Bolt torque & Gasket seating calculations 1

Sec8 Div1 App2 will cover gasket seating, but does not directly address target bolt torque.

For torque, try Sec8 Div1 AppS and ASME PCC-1 Appendix O.

All suppliers give a rosey picture of their equipments's capability I suppose, but they also deal with the consequences in some way or another, whether it is claims or just loss of future sales. Some of the gasket factors published by ASME have been characterized as the "least common denominator" and ASME can't be called to the carpet if 500 gaskets blew out under the code proposed target bolt stress.

- Steve Perry
This post is designed to provide accurate and authoritative information in regard to the subject matter covered. It is offered with the understanding that the author is not engaged in rendering engineering or other professional service. If you need help, get help, and PAY FOR IT.

EN 1591 and EN 13445 App G (same contents) address bolt torque and gasket seating of flanges. Unfortunately the method is mathematically complex and the documents seem to have been written by a bunch of civil servants.

I would agree with the other posts. I would say, that if you are using the ASME App2 calcs, you need to increase your seating/operating loads at least to the hydro-test pressure. Otherwise you will pratically have a hard time sealing the gasket if you perform a hydro-test. I usually go about 10% higher than this for my minimum torque value.

This is not an exact science, but what I usually do is:

1. Take my max oper/seating load for the hydro-test pressure (App 2) calculation
2. Multple 1 by 1.10, this is my minimum required load. I divide this by number of bolts to get my required load per bolt
3. Calculate my bolt tension based on a percentage of bolt material yield stress (40% is my target), or 0.4* Sy*Bolt Root Area. 40% of yield is a ball park number for bolt prestress based on "rule of thumb"
4. I compare step 2 and 3 I make they are not way off, usually around +/-15%.
6. In no case do I allow my bolt stress to be greater than 90% of min yield.


Simple Torque Equation

Tightening Torque = K * D * P

K = 0.15 (for Nickel based Lubricant), this affects torque value significantly

Clamp Load (P) = (Step 2 value / bolts); or (Material Min Yield * 40% * Single Bolt Root Area)

D = Nominal Thread Size


There is also a Joint Integrity Calculation, off flexitallic website (Page 41).

Appendix D in API 6A, is what I usually refer to when recommending bolt torque. It lists a bolt torque equation and some tables.

We typically only use spiral wound, Flexatallic style gaskets, in gas services, in conjunctio with B7 studs, and hydrotest pressures up to 2250 psig, and haven't had any reoccuring issues.

I'm going to pile on to StevenHPerry on this one. His references to ASME Section VIII, Division 1, Appendix S and ASME PCC-1, Appendix O are spot-on.

Torque depends HIGHLY on the lubricant. Make sure that you have good values of coefficient of friction from your thread lubricant supplier.

And torque depends to a lesser degree on the pattern. Actually, what I should say is that the resultant final bolt load (which is really what you are after anyway) is dependent on the assembly pattern. If you follow one of the procedures in ASME PCC-1, you should be fine.

Up until the point where you start yielding the bolts in-service, more bolt load is better(*). Search for papers on this topic from the ASME PVP Conference.

(*) There are exceptions to this rule - mostly where you might excessively deform your flanges resulting in leaks. Most B16.5 flanges don't have this problem. Custom (VIII-1 Appendix 2) flanges may have this issue...

Thanks guys. This has truly been insightful...

In addition the subject we are discussing, I would also like to get your opinion on this appraoch.

How many of you out there would agree to this method of determining bending moments on flanges due to external forces and moments imposed by the piping and further verifying your bolt loads for both the operating and gasket seating conditions?

From ASME VIII Div.1 Appdx 2 we are able to perform flange design calculations based on the internal pressure. However, we are not able to retrieve any information on bending moments due to external forces and moments.

However, from ASME III (NC-3658), we are able to obtain this information. For high strength bolting which the allowable bolt stress > 20000psi

Mfs < = 3125(Sy/36000)C x Ab

Once we obtain this value, we can then account for the external moments using the pressure equivalent method.
Peq = 16Mfs/piG3

Now we are able to further determine P'
P' = P (internal pressure)+ Peq

Upon determining P' we can then substitute P (internal pressure) for P' in the calculation of ASME VIII Div.1 Apddx 2 for bolt loads and stress comparison.

Would you guys consider this approach to be valid at all or just a matter of being stringent with your calculation.

On the other hand, one could say (I'm just thinking out loud here) that if we are in the oil and gas industry the approach of ASME VIII Div.1 is sufficient to meet the requirements of flange design and we need not go further. However, if we are in the nuclear industry, ASME III will govern....

Then again, what other method is there to determine your bending moments due to external forces and moments on the piping?

Ultimately at the end of the day, apart from the gasket seating and bolt torque calculations, I would also want to perform a flange leakage calculation. I know there are various ideas from all sorts of engineers out there but it is still nice to hear answers to a question that I personally put forth.

To those who made an effort to reply, I really appreciate your contribution.

Cheers.

Using the equivalent pressure based on external moment in the Sec VIII method is a practice I think you will find at a number of engineering firms. I believe both ICAS (nee Intergraph) and Bentley now have it implemented as part of their pipe stress softwares.

Equivalent pressure versus the flange rating is certainly more conservative and you are right to think that many if not most oil & gas systems don't need to be held to that standard. However, based on risk, you might want to consider doing so.

On the other side of things, a 1" CL150 water drain line at 100°F may not need to be demonstrated to pass the Sec VIII standard. I forget the details and the source, but someone with some bona fides has suggested that any flange pair where the S_E/S_A ratio is less than a given number is not worth doing any analysis on.

- Steve Perry
This post is designed to provide accurate and authoritative information in regard to the subject matter covered. It is offered with the understanding that the author is not engaged in rendering engineering or other professional service. If you need help, get help, and PAY FOR IT.

PiperTan - you may want to check out this paper:
The gist of the paper is that, using the equivalent pressure method for flanges with spiral wound gaskets, you can go to a P' of 2x rating pressure for Class 300 and below (provided that you assembled them with 65ksi assembly bolt stress). At higher classes, the limiting P' would be somewhat less than 2x.

From my perspective, you would be fine with using 1.5x rating pressure for P'.

The biggest thing to remember in evaluating external bending moments is that you need more assembly bolt loads for the higher bending moments. Otherwise, you risk losing sufficient gasket contact stress on the tensile side of the bending. If you only went to 50 ksi assembly bolt stress (as demonstrated in PCC-1), then my opinion is that 1.5x rating pressure should be used on Class 150 and Class 300 B16.5 flanges, and only 1.25x or less on the higher classes.

Thanks guys...

Cheers