Mechanics Behind Fasteners 8

[Tsiolkovsky]

After reading a brief description on fasteners, their nature of loosening etc etc, its incredible the volume of science behind the simple bolt and nut. I have a few questions regarding this.

On the topic of bolt/nut loosening (of which is a problem on our machinery), I asked myself why cannot castellated nuts together with cotter pins be used wherever possible. After reading up, this configuration is to only be used in "low torque" applications. I ask, why low torque only?

Secondly, I do realize that a big problem is lining up the hole in the bolt so that the cotter pin slips in. But surely this can be corrected by incrementally adding/removing shim washers until a perfect fit is found. However I hear doing this compromises bolt strength as it creates a "weak point" due to numerous shims. TBH, im more concerned about bolt loosening than strength reasons. Also the bolt may stretch in its life or shims compress thus the torque tightness will drop off after time.

Ultimately, how true are the above points and the reasons behind it. I guess castellated nuts cant be a solution or else we would see them ubiquitously on truck hubs, all fasteners etc...

 

[silvermodak]

Nord-Lock is your answer here I'd say.

http://www.nord-lock.com/

 

[Tmoose]

Can you share details of the joints whose bolts are "loosening?" Pictures of failed joint surfaces might be interesting. Details about materials and operating temperatures are important too. If the clamp load is simply inadequate for the applied load, or other features conspire to allow embedment, micromotion of the joint, then fretting and and loss of preload will occur, even if the nut is not allowed to "back off" at all.

I'd be surprised if anything but carefully controlled tightening against gorgeous mating surfaces is used to retain main bearing or con rod caps on Indy or Formula cars. Is your application that severe?

 

[thruthefence]

How about PAL nuts? Thin, stamped, sheetmetal nuts installed on top of a torqued up nut.

What's the science behind them?

 

[flash3780]

When dealing with nuts and bolts, it's all about maintaining the all-important preload that the engineer had dutifully used to calculate the joint separation margin, etc.
Primary bolt retention is always friction, but in any vibratory environment a secondary retention feature is recommended. Multiple methods exist and each has its benefits and drawbacks.

Thread locker (glue) is a pretty good animal, but makes it impossible to go back and check the torques on bolts after assembly.

Locking nuts and bolts are available in many varieties, and work well, but tend to wear out if reused.

Mixed reviews on locking washers; some folks say they work, some say they don't do a thing.

Safety wire and cotter pins do little to maintain preload, but ensure that the bolt will never fall out.

 

[gadkinsj]

The thing that is obvious when you know it and difficult to intuit on your own is that preload can be lost without the nut/bolt ever moving.

When I was a fresh young engineer and the world was still full of unicorns and rainbow's and kittens, I was under the impression that if a bolt is being held in place (like castle nuts, pal nuts, lock-tite, wire ties, nord-lock wasthers, cotter pins, however you please) then it will never loosen up. I mean how could it.

But the earth shakes, machines vibrate, joints heat up and cool down, materials are not perfectly plastic, parts live and breathe, and all these things can induce creep into threads.

 

[Tsiolkovsky]

Okay. I see many, including Tmoose, want to know the real deal behind my scenario.

When I say "machinery", I mean underground mining machines. So yes, the environment is *that* severe. This mining machine has

a standard Berco undercarraige system. i.e. It has the tension wheel, tension cylinder, track set, and finally, "dolly

wheels/or road wheels". In either case, refer to my photograph attachment depicting the seven "road wheels" at the bottom of

the Dozer.

The problem is that the entire machine is undamped so we are talking about a 4 ton machine being tossed and turned like a toy

with the bottom road wheels (more specifically the front road wheel) taking the full impact. Before you know it, at best the

first wheel breaks off at 400 hours and at worst breaks off at as little as 100 hours of operation. The cost of one a piece

is pretty hefty. And that is the problem. Before I move onto my solutions, let me say that by me implementing a solution to

this would be a massive achievement in my junior engineering career. It really is a challenge, and its relevant to my

studies.

