PDA

View Full Version : Is there any sort of factor that relates the strength of rolled vs. cut threads?



abn
08-05-2005, 01:34 AM
I've read a bit about fasteners, and I've seen tensile stress area formulas and such, but never really a term that factors in rolling versus cutting. Is it pretty much trial and error depending on material? Or is there an accepted practice or good reference that you can refer me to that compares rolled vs. cut thread strengths in common materials?

Forrest Addy
08-05-2005, 02:22 AM
"Much" is not an adjective I'd use as a modifier for "stronger" when comparing rolled Vs cut threads. They may be a bit stronger on the average but that's not the point.

I've heard mechanics of many year's experience who, falling hook line and sinker for hot-rod magazine propaganda, declare that "rolled threads are lots stronger." When challenged to define "lots" with numbers or back up their claim with actual studies they gulp like fish because they never stopped to seriously think about it. They simply accepted the assertions and the words of a flunked engineer student staff writer working by the word in a high pressure publication office where deadlines met are far more important than the technicial competency of the prose printed.

The relative strength of rolled threads Vs cut threads is not so much a matter of raw strength as statistical reliability. Most high confidence fastener specs specify rolled threads - for that matter "chipless" machining - to obtain the advantages of cold flow, compressive surface scresses, superior finishes, high product consistancy, and low unit cost compared to a conventional screw machine product.

Modern fasteners and many specialty production parts start as a blob of raw metal mashed into shape in a series of processes that ends with rolling the threads followed by heat threament and surface prep. No chips are made thus it's "chipless machining." 100% of the weight of the raw stock is used to make the finished part. The result is millions of low-cost fasteners whose mechanical properties can be guaranteed to fall within narrowly defined limits and no chip handling and recycling of byproducts are necessary.

Break a zillion rolled-theaded fasteners and their bell curve of failure distribution is narrow and tightly clustered; that is they are very consistant in quality. Fatigue a sample threaded connection and the chipless machine product usually out-perfroms its cut-threaded counterpart.

So the advantage of a chipless machined fastener is not so much strength but cost and uniformity of product. If you're betting a $120M airplane on tens of thousands of mechanical fasteners cost and reliability are of primary importance.

Rolled thread surfaces tend to be smoother and more wear resistant. If a motion producing screw or worm is intended to operate for millions of cycles (in a copy machine application for example) then a rolled thread is probably a better choice than cut.

Again this is an engineering design consideration where many more factors than raw strength are considered and cost of manufacture and longevity of service are carefully balanced.

[This message has been edited by Forrest Addy (edited 08-05-2005).]

SGW
08-05-2005, 07:18 AM
Since, if thread engagement is more than about 1 1/2 (?) bolt diameters the bolt will break before the threads will strip, anyway, simply from a strength standpoint it's largely irrelevant if rolled threads are stronger or not. They aren't what is going to fail.

ahidley
08-05-2005, 07:39 AM
A good example are spokes on a motorcycle. They are rolled.

An analogy is take a tree and put it in a lathe and cut threads around it vs wraping a cable around in a sprial and waiting several years for the tree to "grow" its own threads. The "grain" is not cut in the rolled version and thus more durable. I would prefer that in a spoke for sure.

Sorry for babbling

MechHead
08-05-2005, 07:52 AM
From an engineering/ machine design standpoint, I never account for the better strength and fatigue characteristics of rolled (or forged) threads; simply because if the machine fails, I cannot guarantee that the part being replaced is not going to be machined. Also, looking at the statistical failure models as already mentioned, you will notice that the bell curves of both the rolled and cut threads overlap on the lower end of the rolled threads and the higher end of the cut threads; what this means is that a poor rolled thread (statistically) will fail before a properly made cut thread. This is standard practice among mechanical designers/ engineers.

J Tiers
08-05-2005, 08:26 AM
<font face="Verdana, Arial" size="2">Originally posted by SGW:
Since, if thread engagement is more than about 1 1/2 (?) bolt diameters the bolt will break before the threads will strip, anyway, simply from a strength standpoint it's largely irrelevant if rolled threads are stronger or not. They aren't what is going to fail.

