Assembly Torque Values 4

[miningman]

I'm surprised I cant get the answer to this using Google... maybe I'm using the wrong search words.

I am installing 16 inch diameter steel pipe, to be used to pump water against 1000 feet head, so duty will be approx 450 psi at the pump outlet. Pipe has been engineered as Class 300, which I believe is rated at about 750psi, so no issues there.

What is the required torque value for the assembly of the flanges?? We are in North America, so no references to European standards please.

 

[TGS4]

Your best bet would be to refer to ASME PCC-1. However, it all depends on your gasket.

So, short answer - ask your gasket manufacturer. What gasket are you using?

 

[MTPipeliner]

I might be mis-remembering a conversation I had quite a while back, but I seem to recall that torques values for flange bolting is hard to come by because no one wants to stand behind the data.

I have seem some tables, but I don't think any of them came from a recognized standard.

I think there are so many variables such as cleanliness of the bolts, gasket type, etc. that the answer to how much torque you need is more complex than it initially seems.

 

[SJones]

Google "European Sealing Association Flange & Gasket Guidelines" and don't worry, they use imperial units too! Look at Pages 16 and clause 6.6 on page 29.

Steve Jones
Materials & Corrosion Engineer

http://www.linkedin.com/pub/8/83b/b04

 

[Cuemaster64]

If I were to complete this process assuming that it is a Group 1.1 RFWNK flange as per ASME B16.5 (which has a max rating of 1125 PSI) I would get a company or use a high torque machine to torque the (20)1-1/4" studs and nuts (presumably grade B7 studs and 2H nuts) and proceed to obtain a final value of 690 ft lbs. Not forgetting to follow the bolt tightening pattern. If your company is QC driven then they might also have an exhibit to be filled out for these types of torqued joints. To that end all you would have to do is ask your client what they want them torqued to.

Just how I would do it being only a JM in a Hydro position.

 

[miningman]

Thanks Cuemaster, that was exactly the sort of response I was looking for. I dont have access to all the standards, nor am I am familiar with these documents... thats not the way we tend to work undereground. Your assumptions are pretty accurate.

On a practical basis, I need to prepare some general installation guidelines for my own mining crews... highly skilled in rock excavation and working in dirty, dark, wet confined areas with extremely limited visibility, but who typically never install piping more sophisticated than 100psi victaulic.

I have done work in the oil and gas industry and understand in general terms how to train my workforce and I had found a reference in the Flexitallic gasket web site that suggested a minimum of 500 foot pounds, assuming 30,000 psi bolt stress or 750 foot pounds if the required stress is 45,000 psi. This is close to your estimate.

Service duty will be regular mine water with some suspended solids. Occassional leaks are not a safety hazard, indeed they are to be expected in the environment we work in.

Given this extra info, is 500 ft lbs likely to be adequate or will we have to go to 700+??. Either way, its going to be a major exercise with the amount of pipe we have to install.

 

[SJones]

"European Sealing Association Flange & Gasket Guidelines" are downloadable free. If you have access to the web you have access to the document!

Steve Jones
Materials & Corrosion Engineer

http://www.linkedin.com/pub/8/83b/b04

 

[HeviiGuy]

"I would get a company or use a high torque machine to torque the (20)1-1/4" studs and nuts" ... "If your company is QC driven then they might also have an exhibit to be filled out for these types of torqued joints"

C'mon now; with all due respect, it saddens me that highly-educated people continue to put their faith (and safety!) into something which clearly has dubious or limited, at best, relationship to bolt load.

"Quality Assurance" and "torque" in reference to bolt load should not be put into the same sentence. There are so many variables which affect how much of the input torque actually results in bolt tension, yet many of us choose to remain oblivious to this reality. It's no wonder that in the mining industry, grinding mills throw bolts, autoclaves leak, crushers fall apart and on, and on... We all know that "Torque" is not and CANNOT be an indication of how "tight" a bolt is!

Tightening a bolt by applying "torque" (with a calibrated torque wrench even!) and then walking away, exepecting all to be fine, is almost negligent. Not only can this result in under-loaded bolts, it can also result in yielded bolts! Since "torque" often has such an insipient, unknown relationship to a bolt's load, one must always take the next obvious step: measurement and verification. The easiest way of doing this is by measuring the stretch of the loaded fastener. Any QA process which omits this crucial component will do much more harm than good as it'll only provide a dangerous false sense-of-security.

Ciao,

HevïGuy

http://www.heviitech.com

 

[Ork1]

Heviiguy,
Your last response about bolt measurement is very interesting. It is seldom regarded in large rotating, high temp., high impact type of equipment such as ore milling. Bottom line; it saves lives and equipment.
I was curious to know a source for this information regarding bolt stretch. It is not commonly used and I would like to apply this method myself, with your advice on where to find such a table, graph, etc... I would greatly appreciate it. I work on mills, crushers, etc... myself and I see this type of thing all the time.
Thank you.

 

[miningman]

Heaviguy, I suspect your suggestions are 100% accurate when it comes to crushers and grinding mills.. especially the bigger units that are becoming relevant in the mining industry but I'm just a simple production guy putting in some dewatering lines.

