Pressure Vessel deigned to BS EN 13445-3 designed incorrectly ? 2

[NovaStark]

Hi all,

I currently have an as-built pressure vessel in which I am trying to determine how the minimum thickness was calculated (no documentation showing this was sent with the equipment as far as I can find). Currently the parameters that I have are P=11 bar, T=50C, corrosion allowance = 1 mm, OD = 1600 mm, Joint Efficiency = 0.85 mm and the shall material is P265 GH.

Now I am more accustomed using the ASME code as 99% of the vessels that I have on the plant are designed to this with the exception of this one. So having to use a new code is kind of making me unsure (note that I am just doing this to see what the minimum design due to internal pressure is as in it is already built and tested).

In PV Elite, I entered the above values with the parameters for the tori spherical heads and I get that the minimum thickness as stated on the drawing (7.6 mm) is insufficient.

Can anyone familiar with this code point out what I am missing ?

 

[DSB123]

Is that the thickness for the heads or the shell?

 

[prex]

You stated that the material is P265GH, but in your calculation you use P235GH?

prex
[URL unfurl="true"]http://www.xcalcs.com[/url] : Online engineering calculations
[URL unfurl="true"]http://www.megamag.it[/url] : Magnetic brakes and launchers for fun rides
[URL unfurl="true"]http://www.levitans.com[/url] : Air bearing pads

 

[XL83NL]

Well, first off, you mentioned tori head, but the PVElite calculation you showed is for a shell under internal pressure.

I did the calcs using Excel and EN 13445. Now I know again why I hate 13445. It just doesnt work. It's not written well. I can go on and on ....

Anyway, based on my calc I have found a design stress of 171 MPa. This is based on para 6.2.1:
- minimum yield strength or 0,2 % proof strength at calculation temperature (256 MPa acc. EN 10028-2:2009 table 4) divided by 1.5
- the minimum tensile strength at 20 °C (410 MPa acc. EN 10028-2:2009 table 3) divided by 2.4.
This is different from the 150 MPa your calculation shows.

I then find 6.05 mm (excl. CA and mill tolerance) for the plate. Using 13445-3 fig 5.1 to determine the minimum plate thickness, I need to know the tolerance. Again a reason why I hate 13445; it refers to 10028 which refers to EN 10029 or EN 10051 or EN 10048. Now Im lost. Anyway, for the shell Ive found 6.05 mm ex. all tolerances (CA & mill). For the tori head Ive found 5.16 mm (ex. CA & mill).

Again, I hate EN 13445. It's not user friendly. With ASME, you buy the BPVC, and youre good to go. With EN, you start with 1 standard and end up buying 68 others more just because they're all linked to each other.

All this took me about an hour or so. Just make sure to never thrust programs such as PVElite: you as the designer are responsible and need to verify the balck box called PVElite (remember, shit in = shit out).

 

[DSB123]

XL83NL,
I agree EN 13445 is not well written. That's because it is an amalgamation of British, Dutch, German & French input so it was never going to "hang" together. And yes you need to buy several "codes/standards" to be able to work with it. This again is a way to "rip-off" the engineering Companies thatr have to use it. Once the UK is out of the EU maybe we can drop the EN standards as they are a real nightmare to use - written by a number of countries that do not agree with each other on what should or should not be in there.

 

[r6155]

@ NovaStark
"Joint Efficiency = 0.85 mm", can you explain it?

Regards
r6155

 

[DSB123]

r6155,
Come on you know that is a typo!!!

 

[NovaStark]

@ NovaStark
"Joint Efficiency = 0.85 mm", can you explain it?

Regards
r6155

Sorry that was a typo on my part, JE= 0.85

The material is P265GH which I used (the head for the vessel is stated at S235JR+N which is why I mixed up the two)...which is why the heads also fail the internal pressure calc. Changing the material to 265 in PV Elite does fix the shells however the heads are still slightly over the stated min thickness.

Is that the thickness for the heads or the shell?

I am trying to calculate for both, but I decided to start with the shell as I thought it would be simpler (based on how the ASME calculations are) - the output from PV Elite for the head appears to be quite complicated in terms of how the calculations are done.


Normally I would have done it by hand which would have required the code which I happen to have the section 13445-3 with the formulas however I don't have the other sections as XL83NL would have pointed out that the code constantly redirects you. Thus I thought PV Elite would have been more comprehensive in terms of the calculation so I could have understood the values that were used.

Torispherical Head From 10 To 20 P265GH at 50 °C

BOTTOM HEAD

Design Stress at Ambient Temperature = 170.839 N./mm²

Required thickness = 6.922 mm.
Required thickness in the crown = 5.861 mm.

