Zick analysis, stress at horn of saddle

[jtseng123]

Dear all,
"Pressure vessel design manual" by Dennis Moss, Procedure 3-10, stress at horn of saddle S9 saying ts^2 =ts^2+tw^2. "Pressure vessel handbook" by Megyesy, 14 edition page 89, stress S4 saying ts^2 "may be" thickness squared adding together. (same as Moss, but "may be").

PV Elite is using ts^2 =(ts+tw)^2, which will give less compressive stress. Compress is using the formula from the book that will give higher compressive stress. PV Elite says no over stress while Compress says overstressed in one of our drum. I wonder who is right.

Liner combination of thicknesses by PV Elite seems correct as it is the basic concept of mechanics of material, just like stacking two beams together with both ends simple supports and having loads from top side.

Any thought which equation is correct ?

 

[SnTMan]

jtseng123 Bednar says "The combined thickness of the wear plate and shell can be used in the above formulas if..." and following are conditions on the width and extension of the wear plate.

To me, this would imply the PVE expression ts^2 =(ts+tw)^2. The Moss expression would seem to neglect the term "+2tstw", which could be negligibly small, or not.

No idea which, if any, is correct. I suspect a person would have to go the original paper.

In my class of work I am skeptical of Zick. If the calculation says overstressed, I don't exactly believe it. If it says not, well, I don't exactly believe that either

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[jtseng123]

SnTMan,
2ts*tw is not negligibly small. For both vessel and wear plate = 1", 1^2+1^2 = 2 per design books and Compress, while (1+1)^2= 4 per PV Elite, and then crunching into the Zick stress formula, the former will yield 100% more compressive stress than the latter, that causes Compress to say overstressed while PV Elite saying ok.

Our vendor is using PV Elite and unwilling to add additional stiff ring, threatening for cost impact if added, while our in house using Compress asking for additional stiff ring.

In addition to that, Compress went further to check stress at the edge of wear plate and also indicating overstress that I think it is overkilled, since no paper asking to check stress at the edge of wear plate, and edge of wear plate is not the actual horn of saddle which is a pinch point that has significant compressive stress. The same stress formula shall not be applied to edge of wear plate where it is flexible, not a pinch point. In my opinion stress diminishes quite sharply from the actual horn of saddle to the edge of wear plate.

 

[SnTMan]

jtseng123, I meant "negligibly small" from a math standpoint only. In your case definitely not.

I only have the one reference, Bednar, but looking at the example problem, it does use (ts+tw)^2.

I would generally agree with the statements in your third para.

Regards,

Mike



The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[r6155]

See
1)PD 5500 - 2015
2)Original Zick (attached)

Regards
r6155

 

[SnTMan]

Thanks, r6155. So there's your answer, it seems it is both.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[r6155]

Please see worked example in PD 5500:2015 Annex W, pag 736/1004 to 759/1004 in PDF, and compare with original Zick

In PD 5500:2015 is (tshell + twear plate)^2

In original Zick is (tshell^2 + twear plate^2)

Regards
r6155

 

[jtseng123]

I got the response from PV Elite explaining that they are using equations from Div 2, (4.15.27) and (4.15.28) by adding thicknesses first, and then square.
Compress is using the old Zick paper and way too conservative.