how to apply nozzle load on flange welded on smaller end of eccentric reducer 1

[inammanj123]

I am working on a div-2 horizontal vessel. there is flange, then eccentric cone, then shell. flange is directly welded to smaller end of cone. Piping has given nozzle load, but i am thinking a how the loads can be applied. Following are scenarios:

1- Since flange is welded to cone, we can not use WRC107, OR WRC297. only option is to use FEA. right?
2-Apart from modelling flange and cone assembly on some fea software, i am thinking of applying force and moment on the flange also. I can do this in PVELITE. Its a Conservative approach and flange thickness is also increasing by doing so or first I should confirm this from piping that where should the loads be applied?

 

[TGS4]

Not sure I can help you with the nozzle to shell junction stresses. You may be stuck with Part 5.

Regarding the flange, check out Code Case 2901.

 

[jtseng123]

Rigidity is critical for flange design, so even flange stress is within allowable by FEA, it does not guarantee it will not leak. I believe code formula shall be followed.
My recommendation:
Convert piping moment and force to equivalent pressure, and then added to MAWP of the vessel. If the combined pressure does not exceed the flange rating, your flange will be good. If the combined pressure exceed flange rating, you have two options: (a)increase flange rating (b) Custom design the flange per code analysis.
Regardless increasing flange rating or custom design, the mating flange must match the same. If custom design, watch out bolt diameter and length.

Pay attention that : if your flange is B16.5/B16.47, and the combined pressure does not exceed the flange rating you select, do not use code flange analysis to verify it. Standard flange is exempt from code verification which is very conservative that many standard flanges will fail.

 

[inammanj123]

instead of converting to pressure why not just apply force and moment on flange face? PVELITE has this option.

 

[TGS4]

jtseng123 - your method is overly conservative and, frankly, obsolete. Use the method in Code Case 2901.

inammanj123 - what do you think that your software is doing in the background?

 

[SnTMan]

inammanj120, is this an axial nozzle? If so, just resolve the system of forces and moments into a resultant force and moment at the location of interest, then analyze the cone per strength of materials approach. Analyze the flange(s) per Code Case 2901 or other equivalent pressure method.

Regards,

Mike

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

 

[inammanj123]

Actually it is an desander. 8" inlet FLANGE ,which is welded to ECCENTRIC cone, then a shell(910 OD), and same cone and out let flange is on the other side. My basic issue is piping has given me below loads, now how can i ensure that my equipment is O.K with this these loads. Obviously i cannot use WRC. One way is to apply loads on the flange. Now my question is this the right thing to do, because, i applied the axial force and moment on flange and required thickness increase than that of a standard flange. See the attached calculation report.
Now in the attached calculation i dont see CODECASE 2901 calculation. The report is from PVELITE 2018

FX (N) FY(N) FZ(N) MX(N-M) MY(N-M) MZ(N-M)
80450 53800 20150 15540 19780 20000

 

[inammanj123]

sorry i forgot to attach the flange calculation... it is attahced here in this msg

 

[SnTMan]

inammanj123, I gotta tell you two things, 1) I don't know what a desander is and don't much care, 2) I am not familiar with the Div 2 flange design process.

Having said that, from looking at your output I can see that you have applied some external loadings to the flange, the calculations have accounted for them in some manner, and your flange is overstressed in the radial direction, for the operating case only.

In my opinion, you are unlikely to find another closed-form analysis that will give you acceptable results. You have to deal with what you have. Get the external loadings reduced to an acceptable level. Design a custom flange that meets the given loadings. Prove the given flange and loadings are acceptable by FEA, if this is an option under Div. 2 and the contract.

Don't forget about he cone itself.

Regards,

Mike


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

 

[TGS4]

Get your head out of your software. Look up Code Case 2901. Apply it, outside of your software, either in a spreadsheet (I'm personally partial to Mathcad), or on an actual piece of paper.

 

[SnTMan]

Agree, give it a look.

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

 

[inammanj123]

yeah thx i will prepare one excel sheet and check the output.
But as i am the the client and will be asking vendor to perform a code case 2901 calculation, he will surely comeback by saying why i should apply loading directly on a standard flange? Now to justify myself is there any thing is ASME 2017, which is making it mandatory to check loading on flange face.

One justification which i can give is that since WRC cannot be applied here, so apply the loads directly on flange face.

