For the full academic route -

Do you have access to ASME Sec VIII Div 1 - Appendix 2?

How about ASME B16.5?

Do you know the piping code to which this is to be designed? B31.3 / B31.8 / ...... ?

Shall be designed to ASME B31.8

Thank you very much! I think I have some studying to do.

Bye

Dejan

You can also find relevant information in Appendix E page 118 of this guide (Word Document)

http://engstandards.lanl.gov/esm/mechanical/Ch6_%20D20_AppA_R2.doc

Note, however, that it is for B31.3 but should still be relevant, and without digging into the codes, the only difference may be the bolting stress values (to determine torque)

But the other specifications/requirements should prove useful such as calibration/cleaning/....

Thank you very much...I used LANL before, however the document does not open.

Bye

Dejan

It should open with Word (or at least WordPad)

How about going to Chapter 6 in the attachment of this link:

http://engstandards.lanl.gov/ESM_Chapters.shtml#esm6

and saving the link as a .doc to your computer and then opening with Word?

Guys,

How we can relate the matter to ASME B31.3, Its a pipeline if liquid hydrocarbons then we must refer to ASME B31.4 and if gas then ASme B31.8

Attention in the SKF doc there is an erronated graph which was taken over without critical analysis from a not SKF source.

See ASME PCC-1-2000 "Guidelines for Pressure Boundary Bolted Flange Joint Assembly".

Excellent document.

Thank you very much I l will get my hands on it ASAP.

Bye

Dejan

Agree!!!

Hello Dejand,

Although you've been provided with excellent direction, may I suggest that the most important things that you keep in mind are:

Bolt "torque" does NOT indicate how "tight" a bolt is.

and

A "properly torqued" bolt may, in fact, be too loose or even too tight. Either case could could result in loss of flange integrity.

If you're concerned with ensuring the reliability of the joint (and in covering your gluteus maximus), you really should specify either bolt load in lb-f or bolt stress in psi. Only after having included this in the specification should you specify a torque. It should be made clear that this "torque" value is dependant upon the actual K Factor (friction et al) being identical to that which you had assumed in your calculation. Since this is nearly impossible, it would be highly unlikely that anybody applying the specified torque will achieve the preoad that you had deemed is necessary. The solution out of this dilemma is to also specify the resultant bolt elongation commensurate with the applied preload.Thus, prelolad can be controlled and verified by measuring each bolt's elongation.

Here's a link to a very simple Excel spreadsheet which might help with basic calculations:Bolt Calculator

J,

I appreciate your advise one of my colleague had mentioned the same exact thing. Bolt torque does not mean that the bolt will be tighten properly...

Dejan

Good explanation on the difference between "torque" and "tight". Is there a tool or instrument available to measure bolt elongation, something equivalent to a torque wrench that can be used on the job? If not, and testing can only be done in a lab type setting then what is the equipment/procedure for that? Thanks.

http://www.heviitech.com/Hevii_UT.html

For you the best approach would be a "torque-angle" procedure.

The bolt is first tightened to a low torque with respect to end value and then rotated with an angle which has to be determined by computing and validated empirically so that the bolt is stretched to the right elongation. This approach reduces drastically the dispersion due to friction variations.

If you need more input use the direct message chanel.

Hi!

Thanks. Yes I do need more input since I never heard of your method.

What do you mean by direct Chanel i am new so I may be missing something with regards to communication.

Bye

Dejan

Torque angle?? Granted, this is certainly a better option than blindly torquing the bolts. However,...

it's still fraught with problems:

At what point does one mark the nut in order to begin the rotation? Consider what happens with rotated flanges or gasket variations! Everything that's wrong with the torquing method is, by definition, still wrong with this method.

Here's a consideration: Bolt tensioning is a much more accurate method because the entire friction component is removed from the process: Apply a pre-calculated pressure, stretch the fasteners, lock the load and remove the tensioners. Another benefit of this advanced method is that all of the bolts can be tightened simultaneously thus (in addition to applying accurate loads) the gasket is compressed evenly; (cocked gaskets are a prime cause of leakage).

Regardless of whichever method is used, be it torquing, tensioning, torque'n'turn or even Bubba hanging off of a 10' pipe wrench, peace-of-mind can only come from knowing what the actual residual load is. Surely you want to make sure that the bolts have been tightened properly. Otherwise, why did you bother to calculate the loads? The best way to verify bolt load is by measuring the stretch of each fastener. Any bolts that haven't stretched to the required elongation can be easily "tuned" (apply more torque, tension, turn the nut a bit more or hang a sack of flour around Bubba's neck) until the necessary elongation has been achieved

Even the method you praise is not as accurate as you claim. In fact the rigidity of the parts lead to a non uniform distribution of pre-loads and in big structures special strategies have been developed to obtain a good uniformity . From an other point of view at pre-loading the force loop is NOT through the nut contact surfaces and the bolt stretch is measured under pressure, when the hydraulic pressure is not any more present the force goes through the nut and crushing the surfaces which have not been under load so that after the first cycle after full release the bolt is LESS stretched as during the loading. This is the reason -which you did not mention - that usually 3 cycles are requested in order to guarantee a pre-load as wished. All is a problem of cost and required precision (in fact dispersion to use the right word).

If one uses a torquing approach the torque-angle method is the less uncertain approach, or the yield limit approach can also be considered. Since the torque wrenches are still the most common tools for assembly the method has its right to exist.

Indeed, two or even three pressurizations are necessary to ensure full load. I suppose that if someone can't spend the extra 2 minutes on the flange to perform this, the proven advantages of tensioning over the known inacuracies of torquing aren't worth the effort: One might as well torque and "hope" that the loads are correct. By the way, since tensioners can be used simultaneously on every stud (or at least on 50% of the fasteners), there is a significant speed advantage as well.

