×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

torque to tension
2

torque to tension

torque to tension

(OP)
i need to develop a certain tension in a steel rod with a turnbuckle.  has is the correlation between a torque wrench and the rod tension made?
sure appreciate your help.

RE: torque to tension

Fundamental Mechanics of Materials: PL=EA

RE: torque to tension

oops..  that's supposed to read PL/EA

RE: torque to tension

T = Pd/5 is a typical relationship between Torque and P for Preload (or Tension)

RE: torque to tension

If you need a really accurate tension value, I'd strain gage the rods.  Otherwise, frv is on the right track, but I'd add this to his comments:  1.) support the rods and take all of the sag out of them,   2.) snug up the turnbuckle more than just finger tight,   3.) use the turn-of-the-nut method to estimate your elongation and thus your stress or tension,   4.) work the rods someplace south of Fy.

RE: torque to tension

nut/bolt torque is not an accurate method for determining tension in fasteners.  

RE: torque to tension

...you could bind up you nuts on the threads and get a wildly high torque reading that is telling you nothing more than your nuts are bound up.
Follow the advice of frv and dhengr

RE: torque to tension

ahh... I take that back. Turn-of-the-nut isn't all the accurate either, ESPECIALLY when you're talking about a long steel rod.
If I recall correctly, there is something in RCSC about verifying nut turn on any fastener whose length is more than 10 or 12 times the diameter. I'll have to break out RCSC tomorrow...if I chose to go to work.  

RE: torque to tension

ToadJ:
"...you could bind up you nuts on the"...., "your nuts are bound up."  What's that all about anyway?  I kinda thought you took care of that problem about an hour ago on another thread.  But, you are still the guilty party, as you put it.  Please report your findings from RCSC for  our edification.  I think that's a slightly different problem than we have here, tell us how and why. I don't have my FEM turned on at the moment, so I'll have to await your findings to comment further.  Nor do I have a current copy of the RCSC publications so it'll be interesting to hear what you find.
 

RE: torque to tension

I wouldn't be specifying pre-tension in rods. The turnbuckle should be used to take out the slack.

RE: torque to tension

RCSC is a free download

RE: torque to tension

All I am getting at, is, it will be difficult, in my opinion to use the Turn-of-the-nut method on a turnbuckle located in the center of a steel rod. You will have to some how verify the amount of nut-turn, or how much your turnbuckle has turned with relation to your tension rod. I believe this would entail a solid means of keeping the tension rod from rotating while turning the turnbuckle.
I am not quite sure how that would be done.

Download RCSC and see the comments of table 8.2, which may not even apply since RCSC is for ASTM A325 and A490 Bolts. My reference, in this case, would be in error. See also the commentary of 8.2 with regard to research on bolts longer than 12 db and using the turn-of-nut. This is what I was thinking of when I read the OP.

My apologies directly to dhengr.  

RE: torque to tension

kikflip might have the right idea.
His comment reminds me of guys cranking life lines extremely tight and taking out the catenary....which is dangerous.  

RE: torque to tension

Efsinc:
Regarding turn-of-the-nut (the turnbuckle), put a tick mark on each end of the turnbuckle and on each rod at the same locations.  One revolution of the turnbuckle w.r.t. the rods (average the two tick marks) means a certain amount of total rod extension as a function of the screw thread geometry involved.  Clean the threads well, lub them lightly, try to hold the rods from turning without nicking them up.  You will likely be working your threads well below yield, which is not the case for the slip critical joints and high strength bolts covered in Toad's ref., the RCSC.  But, that's a good ref. for understanding this kind of problem.  Those bolts and joints have been tested extensively, and their discussion about tightening up the plys or faying surfaces, prior to applying the turn-of-the-nut, is akin to my comment about supporting (straightening the rod out) the rod to get any sag out of it, and pre-tightening the turnbuckle in this condition.  The structural high strength bolts are working at .7 or .8Fult. so there is some thread and bolt stretching and yielding going on with them.  And, on larger and longer bolts the thread to body stretch changes, from what's been tested, so they want you to test against a calibrated tensioning device to set the turn-of-the-nut value for a given tension.  Kickflip's comment deserves some thought too, as it takes almost an infinite force to remove the sag from a spanning rod by tension alone, so again, my reason for saying support it straight, then snug up the turnbuckle.