Regards,
Inam

 

[jtseng123]

In my opinion, that is misunderstanding between flange design and WRC107. They are completely two different issues.
Flange can not allow for large deformation to avoid leak, but junction does allow large deformation as long as it is within allowable stress.
WRC107 is to calculate "stress" at the nozzle to vessel "JUNCTION", nothing to do with face of flange.

You can model a B16.5 standard flange with external force and bending moment in PV Elite. Most likely it is going to fail, especially for 150#. Standard flange has been in the world more than 5o years, most likely before code created equations to custom design flange. So you do not use code equations to justify a standard flange, even you have an external loading.

You will need to use code case 2901 to account for external force to check against the flange rating. If not passed, use one Class higher. There is no need to special design a 8" flange.

TGS4 saying my equivalent pressure (which is also part of code case 2901) approach is too conservative. Actually, I have read through code case 2901 earlier, and have decided for our group forget about it. Still go by the easy, simple approach of equivalent pressure, which will not have significant cost impact based on our decades of experience, and guarantee to sleep well.

 

[inammanj123]

Flange Input Data Values Description: FLANGE :

N1

Description of Flange Geometry (Type) Integral Weld Neck
Design Pressure P 9.73 MPa
Design Temperature 88 °C
Internal Corrosion Allowance ci 3.2000 mm
External Corrosion Allowance ce 0.0000 mm
Use Corrosion Allowance in Thickness Calcs. No

Flange Inside Diameter B 147.000 mm
Flange Outside Diameter A 470.000 mm
Flange Thickness t 63.5000 mm
Thickness of Hub at Small End go 36.0000 mm
Thickness of Hub at Large End g1 75.5000 mm
Length of Hub h 98.5000 mm

Flange Material SA-350 LF2
Flange Material UNS number K03011
Flange Allowable Stress At Temperature Sfo 137.90 MPa
Flange Allowable Stress At Ambient Sfa 137.90 MPa

Bolt Material SA-193 B7
Bolt Allowable Stress At Temperature Sb 172.38 MPa
Bolt Allowable Stress At Ambient Sa 172.38 MPa

Diameter of Bolt Circle C 393.700 mm
Nominal Bolt Diameter a 34.9250 mm
Type of Threads UNC Thread Series
Number of Bolts 12

Flange Face Outside Diameter Fod 281.790 mm
Flange Face Inside Diameter Fid 257.970 mm
Flange Facing Sketch 9, Code Sketch 6

Gasket Outside Diameter Go 281.010 mm
Gasket Inside Diameter Gi 258.750 mm
Gasket Factor m 5.5000
Gasket Design Seating Stress y 124.11 MPa

Column for Gasket Seating 1, Code Column I
Gasket Thickness tg 16.0000 mm
Flange Face Nubbin Width w 11.1300 mm

Axial Force on Flange 80450.00 N
Bending Moment on Flange 28129000.00 N-mm

Flange Class 900
Flange Grade GR 1.1


ASME Code, Section VIII Division 1, 2017

Hub Small End Required Thickness due to Internal Pressure:
= (P*(D/2+Ca))/(S*E-0.6*P) per UG-27 (c)(1)
= (9.73*(147.0/2+3.2))/(137.9*1.0-0.6*9.73)+Ca
= 8.8511 mm

Hub Small End Hub MAWP:
= (S*E*t)/(R+0.6*t) per UG-27 (c)(1)
= (137.9 * 1.0 * 32.8)/(76.7 + 0.6 * 32.8 )
= 46.930 MPa

Corroded Flange ID, Bcor = B+2*Fcor 153.400 mm
Corroded Large Hub, g1Cor = g1-ci 72.300 mm
Corroded Small Hub, g0Cor = go-ci 32.800 mm
Code R Dimension, R = ((C-Bcor)/2)-g1cor 47.850 mm

Gasket Contact Width, N = (Go - Gi) / 2 11.130 mm
Basic Gasket Width, bo = w / 8 1.391 mm
Effective Gasket Width, b = bo 1.391 mm
Gasket Reaction Diameter, G = (Go + Gi) / 2 269.880 mm

Original Design Pressure for Flange 9.730 MPa
Axial Force Equiv. Pressure (4*max(F,0)/(pi*G^2)) 1.406 MPa
Bending Moment Equiv. Pressure (16*abs(M)/(pi*G^3)) 7.286 MPa
---------
Final Equivalent Design Pressure 18.422 MPa

Note:
Per ASME NC-3658.1(b), the equivalent pressure is used only
to compute the Hydrostatic End Load H. The original design
pressure is used in the remaining calculations.