This is a Korbbogen Head R = 0.8De, r = 0.154De

The Material is Carbon Steel which affects value of fb:

Buckling Strs at ope. fb = Yield/1.5 = 265.009/1.5 = 176.673 N./mm²
Buckling Strs at amb. fb = Yield/1.5 = 265.009/1.5 = 176.673 N./mm²


Geometry factors per EN-13445 7.5.3.5: - Design Condition:

Pressure including hydro head P : 11.0000 bars
Inside Head Diameter (new) Di : 1586.3999 mm.
Head Crown Radius (new) R : 1280.0000 mm.
Head Knuckle Radius (new) r : 246.0000 mm.
Head Thickness (new) e : 6.8000 mm.
Head Corrosion Allowance Internal ci : 1.0000 mm.
Head Corrosion Allowance External co : 0.0000 mm.
Joint Efficiency Z : 0.8500

Inside Corroded Head Depth [h]:
= L - sqrt( ( L - Di / 2) * ( L + Di / 2 - 2 * r ) )
= 1281.00 -Sqrt((1281.00 -1588.40/2)*(1281.00 +1588.40/2-2*247.00 ))
= 403.659 mm.

Y = min(e/R, 0.04) = Min(5.92196/1281.0000, 0.04) = 0.0046229
Z = Log10(1 / Y) = Log10(1 / 0.005) = 2.3351
X = r / Di = 247.0000 / 1588.4000 = 0.15550
N = 1.006 - 1 / ( 6.2+( 90 * Y )^4) )
= 1.006-1 / (6.2+( 90 *0.0046)^4) = 0.84549


Beta01 = N( -0.1833*Z^3 + 1.0383*Z^2 - 1.2943*Z + 0.837 )
= 0.845 (-.1833*2.335^3 + 2.2124*2.335^2 - 3.2937*2.335 + 1.8873)
= 0.9658

Beta02 = max ( 0.95 * ( 0.56 - 1.94 * Y - 82.5 * Y^2), 0.5 )
= max ( 0.95 * ( 0.56 - 1.94 * 0.005 - 82.5 * 0.005^2), 0.5 )
= 0.5218

Beta = 10( ( 0.2 - X ) * Bets01 + ( X - 0.1 )Beta02
= 10( ( 0.2 - 0.1555 ) * 0.9658 + ( 0.1555 - 0.1 )0.5218
= 0.7194

Thickness Due to Design Internal Pressure: e = max(es, ey, eb) - para 7.5.3.2

Required Crown Thickness due to Internal Pressure, see Figure 7.5-3 [es]:
= P * R/( 2 * f * z - 0.5 * P )
= 11.000 * 1281.0/( 2 * 1708.291 * 0.85 - 0.5 * 11.000 )
= 4.8613 mm.

Computed Thickness [ey]:
= Beta * P(0.75*R+0.2*Di)/f
= 0.7194 * 11.000 (0.75 * 1281.0 + 0.2 * 1588.4 )/170.84
= 5.9220 mm.

Computed Thickness [eb]
= (0.75R+0.2Di)((P/111*fb)*(Di/r)^(0.825))^(1/1.5)
= (0.75*1281.00+0.2*1588.40) *
(11.00/111 *176.67)(1588.40/247.00)^(0.825))^(1/1.5)
= 5.2140 mm.

Computed Head Thickness per EN13445 - 7.5.3:
= max(es,eb,ey) + c + cext
= max(4.8613,5.2140,5.9220) + 1.0000 + 0.0000
= 5.9220 + 1.0000 + 0.0000
= 6.9220 mm.

Which again, I find this kind of weird that they would fabricate a vessel to be below the min thickness. Calculating by hand would have probably given me what the designer intended however I want to try to understand why PV Elite is reaching the conclusion that it is.

 

[XL83NL]

It's a good thing to question the software. Unfortunately there isn't a PVElite subforum here (as there is for COADE's CII), which I find strange as there are so many software subforums, and there is the occasional PVElite question here ...

 

[NovaStark]

It's a good thing to question the software. Unfortunately there isn't a PVElite subforum here (as there is for COADE's CII), which I find strange as there are so many software subforums, and there is the occasional PVElite question here ...

That brings up an interesting conundrum as to how far do I question as I can end up just designing something by hand completely before I enter into PV Elite!

