Correct Torque for Bolt Stress 3

regular,

This topic was discussed under the Boiler & Pressure vessel forum.

As I recall, the discussion was that if you have the common SA193 B7/SA194 2H bolting combination, you should use "industry standard" bolt torques developed by the Crane company and the Flexitalic Gasket Company some years ago.

The suggested stress level for torquing these materials was 45,000 psi.....the suggested stress level for other materials ( such as A 307 bolting) was lower.

My opinions, recollections only...

Search this website for "bolt torque" and you will find more info

Good Luck....

MJC

Check out ASME PCC-1 "Guidelines for Pressure Boundary Bolted Flange Joint Assembly"

The bolt allowable stress is for sizing the bolts. It has nothing to do with the recommended bolt preload and torque. In Section VIII, Div 1, see the discussion in Appendix S. Also, as arto suggests, purchase ASME PCC-1. A generally recommended bolt preload is 50,000 psi. ASME PCC-1 gives you the torque required to achieve this level of bolt prestress.

The 50,000 psi is number that people use for alloy steel though at time 45,000 psi or 60,000 psi maybe desirable. At time I seen 20,000 psi used. Stainless requires a little closer scrutiny as it yield is so much lower.
A point to remember that it is harder to achieve the desired bolt stresses in the larger diameters. Judgement of how to achieve the proper preload has to be considered in the design of a bolted connection.

It was our policy on large flanges to going many small fasteners with a higher preload stress than fewer big mothers with the same preload. The mechanics loved the other Sydney for this.

The main thing is to keep in mind the code allowables of the fastener materials being used.

Guys,

Your response appreciated. But I think you may have missed my point viz. I noted in quite a few cases that the actual bolt stress (for each bolt of the flange) required for the design gasket seating ( or operational pressure) is much lower than the max. allowable stress used during design eg.

Twenty bolts ( 1 in. x sect. area) at 30,000 psi ( allow. stress) = 600,000 psi.
Total bolt stress required for gasket seating ( or operational pressures) could be say about 400,000 psi.
Hence, since there are twenty bolts the actual bolt stress required to meet 400,00 psi would be 400,000 divided by twenty = 20,000 psi.
Then why are we always specifying torque to give a 30,000 psi for the bolt stress ?

You could use 20,000 + 2,000 psi for this particular flange or you could use a stress level up to the limit allowed by the code for a particular fastener material. It falls under the auspices of the engineer/designer as to the what stress level in the fastener he desires or requires. Though gasket seating stresses prevail usually other factors enter into picture, sometimes accounted for and sometimes not. As I posted before your fastener material dictates the maximum bolt stress allowed and you as the designer can use any number between the allowed and required for the fastener preload for a particular flange/bolted connection. A generally accepted value for fastener preload changes with time based on better studies and analysis, experience, and last but not least the changes in gasket materials. A case in point, is when I started the # 1 gasket material was Asbestos, then the rubbers, end even leather. There was no Teflon, carbon/graphite, mica, etc. Preloads listed in my first handbook, yes all engineers had them, start at 7500 psi bolt stress.

It has been found that for certain fastener materials, say B-7, used in coded flanges that a bolt stress of 45,000 psi min, gives much better overall performance on flanges. As you state some people use 50,000 psi while I have seen 55,000 psi. Granted this is not a universally accepted number, but is the generally accepted number for the B-7 fastener material. Again this figure isn't good for the say B-8 material as it would be above the yield of the material and squeezing the hell out of a sheet Teflon gasket.
It is a good maintenance practice to set a desired bolt stress for each fastener material used and construct a table for general use. Doing this minimizes a lot of the problems with tightening fasteners. It is always a better to have a tight joint at the onset than have to try to tighten a leaking flange or replace a blown gasket.
All our mechanics, in the neighborhood of 200, have a mechanics handbook with information on each fastener on site, how to identify it, where it is used, and if any question consult the plant piping and flange standards in the foreman’s office, if it is locked call the shift supervisor.

I’ve found it better over the years to keep the bolt stress on the high side if possible to eliminate unforseen operating or even mechanical problems of installation.

As posted previously the other “Sidney” had ever book and article he could find on fasteners and bolted connections he could find and he told me and others that hardly a month went by that he didn’t learn something new about bolted connections and their operation.

Remember a sound bolted connection is 50% engineering, 40% perspiration, and 10% inspiration and common sense.

