Torque on a nut and relation to axial force? 1

[JsTyLz]

Torque on a nut and relation to axial force?

In relation to the torque, how can one calculate the resultant axial load. I was looking at the same calculation used for a worm gear, but have also seen some examples in the Machinery's Hand Book. Also crucial to the design is that the nut has a nylon insert, so it is a lock nut. The bolt and the nut are also the same grade, so I know there is some interference or clearance between the threads depending on that grade. Basically everything added in, how can I come up with an accurate and MATHEMATICAL proof to my experimental results???

Thanks in advance!!!!

 

[arto]

F=5T/d
See Shigley's "Machine Design" Book for theory

 

[MintJulep]

You can't.

Your post implies that you have done some testing. That is the preferred method of determining the relationship between torque and axial load for any particular arrangement.

Bickford has a good discussion of all of the factors involved in his Book. There are also many many posts here on eng-tips both in this fora and the Welding, Bonding and Fastener fora, as well as FAQs on the subject.

 

[CoryPad]

There are so many simplifications to the equation provided by arto that it is not that accurate.

faq725-536 provides equations to determine preload from applied torque, but there is no treatment for prevailaing torque. Prevailing torque is routinely measured, but I have never seen it calculated based on interference values.

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[rb1957]

the preload is due, theoretically, to the friction between the threads of the bolt and the nut. there are admittedly many factors affecting the precise relationship, unusually in your case temperature will be more significant than typical installations (due to the nylon lock-nut). from this idea you could present your results in terms of a coefficient of friction, possibly as a product of various effects.

you could also talk to the manufacturers of lock nuts to see what they've uncovered.

 

[JsTyLz]

I definitely agree with consulting with the manufacturer, this always seems to cut the design time down as well.

But more importantly is the effects of prevailing torque nuts, i.e. locknuts. These nuts already provide some resistance to assembly torque without any preload. I almost see the prevailing torque as simply the difference betweeen the torque due to preload, and the final assembly torque.

 

[IceStationZebra]

Along rb1957's thinking with temperature, assembly speed can be a big issue - impact gun vs. ratchet. I had a problem with a bolt seizing in a tapped casting several years ago when I used an impact gun, but never when done by hand. The heat generated by the impact caused the bolt to expand and gall on the casting.

I also agree that the only hope you have of getting a mathematical proof is contacting the manufacturer. There are so many variables that you will be lucky to consistently get within +/-15% of your target clamp load.

ISZ

 

[tbuelna]

JsTyLz,

Two ways:

1. Measuring strain in the bolt/stud is the most accurate method. It takes friction and prevailing torque out of the equation. Use a micrometer to measure how much your bolt stretches. This is a common technique for racing engine rod bolts. And they even make special tools for doing it:

2. Torque the nut using a degree wheel. Run the nut down until it contacts the washer. Then turn it "x" number of degrees further. You can work out how much strain you are producing in the fastener based on the pitch of the thread.

And as good practice, always measure the running (or prevailing) torque when you install a self locking fastener to make sure the locking element is still within spec. Even on new hardware!

 

[israelkk]

And how you do it with a blind thread where you can not reach both sides of the bolt?

 

[unclesyd]

You might want to checkout this site for ultrasonic bolt tightening. You get a tension measurement with this equipment.

http://www.surebolt.com/

Here is a bolt, Maxbolt, with a built in tension check.

http://www.vfbolts.com/?gclid=COeGk5X1rYcCFRpoNAodvki16Q

 

[sms]

The best way is to measure the stretch, and knowing the spring stiffness use F=Kx. That is what the ultrasonic device does..

"Why don't you knock it off with them negative waves? Why don't you dig how beautiful it is out here? Why don't you say something righteous and hopeful for a change?" Oddball, "Kelly's Heros" 1970

Please see FAQ731-376 for tips on how to make the best use of the Eng-Tips Forums.

 

[IceStationZebra]

israelkk "And how you do it with a blind thread where you can not reach both sides of the bolt?"

There are two relatively cheap ways I can think of.
1) Use a load cell washer under the head of the bolt. May require a slightly longer fastener to account for the ~0.5" of added thickness.
2) Strain gage the fastener.