Basic Flange and Bolt Loads:

Hydrostatic End Load due to Pressure [H]:
= 0.785 * G² * Peq
= 0.785 * 269.88² * 18.422
= 1053746.125 N
Contact Load on Gasket Surfaces [Hp]:
= 2 * b * Pi * G * m * P
= 2 * 1.3913 * 3.1416 * 269.88 * 5.5 * 9.73
= 126239.234 N
Hydrostatic End Load at Flange ID [Hd]:
= Pi * Bcor² * P / 4
= 3.1416 * 153.4² *9.73/4
= 179811.109 N
Pressure Force on Flange Face [Ht]:
= H - Hd
= 1053746 - 179811
= 873935.062 N
Operating Bolt Load [Wm1]:
= max( H + Hp + H'p, 0 )
= max( 1053746 + 126239 + 0, 0 )
= 1179985.250 N
Gasket Seating Bolt Load [Wm2]:
= y * b * Pi * G + yPart * bPart * lp
= 124.11*1.3913*3.141*269.88+0.0*0.0*0.0
= 146384.672 N
Required Bolt Area [Am]:
= Maximum of Wm1/Sb, Wm2/Sa
= Maximum of 1179985/172, 146385/172
= 6846.037 mm^2

ASME Maximum Circumferential Spacing between Bolts per App. 2 eq. (3) [Bsmax]:
= 2a + 6t/(m + 0.5)
= 2 * 34.925 + 6 * 63.5/(5.5 + 0.5)
= 133.350 mm

Actual Circumferential Bolt Spacing [Bs]:
= C * sin( pi / n )
= 393.7 * sin( 3.142/12 )
= 101.897 mm

ASME Moment Multiplier for Bolt Spacing per App. 2 eq. (7) [Bsc]:
= max( sqrt( Bs/( 2a + t )), 1 )
= max( sqrt( 101.897/( 2 * 34.925 + 63.5 )), 1 )
= 1.0000

Bolting Information for UNC Thread Series (Non Mandatory):
-------------------------------------------------------------------------
Minimum Actual Maximum
-------------------------------------------------------------------------
Bolt Area, mm^2 6846.037 8159.984
Radial Distance between Hub and Bolts: 47.625 47.850
Radial Distance between Bolts and the Ed 34.925 38.150
Circumferential Spacing between the Bolt 77.800 101.897 133.350
-------------------------------------------------------------------------

Min. Gasket Contact Width (Brownell Young) [Not an ASME Calc] [Nmin]:
= Ab * Sa/( y * Pi * (Go + Gi) )
= 8159.984 * 172.38/(124.11 * 3.14 * (281.01 + 258.75) )
= 6.684 mm

Flange Design Bolt Load, Gasket Seating [W]:
= Sa * Ab
= 172.38 * 8159.9839
= 1406457.75 N
Gasket Load for the Operating Condition [HG]:
= Wm1 - H
= 1179985 - 1053746
= 126239.17 N

Moment Arm Calculations:
Distance to Gasket Load Reaction [hg]:
= (C - G ) / 2
= ( 393.7 - 269.88 )/2
= 61.9100 mm
Distance to Face Pressure Reaction [ht]:
= ( R + g1 + hg ) / 2
= ( 47.85 + 72.3 + 61.91 )/2
= 91.0300 mm
Distance to End Pressure Reaction [hd]:
= R + ( g1 / 2 )
= 47.85 + ( 72.3/2.0 )
= 84.0000 mm

Summary of Moments for Internal Pressure: (N-mm)

Loading Force | Distance | Bolt Corr | Moment |
----------------------------------------------------------------------------
End Pressure, Md 179811. | 84.0000 | 1.0000 | 15110258. |
Face Pressure, Mt 873935. | 91.0300 | 1.0000 | 79586568. |
Gasket Load, Mg 126239. | 61.9100 | 1.0000 | 7818638. |
Gasket Seating, Matm 1406458. | 61.9100 | 1.0000 | 87109128. |

Total Moment for Operation, Mop 102515464. N-mm
Total Moment for Gasket seating, Matm 87109128. N-mm

Note: User choose not to perform Stress Calculations on this Standard Flange.
Pressure rating of the flange will be used to check code compliance.

