Lead Screw and Clamp Design

[LennyT]

I need some help designing an I.D. mandrel clamp for a prototype machine.

In order to clamp our mandrel to an I.D. of a tube 20 ft-lbs of torque is needed (measured from a torque wrench). The lead screw is a 5/16-18 B7 All Thread and the torque wrench was turning a 5/16-18 hex nut at the end of the mandrel.

Originally we were tightening the clamp with a wrench, (like above), but we would like to change the design. We may look into a cam lock style depending on what axial force is need to get the equivalent of the 20 ft-lbs of torque form the lead screw and nut. I hope this makes sense and someone can help me out.

Lenny

 

[Cockroach]

The axial force produced is 5330 lbf, assuming your 5/16 - 18 UNC-2A threading are to nominal geometric sizes.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[LennyT]

Wow, that is much greater than expected. Can you tell me or where can I find the formula's so I can try some different sizes. I thank you very much for this.

 

[Pud]

hi Folks,

Has anyone taken into account friction? From a (very) long time ago I recall that some 75% of applied torque is to overcome friction in the threads and under the bolt/nut head. If so, the axial load would then be a lot less. (Intuition would make over 2 tonsf seem high, but it may be so......)

Rgds

Harry

 

[Cockroach]

Yeah, a word on the computation. I never worried about friction which is typically taken as 0.2, i.e. coefficient of friction - metal threads. This presupposes much more information of the problem.

LennyT had only specified thread geometry (5/16 UNC) material (B7 assumed to be ASTM A193 B7 and therefore ASTM A194 2H nut) and torsional load of 20 ft lbf.

Equations can be worked out from first principles noting that the ratio of normal stress to shear induced through torque is approximately 0.577. Actually the calculation should be carried through using the pitch diameter as the mating surface, I would expect the thread load to be slightly higher.

Calculations of this sort are statically indeterminate. Fortunately you provided torque rather than advancement of the nut due to torque. Details of modelling can be found in any textbook dealing with Strength of Materials. I used Beer & Johnston, 1st Edition, problem 2.31 pg 60 and modified the calculation for your conditions.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[LennyT]

Yea because what we are ultimatly trying to do is eliminate the wrench and use a cam lock. Kind of like De-Sta-Co clamp or toggle clamp.

 

[Cockroach]

Well, just to give you an idea of forces generated, there is a problem in Elements of Strength of Materials, Timoshenko & MacCullough, 2nd Edition (1935) similar to yours.

A 3/8 NC bolt is snug fitted with a mating nut, then the nut tightened one quarter turn. The calculation asks for bolt load as a result of tightening. Working through the mathematics, statically indeterminate problem, you get 7588 N or 1705 lbf, the listed answer.

Our problem is similar except we apply a torsion. Making the appropriate allowances, we get 5330 lbf. Again, no consideration is given to friction which I can't see being part of the problem except as Pud correctly points out, resistance to nut turning or advancement of the nut on the pin, but we have no information on that. Our answer is only three times as high than a problem of similar characteristics.

The big question then is as stated, would 20 ft lbf generate a nut rotation greater than one quarter turn? Probably. Then I would expect a bolt load greater than a typical example pulled from a textbook.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[LennyT]

Well with a load that high I can see that using a cam lock with an 8" lever on it would be almost impossible to create that force by a normal human.
...hmmm.

 

[Pud]

Hi Folks,

Another thought....

If you have access to both ends of the studding, you could actually measure the extension of the mandrel. Knowing the material specs, then actual applied tension should be readily defineable. In fact, even if you only have acess to one end you could do it - first in "free" (fingertight) state then after tensioning - assuming other end is fixed, that is.

Rgds

Harry

 

[LennyT]

Good point Harry, I was thinking that myself after looking at some of these examples in my old text books. I can measure the rod at 0 torque or pre-torque behind the nut and then apply the 20-25 ft-lbs. Then a second measurment which should give me a decent indicator of how much displacment. This could be a good starting point because I can't figure it any other way.

 

[LennyT]

Ok, I just did some tests this morning and found that at 20 ft-lbs of torque resulted in .060" elongation and 25 ft-lbs was .070".