In any case, measuring elongation to verify load is the prudent practice. "Torquing" and then expecting a certain bolt load is like being asked what your soup tastes like before you've even put the spoon into your mouth.

Hi Nick,

Bolt elongation is a result of bolt tensioning, do you think bolt also elongates during the torque process??

Well Ash, let me answer your question by asking one of my own...

Q:What happens if you applied torque to a fastener where the nut has been welded onto the bolt?

A:You would apply torque yet the fastener would not elongate.

This is an extreme but, it serves to show that "torque" cannot be used as a metric to define bolt "tightness"; there are simply too many unknown friction variables which conspire against a defined conversion of torque to bolt load.

Certainly, most torque exercises result in some degree of bolt elongation. The problem is that one can never accurately predetermine how much the bolt will elongate at any torque value.

Hi BI,

I was just referring to the Stress- Strain Curve for ductile materials where we can see that deformation here in terms elongation takes place at the yield point and further further to elastic (Temporary) or platic (Permanent) changes. If we agree that the deformation comences at the yeid point then it implies that the torque value is never considered more than the yield stress of the given material, it's always between 50% ~ 70% of the yield stress depending upon the degree of tightness specified to arrest during the Nitrogen- Helium Leak Test.

Cheers,

Ash Bandy

But at every angle value !!!!

As long as you generate an axial force in the bolt -which ever means you use- the bolt will be stretched and the parts compressed.

In torquing the thread angle generates the tension function of rotation so that tension is proportional to the angle this is the reason why an assembly torque-angle is less sensitive to friction.

There are two reasons for doing sequential tightening,

I assume the thread is aboul pipeline flanges, but I have faced a problem with a huge shell (about 5mt diameter)

The first reason is already covered (the initial tightening takes care of

- Thread yields (due to local loading of a few threads)

- Yielding of the flange and nut faces

- Any other possible yields (which include the flange face yield and the distortion removal)

As the bolts are tightened on one side and we go to other side, the bolts on first side were getting over stretched. We had a few equipment failures in this way (way back in early 90s)

to avoid this we have gone for stepped+sequential tightening

Our final torque (or pre-load_ was 40 Tons, with hydraulic tensioner.

we have gone at 30 deg intervals 3 bolts at each position (total 9) tightened to 10 tons, then all to 30 Tons and finally to 40 Tons.

Then sequentially proceeded 3 on each side of the segments (so now 18 bolts) in the above sequence (3 steps) till the whole cycle was completed. in fact we stopped the procedure (shortcut) when approx 2/3rd was covered and then completed the rest by the middle of each left out segment and diversing from it.

No failures subsequently and we have frozen the process and put in our equipment erection manual.

Not a problem!!! :)

I would be more than happy to help you on the subject. In couple of days, I will send you the spec. draft. By looking at your line spec it is obvious that the bolt tensioning shall be recommended in this case. Foremostly, you have to understand to define the torque/ tensioning values like for example 50% of yield or 65% of yield and TPI. The torque/ tensioning depends on the masterial spec of the bolt material and not the flange size. If you find any where the statement asking for torque related to flange size or the flange size, you could take it the statements are wrong... utterly false

Ash,

I really appreciate your offer of sending me a sample spec. I am a bit scared of the task and just have never written a spec that is that detailed.

I will eagerly await your spec.

Thank you!!!

Dejan

Although many ill-informed people do, one does NOT specify a bolt load by basing it on an arbitrary bolt material percentage-of-yield. To do so would be reckless and irresponsible. There are other things that must be considered including minimim and maximum gasket seating stresses as well as the load capacity of the flange itself!

Agreed!!!

Well, we do have to consider the coefficient of friction K and the anti seize suitble to the specified COF

You are partly right since you consider onlybolting for flanges. Do not forget that there are also structural bolting assemblies where the % of elastic limit is considered as an important factor.

hahahahaha

When you consider the bolt tensioning, the following aspects are to be taken into account

a) The yield of the bolt

b) Coeff of friction if torqued.

c) Materials of flanges.

d) The surface condition of the mating flanges and any gasket/ other material on the joint face.

e) The force in flanges in operation.

f) Operating condition

When the flanges are in tension while in operation, the resultant bolt stress will be pre-load+ additional tensile created by the service stress, It must be ensured that this do not cross about 60% of YS

When it is in compression it must be ensured that the resultant reduction in tension is sufficient for purpose.

The first one is common

For the second, I have one example (actual)

There is a hub , bolted (studded ) to a shaft, Axis is vertical (hub above shaft)

Heavy thrust load is transmitted (along with rotational) from Hub, downwards towards shaft.

The studs/ bolt are to produce sufficient joint pressure to drive the hub. If it is not the studs will shear off in no time (had happened)

The retained tension must take care of the

reduction in the stress - due to the joint elastic/plastic deformations due to compression

And increase in stress due to microslip relative rotation (we have later introduce pins to minimise these)

along with other parameters (fatigue, service temperature etc)

This basic load has the property of reducing the pre-tension on the studs.

- Initially the bolts were torqued. Multiple failures of bolts by shear.

- Thread sealant/ thread lubricants - almost no improvement

- tool steel, Thread rolled, MPI - failure reduced, insignificantly.

- Same (but Stud), tensioned - Very huge reduction in failure.

- Tensioned based on elongation check - almost no failure now.

In the last stage what was happening was some times after the studs were tensioned, the nuts are snug tight with spanner and tension released. The Nuts some times bit into the flange surfaces (due to surface machining lines etc)

We went into 3 step tightening (as explained in a prev post) but still retained was a bit varying. This was identified as a result of elongation check.