 

[GheorgheCristian]

Hello, i have a few comments on your calculation and vessel for that matter :
1 - you say that the head's material is S235JR+N . first of all this material is used for structural constructions, not for pressure vessels. EN 13445-3 clearly states which materials are to be used for pressure components for pressure vessels and S materials are not amoung them.
2 - be carefull when you input toroshperical heads in PVElite ( it is not enough to chose torospherical head and input the outside diameter ) From what you can see in your own calculation PVElite is considering "R = 0.8De, r = 0.154De" but according to DIN 28011 ( german standard for torospherical vessels ) Crown radius=external diameter and knuckle radius = 0.1 x external diameter, so be sure to input those values in pvelite for the calculation to be acurate .

 

[XL83NL]

Well spotted George, I noticed the same thing wrt the knuckle radius for a tori head to DIN 28011. Wonder what Intergraph's rationale is behind this.

S = Structural steel
P = Pressure equipment steel
E = Engineering steel
etc....

 

[NovaStark]

Hello, i have a few comments on your calculation and vessel for that matter :
1 - you say that the head's material is S235JR+N . first of all this material is used for structural constructions, not for pressure vessels. EN 13445-3 clearly states which materials are to be used for pressure components for pressure vessels and S materials are not amoung them.
2 - be carefull when you input toroshperical heads in PVElite ( it is not enough to chose torospherical head and input the outside diameter ) From what you can see in your own calculation PVElite is considering "R = 0.8De, r = 0.154De" but according to DIN 28011 ( german standard for torospherical vessels ) Crown radius=external diameter and knuckle radius = 0.1 x external diameter, so be sure to input those values in pvelite for the calculation to be acurate .

1. The drawing shows that the shell is S235 material which when I looked it up also sounded fishy for your same reasoning which is also likely why I cannot select it in PV Elite for the shell material.

2. Inputting the crown radius = 1600 mm and knuckle as 160 mm (0.1 * 1600) I get the following:

Internal Pressure Calculation Results :

European Std: BS EN 13445-3:2014

Torispherical Head From 10 To 20 P265GH at 50 °C

BOTTOM HEAD

Design Stress at Ambient Temperature = 170.839 N./mm²

Required thickness = 9.996 mm.
Required thickness in the crown = 7.076 mm.

This is a Kloepper Head R = De, r = 0.1De

The Material is Carbon Steel which affects value of fb:

Buckling Strs at ope. fb = Yield/1.5 = 265.009/1.5 = 176.673 N./mm²
Buckling Strs at amb. fb = Yield/1.5 = 265.009/1.5 = 176.673 N./mm²


Geometry factors per EN-13445 7.5.3.5: - Design Condition:

Pressure including hydro head P : 11.0000 bars
Inside Head Diameter (new) Di : 1586.3999 mm.
Head Crown Radius (new) R : 1600.0000 mm.
Head Knuckle Radius (new) r : 160.0000 mm.
Head Thickness (new) e : 6.8000 mm.
Head Corrosion Allowance Internal ci : 1.0000 mm.
Head Corrosion Allowance External co : 0.0000 mm.
Joint Efficiency Z : 0.8500

Inside Corroded Head Depth [h]:
= L - sqrt( ( L - Di / 2) * ( L + Di / 2 - 2 * r ) )
= 1601.00 -Sqrt((1601.00 -1588.40/2)*(1601.00 +1588.40/2-2*161.00 ))
= 307.687 mm.

Y = min(e/R, 0.04) = Min(8.99557/1601.0000, 0.04) = 0.0056187
Z = Log10(1 / Y) = Log10(1 / 0.006) = 2.2504
X = r / Di = 161.0000 / 1588.4000 = 0.10136
N = 1.006 - 1 / ( 6.2+( 90 * Y )^4) )
= 1.006-1 / (6.2+( 90 *0.0056)^4) = 0.84639


Beta01 = N( -0.1833*Z^3 + 1.0383*Z^2 - 1.2943*Z + 0.837 )
= 0.846 (-.1833*2.250^3 + 2.2124*2.250^2 - 3.2937*2.250 + 1.8873)
= 0.9256

Beta02 = max ( 0.95 * ( 0.56 - 1.94 * Y - 82.5 * Y^2), 0.5 )
= max ( 0.95 * ( 0.56 - 1.94 * 0.006 - 82.5 * 0.006^2), 0.5 )
= 0.5192

Beta = 10( ( 0.2 - X ) * Bets01 + ( X - 0.1 )Beta02
= 10( ( 0.2 - 0.1014 ) * 0.9256 + ( 0.1014 - 0.1 )0.5192
= 0.9200

Thickness Due to Design Internal Pressure: e = max(es, ey, eb) - para 7.5.3.2

Required Crown Thickness due to Internal Pressure, see Figure 7.5-3 [es]:
= P * R/( 2 * f * z - 0.5 * P )
= 11.000 * 1601.0/( 2 * 1708.291 * 0.85 - 0.5 * 11.000 )
= 6.0757 mm.

