ASME PCC-1 - Gasket and Bolt Loads

[flare9x]

So i follow: ASME PCC-1 Appendix O-3 SIMPLE APPROACH

Condition here is that:

This Appendix outlines two approaches: the simpler
single-assembly bolt stress approach (which is simpler
to use, but may result in damage to joint components);
and a more complex joint component-based approach
that considers the integrity of each component.

So I found after following the simple approach. If one BLINDLY inputs a manufacturers gasket stress, depending on the situation, you can over stress the bolts beyond yield. So found that to be true.

As I am reading I see three things happening:

1. Gasket Stress
2. Bolt Stress
3. Stress to the flange

So one must have adequate gasket squeeze without damaging the components.

This is where the more complex approach comes in because there are checks:

1. Check gasket assembly seating stress is achieved
2. Check gasket operating stress is maintained
3. Check gasket maximum stress is not exceeded
4. Check if flange rotation limit is exceeded

So as i follow the more complex method - during the Check gasket operating stress is maintained step the inputs to the calculation are as follows:

Step 6: Check if the gasket operating stress is
maintained
Sbsel ≥ (Sgmin-O Ag + /4PmaxGI.D.
2)/(gAbnb) (O-8)

sorry might not paste correctly but one can pull it up....

Sgmin-O = minimum gasket operating stress, MPa (psi) - this can be found from the gasket manufacturer
Ag is from the O-1 calculation
Pmax = maximum design pressure, MPa (psi) - Do we use a hydrotest pressure here, or the upper design pressure for that pipe class? In this case its class #150.

Most of my difficultly is arising from: Og or

Og = fraction of gasket load remaining after relaxation

How dos one arrive this value?? How do we calculate the load remaining after relaxation?

K that was one hang up...

Next is Step 8: Check if the flange rotation limit is exceeded.

Here there are tables based on Using Elastic–Plastic FEA and Elastic Closed Form Analysis

My question is: where would FEA and Closed Form Analysis be used? Is it dependent on the type of flange type? Raised face, flat face, slip on, weld neck etc?

Also if your carbon steel type is not listed in those tables what can one do? They have carbon steel grades for SA-105 and SA-182 F304.

So in all of this.. It seems reasonable to think that:

1. Will not over stress gasket past manufactures guidance
2. Will not over stress the bolts past yield and keep in the range of 30 to 60ksi for example - anyone one can see the gasket stress / bolt stress relationships and find this out anyway.

So for 3. how do i make sure I am within the flange stress boundary's and at said gasket/bolt stress combination that will not over stress the flange?

The tables in PCC1 regarding the degree of rotation and maximum bolt stress per flange material and type cater towards specific set of scenarios. What can I do if i have a flat face flange of A 106 versus a weld neck flange of A 105 for example??

Anyway want to figure this part out - sorry for my bombardment but just laying out my current understanding.

One last question - the yield and tensile for A 105 for example is 70,000 and 36,000 - that being said - if we apply compression stress during flange torque... (is it compressive stress acting on the flange material?) how does that interact with the tensile and yield of the carbon steel itself?

Feel have a good grasp on the gasket stress / bolt stress relationship - but now need to add in the different flange classes and see how those affect everything. There is a table-O8 in Pcc1 which shows per class, which is the limiting component in the joint assembly be it bolt stress, gasket stress or flange stress... so flange stress is what I need to get a good grasp on hence my question.


Thanks
Andrew Bannerman

 

[BJI]

I have tried to pick out the questions from the above and addressed them in order below:

Pmax can be maximum operating or design pressure, not the hydrotest pressure.

PCC-1 suggests using 0.7 for gasket relaxation where data is not available. This is determined by relaxation testing for various gasket types. 0.7 represents 30% relaxation, so your bolt stress after relaxation is 0.7*S.

Closed form solutions provide an additional level of conservatism but there is no reason not to use the more accurate E-P results IMO. Slip-on results are available in the closed form solutions only. Flat faced flanges are not applicable to this approach, there is negligible rotation so the flange rotation step can be skipped.

For most carbon steel flanges the elastic modulus and yield will be similar to A105 therefore you can use these tables. F304 is not carbon steel, for any stainless steels use the F304 tables. For higher strength materials, the rotation is basically the same but the localized yielding will be reduced, so I would think it is appropriate to pick either A105 or F304 depending on the elastic modulus of your material.

Step 8 checks the flange bolt stress limit. If you follow the steps in order it ensures the selected bolt stress is not above or below any limits.

Some areas of the flange are in compression and some in tension, and for ductile materials the yield in compression is about equal to the yield in tension. However, the highly stressed regions of the flange are in the hub to flange ring junction and in the hub, and these are tensile bending stresses caused by the 'tensile' load you are applying to the bolts. If you have a look at the appendix 2 flange calculations you are assessing longitudinal, tangential and radial stresses, might be worth trying to understand how flanges are stressed under load first. The PCC-1 bolt limits simply represent the point where the combined localized stresses become increasingly non-linear (GPD point). WRC538 documents the background to PCC-1 Appendix O.

I believe the new version of PCC-1 is due to be released soon, I think there are some changes to Appendix O but not sure of the scope.

HIH

 

[flare9x]

Ok thanks for the explanation.

Considering bolt stress after relaxation - if one follows the simple approach PPC-1, should 30% bolt relaxation be compensated to the initial torque value? That way allowing relaxation to happen on the gasket?

Or go with the initial and go back and re torque it?

2nd option seems really cumbersome.

On further study I realize that the minimum gasket stress per the manufactuer I was looking at was based on a pressure of 300psig.

I looked at the pipe pressure and also the design pressure of the pump that the pipe is attached too. 70psig.

I am apply a minimum based on 300psig. So more than covered. So the minimum operating is below my Sgr value.

Makes sense.

I need to look into FEA and closed analysis now.

In closing all of this really simplifies to Stress = Force / Area hence why the gasket stress / bolt stresses are in a defined relationship.... you can derive both from knowing one or the other.

FYI - I have no formal education but this was interesting - looking to learn more! Perhaps someone can recommend mechanical engineering books I can teach engineering myself without university or college.

Cheers
Andrew