Otherwise, delta = PL/AE still holds true; and at f = P/A = 24ksi, you should get about .1" extension in the rods on a 10' gauge length.  The class and geometry specifics of the screw thread involved will indicate how many turns produce .1" rod extension,  i.e. a 3/4" - 10 UNC thread has 10 threads/inch, or one turn/.1" rod extension.  You might clamp a 5' long metal bar at one end, atop your rod, so it is supported by your rod, and at the other end use a machinist's dial indicator to measure .05" bar movement, then compare the two results for you rod tension estimate.
 

RE: torque to tension

a couple of questions for you guys ...
does "rods" imply rebar ?  'cause certainly you can pretension a free standing turnbuckle (like a tension wire brace)

and IMHO, PL/EA doesn't give you the relationship between torque and preload, 'cause you need to account for friction; no?  

RE: torque to tension

The reason I thought KikFlip comments deserves some attention is, if you tension this rod to somewhere around FY and it receives additional loading, well, then what?
This is what I was getting at with the super-tight life line installations I have seen in the field.
If you crank a life line tight and it approaches being horizontal, the geometry doesn't lend it self well to additional loading from a lanyard. Pre-loading only compounds matters further.  

RE: torque to tension

I understood the ref. to the life line or catenary, and at the extreme you would be talking about a  tension cable structure.  As usual in my old guy fashion, I was offering a first shot at the problem.  We are missing a bunch of info. on the whole problem, and the OP'er. seems to have left the building.  If the rods are 100' long and oriented or spanning in a horiz. position, then we certainly better give some thought to this issue, because just holding the rods near straight will produce more force than he might want.  He has to explain in more detail what he is really trying to accomplish.  If they are above a 45° incline and only 30' long, or hanging vert. then the above approach should be pretty good.  And, to account for some of this uncertainty, not having done any catenary calcs. on this, was why I said work the rod south of Fy.  I was assuming a mild steel rod, not a rebar, maybe 36-50ksi.  Mostly, it has to obey Hooke's Law and work below the yield point, and most of the higher strength steels don't have a well defined yield point.  Otherwise, it seems to me you can preload (prestress) any rod to f1ksi, and then any additional loading just works you further up the stress/strain curve, proportionally, until you get to Fy.  Thus, PL/AE relates the force in the rod to its elongation.  I think RCSC calls this strain dependant tightening.  And, I rejected torque immediately because it is not very reliable for this type of application, for all the reasons suggested earlier, friction, rusty threads, etc.  The ME guys have used torque as an indicator of bolt or stud tension or flange clamping force for a long time, and claim some reliability under ideal conditions.  Most of our nuts and bolts are not so clean and pretty.
 

RE: torque to tension

efsinc,

Why must you develop a certain tension in the steel rod?  What is your application?

BA

RE: torque to tension

dhengr,
yes PL/EA defines the extension of the rod, but how do you connect P to T, the torque applied, ?  wasn't that the original question ??

RE: torque to tension

(OP)
BA and rb and all,
The application is the post tensioning of the bottom chord of an existing damaged bowstring truss.  In addition to the various other repairs, I am intending to utilize a steel round rod each side attached to the heel each end.  I can determine the tension force i want developed in the rod but am uncertain how to verify that the contractor has attained it during the installation.  I am thinking a calibrated torque wrench and turnbuckle system is the way to go but i am not sure. Thus my original question.
Thanks to all for valuable insight.

RE: torque to tension

Friction on the thread can make torque measurements inaccurate as stated earlier.  Also, a turnbuckle is not the best piece of hardware to provide an accurate measure of rod extension because two rods are being stretched simultaneously.

