19 02:25 AM

Thanks but why Caesar gives me lower results than beam theory?

Exemple:
12inch pipe 10S
12 meters long
A312TP304L
ANC at one end and Vertical restrain + rotation locked at the other end
No pressure, no temperature, no insulation, empty pipe

Mid span deflection from caesar and beam theory formula = 1.75mm
But for stress: CAESAR = 12MPa // Beam theory = 18MPa
Posted by: Sigma

Re: Theoretical to CAESAR model spanning study. - 11/21/19 03:26 AM

Vertical restraint plus rotation locked is effective anchor in above case. So, if you follow the beam theory equation for both ends anchored with UDL, you should get bending stress as 12MPa at both the ends matching with Caesar-II.
Posted by: mariog

Re: Theoretical to CAESAR model spanning study. - 11/21/19 04:14 AM

My hand calculation shows 12.003 MPa, so a conclusion is easy: "Beam theory = 18MPa" is not a result of the classical beam theory...

Maybe you wrong use Mmax = qL^2/8 or you made other mistake.
Posted by: Rdlm

Re: Theoretical to CAESAR model spanning study. - 11/25/19 02:40 AM

My bad, it's qL^2/12 for anchor on each sides.

Thanks
Posted by: mariog

Re: Theoretical to CAESAR model spanning study. - 11/26/19 03:17 AM

As a lesson learnt, more important than a conjectural mistake... If I were you I would ask myself what were the odds of strange fact as a FEA package gives correctly the displacement but not the stress calculated.