So we brought in a fourth step, thir step very near the final (say 38T when required is 40 T) and then had a statistical calculation of how much to overshoot to have the final retained.

It was interesting and challenging

PS: service condition means the temperature of operation etc and the temperature differentials which may relax/ increase the tension.

(I think I am off topic since the discussion is on pipe line flanges)

Very very good thanks.

Dejan

"When the flanges are in tension while in operation, the resultant bolt stress will be pre-load+ additional tensile created by the service stress, It must be ensured that this do not cross about 60% of YS"

In engines the heads are fastened with long bolts often tensioned over the yield limit. this is done to avoid fatigue failures. The 60% rule is not always the best and by far not the only one.

The final stress is depend only the application (60% is not a hard and fast rule) but if you are saying that it is already preloaded to 60% in tension and you have additional tensile coming is it ? Sorry not expert in automobiles

But can you see the cycle - as you reach nearer and nearer the YP , the chances of failure by fatigue is increasing isn't it (of course on a certain load it may even have plastic deformation and then the load will reduce)

The 60% (not hard and fast) is taken as the region where you are more likely to remain in the elastic (st line region) of the stress-strain curve during the service condition.

Let me check up my notes (will come back later if found and if wrong or you can correct - some where I had the bolt tensioning data got to search out)

In the bolt fatigue is due to the variation of tensile stress. The external tensioning force affects the bolt in the ratio of the stiffnesses bolt/clamped parts. When the bolt is stretched in the trans-elastic domain the apparent bolt stiffness is near to zero so that the variation is very small and your engine head stays correctly tight over long periods of time.

You are as I said accustomed to bolting in the chemical/oil/power industry with big bolts up to M120 (for reactor) but bolting is also in the car industry, airborne equipment, industrial transmissions, civil engineering aso. For every industry there are optimal procedures considering the kind of loads and environmental conditions and of course the nominal dimension of the thread.

Hi All,

I would agree to all of you but to have a specification as Dejan has asked for, I am almost completed with my homework and tomorrow I would post the draft to all of you, please have a look and comment

As a matter of fact theoritically the deformornation commences at the yeild point but but it is at 100% yield stress to defoemation to happn but whenever there is a torque/ tension applied there is an deformation (elongation) of the bolt which may be negligible, which may be at 50% or 60% of the yield strength.

Cheers,

Ash Bandy

You are TOTALLY wrong deformation start as soon as a tension appears in the bolt and related to it a compressive load in the clamped parts. I would suggest before making a spec be sure of what happens by tightening a bolted assy.

Hi NN,

Then how is it possible to get the flanges or galperti clamps tensioned at 50 or 60% of the yield?? When we talk @ tensioning and practically have seen the bolt being elongated and turn the bolt physically towards the flangle to commence the next step until it is confirmed by "brake torque". I know its the compression load that results to tension but I was just trying to explain the process with tensioning methodology.

Cheers,

Ash Bandy

You are right and as i said and you have rightly pointed I am in a high end of bolting (usually yes as you said it starts from M39(some times) and M92 (max)

And you are correct about joint stiffness (i was looking for the article for that some where I have it must get time to check it up) when I hget it will like to post the link for others to know.

Any you are also right (I don't know why the confusion - the stress strain curve shows that deformation (strain) as soon as the stress is applied even aminute one.

The question is not the deformation but the plastic yield and the amount of it that we are interested in.

Any way GA for the point made.

(Did you notice as i mentioned earlier that we have moved off topic from the OP?

but since this is also referred by another post of AB, we can continue, still the education is on and all gaps are not filled)

We are not "off topic" since we give to the "client" informations which will help him to do a better job by understanding the physics of bolting.

For many reasons -which I shall not describe in detail now but especially because of possibility of heavy over loading- when big bolting is involved the dimensioning of bolts is done so that they stay in the elastic range.

The 60% "rule" is a consequence of an other aspect: very often a bolt is designed only by the axial load, but when torqued the torque induces a shear stress which increases the equivalent loading (see von Mises equivalent stress). In order to avoid this the stress level for ONLY axial loads was limited. Now in the case of bolts which are NOT torqued at assembly above mentioned torque does not exist so that the limit can be higher.

As you see a limit due to a specific technology was extended to all approaches without considering real conditions!!

Hi Guys,

I would like to add on as I had stated earlier the term " DEGREE OF TIGHTNESS", as none of the flanged joints or patented mechanical joints are totally leak proof. When we perform N2-He leak test the degree of tightness would have been defined as per I think API 521. It implies that the torque/ tensioning has to be specified as per the recommendations of API 521 or the class of leak acceptable for the process.

Cheers,

Ash Bandy

Hi BI,

As suggested by you, that's what we do as of now. But here the concern is to spec it, I know in oil and gas 50% ~ 60% of the yield is ok and the joint won't leak but if a question is asked how we have derived it?? We don't really have a one line answer. Do you agree??

Cheers,

Ash Bandy

I respectfully ask you to give your head a vigorous shake.

Of course we have a one-line answer. This is the bottom line after completing the requisite calculations!

Your insistence on choosing a load based on an arbitrary percentage of bolt material yield borders on negligence. "50% ~ 60% of the yield is OK and the joint won't leak" is a ludicrous statement to make! There are myriad other factors which impact the reliability of a joint! How do you know that a load based on 20% of bolt material yield won't damage the gasket or yield the flange?? How do you know that a load based on 70% of bolt material yield is sufficient for internal service conditions, let alone initial gasket compression??

Scary, scary stuff...

