Bolt make-up to ¦3% tension 5

[HotStab]

Guys,

API 17D (Design and Operation of Subsea Production Systems - Subsea Wellhead and Tree Equipment) has included in its second edition a requirement for flange make-up with bolts at 70% yield stress with an allowance of ±3% (see below).

“Closure bolting of all 6BX and 17SS flanges shall be made up using a method that has been shown to result in a stress range between 67 % and 73 % of the bolt’s material yield stress.”

Is there anybody knows a feasible and reasonable practice that could be used to assure a bolt tension in a so narrow limit?

Thanks for helping,
HotStab.

 

[stanweld]

There exist bolt tensioning devices which can meet the requirements. We use them regularly when bolting up steam turbine cases.

 

[SnTMan]

So is this requirement mandatory? If so, how do you exempt yourself from it?

To be brief about it, torque procedures (using torque wrenches) are often qualified to great accuracy thru careful control of the variables (fastener condition, lubricants, etc) by measuring bolt stretch. Don't recall what degree of accuracy can be achieved offhand. As I said these are qualified procedures, in practice bolt stretch is not measured on the assembly.

Maybe youi need to contact

http://www.boltscience.com

Regards,

Mike

 

[Screwman1]

There is no way to qualify any type of a process that involves torque to anything close to 3% tension accuracy. You will never come close to that over more than a few pieces.
The thing that I find semi humorous about this is that the yield strength of the bolt,nut and flange material is allowed to vary by much more than this allowance. Plus, just because you maintain initial tension at that tolerance, what are you doing to control the gasket properties to a tolerance similar and how much preload loss does the assembly have over time. Even if you were to somehow build a flange up and hit that target during the assembly, if you come back in 24 hours, your residual clamp load will be out of spec.
Whoever wrote this spec obviously never actually built up many (or any) flanges in the field, they must be lab based.

 

[CoryPad]

The API document is identical to ISO 13628-4. This document is handled by Technical Committee 67, Subcommittee 4. You should contact these people to discuss your concerns. Here is the contact information:

http://www.iso.org/iso/home/standar...tees/iso_technical_committee.htm?commid=49570

 

[desertfox]

Hi HotStab

This link gives accuracy of various bolt tensioning methods and supports most of what the other threads are saying.

http://www.roymech.co.uk/Useful_Tables/Screws/Preloading.html#Bolt_preload_accuracy

 

[Cockroach]

So how does my mechanic mount a tire? Is he qualified, or just some kid pulled off the street. Should I quiz him on his methodology or assume because he works there, he is qualified. Maybe I should quiz his boss and check his credentials. Are the equipment up to par? Stickers for calibration up to date? Are there recalls on such equipment or changes in procedure to reflect the latest trends in safety, liability and public safety?

Or maybe I'll just take a cab to work and not worry about it. Seems like less work.

Regards,
Cockroach

 

[SnTMan]

Cockroach, if you have a valid contract with your mechanic, he is obligated to meet whatever is in it, as are you. Whether it makes sense or not

Regards,

Mike

 

[ricklts]

Deja vu all over again.

The brain trust(s) that thought up those requirements in API have admitted that there are various errors in the document, subject to comment and revision by the members. There are methods of guaranteeing +/- 3% accuracy, but they are time consuming, and subject to many areas of disagreement. For instance, is the actual Yield/Tensile strength of each and every component within the specified range? Remember, you are stretching a real stiff rubber band, and the stiffer it is, the more pull required to energize it. Also, do you modify your preload procedure per each individual threaded member if the mechanicals vary by more than 3%?

Any torque system is dependent ( 40% of the turning effort is applied to offset friction in the threaded and/or clamped members) on the friction coefficient. Another big dose is offset by angular discrepancies of the threaded members (class 1, 2, or 3, A,B, or C fit?).

Could a suitable hydraulic/mechanical tensioning system generate the required preload/stress, repeatedly and reliably? Yes, but verification would require external measurement (mechanically -depth rods or micrometers- or ultrasonically.

Ask your customer for clarification/guidance, as he is the person who is ultimately the final judge of what/how you are furnishing.


Good luck,
Rick

 

[Tmoose]

yes, clarification by API would be a good thing.

I believe whenever possible A bolt tensioning spec should be more like a field goal.
If it clears the crossbar 10 feet above the ground IT'S GOOD!!
Insisting on a "bullseye" spec without danged good reason is misplaced effort.

The flange joint covered by the API spec is metal-to-metal. The compression of the ring gasket is set and limited by the groove machined in the flange face. It seems apparent that making up the joint and compressing the gasket requires only a portion of the bolts' clamp load, with the rest available to resist external loads that undersea piping will be exposed to.
"The connection make-up force and external loads shall react primarily on the raised face of the flange."

Although it is not stated, I think the API spec with a teeny torque spec tolerance for specifically treated (Fluoropolymer Coating) bolts is probably their effort to achieve consistent clamping on flange joints approaching, but not exceeding yield strength of the bolts. Yes, clarification by API would be a good thing.

For the last couple of decades the folks at API have impressed me with the thoroughness and thoughtfullness of their specs. They consistently provide a road map of what needs to be done to ensure problems are avoided. There are dozens of pages including work sheets devoted to rotor repair and balancing, and shop inspection. Sometimes there may appear to be some "overkill", but I can easily imagine those extra details were included because members of the panel had suffered thru some of the genuine nightmares that result when a component receiving a standard repair at a decent local shop was a little out of the ordinary.

An example of the "extras" is the insistence on runout checks during disassembly and multiple stages of repair and assembly.

Over the last 4 years I heard more than one shop say (after a BIG expensive problem emerged) stuff like "oh we didn't check that" or "we didn't check it after we cut the keyways" or "look at this chip on the micrometer. Someone must have dropped it."
"Gee it never did THAT before." "I've been doing this for 20 years."

I know engine builders who routinely assemble engines without checking actual rod and main bearing clearance, or make any effort to confirm installed valve timing. They feel since it works OK most of the time, they are willing to take the chance they will have to "eat" an engine every now and then (although a few have proven they really intend to deflect any problem back to the customer).