Dear Sir,
All the while, I am using ASME to design my vessel. This is my first time I design vessel using PD5500. OK, for the nozzle design in ASME, it is very straight forward but I found that it is very confusing using PD5500. How to calculate the compensation area for the opening? When to use pad? Is there any maximum limit for the nozzle thickness compare to the shell thickness? For my case, I don't allow to use pad and my vessel thickness only 9.5mm (due sour service). What precaution step I have to look after if I use self-reinforce nozzle in such a thin vessel. Am I violating any part of the code. Is section 3.5.4.9 Alternative "pressure area" design method is mandatory or just an alternative design where we can ignore it?

I have to be very careful with my design as all these items involve long lead delivery time.

Your help is much appreciated.

Thank you.

Regards,
Lee

Hello Lee

Ok, PD 5500 has a very different approach from ASME in the design of nozzles. ASME simply (in effect) replaces the area removed by excess shell thickness nozzle thickness, and if required, the addition of a reinforcement pad. PD 5500 basically determined the stress concentration at the nozzle opening, and then restricts the stress. Perhaps the best way to understand the method is to do a couple of 'hand' calculations, and then the procedure will become clear. However, this is basically the procedure employed by PVElite:

First the programme calculations the ratio of the actual thickness to the required thickness of the cylinder (or head). This ratio is multiplied by a factor 'C'. This ratio is called 'Cers/eps'. A value of Rho has then to be calculated (see PD5500 at3.5.4.3.2). Deriving Rho considers the diameter of the nozzle compared to the diameter of the cylinder (or sphere). Knowing the ratio 'Cers/eps' and Rho, one of the graphs, for example Figure 3.5-10 is then used. You enter the graph at the value of 'Cers/eps' on the vertical axis, move to the right to meet the respective curve for your value of Rho, then traveal downware to obtain the ratio 'erb/ers'. You then multiply this ratio (erb/ers) by the actual thickness of the shell to obtain the required thickness of the nozzle. If the nozzle is thick enough, then the nozzle result is satisfactory.

Now, you asked about pads. Well, in the calculation, the pad is added to the thickness of the shell to obtain a bigger value of 'Cers/eps', thus reducing the calculated thickness of the nozzle.

There is an alternative method available in PD 5500 in section 3.5.4.9 where a similar method to the one you are used to in ASME is considered. This is an alternative method which is availabe also in PVElite in the Tools -> Configuration dialogue screen.

The graph (Figure 3.5-10 etc.) can be worked by assuming a value of the nozzle, and then back calculating the required thickness of the shell. The code allows either method.

If you work from knowing the nozzle to finding the shell thickness, then you might find it interesting to know that the stress concentration factor is actuall 2.25(Cers/eps)/1.1. This is actually the basis on the analysis.

Hi Ray,
Thanks for the quick explanation. I really appreciate it.

Back to the previous question. Is there any violation or limitation between nozzle wall thickness and the shell thickness. I have a case here where my vessel is around 9.5mm thk. and my vessel is not allow to use pad.

Thanks again.

Regards,
Lee

(P/S : Sorry, I am now outstation and I don't have code book to refer. You mention that pad is just to increase the shell thickness. Then how are we determine the pad width as it is not for area replacement. In the other word, is there a set of standard pad width that we need to follow?)

Hello again Lee

I mentioned the value of Rho which must be calculated. In PD5500, the formula for Rho is:

Rho = (d/D)*Sqrt(D / 2*eas)), where:
d = mean diameter of the nozzle
D = mean diametaer of the shell (cyl. or Sphere)
eas = actual corroded shell thickness (with a pad if there is one).

Now, in the case of nozzles that protrude inside the vessel, the value of Rho must must not exceed 2.0, and in the case of flush nozzles, Rho must not exceed 4.0. You will notice that the mean diameter of the nozzle is used in the formula for Rho. This formula gives you the limitation on the diameter of Rho. There are further requirements for nozzles fitted with pads.

As to the question of the diameter of the pad, the code states that the width of the pad (either side of the nozzle) must not be less than the lesser of d/2 or Sqrt(D * eas).

If there is a pad, it must not be thicker than 40 mm or thicker than the shell to which it is attached - whichever is the lesser.

There are some other restrictions also, but if you are able to see a copy of the code, the nozzle reinforcement section should be read for complete clarifiaction.

In PVElite, we have tried to cover all the limitations to safeguard the user, but it is still better if the user familiarises himself with the exact requirements of the code.

Thanks for your reply.

I have a case here.
ID = 1830mm with 9.5mm thk.
Nozzle = 24" LWN without pad and flush.

As per para. 3.5.4.2 (b)(4), D/ens < 200 and no limit is placed on the permissible diameter of a flush nozzle.

During PVElite analyse the calculation, there is a error report that stated that one or more nozzle violation the code. When I check the summary section, at Rho result column stated that the nozzle dia too large.

Am I miss out something in the code?

Regards,
Lee