Hi Niekie,
You bring up good points but as with anything there is a procedure to follow. This might explain further.
When bolting per AISC and RCSC specs (other than snug-tight), we almost always have to do a pre-installation verification test. That involves taking 3 bolts, nuts, washers for each size length, lot number and in combinations as the bolts will be installed.
For turn-of-the-nut methods, we first snug tighten the bolts in a Skidmore. Snug tight is defined in the RCSC as "the tightness that is attained by a few impacts of an impact wrench or the full effort of an ironworker using an ordinary spud wrench to bring the plies into firm contact." I know this sounds subjective but another definition is the tightness required to attain approximately 10% of the minimum required bolt tension. In reality, the spud wrench approach is surprisingly consistent, according to the testing I have been involved with.
Granted, there is always an arm wrestler that tries to break bolts by brute force. But in the end, it doesn't seem to make much difference.
The key is that all plies be brought into firm contact to start from. As we approach the minimum required tension, initial tension values seem to "even out" so that there is a relatively small variation.
From snug tight, we turn the nut an additional amount. Usually 1/3 turn but it varies according to the ratio of bolt length and diameter, sometimes 1/2 or 2/3 of a turn. The Skidmore readings should then be at or above 105% of minimum required bolt tension (41 kips for a 7/8"dia A325 bolt).
At the same time, we record the torque value required to achieve the 1/3 turn, to be used for the inspection torque value later on. Of course accurate inspection depends on the bolts being in the same condition as tested.
Also, we may perform rotational capacity tests, or pay the supplier to do them, where the bolt is over rotated (2X the 1/3, 1/2 or 2/3 turn). During the 2X rotation, the tension may not be less than 1.15 of the req'd minimum at any point.
And the torque must not exceed a value computed as a ratio of measured tension X bolt diameter when going from snug tight to 105% of minimum.
Further, the bolts are inspected for necking down and thread failure after the 2X rotation.
All that is to verify the bolts attain 105 % minimum tension, without undue values of torque, and have sufficient ductility to compensate for reasonable overtightening, without significant 'necking down'.
During installation, after snug tight condition is attained, we mark the bolt, the nut, and the joint so we know starting and final position of the fastener. Inspection is simply to compare the marks to verify the nut was turned 2 flats for 1/3 turn. (The bolt head is held with a wrench to prevent turning while tightening.)
Alternatively, all we have to do is mark our socket at 120, or 180 degrees, line up the socket mark with a bit of spatter, dent, or make a chalk mark, and verify rotation by watching the wrench. Of course that means the inspector has to be there. Overall, the whole process is simple and fast to do.
According to the RCSC specs, if a bolt loosens during tightening procedures, all you have to do is re-establish snug tight condition and rotate the nut again. It is not considered re-using the bolt unless it is first loosened deliberately. Of course, if a number of bolts loosen in the process, then it is wise to recheck that all plies were brought together before proceeding.
It is possible the someone could 'get on with the job' but that can happen with anything. Although, if deemed necessary to do so, then applying inspection torque could be done as a check, but should be done very soon after tightening. The torque needed to turn a tightened bolt will change somewhat as rusting happens.
Chet Guilford
Hi Chet
Thanks for your reply and info. I can see how following the correct procedures can help to eliminate much of the problems. All this does however add to the complexity of the problem, and suggests that possibly people will be prepared to pay for a simpler procedure to both implement and verify.
Obviously this all depends on the criticallity of the joint, and the consequences of failure. (Risk involved)
In the petrochemical industry we have the problem of tightening the flanges on pressure vessels correctly. If you over-tighten the bolts, you can dammage the flanges and gaskets. If you under tighten, you could get leaks. Mostly people just "slog" the bolts to "h_ll and gone", and hope that it does not leak. If it leaks, they tighten it some more! So you can see that we also have our share of problems.
Thanks again for the reply.
Regards
Niekie Jooste
Fabristruct Solutions
A method used by high performance engine builders is to measure the length of the bolt, then tighten it until it stretches enough to load it to near its yield point. This is superior to torque plus turn since it isn't fooled if the material in compression crushes and doesn't depend on predictable coefficient of friction as torque alone does. The disadvantage is that you need access to both ends of the bolt so in an engine you are pretty much limited to connecting rod bolts. Flange bolts seem like an almost ideal application.