I'm still working on getting my procedures together, but ultimately I will instruct my crews to bolt to a certain torque value... and yes, I will insist on the use of never sieze etc... and I will leave the QA/QC stuff to other persons. All Im doing is pumping large quantities of water. Worst case scenario, a few leaks have to have the gaskets replaced... and I personally wont be doing any measurements on bolt stretch or anything like that.

I will follow this thread with interst tho, in case I have to install a crusher at some point in the future. Thanks for your response.

 

[HeviiGuy]

Hello Ork1,

Fastener stretch depends on a number of known variables (unlike the enigmatic "K" factor in the torque equation). In its basic form, it's really quite simple:

? = FiLo/EbAst

where ? = is the elongation
Fi = the required bolt load
Lo = the original effective length
Eb = Youngs Modulous
Ast = Effective cross sectional area

"Ah..," you're probably saying, "... but I don't know what the required bolt load is. All I have is this ambiguous 'torque' spec...". Fair enough. In that case, what one has to do if one can't perform a full joint analysis, is to simply work backwards from the Torque Equation to get a value of Fi :

T = K Dn Fi
Fi = (K Dn)/T

Unfortunately, this is fraught with an enormous problem: one has to guess at the Friction Factor (K) which the designer's voo-doo beads had told him to use. Doing this compounds an already-ineffectual process! All is not lost, though: On non-gasketed joints such as mill flanges or crusher sections, one can assign a value of Fi based on a percentage of proof strength. This has to be done with the assumption that the equipment designer, when performing his flange stress calculation, designed the system with sufficient stiffness.

Interestingly, we have quite a bit of experience with SAG and Ball Mill OEMs who are now desigining their structural fasteners to be preloaded up to 90% of yield!! What's even more interesting (and refereshing ) is that most of them also include bolt elongation values in a column beside the torque specs. They often won't honor future warranty claims if the elongations of all fasteners haven't been measured and recorded. Insurance underwriters have also begun to insist on this for other critical plant equipment.

It seems that people are finally realizing the that the Torque Emperor has been strutting around in the buff...



Ciao,

HevïGuy

http://www.heviitech.com

 

[HeviiGuy]

Miningman,

(I think that you must have posted as I was composing my reply to Ork1). Indeed, you make a very valid point. Some applications simply do not need this degree of control and Quality Assurance. It's all based on risk. One has to ask: What's the economic or human cost of joint-integrity failure on a particular joint? Heck, plant air lines, process water lines, dewatering lines and the like might as well just be beat'n'bashed together with an impact gun (which many of them are!) since the effect of a leak would be relatively minor.

There are a number of different bolting methods ranging from simply having Bubba swing a hammer against a slugging wrench, to using an overhead crane, to tightening by using impact wrenches, torque wrenches, or bolt tensioners, to torque-and-turn, to heat-tightening, to yield tightening and so on. Most of them (although some being quite unsafe and certainly not advocated!) are prefectly suitable for certain applications. The trick is assigning the most appropriate method and level of QA control to each particular application.

Ciao,

HevïGuy

http://www.heviitech.com

 

[johnSchwaner]

Question - is it possible (feasable) to measure stud stretch?

 

[HeviiGuy]

Indeed it is, John. We always use this method when faced by critical or problematic joints. Although it takes quite bit of experience to properly operate the equipment, it's accomplished utilizing specialized ultrasonics. The technology is based on "time-of-flight". In colloquial terms, the process is akin to sonar.

In practice, a signal is sent down the length of a fastener where it's reflected off of the bottom. The time which the signal requires to return to the transmission point is converted into a length. Obviously, one has to have recorded the initial, relaxed length of the fastener in order for the loaded length value to be useful. Variables such as bolt material temperature and stress play a big part in the equation. This is why typical ultrasonic thickness guages aren't at all suitable for this purpose.

As one can see, one of the the beauties of this process is that it can be used on applications with blind holes. Another is that it's effective for a wide range of fastener lengths. It does become a bit of a challenge sometimes but, we've "shot" fasteners from as short as under 2" to as long as 25 feet.



Ciao,

HevïGuy

http://www.heviitech.com

 

[unclesyd]

I agree with {b]Heviguy's[/b], informative posts, but will have to diverge on the topic of tightening flanges by torque values on a routine basis. On our site we tighten thousands of fasteners daily by torque value or in most cases by wrench value. The sizes range from 1/2" - 1" H11 SHCS TO 2 1/4" B7 studs and everything in between. Leaks are extremely rare. All our wrenches are set and checked against a Skidmore-Wilhelm tester, we have 3, weekly. A secret to getting a flange such as yours tight is to get as near the calculated torque value as practically possible and keep the flange faces parallel as there is some freedom in the gasket seating stress required form normal piping systems. Normal meaning piping without tremendous extraneous loads.