Helo,

For example…ASME B31.3 302.3.1 states "The allowable stresses defined...shall be used in design calculations unless modified by other provisions of this code." It then goes on to state that the design stresses for bolting materials are as per Table A-2.

For A193-B7 studs at room temperature, Table A-2 lists a design stress of 25,000 psi. This is much lower than "industry standard" bolt stresses used when bolting up flanges.

For example, ASME PCC-1-2000 "Guidelines for Pressure Boundary Bolted Flange Joint Assembly" Table 1 "Target Torque Values for Low-Alloy Steel Bolting" contains a general note stating that the values shown are for a target prestress of 50,000 psi (root area). I haven't looked at why they have used the root area rather than the tensile stress area, but for a 1/2" bolt 50,000 psi applied to root area corresponds to about 44,300 psi on the tensile stress area. This still substantially exceeds the 25,000 psi that ASME B31.3 would seem to permit for this bolt.

That is just one example, but all the gaskets manufacturers recommend/require bolt loads higher than the code design stresses. Is there a section of B31.3 that permits the use of these bolting materials at such "high" stresses?

Thank you guys!

Dejan

You mean to say there is not "Requisite Calculation" available?? I think at this age it should have been standardized. Regarding, gasket compression do you really know the gasket compression in percentage or any other units?? I am sure the if you take a spiral wound and a NAF gasket for instance, the compression won't be same for both as the material is different. According to your statement it looks like for every kind of joint differing the gasket material should have a torque figure, which is not really the case. I am sorry but it looks like querying is what is your objective rathar than to take the discussion to a resolving direction or leaning something new which can be adhered to.

Cheers,

ASh Bandy

AB,

The opposite I am the one that has to come up with the correct way of calculating and writing the Specification for Bolt Torque/Tensioning for Flanged Piping...I could have worded it better sorry. So which is the right way to calculate the appropriate Bolt/Gasket and use the data to write the Spec?

Bye,

Dejan

You forgot the number of bolts which important for the bolting procedure and also the elements required to estimate the stiffnesses of all components.

"You forgot the number of bolts which important for the bolting procedure and also the elements required to estimate the stiffnesses of all components."

Not at all: These are part of the ASME VIII calcs - which anybody who walks by this flange under load would hope were being adhered to ;-)

Refer the following link.

This is not what I referred earlier but it has very useful information on flanged joints. including gaskets and others. as you scroll through the pages

http://www.mech.uwa.edu.au/DANotes/threads/assemblies/assemblies.html

(the next to next page of this

http://www.mech.uwa.edu.au/DANotes/threads/joints/joints.html#top

talks about the OP requirement.

Another useful site (normally I visit though the site is always under construction and disclaimed is

http://www.roymech.co.uk/index3.htm

Hope these site find use.

I think what ai referred to might have been in my text book (my memory is not what it was ) These sites only I found from my soft library - more may be there but these itself required a lot of effort (see the time diffce betw last post and this)

_{<pat pat pat>}

Interesting

Hello gentleman! Here is the Spec and please do comment.Thanks

Dejan

1 Scope

This Specification covers the hydraulic bolt tension and bolting-up requirements for piping and pipeline flange joints as specified in ANSI-B16.5 and or API-6A for Class 600 raised flat face type flange.

2 Purpose

To provide inspection, assembly and torque values for Piping and Pipeline flanges. So that this flanged connections are not damaged during assembly and maintain their tight seal during service.

3 Standards and References

Table 1: Code and Standards Table

4 Requirements

This Section specifies the minimum requirements for the installation of gaskets and bolting in flanged joints.

4.1 Bolting-up Requirements

The method of tightening of bolts in flange connections shall be done with the application of hydraulic bolt-tension equipment. Use of Hydraulic bolt tensioners requires that the threaded portion of the bolt extend at least one bolt diameter beyond the outside nut face on the tensioner side of the joint.

4.2 Selection of Tools

Tightening of bolts in flange connections is normally done by means of spanners and torque wrenches. This may result in uneven stress distribution in bolting and uneven gasket seating pressure that may lead to flange leakage. When flange joints are leaking additional force is often applied to the bolts in the area where the leak occurs. This may result in more deformation of the flange and its facing, thus increasing the tendency to leak.

An improved method of tightening is the application of hydraulic equipment.

The torque bolt tensioning device, including adapters, shall be suitable for matching boltsize. Each Torque Wrench Device shall be accompanied with a clear manual in language of personnel involved indicating clearly the relationship between read-out hydraulic pressure and moment applied on nut.

All tools used for tightening shall fit the hexagon of bolts and nuts without damaging the width across flats and be suitable for matching nuts in accordance with ANSI B 18.2.2.

5 Inspection

5.1 FLANGE INSPECTION AND PREPARATION

Check the condition of the flange faces for smoothness, scratches, dirt, scale, and weld spatter protrusions. Wire brush clean as necessary. Deep scratches, dent or combinations of defects will require re-facing with a flange facing machine.

All threads and bearing surfaces shall be free from sand, chips or any other foreign material which may influence the torque during tightening.

Check coating on flanges, including inside bolt holes shall be coated to specification.

Examine the gasket contact surfaces of both joint flanges for appropriate surface finish. Raised face flanges shall have concentric or phonographic finish in the range of 125 to 250 microinches Ra as per ASME B16.5.

5.2 gASKET INSPECTION AND PREPARATION

Only new gaskets shall be used. Damaged gaskets including loose spiral windings shall be rejected. Gasket dimensions, gasket material and type, shall be checked to be per specification DEDSNG M-301 and DEDSNG M-401. Extra care shall be taken not to damage the face of the Spiral Wound Gasket

When installing the gasket ensure the flange and gasket faces are completely clean and dry. Check gasket contact surface of both joint flanges for flatness, both radially and circumferentially.

