Calculating Pressure Applied via Threads 3

[gbednar]

We want to calculate the pressure we apply to a piece of tube via a set of squeeze rolls. The rolls are in a vise like stand. Question; If I know the amount of pounds being applied to the the threaded rod (via torque wrench) can I then calculate the force being applied to the squeeze rolls? If I need to be clearer please let me know what info you would need.

 

[desertfox]

Hi gbednar

If the force applied to the threaded rod doesen't get magnified in any way and is the force directly applied to the squeeze rollers then yes that is the force crushing the tube.
The difficulty I see in calculating the pressure on the tube being squeezed is the contact area between that and the squeeze rollers.

regards

desertfox

 

[gbednar]

The rolls have a offset radius of 1.500" and the tube is 1.000" We can squeeze them until there is about 0.06" between the rolls and the tube would measure 0.625 across (shaped like a oval.) On the vise, both jaws move in and out via left and right threads on a single shaft. We are wanting to squeeze the tube until the weld breaks and record the reading from the torque wrench. Hopefully there is a way to calculate the actual pressure applied to the tube. example; 200ft pounds breaks the weld and the thread is 3/4"-16. The tube runs throught the rolls (with bearings in them) approxinmately 4" above the center line of the threaded adjusting shaft. Any ideas??

 

[desertfox]

Hi gbednar

With a 4" offset between threads and centre of the tube you will encounter a bending moment and without seeing the set up I cannot comment on how rigid these squeezing rolls are.
However now you mention crushing the tube till the weld fails, have you considered calculating the strength of the weld on the tube and thus arriving at a theoretical pressure that the weld will fail at.
I can help you along those lines however I would need the type of weld, size and wall thickness of tube and material you are using, in addition how is the weld loaded in service.
Once you have a theoretical figure you can still test the weld the way you have been doing and compare results.

regards desertfox

 

[EdDanzer]

The type of thread (60° V, acme, or square) will have an effect on calculating the force generated by the screw. Friction is the other unknown, and will vary by type and quantity of lubricant. For the theoretical force use a wedge calculation with the helix angle as the wedge angle. The Machinery’s Handbook has a picture and formula to do this. You could put a load cell under the screw to get an accurate measurement if this is an important value to know, otherwise comparative numbers might work one a base line is established.

 

[gbednar]

Desertfox,

The tube is 1.000" in diameter x .110 gage and the material is c1018 Steel. The weld is actually a "forging" We use induction to heat the strip edges then they are pressed together with a 3 roll weld box. The ID and OD beads are removed. This is all done continuously on a tube mill. The weld is located at 0 degrees based on the squeeze rolls being at 90 and 270.

 

[desertfox]

Hi gbednar

I am looking at some formula which may help you with your problem, however the formula I have is for tensile loading of thick and thin rings as your compressing the tube in two places then the mode of failure will be buckling of the tube
which will occur at a lower value than these formula will probably predict.
From your last description it seems that your producing tubing with a seam which if I am correct there should be standards for testing this product?
Can you confirm how this tube will be loaded in service, because if this tube is subjected to either internal or external uniform pressure then the calculation becomes a lot easier and your current method of testing by crushing at two diametrically opposite points is not representative of service conditions.

regards

desertfox

 

[Iskit4iam]

There are way too many variables in the screw equations to do this. You need to calibrate the device. Put a load cell in the device and get a force/torque curve for your fixture.

As an alternative you might consider measuring strain of the sample and calculating the resulting stress at failure. Those equations would be use fewer uncontroled variables.