Bill
You are talking about two different bolting applications; structural steel bolting vs pressure vessels. A325/490 bolts are designed to be tightened per the research council recommendations. Snug tight tensioning for the pressure vessel industry does not apply to structural bolting, they are in no way related.
The answer to jtmiller's question is: After the pre-testing of the bolts ( per the research council's recommendations) further testing is not required unless there is a change in parameters such as equipment,
materials, personnel etc. Some discussion has been focused on whether daily pre-testing is required; other than this, torque wrench verification is not recommended.
Hope this helps,
Vonash
Vonash,
Your remarks refocus attention back to the original question so well I'm a bit ashamed to continue my quest for an answer to a previous question.
Your sentiment deserves to be the last word.
As much as most of us have been guilty of some digress from the original topic (there's been more on the side-track than pressure vessels- I believe this has digressed into a discussion about verification of bolt tension in general) I hope I will be forgiven, and perhaps you can help with my problem.
I'm trying to understand how sandwiching an [easily] compressable layer (a "stale marshmallow" such as a DTI washer with partially compressed nubs) into a critical structural connection makes engineering sense... perhaps the question is best posed to the engineers who specify them... eg:I dont know the pressure/nub compression curve and I'm presuming it to be strictly flat, but I hope it's accounted for in calculating actual service performance of the completed structure.
I sincerely hope my (and others') interest in the various appendant topics we're pursuing won't hide the information jtmiller needs.
Holiday Regards,
d
Been out of the office through holidays so have not followed thread. Hope everyones holidays were good. Lots of good insight above to chew on. Although I have worked in other fields doing NDE, structural steel is where must of my experience lies so that is from where I form most of my opinions. Not to beat a dead horse, but let me enter my last 2 cents on this topic.
I have very limited experience using DTIs so cannot speak authoritatively about the concerns I see listed above. Most of my above opinions about DTIs are from witnessing presentations, reading vendor literature and considering the ramifications. I see some valid concerns listed above and some of the same thoughts have crossed my mind.
It goes without saying, if all 4 methods were installed correctly, all would have bolts installed acceptably. As in all other disciplines, competent, well trained personnel make the difference. That having been said, my preference for installing bolts would be: TCs, DTIs (squirters), turn-of-the-nut, DTIs (nonsquirters), calibrated torque wrench. These opinions are based upon what I perceive the combination of ease of training of personel to install, problems associated with installation, consistency of results and ease of verification combined with the engineers desire to have a properly tensioned bolt. These opinions are further based upon my background with maintenancing, rebuilding, verifying and/or calibrating Skidmores and torque wrenches. I know Skidmores from the inside out and have my opinions as to the reliability of them based upon their state of disrepair. Likewise, proper use of and repeatability of various brands and styles of torque wrenches does not give me much confidence in them, ranking them lowest.
Similar to Chet above, we install shop bolts when they only call to be snug tight. When TCs are to be used (our most frequently used erector's preference), we install a couple of shop bolts into clips and ship TCs to the jobsite in kegs, making the erector responsible for care of bolts and correct installation.
I have heard of verifying bolt tension using ultrasonics also, but have no experience and have not tdone any research on that method for awhile. Does anyone else have any knowledge of using ultrasonics to determine tension?
I think RCSC should get in touch with Niekie's fitter friend to develop another "new" method to install bolts. We've all met one or two of these guys, haven't we?
I might be way off base, but I thought there were ways to use Ultrasonics to check total bolt length(like thickness gaging), which could determine the amount of stretch that has occurred in a particular bolt after pretensioning.
Is this what you had in mind?
John Wright
Might be. Been many a year since I remember seeing the application advertised.
I think they use this in engine building for race applications, checking tension in rod bolts, main brg bolts, head bolts, etc....
John Wright
Here's a link for an ultrasonic bolt meter
http://www.norbar.com/USAcompinfo.htm
John Wright