ASTM A53 (plain old black pipe) does specify yield and tensile strength. The AISC allows its use, the most common use is naturally as columns. According to my 1973 Steel Construction Manual Fy = 36ksi, the same as A36 structural steel.
Yup, this was my response to one of Stu's earlier posts,
"Assuming a stout flange, my SWAG is that failure would occur at the root of the threads near the fitting, where the wall of the pipe is thin due to the threads."
Ned Simmons
No SWAG needed, this is the point of highest stress concentration - threaded fasteners invariably fail at the root of the thread, exactly even with the surface of the female threaded part.
The only question would be, how badly would the stress concentration reduce the apparent ultimate strength of the material? For a sharp-V NPT thread, done with a die head in black iron, my own SWAG would be nearly a factor of three.
That is, to figure out the strength of the pipe under tension, simply look at the cross section area of the pipe, calculate based on the ultimate psi strength for black iron, and then divide by three. Ed's comments about the joint failing by collapsing under hoop stress not withstanding.
My guess is the pipe will break before it collapses inwards. Another SWAG.
Jim
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================
What are you talking about?
Stu
Why not just make the same setup with a short piece of pipe and test in in a press to see what it will break at?
jim rozen jim snipped-for-privacy@newsguy.com
I'm no engineer, and only an amateur machinist. My problem with this is, if there is a taper-thread-induced concern with "pipe" thread, why not disclaim its use for vertical holding whatsoever, rather than go into guesses as to the possibilities? FM
Often amateur machinists a) have a better intuitive grasp on the issues than 'real' engineers, b) know enough to look stuff up when they know they *don't* know the answer, and c) have open minds.
That is pretty much what others have said on this thread before. Pipe threads are not spec'd for tensile strength but facts are this widget is already built. My two suggestions (calculate based on the minimum wall thickness, then divide by three to account for stress concentration, and the other was to simply make up a joint like that, and stress to failure) are only to encourage the fabricator to think a bit more about his machine.
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================
The argument that tapered pipe threads are not intended to carry a tension or compression load is nonsense. (I don't mean to imply you said this, Jim.) If there is pressure in the pipe, then there is an axial force on the joint equal to the pressure x area of the pipe ID.
Ned Simmons
And that may run a couple of hundred psi. Those loads on plumbing pipe from internal pressure are nowhere near the strength expected of mechanical tubing in tension or compression, Ned.
Ed Huntress
Which gives us the lower bound on the axial strength of the threads, but says nothing about the upper bound. Much higher pressures are acceptable with forged fittings and heavy wall pipe. 3000 psi fittings are readily available up to 4" NPT(F), 6000 psi up to at least 2".
I don't see how that's relevant. The question is not what the optimum solution is, but rather, can pipe threads carry significant axial loads. Clearly they can.
Ned Simmons
I will again interject that the real question is, how close to the line is this widget? Black iron hardware store pipe with sharp-V NPT threads, 3/4 inch IIRC.
Best way to figure that is to a) use the minimum minor diameter for that NPT thread and calculate the X-section area, then divide the ultimate number by about three.
Or b) purchase the same fittings from home desperate and load to failure.
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================
Then the job is to quantify strength of a joint that was never designed for much strength. Pipe threads are designed to seal, not to exploit the strength of the material. Straight threads are.
As I said, it may well do the job here. But don't expect it to perform like a straight thread, for the reasons given by several people here. And don't expect to find a spec on it. There may be such a thing, but it's largely incidental if it is.
Ed Huntress
Oh, come on, Ed. What's so different about pipe threads that they don't "exploit the strength of the material"? The sharp thread form likely causes some stress concentrations, but other than that, I don't see much difference.
Sure, they have to seal, but there's gonna be no seal if the pipe is ejected from the fitting with 6000 psi behind it. That doesn't happen in a properly made up joint.
Of course it's not going to behave *exactly* like a straight thread. Why is that a problem? Perhaps the analysis is a bit more involved than for a straight thread on a solid member(though other than taking into account the thin wall at the root of the thread, and the effects of the sharp vee, I'm not convinced of that), but if all the components of every new design had to be used in a mannner that was anticipated and pre-calculated we wouldn't see much progress, would we?
As far as finding a spec, I expect you're right, and I said as much in a post to the OP back on 10/2. My objection is to the several (unsubstantiated IMO) assertions that pipe threads are unsuitable for carrying axial loads.
Ned Simmons
Not knowing all the details of the device, I was reluctant to make a calculation before. Since it's now built, they can't blame us for any mishaps, can they?
I get about .925 dia at the root of the thread at the gage line of a 3/4 NPT fitting. ID is .824 . Yield on A53B pipe is 35 ksi min. Ultimate tensile is 60 ksi min.
.138 in^2 x 35000 lb/in^2 = 4830 lb to yield .138 in^2 x 60000 lb/in^2 = 8280 lb to ultimate tensile
I assume your 3X multiplier is a stress concentration factor, which sounds reasonable to me, so there will be some yielding at the root above 4830/3 = 1610 lb., and failure will occur at 8280/3 = 2760 lb. There's likely a pretty high uncertainty in the second number depending on exactly what happens at the root as it yields.
Ned Simmons
I have not run those numbers but they look good to me.
One point is, my 3X is a stress concentration factor, for sharp V threads that are die cut. I made that up off the top of my head.
Another consideration is, the ultimate 2700 that you suggest is for a purely *static* load. It might be possible to get shock loads that produce what, five or ten Gs?
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================
And therein lies one problem. NPT threads want to get as close to a zero clearance fit-up as possible, so the roots of the male threads are machined sharp V. This provides a stress concentration that would be atrocious in a fastener. Bolts typically have rolled, not cut threads, and in the higher strength grades have radii at the roots.
Still Ned's analysis shows it will probably hold up, as long as nobody jumps on it or anything.
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================
Not that they can't, but rather they are supremely unsuited to it. The sharp V thread form provides on problem, and the fact that NPT threads are invariably die-cut is another.
Die cut threads are rougher and this roughness translates into microscopic tears in the thread - surface defects that provide stress concentrations. That and the sharp V form can weaken the thread by large factors compared with properly rolled threads.
This is why no pipe manufacturer would ever suggest his fittings be used for axial loads, and would never supply a spec for that purpose.
Sure for a railing or something it would work fine, and your own numbers suggest that his lift will not fail unless subjected to a shock loading of five or ten gs.
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================
It isn't the sharp threads, Ned. It's the fact that the walls are weak at each end of the taper, because one side or the other is thin at either end. Only a couple of threads in the middle of the joint can produce the full strength you would get with straight threads, and that isn't enough to exploit the strength of the material. There aren't enough threads where the material on *both* elements is equally thick.
You probably could approximate the strength if you calculated the strength of three straight threads, more or less, of the same size in the same material.
Ed Huntress
Just a minor nit... Wouldn't the strength be somewhat dependent on how tightly the joint was made up? It seems to me that the tighter you made the joint, the more load you'd put on the material and that would have to come out of your strength budget somehow. Not that I have a clue how.
I still maintain that it's a bad application. I think that any flexing could seriously weaken the joint over time.
Why three? The best way would be to use the smallest wall section. And the sharp threads do indeed reduce the strength.
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================
They're *tapered*. That means you don't have full depth threads engaging full depth threads at any point in the joint. Machinery's Handbook says that only NPS threads should be used for axial loading.
Gary
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.