PD5500 G2 - Local Loads on Nozzles / Branch Lines

[LittleNose]

I'm trying to put together a sheet for PD 5500:2012 Annex G2, and I've a few questions regarding the interpretation - first of all:

Section G.2.2.3.2 states that 'phi * r / Cx = Z' and then further down states that the value 'phi * r / Cx + Z - Cphi / Cx' shall be substituted to determine 'M / W'.
I am unsure why it is not simply stated as '2*Z - Cphi/Cx' ?

What am I missing?

My misinterpretation appears to be compounded when looking at the charts relative to these values determined in Section G.2.2.3.2, eg:
Using Longitudinal Membrane Force ()with:
64r/t(Cx/r)^2 = 10
2 * Cx / L = 0.05
gives : Nx * t / W = -0.155, which when using Longitudinal Membrane Force (Figure G.2.2-17) results in Z = 0.4 which when put into the equation mentioned above, returns a negative value.

Any input would be appreciated.

 

[EnergyMix]

Isn't the first one a Z'? Or am I misinterpreting what you wrote? I don't have the original reference to check against.

Want to know the do's and don'ts of Eng-Tips? Read FAQ731-376.
English not your native language? Looking for some help in getting your question across to others or understanding their answers? Go to forum1529.

 

[LittleNose]

Hi EnergyMix

Thank you for taking a look at my post.

Yes, the first one is a Z.
The confusion that I have is due to the initial formula that defines 'Z' as equal to 'phi * r / Cx'
So why then use a formula that includes both 'Z' and 'phi * r / Cx' rather than just '2*Z'

I feel like I'm just repeating myself, but I'm not sure how else to state it. It might be easier to understand once you have a look at the original reference.

 

[MrPDes]

G.2.2.3.2(c) is such a weird and confusing statement that I can't get anything out of it. Use "Suggested Working Form G1" on page G/31 instead.

Find Z and then add 4Cphi/Cx to it.
Then go back into the horizontal axis and find Z + 4Cphi/Cx along the axis. Go vertically up and then horizontal to find Mphi2/W.

 

[LittleNose]

Thank you for the reply MrPDes

Like you, I find the statement quite confusing indeed.

While I can understand that it is a similar approach in the "Suggested working form G1", and described in G.2.3.4 b).
This is for applied moments rather than applied radial loads. Thus I'm not sure how valid it would be.

We are going to run a series of FE analyses, and hopefully they'll allow me to calibrate the results, and maybe understand this term.

 

[MrPDes]

I've had a look at: Pressure Vessel Design : Concepts and Principles (Spence, J.; Tooth, A. S.). It has a slightly clearer explanation there. A very good book.

As I understand it, for the equation (φ1r/Cx + Z — Cφ/Cx).

Figure G.2.2-10 is for line loads (not square pads therefore adjustments are required to get to the edge of the pad. "Z — Cφ/Cx" in the equation "appears" to adjust/set the graph across to the edge of the Pad (where the stress is greatest).

The "φ1r/Cx" in the equation is the distance from the edge of the pad to the location that you wish to investigate.

The reason for going in and out of the graph is so they don't have to include another different set of graphs in PD 5500. If they had included these graphes, stresses away from the graph would be more accurate. Without these additional graphs stresses are conservative but within an order of magnitude. So you FEA might show stresses away from the load of half of what you get in Appendix G.

 

[LittleNose]

Yes, I agree that the reason that the factors are there is due to considering a line load as being applied as an area distributed load.

However, I am still uncertain as to why they write the equation as "φ1r/Cx + Z — Cφ/Cx"
- as "φ1r/Cx = Z" would it not make sense to write the equation as "2*Z — Cφ/Cx"
- note that when Z is small, the equation returns a negative value (yet the charts only show positive values).
- as it is not written this way, and the possibility of negative values, I think I'm missing something.

Thank you for your input, it is proving worthwhile for me.
PS I will be without internet for a few days, but I will respond on my return.