Hi all,
I am supposed to put a Budgetory proposal for the stress
analysis (static & dyanamic) of a 36" flare header pre meter of the length.

I need advise from you about which type of dunamic analysis should we consider? Slug/Modal which will be the pactrical approach?

Option-1
If slug(dynamically) it requires much more manhour than the modal
option-2
Modal and static slug with DLF.

what option is pactrical?
for both option what will be the minimum first frequency(per work practice) to be acheived?
Pls help...

Thank in advance...
S Koley

Subhankar,

Slug analysis ( dynamics) require some data like slug length etc, which are very difficult to get and in absence of authentic data performing a dynamic analysis is of no use.You can assume some data and check DLF spectrum ( Force Spectrum method) and see if it is less than the conservative value ( in static equivalent method,we typically take DLF ranging between 1.5 and 2.0 for suddenly applied loading , in fact DLF for suddenly applied loading cannot exceed a max. value of 2.0)that you have taken for static equivalent analysis.It hardly takes any time to do a spectrum or a time history analysis in Dynamics provided you get authentic data, so that way manhour is no issue.
Many engineering consulatancies use a minimum natural frequnecy requirement of 4 or 8 Hz, although these numbers are mostly based on experience and intended conservatism and not by rigorous mathematical analysis.

Besides these theoretical considerations,practically , few things you need to consider :

1) Try to use "hold down" type guides( not the one we typically use for lines to reciprocating equipments, but a simpler version)and not the conventional one, this helps to minimize the flow induced loading effect.

2) Flare lines are sometimes subjected to "thermal bowing". Check with Process department if such a possibility exists.

3)Generally for Cold Flare lines used in Cryogenic service, analyze for a situation when the line is cold and the Knock Out Drum is at ambient as these K.O.Drums typically will have vaporizing arrangements to boil off any liquid that pours in, thereby maintaining it at ambinet temperature.

Regards

Thanks Anindyada,
But actually the line is existing and client wants to replace
this line. As main shutdown time of the plant is only 7 days,
then they want to lay a new flare header besides the existing
and we have to evaluate the load adequatabilty of the support
(Tressels).
As the time period is short( date for shut down already fixed)
they want only the worst case analysis. so we have to go more
more conservative...which also leads to less less consumption
of manhour.
So my question is..If we carry out modal analysis and acheive
min. 1st feq. above 4hz
and use Slug force(by Density X Area of C/S X Fluid Vel. X
elbow-elbow dist) with DLF(=2) by statically at the elbow for
getting support load....how much that will be pactrical???
Again if you have any authentic backup docs. behind to take
this 4hz value..pls give the details how to get that?
Rgds,
S Koley

Subhankar,

First, I have never seen any theoretical basis for 4Hz, in fact in my opinion, such a value can never be arrived at.Also what happens if one of your exciting frequency is 4Hz or near? I believe that this value is based on a "typically well supported system", the definition of which varies from person to person.

I would prefer to do a static equivalent analysis, by finding out from Process epartment, the density of liquid, density of vapor, homogeneous density ( to get a better view of what is happening, see the wt. percetage of vapor and liquid also, this is typically available in LDT )and compute the slug force. In absence of those data,a very conservative estimate is that the whole cross section or 50% of that ( even that is very conservative)being filled with the liquid bolus.As far as velocity is concerned, I would prefer to take a "homogeneous velocity" cosnidering no slip, although technically it may not be always that way.

Extra long shoes and specially designed axial stops,may be needed if the system has exhibited subatantial slug activities in the past.

Regards

Hi Koley,

When I was going through the "Dynamics of Structures by Raymond W. Clough,Joseph Penzien" ,there was a hint like " The most frequencies of structures one would wish to evaluate in the elastic analysis would be less than 8 Hz". May be the consultancies picked the number "8" from such theories? Would like hear more about this from Anindya.

Regards,

May be, but even Clough and Penzien also has not justified the basis for the number 8.I believe that for most structures a lowest natural frequency of 8Hz indicates a well supported system, i.e. long stretches are "stopped" and change in directions are"guided". Achieving 8Hz itself is very difficult and beyond that either the designing is practically impossible or the cost goes high up without much guarantee on the intended safety.But as I mentioned before, what happens if the exciting frequency is close to 8Hz? Hence for most systems a lowest natural frequency of 8hZ indicates a well supported system but again it cannot be a mathematical conclusion, but a conclusion based on experience on typical systems and their successful design history.Also analyst should check if there is possibility of any exciting frequency close to that value.



Regards