Hi unclesyd,

Your suggestion to use 20,000 + 2000 sounds logical and reasonable.
But my main concern in the example given would be that if I had used the max. allowable bolt stress ( which appears to be the standard when referring to torque versus bolt stress charts), then wouldn't the gasket be over compressed and failed? I will try to illustrate my concern below :

Typically in ASME V111 Div.1 Appendix 2, the total bolt cross sectional area is determined by taking the higher of gasket seating or operational pressure and divided by the allowable bolt stress (which is say 30,000psi). This would give a figure which is say, 2 sq. inches. But in actual, the total bolt cross sectional area may be three times larger. This would mean therefore that bolt stress of 10,000 psi ( or for safety add an additional 2000 psi) would be adequate for the higher of the gasket seating or operational pressure. If subsequently, the field man refers to a torque versus bolt stress chart to fasten the flange and selected a bolt stress of 30,000 psi, then the gasket would have been over compressed.

You are correct in saying that you can over compress a gasket with too high of a fastener preload. This is not desirable or required and is to be avoided by adjusting the fastener torque within it's effective range. A given fastener preload stress needs to be adjusted for this, say for a rubber, sheet teflon gasket material, or a lower strength fastener. The caveat of this is don't use a very high strength fastener for the lower preloads.

This is why a bolted flange connection needs to designed or verified by a person well versed in the art of fasteners, flanges, and gasket materials and not afraid to ask if questions arise.

The design of a bolted connection is an ongoing learning process and can be very simple and quite complex.

What you are saying was at onetime was covered by tables form the flange and an gasket manufacturers. They will still give guidelines and the proper design input parameters for a specific gasket material for you to design a bolted flange, but will shay away from actual design unless formally contracted.

Again, I'll state is all you have to do is preload the fastener to the point where you can assure that given the parameters of gasket manufacturer, code requirements including any additional operational considerations, and what the fastener material is capable of within the boundries of the it’s allowables and the gasket will be seated properly for the service conditions it will see.

This means that you, as the EOR, can set the torque value for a fastener anywhere from what is actually required by design and what is the maximum allowable stress allowed by joursdictional authority that is governing your design.

I want to emphasize that if possible you should make an effort minimize the fasteners, gasket materials and flanges if possible in order to curtail the type situation that you are now in.

Anecdotal:
The problem you are concerned with happened to us with the introduction of sheet teflon and again with the sheet graphite materials. We went from asbestos to teflon and some of the original gasket seating parameters/values for the graphite sheet were too high. We had to scream loud and long to get them corrected to match what we were seeing in the field.

There is some good information on the following website.

There are also a lot good books/publications covering bolted joints just post the fact if you desire to get more information on them.

I don't think anyone really answered regular's original question. I think what he was trying to say was why does Flexitallic etc give recommended bolt stress in the range 40 to 50 ksi for B7 bolts while Appendix II uses Code allowable stress of about half this value? Would this not overstress the flange or gasket?

regular and codeeng-

Yes, it is common to stress bolts well beyond their allowable - but still less than yield. For bolts with a yield at 105 ksi (SA193-B7), it is not unheard of for them to be torqued to 70 ksi (max listed allowable is 25 ksi). I try to keep my recommended torques lower, especially since I typically provide bolt torque guidelines when we're upgrading to CMGC or GMGC (kammprofile) type gaskets which have reduced seating stress and gasket factors.

From a Code interpretation perspective, I agree that it appears odd to routinely call for bolt torques which will cause stresses well beyond those listed as the "allowable" stress. Appendix S provides much insight into this situation. I'd suggest reading the entire Appendix, 3 pages, but here are some selected quotes:

"The maximum allowable stress values for bolting given in Table 3 of Section II, Part D are design values... However a distinction must be kept carefully in mind between the design value and the bolt streess that might actually exist..."

"An initial bolt stress of some magnitude greater than the design value therefore must be provided..."

"In any event, it is evident that an initial bolt stress higher than the design value may and, in some cases, must be developed in the tightening operation and it is the intent of this Division that such a practice is permissible..."

"...the bolt stress can vary over a considerable range above the design stress value."

jt

codeeng- / jte,

Most of the response did not actually address in total my query. However, the discussions and materials that surfaced are informative nontheless. I think I have gleaned from the discussions, a rough answer to my query though I would have preferred if someone could give me a direct answer. Perhaps the eg. below can be clearer to you folks out there who may be able to answer my query.

A
In design of flange, an allowable stress is given for bolts
say 30,000 psi.
B
From the pressures and dimensions of the flange to be designed, the higher of gasket seating stress or operational stress is selected say 400,000 psi.
C
This higher stress 'B' and the bolt allowable stress 'A' are used to compute the TOTAL cross sectional area of bolts required say 13.3 sq. in.
D
The ACTUAL Total cross sectional area of bolts is confirmed to be equal to or higher than that determined in 'C' 2 sq.in.
E
I note, normally, that the actual total cross sectional area used is much higher than that required. say 20 bolts x
1.125 in. dia. = 19.7 sq. in.
F
Thus to provide meet the required stress 400,000 psi 'B' (higher of gasket seating or Operational load), actual stress for each bolt need to be only
400,000 /(20 bolt x 19.7 sq. in) = 1015 psi.

