ASME B16.34 section 6.4 valve joints question

[MuShoe]

Hello,

We have been engaged in a raging debate about interpretation on the calculation of Ag defined in ASME B16.34 sections 6.4.1 and 6.4.2 as it pertains to ball valves. Specifically, we are discussing section 6.4.2 and the bolts between the adapters and body. As the spec reads "Ag = area bounded by the effective outside periphery of a gasket or O-ring or other seal effective periphery".

The question is simple: Does this definition suggest that we take the full diameter (including the flow are) to the sealing face or do we use this diameter less the flow area?

The verbiage "area bounded" seems very clear that it's the full area, but thinking practically, only the area of diametral difference between the OD of the sealing face and the ID of the bore is seeing pressure loads going into axially stretching the bolts.

If we use the full area, some valve designs need to have a ridiculous number and/or size of bolts, whereas using the later approach, the numbers seem more "realistic".

Any thoughts / comments / best practices would be much appreciated.

Jeff

 

[rneill]

Not familiar with the calculations in B16.34 but I reviewed the B16.34 Interpretations that I have (Volumes 2 and 4) and this subject was not addressed. Perhaps someone with the other volumes might be able to take a look for you?

 

[MuShoe]

Thanks for checking those interpretations. We have volume 4 here and I did not see anything in there either.

What are others doing in practice?

Thanks.

Jeff

 

[reverman]

I'm a bit ignorant on Ball valve design but isn't the pressure acting on the full bounded area when the ball is in the closed position?

 

[MuShoe]

Yes, you are absolutely right. When the ball is closed, the pressure is acting over the ID of the seat seal (conservatively, the OD could be used also). However, in the case of a 3pc trunnion ball valve, the adapters have an OD that is larger than the seat seal OD. We are using the adapter seal OD, since the spec is referring to bolting between body and adapters.

Using the seat seal ID or OD would certainly make for a smaller required bolt area. However, as we understand the spec, that's not the OD that the spec is referring to. I'd love to be proven wrong on this one!

 

[rneill]

Have you considered submitting a Code Interpretation. They take quite awhile to get answered but on festering issues, it is worthwhile to resolve them once and for all.

 

[MuShoe]

Have you considered submitting a Code Interpretation. They take quite awhile to get answered but on festering issues, it is worthwhile to resolve them once and for all.

That's a good point. I am looking into this option right now as well, but I'm hoping for more of a quick answer of what others are doing when designing ball valves.

Thanks for the input!

How are others interpreting this?

Jeff

 

[desertfox]

Hi MuShoe

I have no reference to the specifications that your referring to however I agree with you the bolts are tightened onto the gasket or 'o' ring and the area that reacts the bolt load can only be the od of the sealing face less the id of the bore.
To get the load in the bolts it would be conservative to take the system pressure and multiply the od of the seal face to get the total force and I see nothing wrong with that but when you seal the joint with the gasket to get the compressive stress in the gasket it can't be anything else but the cross sectional area of the gasket.

desertfox

 

[MuShoe]

Hello desertfox,

thanks for the reply.

We did some investigation here and noticed that the ASME B16.5 spec for # of bolts and their sizes very nearly meet or just exceed the calculation in B16.34, if using the full effective seal diameter area and not subtracting out the flow diameter area.

Unfortunately, this supports the argument in the way that I did not want it to go. I was really hoping to find an easy and reliable interpretation that would support being able to subtract out the flow diameter area.

Jeff

 

[desertfox]

Hi MuShoe

Perhaps I misinterpreted your first post, I see nothing wrong with using the outer diameter of the seal face for calculating the bolt loads ie (area * pressure) as stated previously this would be conservative in my opinion.
Howver I thought you meant were taking the seal area to be the same and trying to get bolt loads to compress the seal over the full area which didn't make any sense to me.
In the past I have used both the outside area of a vessel and sometimes the mid line perimeter of the seal or gasket face to get the joint bolt loads.

regards

desertfox

 

[MuShoe]

Hello desertfox,

no problem. I tried to be clear, but sometimes these sort of things are difficult to explain in text.

the spec is pretty clear that it's the area bounded by the gasket. I've attached a picture of what I'm looking for. Perhaps that will help.

Thanks for the thoughts and help so far!

Jeff

 

[desertfox]

Hi MuShoe

Its a radial seal then it would be the od of the seal you require for the bolt calculation, I thought it might of been like a gasket sitting on the face of two flanges.

desertfox

 

[MuShoe]

Hello desertfox,

My picture was not completely correct. The seal OD should be the OD of the adapter at the O-Ring groove almost directly below the bolts in question. However, I don't believe this changes your answer. It certainly does not change my understanding or interpretation.

I was hoping to reach the conclusion that we could subtract the area of the flow diameter, but all our reserach and the input from this forum seems to suggest that we can not.

I am still going to pursue putting in a request for interpretation with ASME. Hopefully that will definitvely answer this once and for all.

Thanks for the help everyone!

Any other thoughts would be most appreciated.

Jeff