Withdrawal Capacity of Threaded Rod in Tapped Steel Plate

[NickRomano]

Hi,

I would like to calculate the withdrawal capacity of a 1" threaded rod, 36 ksi, threaded into a 1" thick, 36 ksi, steel plate which has been tapped on site to receive the threaded rod.

Thanks in advance for any guidance in this topic.

 

[boo1]

The critical areas of stress of mating screw threads are:
The effective cross section area, or tensile area, of the external thread.
The shear area of the external thread which depends upon minor diameter of the tapped hole
The shear area of the internal thread which depends on the major diameter of the external thread

Root Area of Thread (in2) = Pi*r**2 = .605 in**2 (assume 8 threads/inch)
Rod fails before threads fails
Most common Specification for the rods are :A36, ASTM A193 (125ksi), and ASTM A307 (60ksi)

 

[racookpe1978]

You're assuming the rod is being pulled only - right?

A single (non-vibrating, non-rotating) steady pull directly perpendicular to the surface plane of the plate with no bending or sideways forces?

Also, remember not to use the full length of the tapped hole: you need to reject (not use) the first thread on the near side (because it has only half-developed thread walls) of the tapped hole and the last thread on the far side (because it too has only half-developed thread wall sides). But a common 1 inch dia nut is thinner than a 1 inch thick plate, so the pull-out theory above remains true. The male threads (the rod) strips before the female does.

 

[NickRomano]

Thanks for your response.

Out of curiosity, what would be the minimum length of thread engagement for the rod to fail before the threads fail? Is there a standard engagement length that I could use as a rule of thumb?

Thanks

 

[jayrod12]

That depends on the strength of the material it is threaded into and the strength of the rod itself. There's no hard and fast rule. That being said, the nuts are designed such that the rod will fail before the nuts will, so take that as you may.

 

[boo1]

For steel a length of thread engagement of 1 x Nominal dia's of the thread

 

[JLNJ]

Odds are that the plate will deform or pull away from whatever it is connected to before the rod pulls through the threads. Especially if the rod is of high strength material.

 

[racookpe1978]

Well, the original statement did call for a 36 ksi plate and an equal 36 ksi steel rod.

 

[UcfSE]

You can look in a mechanical engineering reference for thread stripping strength. This strength will be based on the materials involved, and the thread engagement. Knowing your loads, you'll be able to figure out how much thread engagement you need. Check also the connected material per boo1. You need to know the geometric properties of your threads, such as UNC or UNF.

 

[oldestguy]

Why not run a test? Any materials testing lab can do a test and then you will know. It should be quite cheap to do. You can play with partial penetration also. As a starter use a bolt and a nut. For checking the thread depth variations, thread a rod in a lathe and over penetrate the tool. Also cheap to do. With partial thread depth engagement there may be bending of threads and no shearing as such.

 

[rb1957]

"Root Area of Thread (in2) = Pi*r**2 = .605 in**2 (assume 8 threads/inch)
Rod fails before threads fails"

maybe i'm especially (or perhaps just normally) stupid today, but i don't see this, at least not as presented.

ok, so the rod has 0.6in2 working in tension
and the plate has something like pi*d*t = about 3in2 working in shear

same material rod and plate ...
so the rod can support 0.6Ftu
and the plate 3*0.57 = 1.7Ftu

so, yes, the rod is more critical than the plate

Quando Omni Flunkus Moritati

 

[racookpe1978]

But you are assuming the "rod" is failing across a circular cross-section as if the rod were being pulled through. Instead, it is the root of the spiraling thread that fails

 

[rb1957]

yes, i've always considered the tension area of the bolt as a disc, diameter = thread root

Quando Omni Flunkus Moritati

 

[oldestguy]

When I strip a bolt in a nut, the resulting appearance of the bolt is not a nice clean cylinder, sheared along the roots, but with the threads that are still quite visible, almost usable. That's why, do a pull-out test and then come up with a formula or explanation for that failure.

I'll bet some will say that two structural plates riveted together develop their usable shear resistance by the cross section area of the rivets. Wrong. Its the friction between the plates. Probably applies to bolted joints also. Not my theory: Prof George Winter Cornell U 1950 structural Dept Head.

 

[Agent666]

As a conservative guess if you know the strength of the normal nut material used and plate strength and nut thickness, then you could multiply nut thickness by the ratio of nut strength divided by plate strength. Just use either ultimate tensile strength or yield strength ratio (whichever is greater).

In my mind this would achieve the same intent as the usual high strength nut provided the bolt is threaded right through the plate. I.E. that the bolt yields first.

 

[Agent666]

[link]http://www.scnz.org/content/steel-advisor/docs/ERC1003.pdf[/url]


I randomly came across this today which is relevant to the discussion, there are some equations for the length of engagement. Hopefully it helps.

 

[racookpe1978]

11 Apr 14 20:09
When I strip a bolt in a nut, the resulting appearance of the bolt is not a nice clean cylinder, sheared along the roots, but with the threads that are still quite visible, almost usable. That's why, do a pull-out test and then come up with a formula or explanation for that failure.

But the reason that happens - that pattern of stripping out threads in the threaded rod that you describe - is NOT because the formula assuming equal root lengths and equal material strengths is correct, but because the threads are NOT "perfect 100% engagement-100% cut to diameter" threads. For tapping, the tapped female threads are first drilled over-sized so the diameter of the drilled hole is greater than the theoretical diameter of the minor diameter of the thread. (As I recall, 68% - 75% thread engagement.) This is to allow room while tapping for the shavings to get removed back into the "curl" of the tap. If it is not done, the tap tends to break off in the drilled hole.

Thus, the "real world" thread in 98% of the tapped holes is "loose" and only the tip of the male threads on the threaded rod or bolt actually engage the female threads. Since the male bolt threads are pointed, this means that the root of the male threads where they are actually engaged is significantly less than the root of of the female threads in the tapped hole.

So, combine a smaller male thread diameter with a smaller root length on the male threads means that the force created inside the bolted joint is divided equally in half: half into the small cross-section in the male threads, and half in the larger cross-section area of the female threads. Obviously, the stress in the male threads are closer to the material yield point stress, and the male threads will always strip first.

But, they will strip the at the "tips" of the male thread, not across the (unengaged!) root of the male threads.

Now, if you deliberately call for a designed joint with a "perfect" very-closely-fitting thread specification, you are increasing costs significantly (remember that greatly increased probability the tap will break off?) and slow productivity down by requiring much higher tool costs, but you are also increasing the chances that the stripped threads and galled threads on disassembbly will be in the very expensive female threads!