Estimated Finished Weight of Flange at given Thk. 104.1 kg
Estimated Unfinished Weight of Forging at given Thk 196.5 kg

ANSI Flange MDMT including Temperature reduction per UCS-66.1:

MDMT of ANSI B16.5/47 flange per Matl. Specification -46 °C
Flange MDMT with Temp reduction per UCS-66(i)(2) -48 °C
Flange MDMT with Temp reduction per UCS-66(i)(3) -104 °C

Where the Stress Reduction Ratio per UCS-66(i)(2) is :
Design Pressure/Ambient Rating = 9.73/15.32 = 0.635

Note:
Using the min value from (i)(2) and (i)(3) above as the computed nozzle flange MDMT.

PV Elite is a trademark of Intergraph CADWorx & Analysis Solutions, Inc. 2018

 

[inammanj123]

I modeled a standard 8" 900 rating flange and applied load and used the kellogg method. In calculation im getting an equivalent pressure of 18.422 Mpa(highlighted red in above calculation), but in the calculation which follows pvelite is not using this pressure as operating pressure, it is just using it it to calculate hydro static load. shouldn't this 18.422 Mpa BE COMPARED AGAINST THE AGAINST THE FLANGE RATING PRESSURE, WHICH IS 14.235 MPA or its OK the way pvelite is doing ????

NOTE: i modeled the flange as DIV-1 , because pvelite option "ANSI pressure reduction option" does not work in DIV-2. Does this means with div-2 i should
a) perform flange calculation of 4.15 of div-2(in which case thickness is increasing) or
b) separately prepare a code case 2901 calc but do not a perform 4.15 calculation.

thanks,
Inam

 

[SnTMan]

Don't exactly know what PVElite is doing* there, but...

Using an equivalent pressure method for a custom design flange, the equivalent pressure is used as the design pressure, the flange is designed accordingly.

Using an equivalent pressure method for a standard flange, the equivalent pressure at the design temperature is compared to the flange rating. It makes or it don't.

*Actually I do know what PVE is doing, it is taking the input data and running every possible calculation on it. It's just software, it has no real knowledge of the physical situation, whether the calculations make sense or not. That's what engineers are for.

Regards,

Mike

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

 

[TGS4]

Actually, I have read through code case 2901 earlier, and have decided for our group forget about it.

If I may ask - why?

You are setting up an equation such that:

Internal Pressure + Equivalent Pressure <= Rating Pressure

What Code Case 2901 gives you is:

Equivalent Pressure <= (1+F_M)*Rating Pressure - Internal Pressure

or essentially

Internal Pressure + Equivalent Pressure <= (1+F_M)*Rating Pressure

Which, when F_M is greater than 0 gives you more to work with. Going up a pressure rating class is a poor design choice when many other options exist. The is an especially poor design choice considering the personnel that have to run and maintain the plant after its designed.

 

[TGS4]

inammanj123 - your software is "analyzing" the flange as if it were a custom flange in VIII-2. That is inappropriate for a standard flange.

(I know, I know, it's done all of the time. Doesn't make it right, though.)

If you were to actually perform the CC2901 calculations, you would find F_M = 0.5. If you design pressure = 9.73MPa, and your equivalent pressure = 8.692MPa, and your rating pressure = 14.235MPa, then the inequality that you would be evaluating would be:

Design Pressure + Equivalent Pressure <= (1+F_M)*Rating Pressure

Substituting in all of the values
9.73MPa + 8.692MPa <=? (1+0.5)*14.235MPa

18.422MPa < 21.35MPa

Why are you making this hard on yourself?

 

[Geoff13]

When I get piping loads that come close to failing the flange, my first reaction is to suspect something wrong with the loads.
[ol 1]
[li]Most common is did I mess up the loads? Usually this would be a unit conversion error between what the customer gave me and what I entered into my calcs. Or converting from customer global axes to the nozzle local axes.[/li]
[li]Did the customer just give me some "standard" loads? If so I would ask if they would rather pay extra for a stronger nozzle from me, or send me loads from a vessel specific piping analysis.[/li]
[li]If they are already vessel specific loads, are they based on a "good" model of the piping system? The piping analysis often assumes an infinitely rigid vessel connection, and the loads could drop significantly using realistic stiffness coefficients. The problem is you'd need to give them these coefficients, and that would most likely come from FEA of your nozzle configuration. For my company a thicker 8"Ø flange would be cheaper than FEA modeling.[/li]
[/ol]

 

[TGS4]

Geoff13 - fair enough. You need to understand the entire value proposition. But really - that CC2901 calculation that I did above took my longer to type in than to do it on my hand-held calculator. Why do people need to make this so difficult?