We had an A.I. opinion that the thickness of the ellipsoidal heads should satisfy the requirement of Code paragraph UG-32(b) for the unstayed heads. These heads are welded to the cylindrical shells by a groove weld. We checked this point through two softwares (Compress and PVellite) and this check is not performed by both of them. Would any body tell us who is wrong in this case and why?

Regards,

A. Zaki

Both are correct.

If you check the formula for hemi heads, the only difference is L (inside speherical radius) as compare to formula for ellipsoidal head which uses inside diameter (D).

Since value of L (radius) will be always less than diameter (D), thickness for spherical head will also be less and therefore you do not check it almost in every case.

If required by your AI, you can sumit a manual calculation for the same, no harm in it.

Regards

Sanjeev

Sanjeev, are you sure about it?
1. The formula for torispherical heads is slightly different
2. You forgot the weld efficiency

I haven't come across it, but I think I can imagine a design when the thickness would be driven by UG-32(b). Anyway, if someone doesn't get a simple calculation from a software and it makes a problem... how about going to a Tesco and buying a calculator?

_________________
Konrad Anikiel

Konrad,

Agree with you that in some cases thickness may be governed by UG 32, any way my reply was with the following consideration;

1. Formulas given in UG 32 are applicable
2. 2:1 ellipsoidal head is considered
3. shell to head joint is butt welded (even considering radiography as NIL, JE = 0.6, UG 32b will not govern)

Regards

Sanjeev

Sanjeev, I absolutely agree with you, there's the Code and there's the reason
Anyway, if there's a compulsory requirement in the Code, the AI has his right to ask for demonstration of checking it. If it hasn't been checked, it doesn't necessarily render the design erroenous, but it renders the design engineer not fully credible. Very thorough checking should follow.

_________________
Konrad Anikiel

@ahmed,

The AI has the duty to check if the Manufacturer has a design calculation available which fulfills all applicable Code calculations. If the formula is applicable, the Designer shall use it.When it is not considered, the AI can not sign the MDR.
The AI does not need to evaluate if the vessel design is safe, or perfect, or beautiful, only Code compliance is his job.

In many cases the Designers prefer to rely on computer software. Note that the FOREWORD of the Code mandates the User of a program to verify that it meets Code. Evidence of such verification belongs to the documents to be presented during ASME Joint Reviews, and requested by the AI for his monitoring of the QC program.

Whatever the AI may ask, he should be able to point at a Code paragraph.

DK

My thanks to Sanjeev, Konrad & Dr. Kolebel for their valuable contributions to my post.

For Dr. Kolebel,

I have this query that raised from many situations similar to the one described in my post and dealing with many ASME inspectors in such situations.

As an example for my post, Obviously, by examine the requirement for UG-32(b), Every one agrees that in all cases it will not lead to a thickness greater than that required by head equations [e.g. UG-32(d)] that applied by computer software.

So, When the A.I.S. get this confidence, They followed different manners for situations like this. Some of them consider their confidence is sufficient and there is no need to reflect this in the design reports. On the contrast others, insist on following an opposite way by adding such check to the design report for every design. So, "Which one of these two parties considered fulfilling the Code requirements?"

Best Regards,

A. Zaki