5.3 STUD AND NUT INSPECTION AND PREPARATION

Check thread on stud and nut for damage such as rust, corrosion, and burrs. Nut to flange contact surface shall be clean and smooth. Studs area for nuts to be lubricated with approved lubricant. The same lubricant shall be used consistently on all joints. See Table 3 in Appendix 1.

5.4 INSPECTION OF FLANGE JOINT FIT-UP

Prior to installation of the flange gaskets, fit-up of the flange joint shall be visually inspected for proper alignment and fit-up. The requirements for the flange joint fit-up shall be as specified below.

5.4.1 Flange Face Alignment

All flange faces at rest prior to bolt-up shall be parallel to within 1/16 of an inch per foot across any diameter.

5.4.2 Flange Bolt Hole Alignment

The maximum allowable offset of the bolt holes in flanges at rest prior to bolt-up shall be of 1/16 an inch.

5.4.3 Gap Between Flange Faces

Flange faces at rest shall be practically matching without bolting. In no case shall the gap between flanges at rest exceed 1/16 of inch over the minimum dimension required to allow proper insertion without damaging the gasket installation of the flange gaskets.

6 Installation Procedure

Install gasket and all bolts and nuts before any tensioning. Controlled torque wrenching shall be carried out using Hydraulic Bolt Tensioners.

The Bolt Tensioners operate by hydraulically "stretching" the bolt to a pre-defined limit after which the operator is then able to hand-tighten the nuts. The hydraulic load is then released and the bolt remains tensioned. The advantage of tensioning (stretching) against torquing is that the process is not dependant on the type of lubrication used and eliminates the effect of friction under the nut and between threads. Accurate bolt tensions are therefore obtained.

To pull down the flange evenly, several bolts can be tensioned at the same time. All the bolts will eventually be tensioned at the same time. All bolts will eventually be tensioned after successive "passes" of the bolt tensioning equipment.

Note that the use of the bolt tensioning equipment usually requires the studbolts to protrude past the nut by an additional bolt diameter. Obstructions such as pipe supports and instrument tappings may prevent the bolt tensioning equipment from being fitted over the studbolt. In such cases, hydraulic torque wrenches will then be used to tension the bolts.

6.1 INSTALATION OF FLANGE GASKETS

The following steps shall be used when installing flange gaskets:

Step 1: Align the boltholes in the flanges by the use of a drift punch or other suitable means. The alignment of the bolt-holes shall allow the stud bolts to pass freely through the bolt-holes.

Step 2: Insert two bolts in bottom flange holes in vertical flanges. Align the flanges to have a gap sufficient for installation of the gasket.

Step 3: Place the Spiral Wound Gasket centrally within the flange assembly and clamp in place. Use of grease, pastes, adhesives or gasket compounds to hold the flange gasket in place is not allowed.

6.2 BOLT TORQUING PROCEDURE

Check and ensure that the general requirements of Section 4 have been satisfied.

To achieve uniform joint tightness and stress distribution the bolts shall be tightened in three stages as follows; 30%, 60% and 100% of the torque values stated in Table 3, Appendix 1. At each stage of the tightening, bolts shall be tightened in a controlled sequence as follows:

Table 2: Bolt Tightening Sequence

Note: Bolts numbered clockwise around the flange

Finally the bolts shall be chased round using the 100% torque value stated in Appendix 1, Table 3 until no movement occurs.

NOTE: Bolts must be tightened in the cross-pattern tightening sequence and employing the incremental rounds of tightening as prescribed in Section 6.2 of this procedure. If this is not done, the flanges may become cocked relative to each other, an indicator of non-uniform gasket loading and potential joint leakage. This is particularly true the smaller the flange bolt circle and the fewer the number of bolts.

DIAGRAM 1: BOLT TIGHTENING SEQUENCE EXAMPLE

Bolt Tightening Sequence: 1-7-4-10 2-8-5-11 3-9-6-12

The bolt-up sequence indicated in Diagram 1 is as an example applicable for all types of fastening tools, orientation can be changed to suit very minor differences in distance between flange facings resulting from fabrication tolerances, not out-of angle welded flanges.

6.3 BOLT TENSIONING REQUIREMENTS

The torque bolt tensioning device stretches the bolt shanks by means of hydraulic oil pressure. While the bolts are elongated, the nuts shall be run down the thread by hand, until they bear against the flange. When the hydraulic pressure is released all of the bolt extension is retained over its active length between the nuts, and the required load is obtained assuring high integrity of Flanged Joints.

A Hydraulic Bolt Tensioner or Jacking Tool screws onto one end of the bolt, therefore the bolt needs an additional length of thread protruding out of the hexagonal nut at one end by about 1 to 1.5 times the bolt diameter. After the tightening procedure is complete the extended thread lengths shall be protected against corrosion and damage to allow the nuts to be subsequently unscrewed. Grease filled rubber or plastic caps are recommended for this purpose. The hydraulic pressure is supplied by a high-pressure oil pump via a distribution ring main to a number of jacking tools. The number of tools used depends on the application or the total number of bolts to be tightened, refer to Section 6.4.

6.4 hydraulic BOLT TENSIONING PROCEDURE

Check and ensure that the general requirements of Section 4 have been satisfied.

Before starting work check that the tensioning tool is not too large to fit on the nuts and flange.

The application of hydraulic jack bolt tensioning equipment shall be in strict accordance with the manufacturer's instructions and safety procedures.

Personnel using the equipment shall be properly trained in its application.