Nevertheless, onto my thinking of how to go about this... Any criticism you have, please dont be afraid to hammer home some

points, I will love any help.


I see many problems:

1) Bearing surface embedment. Being in the mines, it is inevatible that dust grit or flakes of paint (from new road wheel)

will be ever present between the mate. So that when you tighten the bolts, the joint will "relax" quickly after that? Apart

from the obvious solution to clean and make sure no grit is on bearing surface, I see no other way around this. And just by

the way, Im not too sure as what literature means by which mating surface. They only seem to mention bolt head- part

interface. What about the part 1, part 2 interface? And the nut-part interface? I will provide a sketch if anybody requires

to understand.

2) Undertorquing. Using simple spanners, how can the technician "know" what is the right tightness. If they undertorque, the

preload will be low and the bolt may loosen quicker. Yes, I know I know, even without the bolt loosening, the preload may

drop but ill get to that now.

3) Overtorquing resulting in tapped hole deformation/stripping. They may at times use a pipe extension and tighten so much

that the thread on the machine strips. By the way, the road wheel bolts are bolted into a tapped hole and therefore do not

have a nut on the other side. If the thread is damaged (deformed or stripped), it may loosen quicker, or it may deform and

thus loosen pre-load without ever loosening.

4) Overtorquing resulting in bolt necking. In a desperate attempt to keep it as tight as possible, the bolt may neck. To be

honest I have not seen or heard of this. I even wonder if this is a possibility because the bolts used are 10.9, and the

thread into which they bolt into (directly into the machine) is mild steel. So the thread will always be the weaker point and

the thread will strip due to overtightening before the bolt necks?

5) My worst fear is the invetable and what maybe gadkinsj was pointing out. The tapped hole into which the bolt goes into,

simply gets deformed over time no matter how the hole is torqued. Due to the severe impacts, the hole plastically deforms

each time. Its mild steel afterall. This worst nightmare of mine may be true. Upon inspection of undercarraige with road

wheels broken off, the holes looked simply terrible. The steel almost looked like chewing gum on the surface. If this is

inevatable, I can surrely to some degree mitigate this?

Ultimately the bolts break due to loss of preload which is a combination of bolt loosening and surrounding deformation.

One of my solutions, and correct me here, is to use a torque wrench and adjust the bolt to a value *JUST* high enough before

the bolt necks OR before the surrounding thread strips or deforms. Like I mentioned before its one of these and most prob the

thread because its mild steel. But what value to use? Well... solely judging on the bolt (M12 and 10.9) the literature says

that a guideline value of 129 N.m ("dry thread") is to be used considering failure to shear and tension combined with a 90%

of the actual yield value (safety factor... if you will). But my problem with this is that I cannot only consider the bolt yeild strength. What about the thread which should give way first??? Infact the bolt science website (very comprehensive by the way, I recommend this site to all of you... its simply biblical in its breakdown of bolt mechanics) for some reason never

mentions how to take thread strength into account. If I can know what the thread can stand, then I know the limiting factor we can then tighten it to that limiting factor. Further, if it is indeed inevitable that even this value loses pre-load after many hours (initial pre-load loss due to embedment), we can pick it back up to that value after, I dont know, say, 50 hours.

It is at this very point that I had a heated debate with one of my colleagues. It was the typical "young engineer" (me)

versus the "old veteran" (colleague) of the company.
I argued my torque wrench idea that it may have *some* help with this and he flat out rejected it. His argument is that there

are too many variables. By this he consistently pointed to the fact that "we cannot know the value" "too much dirt" "what

about deformation over time?" etc etc. Im sure he has a point to a degree. But I find it hard to accept prefering the use of

a spanner over a torque wrench. Atleast with the wrench , you can consistently hit *a* value and not guess. Yes I understand

that thread fricion has play here and grit etc but this can be put under certain control. The colleague mentions that "with

nordlock washers, using a clean thread/surface, using loctite and tightening the hell out of the bolt, the road wheel WILL

NOT COME LOOSE". I cannot beleive this as im sure thats been done and never worked therefore I am to conduct this procedure

and see if it holds out. Ultimately, how true are points 1 through 5? And how does my line of reasoning hold? as you see, its very interesting and stimulating topic

 

[Tsiolkovsky]

Hmmm, so much white space in that post, I assure you I never hit "enter" so many times, so excuse the gaps. Unfortunately I cannot find the edit button. Old school forum layout uh?