ISZ

 

[mechengdude]

To calculate axial force use the following formula:
M = K*D*F
where: M = torque, K= "friction factor" (.12 to .2), D = nominal diamter, F = axial force

Source: What Every Engineer Should Know About Threaded Fasteners. A. Blake, Lawrence Livermore National Lab.

Regarding lock nuts, they will increase torque required without payback in axial force. The amount will depend on what type of locknut. If you are using a locknut with the "nylon" type insert the Long-Lok Fastener Catalog has a chart giving "Empirical Torque Adders" on page 9.

http://www.longlok.com/pdf/handbook.pdf

 

[tbuelna]

Another less expensive (and consistent) method is to use a PLI (Pre Load Indicating) washer.

http://www.spstech.com/aero/prod_lit/pli_brochure.pdf

 

[rb1957]

i didn't think the OP was asking for ways to measure preload, but wanted ideas on the factors affecting the relationship between torque and preload ... "how can I come up with an accurate and MATHEMATICAL proof to my experimental results?"

 

[IceStationZebra]

Here is a list of some possible factors. If you wanted to incorporate these into a formula you would need to have a K factor for all of the possible combinations. And once you have determined your particular K factor you have already physically proved what your torque vs. clamp load relationship is anyway. That is why most everyone says that if you are that concerned you need to measure your particular joint.

As an example: I measured the amount of torque necessary to generate the same stretch in a wheel stud with two flange nuts that where almost identical except for the flange OD. The smaller nut ended up needing about 160% of the torque of the larger nut due to the decreased surface area thus more surface friction.

ISZ



- coating type (nut and bolt)
- thread classes
- non-standard self-locking threads such as Spiralok
- foreign material (for lack of a better term )
- pre-applied thread sealer, loctite, etc.
- lube and/or corrosion protection present
- contamination such as paint, machining debris, rust, etc.
- nut type
- std or prevailing torque
- nut height
- if prevailing torque, what kind?
- if prevailing torque, first installation or reuse?
- rotating member's surface contact/friction (rotate head or rotate bolt)
- head type: flange head, serrated flange head, hex head, socket head, etc.
- washers present?
- if present, what type (flat vs. flat w/teeth vs. lock, etc.)
- installation speed
- constant rotation (air ratchet) vs. non-constant rotation (by hand with wrench or ratchet)
- clamping surface
- surface finish (as cast, hot rolled steel, etc.)
- surface material (steel, AL, CU, etc)
- surface coating (paint, plating, etc.)

 

[JsTyLz]

All,
Thanks for the posts, it has really helped. I will keep posting as I get farther into the analysis... which brings up an interesting point. I have experimental results, but I wonder if what I have tested is truly the results I am looking for. I think it would be more advantageous to test the bolt and nut seperately, sort of as a control, or at least the the level I need the control to be at.

Also, I know different thread classes have different thread interferences, i.e. 2A bolt with a 2B nut. Or, 1A with 3B. etc. I think it would be better to minimize thread friction by ensuring clearance between the bolt and nuts major OD/ID. Any thoughts???

 

[mmart]

You would probably go a long way towards a solution if you have access to an Instron or Tinius Olson test machine, then you could run sample lots of different combinations in quick order and let the machine spit out the stress/ strain and load results.
just my 2 1/2 cents

 

[IRstuff]

You might take a gander at an Excel sheet posted at Excelcalcs:

http://www.excelcalcs.com/component/option,com_remository/func,fileinfo/id,245/

Purpose of calculation: For an applied load on a screw thread calculate the axial thrust of a nut.

TTFN

FAQ731-376

 

[tbuelna]

rb1957 is correct. I guess I should have read the original post more carefully.

Correlating your analysis and experimental data is simple enough. You just need to correct for friction.

Unfortunately, friction can vary WIDELY.

As an aside, when I submit a design to my analyst for checking, if friction helps my loading situation, he will claim that as far as he's concerned, friction doesn't exist. But if friction makes my loading condition worse, he will be adamant about including friction effects. The analysts always make sure to cover their butts!