6.4.1 Installation of Service Flange Bolts

Installation of the flange bolts shall be performed just following completion of the visual inspection requirements specified herein. Installation of the flange bolts shall be performed in accordance with the following steps:

Step 1: Just following installation of the flange Spiral Wound Gasket, select the required flange bolt components in accordance with the appropriate pipe material specification in DEDSNG G-301 "Pipeline Components Data Sheets" and DEDSNG G-401 "Piping and Piping Components Data Sheets"

Step 2: Install the bolts in the flange assembly so they have an equal length of threads exposed on each end. Bolt holes in the flange assembly shall be aligned so that the studs will pass freely through all of the flange holes.

Step 3: Apply the specified bolt lubricant shown in Appendix 1 to the bolt threads over the length of the thread engagement, and the face of the nut that will contact the flange.

Step 5: Install all studs and nuts finger tight so that the studs are centered between the nuts and have an equal number of thread projection at each end of the studs.

Step 6: Number each bolt according to its position in the flange as shown in Diagram 1, Section 6.2

Step 7: Tighten bolts with a calibrated non-impact type torque wrench in accordance with the stud bolt tightening sequence shown on in Diagram 1 and Table 2.

Note:

The hydraulic Pressures "1^st Pass" and "2^nd Pass" listed in the tables were given by Hedley Purvis Limited in Appendix 2 and by Hydratight Limited in Appendix 3 and by Hydratight Limited in Appendix 3 and Appendix 4. The following procedures explain how these values should be applied depending on the number of hydraulic tools employed:

6.4.2 Procedure "A"- 100% Tensioning:

Set up hydraulic tools on all flange bolts simultaneously.

Apply "2^nd Pass" Pressure and tighten nuts with tommy-bar. Release pressure.

Repeat step (ii). Tightening completed.

6.4.3 Procedure "B"- 50% Tensioning: (i) Set up hydraulic tools on first half of alternate flange bolts, which shall be positioned equi-spaced around the flange. (ii) Apply "1^st Pass" Pressure and tighten nuts with tommy-bar. Release pressure. (iii) Repeat step (ii) (iv) Transfer hydraulic tools to second set of alternate flange bolts (v) Apply "2^nd Pass" Pressure and tighten nuts with a tommy-bar. Release pressure (vi) Repeat step (v) (vii) Transfer four hydraulic tools back to the first half set of flange bolts. (viii) Conduct a "Break Loose" check.

Apply pressure and increase slowly whilst holding the tommy-bar in the "sockets" and applying pressure in the loosening direction.

At the moment the nut becomes loose, stop the pump and note the pressure. This is termed the "Break Loose Pressure".

If this is equal to or greater than the "2^nd Pass" Pressure then the tightening is complete.

If this pressure is less than the "2^nd Pass" Pressure then repeat steps (i) thru' (viii).

6.4.4 Procedure "C"- 25% Tensioning (i) Set up hydraulic tools on the first set of bolts, equi-spaced around the flange. (ii) Apply "1^st Pass" Pressure and tighten nuts with tommy-bar. Release pressure. (iii) Repeat step (ii) (iv) Repeat steps (i), (ii) and (iii) on each successive set of bolts around the flange until all bolts have been tightened, using the "1^st Pass" Pressure on each set. (v) Transfer the hydraulic tools back to the first set of bolts. (vi) Conduct a "Break Loose" check (see procedure B, (vii))

If the "Break Loose" pressure is equal to or greater then the "2^nd Pass" Pressure then tightening is complete.

If the "Break Loose" pressure is less than the "2^nd Pass" Pressure then repeat steps (i) thru' (iii) (vii) Continue to transfer the hydraulic tools to each successive set of bolts around the flange, repeating step (vi) on each set of turn.

To summarize, the hydraulic tools must be continually transferred to each successive set of bolts in turn and tightened to the "1^st Pass" Pressure until a "Break Loose" pressure equal to or greater than "2^nd Pass" pressure is achieved.

It is important that the "1^st Pass" Pressure should not be simultaneously applied to all bolts as this may cause overstressing of the flange.

When alternative manufacturer's equipment is proposed, it is the alternative manufacturer's responsibility to determine the 1^st and 2^nd Pass Pressures that will achieve the recommended bolt stresses listed in the tables without overstressing the bolt and the flanges, taking cognizance of the aforementioned procedures.

NOTE:

Hedley Purvis Ltd "1^st Pass" Pressure equates to "Pressure A" supplied by Hydratight Ltd and Hedley Purvis Ltd "2^nd Pass" Pressure equates to "Pressure B" supplied by Hydratight Ltd.

6.5 THREAD LUBRICANTS AND PRESERVATION FLUIDS

All bolts shall be treated with a preservative lubricant using only the approved grade below:

Molykote 1000 (combined friction coefficient for application nut face and threads=0.11)

The use of lubricants with a lower coefficient of friction than that shown (i.e.< 0.11) can lead to excessive bolt stress being applied by wrenches or torque wrenches, and yielding or failure of flanges or bolting may result. Molybdenum Disulphide greases shall not be used under any circumstances as this can result in stress corrosion cracking.

Note that Molykote 1000 does not contain molybdenum disulphide.

When the bolt tightening procedure is completed, to prevent corrosion a liberal coating of grease should be applied to the nuts up to the back of the flange, and to the stud ends protruding from the nuts. This is particularly important where low alloy bolts are used on unpainted stainless steel flanges.

6.6 CLAMP CONNECTORS

This standard does not contain bolt torques for clamp connector bolting. Using Dow Molykote 1000 bolt thread lubricant with a coefficient of friction of 0.11, the bolt torques required to achieve the required residual bolt stress shall be obtained from the particular clamp connector manufacturer.

