Gasket Seating Area

[chris456s]

Two questions relating bolt load in the gasket seating condition to pretension.

1) Notice in ASME PCC-1 App O-1.3
the gasket area is listed as pi/4 (OD^2-ID^2), which makes sense
In ASME BPVC VIII.2 and VIII.1 gasket area in load equations is (pi)bG, which also makes sense
Both expressions are equal if b = width and G = mean diameter
However, in some cases b can be half width according to BPVC (and 1/8 width for RTJ).
This would make the expressions not equal
Also, G is the reaction diameter which may not be equal to the mean diameter in some cases.
How does this make sense? Is PCC-1 not supposed to agree with BVPC?
It seems like they should since both equations arise when calculating bolt load in the gasket seating condition (BVPC for design and PCC for installation).

2)
a) If we look strictly at PCC-1 App O (using pi/4 (OD^2-ID^2) as area) and calc the bolt prestress to seat a 2" #600 RTJ we get:
bolt root area, Ab = 0.202 sq in
number of bolts, N = 8
gasket area, Ag = 4.47 sq in (R-23 from B16.20)
gasket seating stress, y = Sgt = 26,000 psi (Stainless ring joint BPVC VIII.1 Table 2-5.1)
assembly bolt stress Sbsel = (Sgt)(Ag)/(N*Ab) = 72ksi
This is 90% of A193 B7m yield...seems high??????

b)If we look at the bolt load calc's using BPVC VIII.1 APP 2-5
ring width, w = 0.438in (B16.20)
b0 = w/8 = 0.05475in
effective width, b = b0 for b0<0.25 = 0.057475in
gasket reaction diameter, G = mean diameter for b0<0.25 = 3.25 (B16.20)
minimum initial bolt load required for gasket seating, Wm2 = 3.14bGy = 14,526.8 lbf
bolt stress, Sb = (Wm2)/(N*Ab) = 9 ksi
This is an eight of the PCC-1 calc but they are both supposed to be bolt stress in the gasket seating condition.
How does this make sense???

 

[Vikko]

Any code is different. ASME VIII div. 1 Apx. 2 is based on Taylor Forge method and it was used for long time and has given good results, but it is not good to calculate actual bolt stress and preload or torque for that matter follow PCC. Taylor Forge method shall be used for the joint design only. Why did they used different gasket width for different gasket type in Taylor Forge (ASME VIII div.1 tab 2-5.1/2) ? Maybe, I think they thought that with a metal ring RTJ the contact surface area is different (there are inclined surfaces in contact with the RTJ), see the scheme of a (RTJ) metal ring joint. (Also b is corrected by a formula from the actual b0 and they both represent half width.)

* for geometrical theory the exact area of circular ring is and only is: A=pi/4 (OD^2-ID^2)
* the fgormula pi * (OD+ID)/2 * (OD-ID)/2 (or the same like pi*G*2b) is and only is an approximated formula for that area and that is the formula to be generally used in the Taylor Forge method and many other engineering applications, where this approximation is valid (it is valid if the width of the ring is small in comparison to the diameter).

 

[chris456s]

Vikko,
thanks i have a better understanding now. ASME VIII.1 is for design not actual preload
However,
1) A=pi/4 (OD^2-ID^2) is exactly equal to pi*G*w (if w is width) not just approximate. see attached proof
2) If b represents "half width" why does the ASME VIII.1 call it "effective width" and say the following in Appendix 2-5

"The minimum initial bolt load required for this purpose [gasket seating] Wm2 shall be determined in accordance with eq. (2).

Wm2 = 3.14*b*G*y "​

Notice that y= seating stress. So 3.14*b*G is area. It doesn't say "3.14*2b*Gy", (just one b)
If this is for design and b is the half width, then the initial bolt load Wm2 calc'd in accordance ASME VIII.1 App 2-5 would represent half of the the load actually required to seat the gasket and would therefore be non-conservative. As you said these are time tested methods, so this interpretation must not be correct. But math is math, so what am I missing?

 

[Vikko]

1) I was wrong, you right.
2) check the tab. 2-5.2 and definitions in 2-3:
N and w is the width according to both the definition and the scheme, while b is called the seating width and it is generally half of the real width.

You are right if you consider only the formula (2) in calculating the assembly seating condition load W=pi*G*b*y : in this formula b represents the contact surface in seating condition.

But not if you consider formula (1) for the operating condition or also check the definition in 2-3 of "total joint contact surface compression load" Hp = 2b * pi * G * m * P : according to this formula 2b represent the width, in fact the contact area here is 2b * pi * G.

Also consider that b is corrected (minor) than b0.

So in operating condition you consider N which is the actual width, than b0 is just the half of it usually, finally the correction that reduce a little bit the area and you have b.

In the seating condition the method consider much less contact surface area.

"If this is for design and b is the half width, then the initial bolt load Wm2 calc'd in accordance ASME VIII.1 App 2-5 would represent half of the the load actually required to seat the gasket and would therefore be non-conservative."
I think that the y value is too high sometimes, so they chose to consider less area. These m, y values were set in Taylor Forge and then were replyed in all the calculation codes with no changes up to today, but they are like magic values. They are good to design the flanged joint because we know that this design will work well, but they are not rocket science, and when design pressure rise the taylor forge method does not even work well because makes the design too big.

 

[SnTMan]

Question of why b vs 2b was asked in thread292-479812

No obvious reason seems to be known.

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