Computed Thickness [ey]:
= Beta * P(0.75*R+0.2*Di)/f
= 0.9200 * 11.000 (0.75 * 1601.0 + 0.2 * 1588.4 )/170.84
= 8.9956 mm.

Computed Thickness [eb]
= (0.75R+0.2Di)((P/111*fb)*(Di/r)^(0.825))^(1/1.5)
= (0.75*1601.00+0.2*1588.40) *
(11.00/111 *176.67)(1588.40/161.00)^(0.825))^(1/1.5)
= 7.8364 mm.

Warning: Head thickness is less than the buckling thickness:
Warning: Actual Corroded thickness = 5.8000 mm.
Warning: Calculation abandoned!

Now the section in red I can see the designer using that however that won't seem to be satisfying the code requirement as it isn't the largest calculated thickness which is greater than 6.8 mm indicated by the drawing that I have.

 

[r6155]

@ NovaStark

Have you actual thickness of head and cylindrical shell?

Have you dimensional control of heads?

Actual calculations shall be made without adding corrosion allowance, then compare with actual thickness.

Regards
r6155

 

[NovaStark]

@ NovaStark

Have you actual thickness of head and cylindrical shell?

Have you dimensional control of heads?

Actual calculations shall be made without adding corrosion allowance, then compare with actual thickness.

Regards
r6155

I actually do! I initially set about this task because I got some thickness readings on the head and shell. The shell was between 8.5 - 9 mm and the head was an average of 8 mm. So I wanted to see how the code calculation for just internal pressure would be to give the min. thicknesses given on the drawing (which PV Elite says is insufficient - thus my query).

 

[XL83NL]

Now the section in red I can see the designer using that however that won't seem to be satisfying the code requirement as it isn't the largest calculated thickness which is greater than 6.8 mm indicated by the drawing that I have.

Correct, however, Im not sure what the code says about the buckling failure mode. Is buckling a failure mode related to overpressure bursting?

 

[GheorgheCristian]

NovaStark, from what i can tell from your thread , your company bought a old pressure vessel and now you want to see if it holds up to your installation. Am i right ?
If that is the case , r6155 is right about "Actual calculations shall be made without adding corrosion allowance, then compare with actual thickness." That is also how you calculate the life expectancy of the vessel ( also taking into consideration the anual corrosion rate) .
Another piece of advice : in your calculation use P235GH ( it's got the same yield strength as S235JR and almost the same tensile strength ). I know that your required thickness will increase.
My judgment is : if the required thickness (without corrosion allowance ) is more than your measured thickness do everything you can to NOT use that vessel ( double check the PVElite calculation with your own hand calculation as well ). I know that EN has safety coefficients in it's calculation but we are not playing with people's lives here .

 

[NovaStark]

Correct, however, Im not sure what the code says about the buckling failure mode. Is buckling a failure mode related to overpressure bursting?

I'd have to check the 13445 code but I don't have all of the pages for it.

NovaStark, from what i can tell from your thread , your company bought a old pressure vessel and now you want to see if it holds up to your installation. Am i right ?
If that is the case , r6155 is right about "Actual calculations shall be made without adding corrosion allowance, then compare with actual thickness." That is also how you calculate the life expectancy of the vessel ( also taking into consideration the anual corrosion rate) .
Another piece of advice : in your calculation use P235GH ( it's got the same yield strength as S235JR and almost the same tensile strength ). I know that your required thickness will increase.
My judgment is : if the required thickness (without corrosion allowance ) is more than your measured thickness do everything you can to NOT use that vessel ( double check the PVElite calculation with your own hand calculation as well ). I know that EN has safety coefficients in it's calculation but we are not playing with people's lives here .

Typically what we would do is take the actual thickness and subtract the min thickness required for internal pressure and then divide by a corrosion rate. We'd use an alert thickness of the nominal thickness - mill tolerance - corrosion rate to know when to start planning for replacement (for piping). If the calculation is incorrect, then I am not even sure how this pressure vessel made it onto the site and is in service!

Strangely enough, there is a smaller but similar pressure vessel (same service, slightly lower pressure) that is designed in accordance with directives EC 97/23 and AD2000 rules. I can't find these codes to see if they have a min. thickness calc or if the wrong code was written on the first vessel.

 

[medmed89]

Hello, I have done the calculation thickness envelope and elliptical bottom according to codap 2005, but I have a problem with the choice of calculation pressure, is what pressure of service or test pressure

thinks