FEA STRESS INTENSIFICATION FACTOR 1

[howmansito]

I am new with the FEA for piping designs (CAESAR), and I'm working in a branch connection for a High Pressure Relief, that runs over a Header = 36" diameter, 0.5" Thk, API 5L Gr.B, and takes the branch reduction with a weldolet = 8" diameter, 0.25" Thk, API 5L Gr.B.

The ressults in all EXPANSION cases indicates failure, the Tamb=90F and the WorkingT=110F.

For testing procedure I only modeled the branch connection with a 4" nipple running completely free and this even indicates failure, that is because i can't understand wath is going wrong, if the branch connection is totally free, it should moves with the header and the SUSTAINED case indicates 40% of the allowable stress, why the expansion case Fails???Even if I change the Branch reduction diameter to a large one, the CAESAR shows the same results!!!.

When i saw the sif calculations indicates SIF=4.386
But i dont know how much can i reduce that factor.
Is necesary indicate a SIF for the header different to the SIF in the branch?. in that case what value could be conservative???

At the momment the best way to resolve the issue is changing the type of connection from weldolet to a Reducing Tee (36x20) + two concentric reducers (20x12 & 12x8), but is not apropiate from the design view.


Thanks for the comments.

(Note: I apologyze for the writting, the english is not my native language)

Regards.

 

[BigInch]

Hard to say without knowing the configuration and other details of your piping system. Have you tried increasing the wall thickness of the 8" pipe?

What would you be doing, if you knew that you could not fail? Ans. Bonds and derivative brokering.

 

[LSThill]

howmansito (Petroleum):

checking SIF SEE ATTACH using PRG Paulin Research Group Tony Paulin is the Author of CAESAR II, SOFTWARE.



L S THILL

 

[howmansito]

Ok, thank you BigInch, Increasing the wall thickness to the Branch connection no makes any difference, when CAESAR calcules the SIF only takes the Header wall thickness for the calculation, taking on count that the Header is a 36" diameter Pipe, increasing its wall thickness makes expensive the developing, because there are a lot of this type of connection in the project and I think that changing the piping class for the 36" pipe because the small connections should be the last option.

Thank you LSThill (too), I saw your attach and is very helpfull, no much time ago I try with the PRG but i couldn't understand the input/output and in that moment I got very high values with that metode and now i don't have license to learn and try again.

In CAESAR I can calculate Header SIF(i), Header SIF(o), Branch SIF(i) and Branch SIF(o) and I can see if the connection is working or failing, but in the PRG I don't know what means Kib and Kob and how to see if the connection works or fails.

In the case of getting a PRG License do you know if i can take the values from the FESIF and use them in the CAESAR (really I feel more confortable with the CAESAR - I have been working with this for more time).

Again thanks for the Repply.

Best Regards.

 

[BigInch]

Stress will usually be smaller with an increased wall thickness, even if the SIF remains the same.
Do you want it to work, or do you want it cheap.

What would you be doing, if you knew that you could not fail? Ans. Bonds and derivative brokering.

 

[LSThill]

howmansito (Petroleum):

YES: In the case of getting a PRG License do you know if i can take the values from the FESIF and use them in the CAESAR.

Call PRG

Phone: 281-920-9775
Fax: 281-920-9739
Support: 281-310-0852

OR:

send e-mail to
E-mail - Sales Support: sales@paulin.com
Technical Support: support@paulin.com
General Questions: info@paulin.com

 

[LSThill]

Fred Hendrix said: the link below documents the methodology for building a Caesar II intersection with additional SIF data.

http://www.paulin.com/WEB_USING_FESIF.aspx

While the document references our software, the methodology should be independent of the source of the SIF (FEA, experimental, analytical).

FEPIPE can certainly be used for SIF calculations but it might be overkill. NozzlePRO should be sufficient for the original poster's stated requirements.

 

[ADGsfhjkl]

Dear howmansito:

For your issue, can you attach your C2 model?
For SIF, I can get the SIF by NOZZLEPro for you.

the discussion for C2 need model .

good lucky.