</font>

Does not the bolt typically break through the stress riser formed by the thread?

Is not a rolled thread (almost forged, btw) more likely to have a good form free of sharp corners than a thread cut with a nice sharp new die or cutter?

MechHead
08-05-2005, 09:14 AM
It's all about the grain structure in the steel. Going back to the tree analogy from before, with cut threads, the grain is interrupted, just as if you were notching a 2 x 10 for stair treads. With rolled or forged threads, the grain flows around the sharp corners, radii, etc. This is the basic principle on why forged parts are better for fatigue strength than others (eg. crankshafts).

Mcgyver
08-05-2005, 09:17 AM
SGW I don't think its so much a matter of the threads stripping thats the issue, its more the stress raiser present in cut thread as MechHead and JTiers said. There's a second distribution need here, and that's then when they break. its my understanding that the rolled thread is less subject to fatigue

[This message has been edited by Mcgyver (edited 08-05-2005).]

MechHead
08-05-2005, 09:33 AM
Yes, anytime you have a geometry change in a part with stress flowing through it, it creates a stress concentration. That is why we try to design mechanisms and devices that avoid bending stresses in threaded parts.

sauer38h
08-05-2005, 09:58 AM
<font face="Verdana, Arial" size="2">Originally posted by abn:
[BOr is there an accepted practice or good reference that you can refer me to that compares rolled vs. cut thread strengths in common materials? [/B]</font>

In thirty years in the business I've never seen such a reference, even in critical applications like pressure vessels or aircraft/aerospace. This is not to say that there isn't such a reference somewhere, but the people who actually do this stuff for a living aren't much concerned with it.

All threads are stronger than they theoretically should be. The standard thread forms certainly look like they're designed to fail, especially in fatigue - after all, a thread is just one long stress raiser wrapped around an axis. But in practice, they're not so bad. There are thread forms with modified root profiles (larger radii) intended for increased fatigue life in aircraft applications, but I've never seen them actually specified for any application.

The usual explanation for the unexpectedly good performance of threaded fasteners is some sort of hand-waving about work hardening, blah blah. That remains unconvincing as work hardening is dependent on more factors than the simple fact that a thread has been formed in the part. These hand-waving exercises generally come down in favor of rolled threads. I lean toward rolled threads myself, but not because of any actual advantage - it's because the machines which do the rolling are so "totally rad" - although I've encountered threaded parts with huge errors in lead (ie, tpi), which I assumed were due to improper setup of the rolling machinery.

The most common failure mode of threaded fasteners is torsional yield when installing. This has nothing to do with the thread profile, or how the threads were formed.

Oh, and we don't use the phrase "trial and error." Something like "empirical data" is much preferred. Everything in materials and machine elements has been tested to death, starting with Whitworth threads in the days when steam locomotives were the acme of high tech. Look up Hertz stresses in spheres if you want to see structural calculation and measurement run amok.


[This message has been edited by sauer38h (edited 08-05-2005).]

precisionworks
08-05-2005, 11:57 AM
The weakest area of any threaded fastener is the minor diameter of the threads. Since the thread dimensions of a cut thread and rolled thread fastener are identical, there is absolutely no difference in strength.

One could argue that the work hardening which occurs during the roll threading process may even make the fastener stronger. Also, cut threading interrupts the natural grain structure of the round bar whereas roll threading reforms it.

The exception to this is on wooden threads, where cut threads outperform rolled threads every time &lt;LOL&gt;.

Carl
08-05-2005, 12:19 PM
Seems to me most if not all threads on steam engine staybolts were cut threads, yet almost all boiler explosions were a result of human error (i.e. low water level) and not a result of thread failure. I don't think I can recall ever hearing of a well maintained boiler exploding due to staybolt thread failure (staybolt failure due to lack of maintenance is another matter all together).

Paul Alciatore
08-05-2005, 12:29 PM
Precision said that it is the core of the screw, below the minor diameter that ususlly fails, not the threads stripping off. Usually. But I think he is incorrect in saying that there is no difference because the minor diameter is the same.