Which therefore, would mean that I do not need to torque each bolt to 30,000 psi let alone exceed the 30,000 psi allowable.

Regular-

Just working with the bolt numbers above, I'd have a few comments: If you're trying to get 13.3 sq.in. you must have (20) 1 1/8" bolts; (20) 1" bolts won't be enough. Remember to use the "tensile stress area" of the bolts, not the "gross area." The gross area is based on the nominal diameter disregarding the threads. The tensile stress area takes off the depth of one thread (the opposite side, by definition of a thread, is always there). For what its worth, the "min root" area is essentially the gross minus two threads, i.e. the area based on the diameter to the root of a thread. If you have a copy of the green steel book, Allowable Stress Design, 9th ed., from the AISC take a look at page 4-147.

Using the tensile stress area, your (20) 1 1/8" bolts have an area of 15.26 sq.in. while (20) 1" bolts would have an area of 12.12 sq.in. You could go up to (24) 1" bolts for an area of 14.54 sq.in. Given a choice, I agree with unclesyd in his previous post to give preference to more small diameter bolts over fewer large diameter bolts. Thus, I'd choose (24) 1" bolts for your connection provided there was reasonable clearance.

What bolt torque you specify needs to be not only based on calculations, but also on experience. It is the experience of most folks in industry that a tightening bolts can stop a flange leak. They will be leery of lower bolt stresses which can come up when using GMGC or CMGC type gaskets. They're used to double jacketed or spiral wound gaskets which would require higher preloads. Arguably, gasket manufacturers know something about bolted connections. Take a look at choose "Product Caltalogs" from the left side, and download the GMG 1-1 catalog listed under "metallic gaskets." Take a look at page 30 of the catalog (32 of the .pdf). Their "preferred" torques are set to provide 60 ksi bolt stress. Keep in mind that these are limited to typical piping flanges. Most of the time I'm calculating bolt torques for larger diameter flanges for exchangers.

What I do is to calculate the required bolt torque based on Wm1 or Wm2. Then I increase it by 150%. Then I compare it to 30% yield. Whichever is higher is what I'll recommend. Since 30% of yield for a 193-B7 stud is 31.5 ksi while the allowable at ambient is 25 ksi, I wind up always "overstressing" the bolts...

jt

jte,

Many thanks for the info. Will read up on the Garlock pdf.

jt -we always used the minor-diameter-area [i.e., per the Taylor Forge Chart] for flange calc's, rather than the tensile stress area per B1.1


& used ~"Pi" bolt spacing to prevent scalloping.
[i.e. #bolts ~ BCD in inches ==> ~3-1/8" between bolts]
round up to the nearest multiple of 4 so's you can straddle cL's

Regular,
I am on the maintenance engineering side of the house and like you, I too was looking for a straight answer on this issue. After reading all the references, I believe from a design standpoint you need to get away from the large bolts.
My personal interest comes downstream in the process. I have a 50 psig MAWP pressure vessel with a 24 inch manway. A standard 24" 150# flange ring with 20 1-1/4" SA-193 B8M bolts was welded to the neck. The vessel specs call for a spiral wound gasket. Wm2 governs and comes to only 15,300 psi per bolt. It is now apparent to me that here too, smaller bolts should have been used. Our bolting spec calls for this size bolt to have 690 ftlbs of torque to achieve 40,000 psi indicated bolt stress. Every time we access this vessel, the technicians are hanging 35 feet over the water with too short a lever and no where to stand wondering why we need so much torque. I am inclined to cut back on the torque but deviating from the spec could open up a lot of extra work that specs are written to avoid. Not a popular decision; make sure the design is maintainable.

Though it is not a good practice to have every flange a separate entity you could reduce the stress on the bolts. We used to seat spiral wounds with 30,000 bolt stress but you had to be careful in assembly.

It can be difficult at times to seat a spiral wound gasket with a 150 Class flange. Our standard was essentially 300 Class flange dimensions aor a 300 class flange for the spiral wound gasket even though the 150 would work. This was one effort to keep from cupping the flanges, this also put us on the good side of A193 B8 Cl 2 fasteners sizes, and this allowed us to take a cut or two to clean up a corroded flange. If we made the flange or had one drilled, as stated above we used many smaller fasteners.

I thought I had some calculations with some numbers and sizes but they are in hiding. I'll check tomorrow and get a few numbers.