 

[Duwe6]

There are too many unknowns in any given bolt-nut-lube combination to calculate it empirically. Get a Skidmore-Wilhelm tension device and 'ask the bolt' what torque is required for a proper preload. Try about 80% of specified min. yield. This procedure is well documented in RCSC's A325/A490 spec, and is available on-line, for free.

"Too much dirt" is a pitiful excuse. The mechanics should be rinsing out the tapped holes with brake cleaner or other clean-drying solvent. If they are gritty after rinse, use a bristle brush to clean the threads. Soap & water will work fine.

Buy the mechanics a hydraulic or air-over-hydraulic powered torque wrench. Easier to use than a spanner with pipe 'cheater', so they will be inclined to use it. THEN, somebody from Eng/QC needs to go witness the torquing. It is silly to expect a worker to perform a hard, exacting task correctly and without shortcuts when nobody considers it important to go witness & inspect the work.

 

[Tmoose]

Did I miss the attachment?

 

[dinjin]

http://www.berco.com/wps/wcm/connect/bercositelibrary/it/home

Is this the manufacturer of the under carriage? I would give them a call. Are the depth of thread penetration at least 16mm deep? There are calculations for designing or ensuring proper thread depths for different materials. I think I would try using 85 percent of proof load to see if you can get better life. 100 hours to failure sounds like an underdesigned parts or extreme abuse of equipment. Are you certain that the bolts are certified and not Chinese copies?

 

[dinjin]

BERCO Deutschland - Process technologyBERCO, manufacturers of undercarriage components for chain driven earth moving machines and member of the ThyssenKrupp AG Group, have invested more than 400 million Euros over the past years in modern process technologies - which has also allowed enhanced production of mining components, products subject to extreme demands.

BERCO operate a total of seven fully automatic production lines in the four Northen Itally plants.

Why not give them a call?

http://www.berco.co.uk/pages/aunder.html

They did have a dealer here in the States and I think it was in the Carolinas. Maybe it was only replacement parts.

 

[Tmoose]

How many hours did the factory built machine survive before the first "loosening" occurred? Was it after some field maintenance was done?

I expect a professional, conservative design would have plenty of thread engagement (over two bolt diameters) and grip length (5 or more diameters), and generous joint footprint". IF the mating faces between bolted components are (still) flat, AND the fasteners bear on hardened washers AND the threades don't bottom in the tapped holes AND sufficient torque is applied to achieve clamping greater than the service loads, then the joint will be dead reliable without fancy ramped washers or desperate measures like safety wire, cotter pins, jam nuts, or tack welded nuts. If any of the four prerequisites are missing then the joint will loosen and cascading damage will occur.

++ How many diameters of engagement do the failing fasteners have? Thread engagement is often described relative to bolt diameter. A 14 mm tapped hole that engages a 28 mm long section of the bolt's thread has "two diameters" of engagement. Generally 1.5 diameters engagement in mild steel allows the full strength of a Grade 8 fastener to be utilized (full torque, etc, and thus clamping). The torque applied to the 7/16" main bearing cap bolts of Many early Chevy V8 engines was over 60 lb-ft into cast iron. Main cap loosening is pretty much unheard of in any service.

Deformed tapped holes suggests significant damage to the mating faces of the components that are supposed to be clamped together. Plus the threads need to be restored with an insert of some kind. "High spots" adjacent to the wallowed out deformed tapped holes will prevent proper clamping and support. Here is a picture of a VW crankshaft damaged when the flywheel worked loose. 4 hardened dowels were no match for the alternating forces the crank and flywheel endure. A typical cause is the large bolt (called a nut) that secures the flywheel to the crank was not tightened correctly during a clutch replacement. A proper repair requires refacing the end of the crank and mating flywheel surface flat and smooth.

I'd say if the mating faces are not flat and smooth, and proper torque, etc is applied, your next assembly is doomed to failure by loosening.

 

[Tmoose]

"Here is a picture of a VW crankshaft damaged when the flywheel worked loose. 4 hardened dowels were no match for the alternating forces the crank and flywheel endure."

http://www.aircooledtech.com/8_doweling_crank/4-dowels2.jpg

Here's a Toyota with loose flywheel damage. The raised metal at the edge of the damaged holes must be removed.

http://farm5.static.flickr.com/4001/4492760185_f7908f0595_o.jpg

 

[Tsiolkovsky]

There are too many unknowns in any given bolt-nut-lube combination to calculate it empirically. Get a Skidmore-Wilhelm tension device and 'ask the bolt' what torque is required for a proper preload. Try about 80% of specified min. yield. This procedure is well documented in RCSC's A325/A490 spec, and is available on-line, for free.

I know what you mean, but before I can even consider tightening that bolt to *its* maximum, what about the chassis maximum strangth? Last time I tightened a 10.9 bolt on another tap (not the ones in question), I felt the threads on the chassis stripping (it started to get easier to tighten the more I tightened). So first I need to know can the hole take it?

"Too much dirt" is a pitiful excuse. The mechanics should be rinsing out the tapped holes with brake cleaner or other clean-drying solvent. If they are gritty after rinse, use a bristle brush to clean the threads. Soap & water will work fine.

Thanks for this. We will definitely make cleaning the threads a bigger priority.

Buy the mechanics a hydraulic or air-over-hydraulic powered torque wrench. Easier to use than a spanner with pipe 'cheater', so they will be inclined to use it. THEN, somebody from Eng/QC needs to go witness the torquing. It is silly to expect a worker to perform a hard, exacting task correctly and without shortcuts when nobody considers it important to go witness & inspect the work.

You mean the guys on site torque it to some low set value and then used a torque wrench once QC comes by?

Did I miss the attachment?

No you never, I forgot to attach it. See attachment on this post.

http://www.berco.com/wps/wcm/connect/bercositelibrary/it/home

Is this the manufacturer of the under carriage? I would give them a call. Are the depth of thread penetration at least 16mm deep? There are calculations for designing or ensuring proper thread depths for different materials. I think I would try using 85 percent of proof load to see if you can get better life. 100 hours to failure sounds like an underdesigned parts or extreme abuse of equipment. Are you certain that the bolts are certified and not Chinese copies?

Yes thats them. And I would love for you to forward me to these calculations of depth penetration so I can check if its in range. Like I said I cant proof the bolt yet because I dont know if the chassis/tapped hole can handle what the bolt can handle. Like I said, the thread might strip. How do I calculate for that? What I know is that material is "mild Steel". and thats that. So how much can a mild steel thread take? And these bolts are most surely the real deal. BTW, how would I check if they were poor quality? I hear, examining the nature of the crack along the bolt reveals a story.

How many hours did the factory built machine survive before the first "loosening" occurred? Was it after some field maintenance was done?

HA! Why you ask? The one technician complained that after the first new road wheel is lost, after that, it NEVER performs as well after new ones. Secondary and tertiary replacement wheels etc etc always break off quicker! Apparently after the machine rolls out the workshop all new, they do NOT torque it to specific values, so its not the torquing thats an issue. It surely can only mean that the thread on the chassis gets deformed more and more after time. But how do I minimise this?

I expect a professional, conservative design would have plenty of thread engagement (over two bolt diameters) and grip length (5 or more diameters), and generous joint footprint". IF the mating faces between bolted components are (still) flat, AND the fasteners bear on hardened washers AND the threades don't bottom in the tapped holes AND sufficient torque is applied to achieve clamping greater than the service loads, then the joint will be dead reliable without fancy ramped washers or desperate measures like safety wire, cotter pins, jam nuts, or tack welded nuts. If any of the four prerequisites are missing then the joint will loosen and cascading damage will occur.

Now we are talking. I will measure this as soon as possible and let you know. But roughly, the thread engagemnt is 2 bolt diameters. I will get back on this. Any calcs or standards on this by the way? What do you mean by "grip length"? I think what you mean is over the yellow portion of my attachment? If thats the case its about 2 bolt diameters aswell (but not sure what you mean. The joint footprint is something I cant change. Thats a function of the road wheel. Look at my attachent where I outlined in red. Thats the footprint. Although it seems like a lot, on the edges, around the bolt, its really bad. Very thin areas around there. while were at it, make it hard interface and clean for no embedment. Bearing on hardened washers im not too sure about, that depends on what the nord locks are. You mention torquing correctly but once again, like I said in the replies above, what can the mild steel chassis thread take? I cannot find a way to determine this.

++ How many diameters of engagement do the failing fasteners have? Thread engagement is often described relative to bolt diameter. A 14 mm tapped hole that engages a 28 mm long section of the bolt's thread has "two diameters" of engagement. Generally 1.5 diameters engagement in mild steel allows the full strength of a Grade 8 fastener to be utilized (full torque, etc, and thus clamping).

As above, ill check this value out, but it looks roughly like 2.
You said 1.5 d's allows grade 8 to be used. Any where I can see how you got this? :-D Seems very "out of the blue". You see thats what I mean, I dont know the calcs or if its too complex to calc then what standards to follow? Any literature? Also when I mention alot how "can the thread take it". By that I mean, overtightening the bolt resulting in hole tapping to strip so I dont see how this can be a funcion of how long the thread engagement is.

 

[Tmoose]

Tsiolkovsky said - "You said 1.5 d's allows grade 8 to be used. Any where I can see how you got this? Seems very "out of the blue". You see thats what I mean, I dont know the calcs or if its too complex to calc then what standards to follow? Any literature?"

=================

Not quite "out of the blue". More like, consistently found when ambling aimlessly all alone through the forest sniffing and foraging for scraps of knowledge under rocks and logs.

For charts of what UNBRAKO felt in 1996 were reasonable tapped hole thread strength to consider for a quality socket head capscrew into brass, aluminum, cast iron and mild steel, look at page 66/68 here -

http://www.dalecompany.com/pdf/UNBRAKO.pdf

Note their comment on page 64/66 about using figgerin' and cypherin' to come up with tapped hole thread strength being "not entirely satisfactory". Thus what they chose to present in the technical section of their "engieering guide" were tables created using empirical data from tests done with non-digital threaded test pieces, real bolts, and torque wrenches of some type. Seems like something I could even do at home if I put my mind to it.

For an online discussion with a rule-of-thumb allegedly from a machine design handbook, look 2/3 of the way down here-

http://store.curiousinventor.com/guides/Metal_Working/Screws/

Attached is an image of a mossy rock, I mean, the worn cover and a pertinent page from a hard cover machine design handbook I bought new from the Lowell Tech (now UMass Lowell) bookstore for a mech engineering class I was taking nights. I paid the then shocking price of almost ~20 US Bucks for it. Guess the copyright date.

26th Edition of Machinery’s Handbook has some formulas for calculating such stuff, too.

 

[Tmoose]

"HA! Why you ask? The one technician complained that after the first new road wheel is lost, after that, it NEVER performs as well after new ones. Secondary and tertiary replacement wheels etc etc always break off quicker! Apparently after the machine rolls out the workshop all new, they do NOT torque it to specific values, so its not the torquing thats an issue. It surely can only mean that the thread on the chassis gets deformed more and more after time. But how do I minimise this?"

=========================================

My suspicion is after the first failure the mating surfaces of the cap and frame are beat up deformed and nasty, and the threaded holes are all wallowed out. Timidly tightening the bolts to clamp the uneven surfaces together means the impacts quickly buff down the high spots, leaving the bolts loose and subject to motion and bending, doing even greater damage to themselves and the badly damaged frame and bearing housings.

If the original bolts were torqued tight enough, then they never would have loosened.

Is the contact surface of the bearing housing really a raised area around the perimeter, NOT a full contact flat surface?

This is what I think it should look like. Big, flat and smooth, but without the dowels which serve only to locate it, not to handle loads.

http://glandpackings.com/image/main_bearing_oil_sal/LowerBearingCapSealIn.gif

I'm confused about

 

[Tmoose]

what I meant to say is, I'm confused about the comment "Apparently after the machine rolls out the workshop all new, they do NOT torque it to specific values, so its not the torquing thats an issue."

What is the event the triggers the first field repair? If it is the wheels "breaking off" then I'd work to understand the suitability of factory assembly method and the design. Certainly some operators can break almost anything, but I don't think wheels routinely break off Caterpillar and Case off road equipment.

Until REAL info is available might consider alternatives for a about 30 seconds. I REALLY think No official "torque" by the OEM is unlikely, but I wonder if that could be a misconstrued understanding of a "turn of the nut" spec, or, that they just buzz the bolts down as hard as some generic impact wrench will tighten them.

Regardless, unless high quality fasteners are tightened thoughtfully, the full potential of the design will not be available.

 

[Tsiolkovsky]

Wow, great, great contribution here Tmoose, I am overwhelmingly grateful, you really are clued up on this. I will reply to your above suggestions in due time.
That UNBRAKO technical section is *JUST* what I was looking for.
Okay I will start going through that theory plus measure the parameters you mentioned and I will reply here soon to give feedback. Thanks again for that and everyone else. Heres hoping, solutions will be found!

 

[Tsiolkovsky]

Such colossal depth of information in this manual. I dont even know where to start. What a gem.

Anyway, refer to the charts on pg 65 onwards...

In the material lines, what do "SS"values refer to? Shear Strength? As a matter of interest, what does "Rc"mean?

Secondly; Does theory mean to say that (should the load along the thread be uniform) the longer the thread engagement is, the less tendency to strip meaning that if my thread engagement was infinitely long, no matter how soft the material, the bolt will break before thread strips? i.e. A very high grade bolt in the softest of Steel tapping with sufficient length of engagement would make the bolt break first from torquing?

 

[Tsiolkovsky]

Hi all.

I had another heated argument with the same party regarding the torquing of these road wheel bolts into the tapped hole. His message was this:
Some holes were drilled into to get broken bolts out in the past. Some taps are corroded. Some housed loose bolts and these loose bolts combined with the impacting road wheel, deformed the thread.
He also mentioned how its impossible to know the condition of the thread with respect to other tapped holes. And therefore you cannot apply one given torque to different taps even if you clean the taps. secondly he says after assembly, my theory of applying a predetermined torque doesnt work. After many hours the tap deviates from theory. I dont even understand whats going on here to be honest. So he mentioned to just tighten the bolt like you would any bolt because torque specifics means nothing in an old tap. I cannot find any theory or reasearch into old vs new threads/taps. To be honest, I thought the only thing that makes a tap "old" is corrosion and thats it. My thinking here is if you dont exceed the shear stress of the tap, it will behave elastic and wont plastically deform.

So basically, can threads get damaged (apart from corrosion) to the extent that torquing to specific value means nothing? even if you dont exceed shear stress? (i.e. overtightening the bolt until the thread strips). If so, how exactly does it get "damaged" if its neither of these. He said drilling out broken bolts in past damages the tap. I read about this and it seems true but in that case wont an insert fix things up and so torquing it in new insert will tend to approach what theory says.