As explained by ahidley and MechHead, the rolling process produces a grain srtucture that tends to follow the outline of the thread. The grain is compressed at the root of the thread form and is therefore stronger there than a cut thread root which is simply the original grain srtucture.

Also the cutting process can, no make that WILL, leave small grooves in the surface. In the root of the thread, these are additional stress risers (in addition to the basic geometry of the thread's Vee. So you have a double dose of stress amplification there. Like scratches in a sheet of glass, that's where the breaks always start.

Is it a big difference? I have never seen actual figures but I doube it is 100% or even 50% better. More likely 10% to 25%. Perhaps only 5%. I suspect it will be more of a factor in an application that is subject to constant vibration (airplanes, heavy machinery, auto, etc.)

Paul A.

Kdahm
08-05-2005, 01:57 PM
Speaking as a non-structural civil engineer here (Which means I don't know squat)...

Any design, where the difference in strength between a rolled thread and a cut thread is critical to the success of the mechanism, has an insufficient factor of safety. Any failures in other areas such as loose or misplaced bolts, small material flaws in the bolts, variability of the materials used, or incorrect guesses about the loads applied will result in the properly installed fasteners failing. The difference is small enough that adding an extra bolt or increasing the size would normally have a much larger factor of safety.

That said, there are some applications where the rolled thread is sufficiently stronger that it makes sense to specify them. Cheap insurance, and good engineers are belt and suspenders types anyway. Design using cut threads, specify rolled threads, and things work out.

****************************
Whats the difference between mechanical engineers and civil engineers?

When something in a mechanical engineering design moves, it's working.
When something in a civil engineering design moves, it's broken.

Karl



------------------
At a certain point in the course of any project, it comes time to shoot the engineers and build the damn thing.

MechHead
08-05-2005, 03:17 PM
As to designing with threads either rolled or cut: There is a difference, I NEVER design with specs for rolled threads, and yes, that is written in stone. If anybody does, and puts a product out on the market relying on that, they are asking for it. You assume the weaker breaking strength of the cut threads, add an extra bolt(s) if you need them, go home, and have a nice, sound sleep.

Your Old Dog
08-05-2005, 08:50 PM
<font face="Verdana, Arial" size="2">Originally posted by MechHead:
Yes, anytime you have a geometry change in a part with stress flowing through it, it creates a stress concentration. That is why we try to design mechanisms and devices that avoid bending stresses in threaded parts.</font>

If I had to guess I'd say a rolled bolt done on well seasoned tooling would likely be stronger than one on a fresh tooling unless they take the time to polish out the tooling.

That's why they polish expensive wrenchs. Not to make them more slippery in oily handds. Polished parts, parts sans striations, can tolerate more stress then non-polished out pieces. Want to make metal take more punishment before failing? Polish it. A rolled thread is more like a polished surface then a cut thread.

Any bolt I've seen fail has always been at the minor diameter. But both rolled and cut bolts have minor diameters and I would expect them both to fail at the same place. But the rolled should take more punishment before failing.

That hurt. I got a headache http://bbs.homeshopmachinist.net//biggrin.gif



[This message has been edited by Your Old Dog (edited 08-05-2005).]

wierdscience
08-05-2005, 09:12 PM
Machinery's handbook 15th edition states that rolled threads are 10-20% stronger than cut threads and may test much higher when tested for fatique resistance.

jburstein
08-05-2005, 09:13 PM
"Whats the difference between mechanical engineers and civil engineers?"

MechE's build weapons, CivE's build targets.

-Justin

Your Old Dog
08-05-2005, 09:16 PM
jburstein

Nice http://bbs.homeshopmachinist.net//biggrin.gif

Paul Alciatore
08-06-2005, 11:43 AM
<font face="Verdana, Arial" size="2">Originally posted by wierdscience:
Machinery's handbook 15th edition states that rolled threads are 10-20% stronger than cut threads and may test much higher when tested for fatique resistance.</font>

The 26th Edition essentially says the same thing. So there is the factor. It also says that rolling increases the surface hardness and thereby the wear resistance. It says many micrometer threads are rolled.

I guess my gut guess above was close.

Paul A.