6.6.1 Clamp Connector Tightening Procedure (i) Note that clamp connectors are particularly sensitive to cleanliness. Check that the sealing surfaces of the hubs and seal ring are perfectly clean and completely free from scratches, corrosion or any other damage. Superficial marks or superficial corrosion on the hub sealing surface may be removed by light application of emery cloth. (ii) A light lubricant may be applied to the sealing surfaces immediately before assembly to assist the correct location and alignment of the seal ring. (iii) Apply Dow Molykote 1000 to the bolt threads, nut threads and the domed surface of the nuts. (iv) Assemble the joint fitting with the nuts with the domed surface of the nuts facing the domed seating on the clamps. On horizontal piping the clamps should be orientated with the bolts at the top and bottom of the joint to allow water to drain from the clamp to hub interface. (v) To ensure even tightening of the two clamps, the bolts should be tightened diagonally in a criss cross pattern. Apply the clamp manufacture's recommended bolt torque in 3 stages, setting the torque setting to 30%, 60% and 100% of the recommended torque at each stage. Finally apply the 100% bolt torque value to each bolt in a clockwise direction. The gap between the clamp lugs should be the same at each set of bolts.

7 Hydraulic Torque Wrench Requirements

7.1 Torqu wrench requirements

All torque loads shall be applied by the use of a suitable calibrated non-impact type torque wrench. All wrenches torque shall be calibrated at the start of each work shift. Use the final torque values shown in Appendix 1, Table3. Note: do not exceed the specified final torque values shown in torque Table 3, Appendix 1, this could over stress the flange gasket or flange bolting.

Wow, Dejand: great job! :-) You've certainly covered most of the issues. However, there could be some confusion. Perhaps I can help:

Your reference of a "bolt torque tensioning device" highlights this confusion: One cannot apply torque by using a bolt tensioner. One can only apply bolt load. "Bolt load" is that which one needs to clamp a joint. "Bolt torque" only means how much resistance is felt when turning a nut. As indicated in one of my earlier posts, one can apply a great deal of "torque" (even in exces of that which was specified) yet the the fastener may still be loose.

You've specified that the bolts should be tensioned except where obstructions prevent the use of tensioners. You say that in those cases the bolts should be torqued. This is understandable but often unnecessary for standard flanges. You simply have to ensure that the flange is designed accordingly. It seems that the flange in this example is of a standard design. Thus, you've got nothing to worry about. Otherwise, if you allow torquing, you open yourself to abuse by the installation contractor, especialy if he is more comfortable with this inaccurate beat 'n' bash method.

Nevertheless, if you feel that you have to include torquing, include the procedure as an appendix, not in the body of the spec. This will make it clear that you want the flange to be tensioned.

Speaking of the appendix: you've indicated that this is where the "bolt torque" values are. Once again: The term "bolt torque" is absolutely meaningless in the context of bolt tightness. Somewhere, you've already determined the required preload needed by each bolt for proper gasket compression and service conditions, right? Why, then, would you want to read the tea leaves, throw the chicken bones, consult with the spirits or otherwise take a wild guess at what the "K" factor might be during actual assembly? You've got no control whatsoever over what this would be! Let me give you an example to illustrate the folly of this:

A bolt load of 111,000 lb-f on a 1.5" B7 fastener would require a torque of 2,500 ft-lb when assuming a K factor of .13. If the K factor is actually .18, the resultant torque would be 1,800 ft-lbs! Thus, depending on which K Factor value is correct, one can either end up with bolts that are too loose (and a leaking flange) or, bolts that are too tight (and a leaking flange). It throws a kink into the requirement for a "calibrated torque wrench", huh? What's the point of ensuring that the torque output of a wrench is within plus or minus 3% if the accuracy of the torque-to-bolt load is plus or minus 40% (!!) ?

You need to specify bolt load

If you want to include torque values in the appendix, do it only as a sub-section in the torquing procedure. In doing so, one MUST specify that residual bolt loads MUST be verified by bolt elongation measurements. Not including this critical QA step blows all of your previous attention to detail and safety out of the water.

In your tensioning procedure, you refer to two certain manufacturers of bolt tensioners. This isn't considered good form. General bolt tensioning procedures are common to all standard tensioners. Some companies frown on this practice as it leaves them open to charges of conflict-of-interest or supplier kick-back. I'd suggest that you keep the spec as generic as possible. Indeed, the load capacities of tensioners varies between manufacturers. However, as long as you include the generic formula to determine operating pressures, everything will be covered. You'll also need to include the fasterner grip lengths so that "A" and "B" pressures can be calculated if 50% tensioning rather than 100% tensioning will be used.

Yes: This is a detailed process. However, so would be the clean-up after a catastrophic leak on a critical flange where the appropriate degree of control hadn't been exercised.

BoltIntegrity,

Thank you very much for replaying. Now that you point out the confusion on the bolt torque tensioning device I see that you are absolutely right. I will take your advice and make those changes.

Thanks,

Dejan

Wow, you guys are into flange bolting in a big way as it's so important to your industry.

The simplistic explanation that I received as to the reason for "60%" tightening was that if the bolt/stud/fastener spacing had been correctly designed, then that value would be such that changes in the load condition would be away from the stress/strain/elongation zone for the material and so FATIGUE would be avoided in the fastener.

I read hints to that in the responses from Ash Bandy and Nick Name and defer to their expert knowledge.

Obviously you cannot specify a tightening condition based only on material and diameter. It's interesting that no-one has yet added the complexity of thread pitch and the mechanical advantage represented by finer pitch at the same "torque" on the fastener.

Love this site!!!

Opinions are divided on the mechanical advantage of fine pitch.

