Results 1 to 20 of 43
Thread: Thread pull out load charts
11-06-2011, 06:09 PM #1
Thread pull out load charts
Its easy to find bolt sheer and pull out data, but I am having a hell of a time finding one for female threads in various materials (steel, Al, etc.)
Basically I am looking for the difference in pull out for a 1/2-13 in 1/2" steel plate *against* 1/2-13 in 3/4" steel plate. I am basing both on a 80% thread.
Anyone know of any tables or charts I can use for this?
11-06-2011, 06:42 PM #2
Both of those examples will fail under tension at exactly the same load.
Once you go above one or two threads in the female thread, the male fastener
is the limiting part. If you were in, say, nylon, then the numbers would be different.
11-06-2011, 06:43 PM #3
A nut doesn't have a sheer strength nor a pull out in the way you probably mean it. An internal thread will strip if too much torque is applied and it will also depend on thread length. The two lengths you give are 1/2" and 3/4". My opinion is that the bolt will as good as always be "the weakest link in the chain" - assuming of course both are within normal tolerances.
This guy might be able to help. Bolt Science Web Site
11-06-2011, 06:45 PM #4
11-06-2011, 07:56 PM #5
11-06-2011, 09:33 PM #6
"pullout strength", which may be greater than tensile strength of the male fastener, will theoretically be pi x PD x % thread x length of engagement x shear strength of material, for a vee thread..
But it will vary a lot with cut vs rolled (formed) thread, grain orientation of material, and Young's modulus of male threaded part. Threads will wear and become weaker if the assembly is disassembled/reassembled/overtorqued/corroded.
If it is important, test, and choose suitable safety factor, and test again after ANY change in process or material.
Mark McGrath liked this post
11-07-2011, 12:54 AM #7
If it is important, test, and choose suitable safety factor, and test again after ANY change in process or material.[/QUOTE]
I'll certainly +1 that. Test and you'll know. Theory is great as long as you remember that in theory the Titanic couldn't sink.
brewertr liked this post
11-07-2011, 02:31 AM #8
Well these little added details can be important. You're right. You probably won't find a table that gives you reliable data for this specific case.
Assuming the tapped threads are champhered to the major diameter and the steel is 36,000 PSI yield in shear strength: work out the effective shear area and do the math.
I'm gonna take a WAG and call it about 20,000 lb per bolt. Tensile stength of a 1/2 - 13 Grade 8 bolt is listed as 21,300 ln. The plate MAY be more likely to fail but it's really too close to call. Materials have a range of speciified properties and when they are this close there is no telling which will fail first. The failure mode of a Grade 8 bolt in your tapped hole in A36 plate may be pull-out it may be a tensile neck-down to failure but it will be close.
The best an analysis will give you is an estimate. If hard data is needed you have to conduct physical tests to failure of samples from the actual plate and fasteners under consideration.
11-07-2011, 02:55 AM #9
I have heard this "two threads are as strong as the bolt" platitude for all these years and my personal experience doesn't support it. For a concrete example, one time I mounted a bracket on the end of a hydraulic cylinder with the typical four rods through end caps and nuts. As I recall, the threads were 1/2-13 and the nuts were typical for that size, so there would have been six or seven threads engaged, maybe less. I would have to drive across town to check it, but I don't think the bracket was more than 3/16 thick so there would still probably be four threads engaged. It is on a molding machine clamp cylinder that is cycled, but within the pressure rating of the cylinder. After a few days running, it stripped the nuts out. The two threads line also neglects to specify whether they are coarse or fine threads, which obviously makes a difference. I also hear that after a couple of threads, clearance prevents loading of the threads beyond a certain point. Once you pass the yield point of the first threads, load will be transferred to ones farther on while the first ones will be contributing close to yield force if they are only deformed slightly. I know not what course others may take, but as for me, I will continue to look for at least a bolt diameter of engagement length.
11-07-2011, 04:14 AM #10
Take a standard bolt and the standard nut for the bolt. Measure the length/height of the nut. Anything less will be weaker than "recommended".
Making the nut (or plate in this case) longer (or the plate thicker) only means the chance of the bolt being the first to give up is definately greatest.
11-07-2011, 07:06 AM #11
This become a little more complicated when the bolt and thread part are made of very different materials. For example - steel bolt in 6061 aluminum threads. Would be nice to have some sort of a chart for ballpark figures.
Would it be safe to just compare shear strength steel vs alum and multiply steel nut thread length by that factor?
Thread length in alum = strength of steel/strength of alum x length of steel nut ????
11-07-2011, 07:56 AM #12
Do some math using the 2 threads idea and I think you will soon find that it's all BS
Do the same math using 1.5 times the bolt dia in thread engagement and you will see that it is the rule of thumb...
So, your 1/2-13 bolt developes max strength witn 3/4 inch of threads holding.
Then there's the thread length I've seen on Straight Side Press tie rod instalations that blow the 1 1/2 times dia all out the window...
And to the original poster... Why are you not looking into the Machinery's Handbook for the table of safe loads for threads?... it'l save you a lot or hassle.... or are you trying to pull the threads out?...a drill will do that a lot faster.
11-07-2011, 08:26 AM #13
Go read the last 2 pages here (or google it for dozens of more examples):
The first thread in their 7/8-9 grade 8 example carries 34% of the load even though w/ simple math w/ 6 threads would predict only 16%. Thats a 2:1 error. 1.5x engagement w/ their bolt would have resulted in 12 threads and a would be ~4:1 overestimation error.