It would be better to snug everything up, place lock nuts at the dead ends, then apply torque to a single nut, first marking where the starting position is, then measuring the elongation.   Continue tightening until the desired elongation has been reached.

BA

RE: torque to tension

(OP)
BA,
thank you.  If i am understanding you correctly, the idea is apply a little tension to the rod to pull the truss heels towards each other to snug everything up, to allow all the connections to slip as much as they might, letting all the bolts move to the sides of their holes etcetera.  The point reached after this initial 'snugging' movement i would call effectively the point of zero tension in the rod.  then lock everything in place at this point except for a single turnbuckle.  from this point i would require the number of turns of the turnbuckle to introduce the elongation necessary to achieve my design tension.
am i warm?

RE: torque to tension

Warm, yes and your method would work if you know the pitch of the turnbuckle threads and don't lose track of the number of turns, but I would prefer a visual aid while tensioning.  That is why I would mark the position of the nut at the point of zero tension and continue tensioning the rods simultaneously each side of the truss until the elsongations matched the required elongation.

Finally, put a lock nut on each live end.

BA

RE: torque to tension

If you have a wind uplift issue, obviously that has to be accounted for in you repair design.  I wouldn't think that was the cause of broken bottom chord members though, unless your analysis shows how that happens.  You have gotten good advice on the wood design and engineering forum, on your other thread.  I believe I have participated in a couple other threads on this same subject in the past year or two.  But, I'm having some trouble convincing you, on this thread, that a torque wrench may not be the way to go on tensioning these rods.  Assuming that you have read and digested the above posts.  To more effectively reinforce the trusses, I would consider harping the two rods at each bot. chord.  Pick two harp points under bot. chord panel points, and if you can gain access, hook the rod end reaction saddles up over the top corner of the top&bot. chord end joint.  That way these saddles are working in bearing on the wood rather than added shear in the end side plates, although that might work too.  By harping the rods you add rod tension at the bot. chord over the middle third of the truss and you also add an upward component at the third points (panel points) with the two affects acting as prestressing on the truss.  By the way, postensioning in conc. is accomplished by measuring wire extension, not by any torque method.  I've suggested several ways to gauge the elongation in the rods and thus the tension.  If there are a bunch of these trusses, maybe you would strain gauge a few rods to prove your less costly methods.
 

RE: torque to tension

(OP)
dhengr,
i think i have been disenchanted with the use of a torque wrench.  The direct measurement of rod elongation as discussed in the 2nd 3rd and 4th posts directly above i think will form a part of the final plan.  
Regarding the use of a strain gage, i think this sounds excellent, although, honestly, i do not know how to specify one, how to use one, nor how to direct a contractor to obtain and utilize it.  the last one i remember was in college some years ago.  if you would provide me some direction on this i would be most grateful.
 

RE: torque to tension

Efsinc:
Read my two posts from 28FEB, for that matter make of hard copy of both of your threads and really study them.  You must read btwn. the lines a little bit to see what applies to your specific problem, because until your last couple posts we were guessing at what you were trying to do.  But, there are good bits of advice and applicable ideas in a number of the posts in the two threads.

Your local testing lab will do the strain gaging for you, the same people who test your soil, conc., steel and welds, etc.  Or, talk to your local Uni. Prof. and get one of his/her Grad. students to do it as a class project.  They have the equip. to do this and are experienced at it.  It's not hard to do, we just aren't proficient at it.   Look at my 28FEB11, 14:17 post for answers to several of your questions, and the last para. offers another fairly simple means of measuring rod extension.  I don't know that I would strain gage on every truss, you are just trying to prove your simpler methods for reliability, if there are many trusses involved.  This is certainly not an exact process, we are talking +/-kips, not 100's of pounds, and further truss relaxation will change things in fairly short order.

