RTJ Flange Bolting 6

[tgstodd]

I have an application involving a special flange (based on ASME 16.5 1500lb) which has a spacer in the flange pair. There are ring type joints either side of the spacer. The problem I have is that very high bolt torques (and preloads) are required in order to achieve the 3mm offset at each ring as specified in B16.5. The bolts are now over stressed compared to the B31.3 allowable stresses as are the flanges based on an ASME VIII Div1 Appx 2 calculation. With "regular" bolt torque values each ring joint has an offset of approximately 5.5mm.

Can anyone shed any light? This is a common scenario for example wafer check valves with RTJ faces.

Secondary questions are:

A, Do B16.20 RTJ's deform plastically under gasket seating
B, Are the ASME VIII Div1 Appx2 seating stresses reliable?
C, Secondly are the B16.5 gaps typically attainable without using high torques?

Regards,
Tim

 

[XL83NL]

A sketch would be very helpful, to me at least.
If Im not mistaken, RTJ tend to strainharden after torqueing, especially when re-torqued over and over.

 

[TGS4]

You are aware that the "allowable" stress values for bolts are only to be used when designing a non-standard flange? For assembly bolt stresses, I would recommend that you refer to VIII-1, Appendix S and ASME PCC-1.

Unless your bolts will exceed the proportionality limit, there's very little risk of failure...

 

[gasketguru]

The RTJ will deform to some degree (it should be at least 20 Brinell softer than the flange), but the gaps mentioned in the flange standard are mostly an approximate value (try calculating the nominal gap for your particular ring size with everything on exact size, then with the ring and groove on max / min tolerance). The load requirement in practice for two rings in series will be a bit higher than for one on it's own, even though the load should transfer right through. The likely load needed will probably be higher than the ASME VIII calculated value, especially on gas/vapour services as the code calc is flawed and often under-designs. - Obviously it will depend on size, class and material, but I might expect the bolt load to be in excess of 40ksi typically. Also, be careful if you have any joints with stainless flanges & rings and carbon alloy bolts as differential thermal expansions can give problems.

 

[tgstodd]

Thanks for your replies.

The bolt pre-stress required was around 90ksi the flanges were well over-loaded. We have decided to use standard asme torques and target a larger gap in order to protect the non-standard Appx 2 flanges.

 

[LittleInch]

I think you have fundamentally misunderstood how these flanges work. I has a look through 16.5 and couldn't find any reference to 3mm. If you have a section where that is started then please post it so that we can review and comment.

This is a commonly used gap when working out piping distances, but doesn't mean that is what you work to when sealing. It's all about bolt torque/force and ring sealing forces, not distances. Nothing is that precise in the world of flanges and rings for dimensions.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[TGS4]

How did you determine that the flanges were "overloaded"?

What's a "standard ASME torque"?

Have you read and understood Appendix S?

 

[ricklts]

tgstodd,

Unless the center/spacer flange is threaded to hold the studs in place, normal bolt loads should suffice to energize both mating RTJ gaskets, it will just a little longer for the bolt loads to seat both gaskets. But your joint can successfully be joined using the normal bolt loads/toruqes. Dependent on bolting materials, flange materials, operating conditions, somewhere between 40% and 75% of bolt yield strength is usually acceptable for most applications.

To echo what was said earlier, please do read ASME Sec VIII, Appendix S. It relates specifically that design loads are just that, design, and are not sufficient in most cases to successfully seal a bolted joint in service conditions.

I am not familiar with the 3mm offset you are referring to, unless is a customer specific requirement? Pictures???

Rick

 

[tgstodd]

I regret to say that I don't have access to the full ASME VIII Design Code and haven't read Appendix S. I do have ASME PCC-1 and have been reviewing that today. As such I have seen the guideline relating to a proof load of 40-70% of yield stress. The torque required to achieve the designed gap resulted in a pre-load of around 90%.

I realize this isn't the code defined methodology. The equipment purchased is supposedly manufactured to a +/-1.5mm overall length. This length includes 4 RTJ joints so it was critical for us to achieve the design 3mm gap in each joint to meet the overall length for our package assembly. The 3mm gap comes from table 5 of ASME B16.5 (3 and 6 inch 1500lb flange).

The flanges were deemed overloaded based on running an appendix 2 calculation substituting the actual bolt pre-load for the code values. I made this decision in order to check the flange stresses in the non-standard flanges with the high bolt pre-load.

In truth I don't know what the standard "ASME" torques are. Our fitters stated that the required bolt torques were higher than what they typically use for standard ASME flanges.

I need to convince my company to purchase ASME VIII. It is a hard sell as we only design items to B31.3 at the moment.

 

[XL83NL]

Tell your boss you cant do B31.3 work without VIII-1 because B31.3 refers to VIII-1 for (amongst others) external pressure calcs.
Next, read appendix O and K of PCC-1. Especially the 2013 edt which will be out soon contains some useful updates.
Also purchase ASME II part D (as you might need it for B31.3 work), and try to get a hold of some of W. Brown's papers presented by PVP conference proceedings; they present the latest design technology for BFJ's, esp. on this topic. Esp. last PVP conference in Paris contained some useful papers on this subject.

 

[LittleInch]

Table 5 of 16.5 states clearly that the 3mm gap is "approximate". This means it is not a defined amount and to manufacture something with 4 RTJ joints to a +/- 1.5mm lenght is totally wrong and the root cause of why you are trying to literally squezze the life out of a solid piece of metal (i.e. the RTJ ring). Go back and do it properly by allowing a green length to be cut / measured on site after bolting up the flanges to the correct torque.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way