Torque Specification for a threaded connection

[Micromatic]

We are changing a design from a welded connection to a dry threaded connection. The welding process causes the part to warp. I have designed the connection to be a 1-14 UN male to a 1-14 UN female. The male thread is made from 4140 PH material with an outside diameter of 1.25 inches. The female is made from 4140 (heat treated to Rc 52-56)with an outside diameter of 1.75 inches. I found a simple formula for basic torque calculations and from them, I came up with a torque specification of 34,452 in\lbs. This seems way to high. If there is a way to properly calucate the torque, please let me know.

My calculations:

Based on a permanent connection: F = 0.9*At*Sp
Where;
F = Torque
At = Tensile Stress Area
Sp = Proof Strength

My assumptions are:
Yield Strength of 4140 PH = 61,000 lb/in2

Thank you,

Wiley

 

[RossABQ]

It appears your formula is calculating the MAXIMUM torque the bolt should see. I believe the calc you want is to relate the clamping force the bolt generates to the loads it needs to resist. But I'm not a fastener guy...

 

[Micromatic]

Okay, I have reviewed the calculations and what I calculated was the Pre-load tension required. I needed to calculate the torque with the following formula.

T = k*Fi*d;
T = Torque
k = contstant coefficent based on threaded connection
Fi = Pre-load
d = bolt diameter (nominal)

To make sure I have a good torque value, can anyone give me the yield strength of 4140 PH?

Thank you,

Wiley

 

[RossABQ]

Isn't it in your opening post?

 

[Micromatic]

No, the formula that I used was for Pre-load only. I needed to calculate the torque required based on the pre-load. I made a simple first year engineering mistake, not doing enough research prior to doing the calculations.

To expand on my calculations, my preload for the treaded connection is 31,732in\lbs. Based on the ultimate yield strnegth of 4140 PH is 61,000lbs\in2.

With this information, the torque that I calculated is 6346 in\lbs.

T = (.2)*(31732)*(1 in)

Thank you,

Wiley




Thank you,

Wiley

 

[Micromatic]

To make it clear, at this point, I am looking for someone to check my method of calculation and the value of yield strength for 4140 PH at 61,000lb\in2.

Thank you,

Wiley

 

[desertfox]

Hi Micromatic

Firstly according to this site 60.333lb/in^2, (417.1MPa) is the yield strength for 4140PH in the annealed condition.
If you click on the more button you can find other values for heat treated conditions.

http://www.efunda.com/materials/all...fm?ID=AISI_4140&prop=all&Page_Title=AISI 4140

The formula you are using is very approximate and you could quite easily be +/- 25% either way with your preload.
Your preload should be based on the external load your joint will see in service and the strength of materials within your joint and not just the maximum load you can put on the threads which is what your doing now.
If you exceed the yield strength of the materials clamped by the screwed joint because you have put the maximum torque you can on the threads then you will not achieve your desired preload because the joint material will give.
Have a look at this site it may help.

http://www.roymech.co.uk/Useful_Tables/Screws/index_screws.htm

If you require further help with this then we could do with more information about the joint materials, service loads etc and why you need such a high preload.

desertfox

 

[drawoh]

Micromatic,

Check your old Metallurgy textbooks. Rockwell hardness should map directly to tensile strength. For Rc52, I am getting 260ksi. I assume this is ultimate strength.

This assumes of course that your part is hardened all the way through.

There is no hole through this thing is there? I am trying to visualize your part. Hopefully this is a dumb question. (-:

JHG

 

[hydtools]

Maximum preload should be based on .75*yield strength. You do not want to preload to yield. If you are just making a connection, there is only the thread section to yield.
Your required preload should depend on your application loading and load cycling, if any. In the case of your joint, just use max. preload base on .75*yield.
As noted above, preload by defining torque will have a wide scatter. Especially dry. A light lube will reduce the scatter and change your calculation for torque.

Ted

 

[MintJulep]

Your formula in your first post doesn't work on units.

If you have area x psi the result is force, not torque.

At any rate, your approach is all wrong.

1: Determine how much preload (clamping force) you need to keep your thing together, and how much variation you can tolerate.

2: Determine how many fasteners you want to use.

3: Select a fastener size based on 1 and 2.

4: Determine the nominal torque that will produce the needed preload.

5: Determine the expected variation in preload as a result of using torque control. Make sure that the lowest expected preload is not less than the lowest preload you need.

 

[chicopee]

Micromatic, be careful about torque unit. It should not be "in/lb" instead it should be "in-lb".

 

[tbuelna]

Micromatic,

Besides what MintJulep points out above:

a) 4140 CrMo alloy steel is not a "PH" (Precipitation Hardening) alloy. It's mechanical properties are improved through quench and tempering. 1" diameter air-melt 4140 bar, oil quenched and tempered at 1200 degF, will have a tensile yield strength of at least 122 ksi.

b) When calculating stresses in a threaded body, be sure to account for all of the combined loads the body is subject to during installation and service. Not only will it have tension, it will also have torsional loads applied during tightening. And there may be bending present if the head or nut bearing surface is out of alignment.

c) Be sure to apply a satisfactory Kt factor in your stress calculations for any areas with small radii or sharp transitions.

d) Wrenching torque is a relatively poor way to preload threaded fasteners, due to the wide variations in the friction characteristics of the threaded joint. A much more reliable method is torquing to strain, using bolt stretch measurements or a torque angle procedure based on thread pitch.

Good luck.
terry

 

[racookpe1978]

I'm surprised you first specified a "dry threaded" bolted connection: You MUST use some ofrm of anti-sieze or lubricant uniformly and consistently applied to CLEAN and WELL-MACHINED threads (no burrs, rust, chips, grit or deformed surfaces) to even get close to creating predictable clamping values from a threaded connection.

First, and repeating what was said above: Determine what your total clamping force must be, then add a margin. (Can you add pins or posts or keys so the two parts can't rotate?) Determine your pre-load force on each bolt so the bolted connections NEVER pass out of compression.

Then arrange your mechanical connections so you can actually measure "total stretch" through every one of the bolts. Accept each bolt when it has stretched enough to meet your preload. DON'T rely on torque or numerb-of-flats or other approximations of this is a critical joint.

If possible, use an internal heater (or externally heated bolt that is assembled hot) so you can use heat to set tthe pre-load you need.

 

[Tmoose]

With the diameters not that much larger than the 1 inch thread, I envision the connection is a single thread between two cylindrical parts, like pipe used to drill wells.

http://www.surplusequipment.com.au/images/equipment/Sample_2_Drill_Pipe.jpg

Notice also the piloting male/female diameters to enforce concentricity.

The accuracy of the faces (perpendicular, concave (not convex) will govern how straight (unkinked) the assembly is after things are tightened up. The threads must be concentric and colinear as well, but are poorly equipped to win the struggle that results when faces first meet in a V.

 

[drawoh]

Tmoose,

I take it you are trying to answer my question.

All the calculations above are based on the stress area of a threaded connection. If there is a hole through the middle of the male thread, the stress area assumptions, above, are wrong.

JHG

 

[HEC]

From Shigley 1st metric ed Eqn 8-16 p309 the torque required is

0.2Kid
Where
0.2 is a magic number determined from emperical means based on a lubricated thread.
Ki is the bolt preload force required (kN)
d is the bolt diameter (mm)
Torque is in Nm


Mark Hutton

 

[GregLocock]

That's a fine equation but the OP specifies a " to a dry threaded connection". As such the 0.2 value is suspect.

Cheers

Greg Locock

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