Measuring bolt stretch by micrometer was the norm when I entered the petrochemical field in the 1950's and is still used today, though sparingly. Our chemical group had technician whose job was ensure the integrity of any and all flanges using the best practice. You will find the bolt holes on some of our older exchangers are still numbered.
The normal procedure was to applied 1/3 the torque value, measure, then apply full torque in 2 steps, final measure with adjustments. As torquing normally understates the preload it was very easy to ratchet up to the required stretch

Here is the Sure Bolt site that has a ultrasonic measurement system that works extremely well.

http://www.surebolt.com/Flange-Design.htm

HeviiGuy,

Here is a interesting paper from Superbolt concerning fasteners. I noted in a previous post that this was eluded to by Mr. Biach, Biach tensioners, several years before this article.

http://www.superbolt.com/articles/virgin_bolt.html

 

[HeviiGuy]

Hello Uncle Syd,

Thanks for the comments and links (I'm quite familiar with them ;-) ) I'm also very familiar with the Surebolt equipment. It's certainly good but, we don't use these particular devices in projects where were involved. In our opinion they're much too cumbersome to use in light of the new generation of equipment which has superseded this.

Ciao,

HevïGuy

http://www.heviitech.com

 

[gasketguru]

This is pretty simple - we all know that torque only measures the turning force on the nut, not the actual bolt load. Hydraulic tensioners are also pretty unreliable once you let the oil presure down. Final loads from either torque or tensioner are +/- quite a wide scatter.

Check out these guys -

http://www.rotabolt.co.uk/

They convert the bolts with what is effectively a mechanical strain-gauge pin down the centre and set the cap to lock-up at whatever the desired axial load is. Much quicker to install than using ultrasound, so the cost of the bolt conversion pays back in reduced downtime and installation costs. Also, you get some condition-monitoring in-service as if the bolt loses load, you just twiddle the cap. They are used a lot in offshore cranes as well I believe.

 

[HeviiGuy]

Wait a minute: "Hydraulic tensioners are also pretty unreliable once you let the oil presure down". With all due respect, gasketguru, I don't think that you fully understand bolt tensioners...

Bolt tensioners impart an axial load to a fastener by directly stretching the fastener rather than relying on the "inclined plane" action of a nut traveling along a bolt's thread as it's turned. Friction doesn't enter into the process at all as it does with torquing. Granted, there is some thread relaxation involved but, this is well known and controlled by the proper pressurization process. There is nothing "unreliable" about bolt tensioners at all.

Not wishing to offend or to be improper by posting a blatant commercial link, if anybody would like further information, I could provide them with a link to a non-commercial, generic explanation of bolt tensioning.

It's great when one has a number of options to choose from. In our business we have to be impartial to every option as we need to choose what's best for our client's specific problem.

The rotabolts have been around for a long time. In some cases they're adequate However, there are two significant issues with them:

One can't tell if the bolt is over-tightened.
More seriously, they don't provide a definitive indication of bolt load as actual measurement would.

One must place one's faith in the initial set-up of the "rota" by the manufacturer. Earlier this year, we were in central Asia, working with a client who had similar "indicating bolts" on a key piece of critical equipment. Although many locked close to the desired load, quite a few showed a discrepancy of up to 15% - with a couple even beyond 20%. In fact, some even indicated load when the nuts had been removed. This would have been fine on a less-critical application (where we've sometimes recommended them) but not on this particular one where forced shutdowns due to bolt problems had routinely resulted in hundreds of thousands of dollars per hour in lost revenue.

Remember: "The trick is assigning the most appropriate method and level of QA control to each particular application."

Ciao,

HevïGuy

http://www.heviitech.com

 

[gasketguru]

I'm not sure what problems you had with "indicating bolts" from whichever supplier, though as you say it may depend on how you set them up initially, but installed measurement is the key to success.
How did you measure your 20% inaccuracy? Did you calibrate the bolts out of the installation initially and measure extension over a given clamp length? (Even modulus can vary slightly between material batches.) With tensioners, I had a problem a while back where the calculated tensioner over-pressure was around 25% but it took more than 100% overload to achieve the final bolt extension and applied load.
These overload values are based upon around 2-4 thou' (typically) of relaxation per inch, depending on the tensioner size, but I've never seen any research papers to prove it - I would still say that they are indeed highly unreliable in terms of accuracy, but pretty good in terms of applying plently of load in one go fairly evenly around the joint.

 

[HeviiGuy]

The "accuracy" of these fasteners was verified by actually measuring stretch - best way to do it short of attaching strain gauges onto the bolts ;-)

I dare say that if you had to apply "100% oveload" to acheive a certain residual load, there must have been a significant problem with your method. We have some very detailed generic instructions in the safe and effective operation of bolt tensioners which should help. I'd be happy to have someone send a copy to you, if you would like. I'm sure that this will go a long way in ensuring that you obtain results typical of this technology in the future.

gasketguru: I'm quite confident that qualified individuals who have had any experience at all with bolt tensioners would find your incorrect phrase "highly unreliable in terms of accuracy" smacking of a nefarious and blatant ulterior motive :-(

Ciao,

HevïGuy

http://www.heviitech.com