Infact most of the calculations show the otherway round - ie, the basic strength (the thread strip strength) of coarse pitch are better than the fine pitch.

See one good link to a calculator for joint design

http://www.engineersedge.com/calculators/machine-design/bolt-preload/bolt-preload-calculation.htm

also

Check on www.boltscience.com

http://www.boltscience.com/programs/fastenerdemo.exe

this is a good program, along with some other nice softwares. The fastener program (of course the demo version) i have tried out and found quite impressive.

Fatigue in bolts is limited by the variation of stresses due to variation of applied loads. According to many results the pre-load is not influencing the fatigue life. It is in fact logical, if you look at the thread root the radius is relatively small and generates a strong stress concentration. It is also a 3D stress pattern due to streching and torsion at same time. The stress increase factor in comparison with mean stress computed with the equivalent stress area is around 3.

For fine threads the radius is smaller and the stress area bigger so that the yield limit is present at the thread root even at small torques. Fatigue will start at the root and is correlated to the hystheresis loop which is partly in the transelastic range, this loop is proportional to the load variation and not to the preload. As I mentioned in an other comment the 60% limit is due to the differences between the "classical" and simplistic computing based on axial load and stress equivalent section and the true stress equivalent (for instance according to the von Mises equation) level when the torsion is also considered.

There are other reasons to avoid fine thread in heavy loaded assies, the load is not transmitted even on all threads in contact, the first thread takes over 40% of total load. A fine thread has a smaller contact flanc area and a smaller shear section so that for "same torque" the specific loads will be higher and the risk as well.

Gentlemen,

Great posts from everybody.

I must apologize profusely. I have let some of you down because you still refer to "bolt load" as "bolt torque". I'll try again:

Bolt torque in terms of how 'tight' a bolt is or should be, is utterly meaningless!!

"Torque" (ft-lb, Nm etc) is what you apply when trying to turn a nut.

"Load" (lb-f, KN etc) is how 'tight' the bolt is. *Load can also be described as bolt "Stress" (psi, MPa etc) or, as percentage of strain.

For the reasons described above (obviously not explained well-enough - I apolgize again), the terms "torque" and "load" are not interchangeable!!!

The problem is that MOST tightenings are still obtained by "torque". In the range of small threads 100% and in bigger threads 75% or even more.

The sensitivity of preload versus friction lead to several "tightening" procedures which reduce the influence offering a better reliability of preload/torque. Those procedures are not any more based only on the torque measurement but use also other process parameters mainly the angle. Since mechanical properties as pitch or Young modulus do not change in a significant manner over stress level there is a good almost linear relationship between angle and preload (although somebody doubted that turning will lead to a strech). Depending on the choice of parametres the influence can be reduced to 10%...15% of the one on torque only.

Only to mention 2 of them:

- torque / angle

- yield streching

both used with success in series assemblies with a high critical rate as automotive or aeronautics.

I would very much like to know how to use streching on the wings of a plane where there is so limited place and the number of bolts per wing goes from 160 to 2500! We cannot use all approaches good for an industry in all others on the contrary we have to use the best approach for the considered case.

We have not to be aside reality. If we like it or not "torking" is THE major part of "tightening" in the bolting assembly technology all domains included.

All critics will not change the face of the world and even there where streching is more recommendable not always can it be used as dejan very well put it in the spec.

What ever "preachers" say torquing has its reason to be as well as streching. We have to be pragmatic not hang on only one idea and fight against wind mills, better use them for wind power harvesting.

.

An addendum to those who still don't quite understand:

When applying "torque" (ft-lb, Nm etc) one does indeed apply "load" (lb-f, kN etc). However, until one measures the residual load, one has no idea how much of the "torque" was actually converted into the all-importand "load".

.

May I use your sentence since I like it,

An addendum to those who still don't quite understand:

The problem is the requested uncertainty of the preload. As I mentioned above in a previous comment there is a real proportionality between angle and preload via the thread and the pieces&bolt compliances. The rate ΔT/Δφ gives an indication about the total friction. It is thus possible even in the case of tightening by torquing to obtain with MODERN equipment a very correct uniformity of preloading and also know how sterssed are the bolts. Of course this is correlated with a certain degree of uncertainty but preloading has NOT to be with zero dispersion it has to have a correct dispersion according to use, environment, aso. We have to be pragmatic and use what is most economical and not let us be impressed by exagerated examples.

Of course streching has advantages NO ONE disagrees but since it is not possible to use it every where we have to be aware that other techniques bring as well the precision we need.

Wait a minute! There seems to be a serious and unfortunate continuing misunderstanding by some...

I am not advocating one bolt-tightening method over the other! In certain applications, bolt tensioning is indeed the best option. In others, it may very well be simple torquing or, as in factory-clean automotive plants, torque-to-yield. And of course, torque-angle has applications as well. The bolt tightening method and its associated degree of accuracy and control is all dependant upon the environment and consequence of potential failure. It's up to an experienced and well-versed engineer to choose and specify the proper method from the various options available to him or her.

Being "well-versed" starts with a fundamental and vital understanding that "torque" is not a measurement of bolt load! We will all live in a much safer world once we begin to understand and use the proper terminology. Otherwise, the myth perpetuates and people continue to wonder why sometimes their joints fail...even though they've applied the proper "torque" ;-)

Guy's,

i am questioning the need of Section 6.6 Clamp Connectors. Do I need Clamp Connectors in my Spec? Are they applicable?

Thanks

Dejan

Hello Dejan,

Quick answer: No. They're not part of this assembly and thus, are not applicable.

Thanks, those clamp connectors act as a Flange?

Dejan

Yes: Here's a link to one vendor of clamps.(Oceaneering/Grayloc is another)