The best point to remember out of the paper:
"Caution: It appears that one could theoretically increase the thread strength by increasing the length of engagement. However, as illustrated in the Load Distribution chart above, the first thread will be taking the majority of the applied load. For carbon steel fasteners (including tapped holes) the length of engagement would be limited to approximately one nominal diameter (approximately 1-1/2 times the diameter for aluminum). After that, there is no appreciable increase in strength. Once the applied load has exceeded the first thread’s capacity, it will fail and subsequently cause the remaining threads to fail in succession."
11-07-2011, 10:52 AM #14
I've seen many threads in tension in FEA models in the past years, and yes the first turn of the thread in full engagement does show higher stress, but each thread in engagement thereafter takes a slightly lessor amount of the load all the way to the last thread in engagement. But the addatives of each turn of thread of engagement aids in increasing the ability of hold loads much more than one turn will by itself. So having said that, the formula as magneticanomaly stated is almost correct.
According to "Shingly's" book for basic shear stress in a thread, or reversed, what load will it handle at it's failure point, it is (pi times basic pitch diameter times number of threads per inch in full engagement)divided by two. This is basic shear area of the thread in engagement. Basic shear of steel is .6 times ultimate yield. This times shear area=the load it will handle at shear with no factor of safety added. Singley's formula is based on a "square" thread form. There are other formulas out there that take in account thread flank angles, coeficient of friction, external thread or internal thread material strength, etc, etc.
Again, I'm not a license engineer, but many come to me for an answer.
11-07-2011, 12:14 PM #15
Real world example of both side of the failure case.
All testing done on an Instron tensile test machine.
5/16-24 thread, bolt threads are rolled, nut threads are tapped.
Nut is ~6 threads deep, a little less than 1 D
Inconel 718 bolt HRC 44
Inconel 718 nut HRC 31
Nut fails at 10,100 Lb load (jumps thread)
Nut is now .022” shorter and .026” larger in diameter.
Bolt is ~0.020” longer
Inconel 718 bolt HRC 44
Inconel 718 nut HRC 44
Nut never fails at up to 13,000 Lb load, but bolt is obviously yielding so test is stopped.
Post test examination shows nut has not deformed at all when measured to
0.0001” accuracy, and threads gauge as new.
Bolt is ~0.100” longer
If I need to hold 8000 Lbs, (which I do) than either nut will work fine.
11-07-2011, 01:13 PM #16
Any chance we can get some more information? Plate and bolt material(s) would certainly help. It might result in somebody a giving a (theoretically) precise answer. If the plate and bolt material are more than just a little different I can't imagine any chart being found. Then it'll be test.
I'm assuming you do want to know for a specific purpose and not just thirsting for knowledge
11-07-2011, 01:40 PM #17
One reason we build machinery out of metal instead of brick or glass or wood, is that metals (generally) are ductile as well as elastic, , and work-harden. That is why long thread engagements are stronger than short, in static tension. The early threads deform, so that threads farther down see stress, too, but the threads near the top are still carrying near their yield stress.
But this is not good design practice, because deformation repeats with each dis/re-assembly cycle, with interference between deformed threads further distorting metal. Pretty soon parts won't go together any more, or you observe, "Threads are "stripped"! When did that happen?" They are often not actually stripped, but just chewed out by portions of the male and female threads no longer having matching forms and pitches.
4GSR must be misquoting Shingly, because shear strength of a thread is independent of pitch. Fine threads IN THEORY are stronger because PD and root area are larger. But manufacturing tolerances, and rust, and wear, are more critical because .001" is a bigger % of engagement on a fine than a coarse thread. I stand by my formula, but I will say that "% engagement is a half-assed proxy for the percentage of the theoretical cylinder that is actually there to carry shear.. And threads can fail in flow, and by bursting of the nut.
That is for-why we TEST and use SAFETY FACTORS. Only to God is engineering an exact science, We down here are just fudging about with rules of thumb.
That bolt-doctor website is very good
We're having fun, but OP still has not said what he is building.
11-07-2011, 04:11 PM #18
For the materials involved here, either thickness will give the same ultimate
strength and the failure will be, in both cases, that the bolt will fail at the
point where it enters the first thead in the plate. That's where the stress
Attempts at calculating the ultimate strength of the bolt will fail because
they will not consider that stress concentration factor which depending
on the type of thread and manufacture, can be factors of nearly two.
If you really care about this, test one to ultimate fail on an instron. Anything
else is simply going to be hand-waving.
Read Caroll Smith's excellent treatise on this: "Nuts, Bolts, Fasteners and Plumbing."
It's a good bible.
11-07-2011, 05:09 PM #19
Reminds me of a sign that is on our vintage Riehle testing machine.
"One test is worth a thousand expert opinions"
11-07-2011, 06:29 PM #20
A36 plate .50 thick /w 1/2-13 tapped (cut) threads /w grade 8 bolt installed
A36 plate .750 thick /w 1/2-13 tapped (cut) threads /w grade 8 bolt installed
this is a straight pull load of 10K with load moments of up to x4 (40K)
There are 4 bolts per assembly with equal pull load.
Think I'm headed back to a destuctive test, already did one, but without this part attached.
Was hoping for an easy answer, but guess there isnt one.