Fastener preload.

[mcaskey]

Can any one tell me an easy way to measure the preload on a small screw. I have a #10 screw used to assemble 2 housing halves. I am looking to switch to a thread forming screw to elliminate the secondary opperation of tapping. The dillima comes from the driving torque of the thread rolling screw. It takes 16in/lbs to drive the thread rolling screw and my #10 has an assembly torque called out at 20in/lbs. The thinking around our office is that this takes away from the end clamping force by using most of the torque while driving. The screw manufacturer however disagrees with this and says that the torque will not effect the clamp force.
So I need a quick, cheap way to measure this to decide which school of thought to go with.
Any suggestions??

 

[MintJulep]

Measure elongation, or drive into a test assembly that includes a load cell in the joint stack.

 

[israelkk]

If the torque to form the thread is 16lb-in than you have to use 36lb-in to get the net 20lb-in which will guaranty the clamping force. You could first assemble the screw using 20lb-in and form the thread. Then release the screw one turn and re-tighten it at 20lb-in

 

[rb1957]

i'd side with the screw manufacturers; sorry israelkk, i don't think the torques add (as you say at the begining of your post), rather they are sequental operations (like you say at the end !).

it's going to 16 ftlbs of torque to form the thread; maybe it'll take more to form the thread, maybe less. once the thread is formed, who cares about this. as israelkk says, back off one turn then tighten to 20 ft.lbs and you'll have the required preload.

 

[mcaskey]

I can see this from both perspectives. On one hand you have the amount of torque to drive a pre-threaded fastener, which uses the standard torque coefficient. On the other hand you have a thread forming fastener that requires added torque to drive, thus the question. Does the preload drop due to the subtracted drive torque or is the drive torque irrelevant to the preload?
These two oppinions are what I am currently trying to decide between (my boss & screw manf.)

Now the question becomes: How do I measure preload with out a load cell, (no budget)??

 

[Drej]

On the torque issue I would say it really depends on the procedure/materials/component quality/person used to tighten the screw. Friction will play a big part as it always does in fasteners, and equally it's really difficult to quantify with any level of justification. Same goes for difference in hardness of materials (between the screw thread and the part). As mentioned above, why not drive the screw down and bottom it using a torque figure of 20 in-lbs, then release the screw by a few threads (3-4 to be sure, more if the application is critical) and retighten to the 20 in-lbs figure. I would imagine the correct/level lubrication on the screw threads to be essential here.

Cheers.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[Drej]

Calculate the preload in the normal way after you've tightened it to the procedure above:

Preload = Torque/(Screw Diameter x 0.15)


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[rb1957]

you could completely resolve the question about summing the torques by removing the fastener after the thread forming operation. i think the thread forming torque is doing just that (ie forming the thread) and not stretching the screw (ie preloading). i think backing off one turn is sufficent to unload the screw.

 

[israelkk]

rb1957

I was assuming that the 16lb-in to form the thread is a fact. Therefore, to get the clamping force equivalent to a non forming screw you have to use 36lb-in (16lb-in goes to form the thread and the 20lb-in to create a clamp).

However, if the 16lb-in is varrying too much then you have to form the thread first and then back off one turn and re-tighten the 20lb-in.

 

[rb1957]

but you'd never apply the two torques at the same time. first you'd form the thread, using 16 ftlb of torque, then as the head of the screw hits the plate the screw will start to stretch giving more resistance increasing the torque to 20 ftlbs. you wouldn't need to apply 36 ftlbs.

 

[Drej]

rb1957 - You wouldn't apply 16 units of torque either to form the threads, you would just go ahead and form the thread as necessary.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[rb1957]

yeah, whatever it takes to from the thread, then tighten to the required torque

 

[mcaskey]

Thanks for all of the good tips.

 

[MintJulep]

Ok, try this:

Measrue the torque needed to drive a screw the first time while forming threads.

Remove the screw and measure the torque needed to run it down into the now threaded hole.

Are they the same? On the first shot the screw is cutting threads - always. The torque needed to cut the threads needs to be subtracted.

This is the same as with a prevailing torque fastener, the prevaining torque needs to be accounted for.

 

[Drej]

MintJulep How does what you've say relate to "measuring the preload in the fastener" as per the OP question? And how would you "measure the torque" as you've described in either case?


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[Tmoose]

Like others have said Its pretty commonly stated that prevailing torque of a locking nut must be added to the desired torque. In some materials those self tapping/thread forming screws seem to have a lot or prevailing torque on their own.

It seems to me the only way the prevailing torque of thread cutting would NOT be additive is IF the thread cutting stopped when the screw first seats.

After the instant the screw seats there is still some thread cutting to be done if the screw is to advance the 1/8 turn (or whatever it takes to tension the screw). I'd note that whether the screw is mostly stretching or a gasketed joint is mostly compressing the screw must still "turn more."

One test beats an hour of my intense speculation.

 

[MintJulep]

Drej,

Ok, it doesn't really speak to the "how to measure preload" question, but the thread has diverged to discuss the "does the effort of cutting the threads need to be accounted for?" question.

Torque-o-meters are available, however several simple (albeit rather crude) methods are:

Use a beam type torque wrench.

Use a micrometer-adjust type torque wrench, start with it set as low as it will go and increase incrementally in small amounts. Note the lowest setting that results in rotation before the click.

Use a force-o-meter at the end of a common wrench.

 

[Barry1961]

You should be able to measure the clamping force with a cheap fish scale you can buy for probably under $5 US.

The difference between the two sides of debate is great enough you won't need anything fancy. The shape of your parts will dictate the type of linkage or lever you will need to hook the scale to your part.

Hook up the scale and measure the amount of torque required to move the scale.

If the original clamp spec was to ensure the screws would not come loose you made need to change the spec for the new style of screw.

Barry1961

 

[mcaskey]

I believe that I failed to mention that these are being driven into a blind hole. This, in my oppinion, creates an instance where, as Tmoose stated, the screw is still turning and cutting threads at the same time that it is stretching and being placed in tension.
I have machined a jig and by using 2 die springs and a scale have found that the tapped hole compresses the springs more than the thread forming screw at the same torque. So I now believe that the drive torque should be additive.

 

[MintJulep]

mcaskey,

Since you have the test jig now you might as well go the extra step and test to determine the torque you need for this application to get the preload that you want.

More better and no harder than figuring out what torque you need to add to the "standard" 20 lb-in.