As other's have said, you will not pull the trusses back together with these rods, you must jack the trusses up and shore them to do most of that, and bring the bot. chords back above a straight line.  Then, your thought that you start snugging everything up with these rods is a sound idea, but this is also a judgement call, not an exact number.  You have already studied the condition of all the truss joints and you watch them settle back into position; you watch the truss length, movement at bearings, etc.  And, you called it 'zero tension in the rod' at this point, but that's no longer really true, I'd call it your started point for the final prestressing, the starting point for your elongation measurement.  BA's comments on nuts and locknuts at the ends is good because that gives you three points at which to do the tensioning, and this will be helpful particularly if you harp the rods.  Of course, this assumes you have some working access at the ends of the trusses.
 

RE: torque to tension

if you don't like the torque wench (i think it's a reasonable tool, just you need to be a little careful in using it), you could use PLI (PreLoad Indicating) Washers, which crush at a predetermined load (and have much less scatter in the load than a wrench, but cost a lot more).

thinking about it, you probably can't use these ... you're tensioning up the rod, to snug up everything, and want to understand how much load you've up into the rod.  P = 5T/d is a rough estimate.  You could dble check this with witness marks on the rod, to measure the extension (but i expect this'll be so small that the error in measuring would be like the scatter in the torque wrench).

RE: torque to tension

dhengr,

I'm not so sure harping the rods is a good idea.  We are talking about a bowstring truss.  Under uniform snow load, the top chord is in almost pure compression, the bottom chord in almost pure tension and the web members are almost unstressed.  Do you think it is wise to introduce upward reactions at the third points?

If the rods are straight along the bottom chord, there is a danger in introducing too much tension in them.  We know that the bottom chord has been damaged, but we don't know where and how badly damaged it is.  Is the bottom chord capable of taking the compression which the rods are adding?  If the bottom chord buckles, nothing useful has been accomplished.

Perhaps a better plan is to stress the rods until the bottom chord is raised to the same level as the undamaged trusses.  Then, when the snow arrives, all trusses will sag more or less uniformaly.  Does it snow in California?

The rods should be periodically supported from the truss to prevent sag under self weight.  

BA

RE: torque to tension

(OP)
wow....you all make me proud to be in the engineering community.  i'll post more as i proceed.

RE: torque to tension

efsinc,

There is another issue which has not been discussed.  If the existing bottom chord is totally disregarded, the addition of the two rods results in a tied arch, not a truss.  That means that unsymmetrical snow load will cause bending in the top chord which was not present with the truss.

BA

RE: torque to tension

rb1957....   From sad experience we Structural Engineers just don't trust torque wrenches on semi-rusty bolts, nuts and faying surfaces, in highly stressed joints.  You ME's and Aero's have more faith, but usually under more optimal conditions.  I'd need to know more about the origin of P = 5T/d as a rough estimate to comment on that.  I think most of your torque values are based on considerable testing over fairly specific sizes, joint configurations and load ranges, etc.  Intuitively, torque and bolt tension should be related, if all the variables which can foul things up are adequately controlled, but there is plenty of literature and testing out there which muddies that water.  Once we grind up a little rust and even out any imperfections in the mating thread profiles, by pre-tightening the nut, and pre-clamping the plates, turn of the nut (or the bolt) is pretty reliable, and it eliminates a few of the variables, still using essentially the same logic, intuition and calcs.  And yes, some sort of witness marks should be used to measure turn angle.

BA....    All good points, and you may well be right.  But as is usual, we were asked a fairly complex design question and have been given a dearth of meaningful and important info.  At least we have peaked Efsinc's interest now, and he's back in the bldg.  I've been working on the assumption that the trusses were overloaded, the bot. chords over stressed and some small uplift might be helpful, but we can't really answer that right now, in either need, eccentricity or in magnitude.  For all the trouble I seem to get into when I talk about computers and computer programs, I would like to see a couple simple computer runs on this truss, under the different loading conditions, along with its construction details, member sizes and dimensions.  I don't know where the structure is or what the load conditions are.  In the U.S, that will be another 30 or 40 pages of ASCE 7, Min. Loads reading and calcs.
 

RE: torque to tension

if the rods are temporary maybe install a load cell in the rod?

RE: torque to tension

(OP)
there is no snow. just palm trees and movie stars and a heavy tax load condition.

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources