We’ll admit most of us are more comfortable with solder and software than mechanical things. However, between robots, 3D printers, and various other mechanical devices, we sometimes have to dig into springs, belleville washers, and linear actuators. Unless you are a mechanical engineer, you might not realize there’s a lot of nuances to something even as simple as a nut and bolt. How many threads do you need to engage? Do lock washers work? [Engineer Dog] has a post that answers these and many other questions.
The top ten list starts off with something controversial: split ring lock washers don’t work. The original post cites a paper that claims they don’t except in very special circumstances. However, he updated the post later to say that some people disagree with his cited study. In the end, you’ll have to decide, but given there are other options, maybe we’ll start using those more often.
Speaking of other lock washer options, have you ever seen a Nord-lock, a serrated flange nut, a castle nut, or a slotted nut? He also mentions chemical lockers and nylock nuts, which we have used before. How about a DTI smart bolt?
The post also repeats a rumor we’ve heard before that Phillips head bolts are made to cam out — that is, have the screwdriver jump out of the hole — on purpose to avoid overtightening in aluminum aircraft panels. A few cam outs and you have a stripped out head. It isn’t clear it is really a design feature, though, as the original patent from John P. Thompson doesn’t mention it. A later patent does seek to reduce the amount of “throw out”, though, so it may have been an accidental feature.
We talked about lock wire in the past, which is yet another way to secure critical fasteners. We’ve also had a talk about Nord-lock and jam nuts. Want to hold your stuff together? You need to know this stuff.
” However, he updated the post later to say that some people disagree with his cited study. In the end, you’ll have to decide, but given there are other options, maybe we’ll start using those more often.”
Woah, woah: none of those “counterarguments” actually are intending to say that split-ring lock washers actually work! Both of the testing still shows split-lock washers loosening incredibly fast. The testing videos are actually intended to advertise other kinds of locking methods which *actually* work.
Those results only show split-lock washers loosening a little slower than a flat washer or equivalent to a toothed nut, and that’s of course possible depending on what the mating surfaces are, the vibration profile, and the torque on the nut itself.
At least for me, the “fastener bible” is always the NASA fastener design guide, which explicitly calls out split-ring lock washers as useless. It *also* calls out toothed star washers as mostly useless as well, because they work by damaging the mating surface (so obviously if the mating surfaces are too hard, they won’t work).
I agree that “toothed star washers” are not to be used for proper/permanent fastening of things. But they do work to make proper electrical contact between two metal parts. The damaging of the surface of the material it is mounted to, allows for a good electrical contact, saving you the trouble of scraping of paint or rust. Because the washer digs into the material the last point of contact is as good as completely sealed from the air preventing corrosion at that electrical contact point.
A common problem with these washers is that in many cases they are overturned, totally ruining them, sometimes flattening them to the point of looking like a plain washer. In that case, use stronger washers or if not available fire the person who’s over-tightning them.
That’s a good point about the electrical contact, and now that you mention it, most of the places I’ve seen these have been in cabinet grounding kind of situations. So it makes a lot of sense, they are being used appropriately.
But I have to defend the plebe who over-tightened them! I didn’t over-tighten it until I had to re-tightened it twice in one journey! :P
Link to the NASA fastener deign guide : https://ntrs.nasa.gov/search.jsp?R=19900009424
I’ve never seen this but it’s interesting, and has information that seems intuitive. I wouldn’t take it as law simply because it’s printed and says “NASA” on it, though. It definitely lacks supporting data.
If one assumes these documents are used as resources within NASA, them not being discarded after 30 years suggests they’ve been field proven even in the absence of data in advance.
The reference section, though brief, gives several sources that are well referenced and well supported, such as the McDonnell-Douglas design handbook. As the NASA design manuals are requirements documents, not research publications, they are generally supported by a lot of other research, much published in the open lit, and failure investigations/lessons learned reports.
Well, yes, actually no. What those “counterargument” show is exactly that lock washers DO work – admittedly incredibly poorly compared to what one might ignorantly expect, but nevertheless still somewhat better than a simple washer. You’re still better off using them than not using them, especially considering not every application will see strains as severe as the Junker test. If what you need is absolutely no loosening ever “or else”, you’re not supposed to rely on a fscking friction washer of any kind – use a castellated nut or a proper, keyed tab lock washer or other equivalent.
The industry does not agree with you. You are not better off using them. Porsche learned this the hard way on drive shafts.
Tests show a nut with helical spring “lock” washer actually loosens =faster= than a nut with no locking method at all!
https://www.boltscience.com/pages/helicalspringwashers.htm
I’m the link takes me to a site with “your device is infected” spam.
Use caution.
Which link? They all look ok to me.
I’d guess the engineerdogs one. try and visit it with the sketchiest PC you can find?
hmmm…
“…split ring lock washers don’t work..blabla… except in very special circumstances”
Ehmm, even if they work in very special circumstances, then they DO work.
That something doesn’t work under all circumstances doesn’t make it useless. You just have to know when to use what.
Some may say that nails are useless because it is proven that nails can be pulled out much easier then screws, but nails are still used and for many good reasons. No matter how much “research” you do in order to make your point, you must always be prepared to look at something from all angles (not only the angles that are important to you), because if you don’t then the research isn’t complete and therefore useless itself.
PS: I’m a satisfied user of split ring washers for many decades and they always served me well. Their main reason is to apply a constant pressure to the thread making it more difficult to loosen by various kind of forces, twisting, bending, vibration temperature cycling etc. I’m not building motorcycles, airplanes, helicopters or rockets, just simple mechanical constructions with lot’s of knops, handles, hinges, electric motors and solenoids a few connectors. So therefore my field of view is very limited, but more then wide enough to say that split ring whashers work fine for me.
I actually do build helicopters, and there are many places that we use split washers without incident. They are mostly useful in instances where you only have tension on a bolt, and maybe vibration or heat cycling. We even use them on hard stainless and titanium parts. They are especially common on MIL-SPEC connector strain reliefs. The NASA manual specifies situations where a fastener is subjected to torsional forces, where you would want to use something more secure, as well as a secondary locking mechanism, such as a castellated nylock nut, and a cotter pin.
Ehmm, even if the research doesn’t cover selected, very special circumstances, then it is still ok to draw conclusions that apply to most cases, most of the time.
That some research leads to conclusions that don’t cover all circumstances doesn’t make it useless. You just have to know when to dig deeper.
By your logic nobody could ever draw any conclusions because they’d never be able to consider all cases
I really like those articles that cover so basic subjects like locking nuts, wire crimpers etc. They fill up holes in my knowledge and sometimes demistify what “experienced” colagues tell me.
Also “hard lock nuts” are worth checking out. They use wedge principle.
Hi Janusz,
I didn’t knew these existed, thanks for mentioning them.
You welcome.
I found them only on 11kV terminals (up to 6 MVA) in my workplace. It took me hours to find them on the web when requesting spares. Just to be clear – I have never had oportunity to make any comparation between this and other solutions.
Thanks for introducing me to the DTI bolts.
But one question not answered on the link is…
If tension and torque are different (of course) how do I know what DTI fastener to use if all I have is a specified torque.
You don’t. There are too many consumers of torque that aren’t tension to tell. Friction in the threads and friction under the head are too large to ignore and can take, in the worst case, all the torque and produce zero tension.
He leaves out one interesting trick for checking if a bolt is properly tightened. If you have access to both ends of the bolt, you can use a micrometer to measure the length and tighten the bolt until it stretches a target length. This is used on race engine connecting rod bolts, which are very highly stressed and easily measured.
That idea is great, but it assumes you know 1) the specs of the nut and bolt, 2) the torque required by the job at hand, 3) how much stretch is the ideal amount and how much torque it takes to achieve it.
Single most useful article/post I’ve yet seen here — especially sections #2 (double-nutted joints) and #3 (fatigue life).
Thank you!
My go to is a Stover Lock Nut. Did rigging design for lighting grids for theatrical locations for years, and any place I used them, I had the owners ask what they were. A bunch of satisfied customers.
Because a night club vibrates like the dickens!
Those look like fancy Crimped Lock Nuts.
http://www.leftlanebrain.com/shop-tips-tricks/4-styles-of-crimp-lock-nut/
Ive seen these used on vehicle assemblies where you see lots of heat cycling (exhaust) and high shock loads (suspension and shock/strut mounts).
When installed correctly, they take a great deal of force and a large breaker-bar to get loose, but they also never come loose in some really difficult installations.
Pro’s and Con’s I guess.
Indeed. Make sure you’re ready to install them because it’s a bear to remove without damage to your threads. One of the techniques I used was to do fit and finish of the item I was installing using standard nuts, then cap the threads with the stover when the install was complete. Nuts used to prevent upper crawl got a few drops of loc-tite wicking version to keep them from loosening.
if I am working on a through hole, I have got to the point that I just gob the threads with a drop of the 10 for a buck packs of superglue. It is far less expensive than loc tite, and most of the time you are able to just snap the hardware to remove it. I used to think that was unthinkable as using superglue to lock threads. Than one day I broke the internal chain on my 1952 merry tiller. The side case was held on with about 30 1/4″ nuts and bolts that had as far as I can tell lever been off of it. I fought with a dozen nuts and bolts or so, before I managed to snap one by accident. Than something clicked in my head. I was not going to re-use this crappy old hardware and snapping them was a lot faster than politely taking them out. So now I thread the bolt over a drop of superglue when I put them together, and over tighten until they snap to remove them. At least with through holes. It sounds odd, but if you have a lot of hardware to remove and you are not planning on re-using, give it a shot. You will be surprised at what a time saver it is.
I remember doing exactly the same thing with a load of 1/4 inch nuts and bolts in a land rover hard top, tge first fer undid until they were lose and floppy then sheared, so I just tightened the rest to shear them, like you say, an excellent time saver.
“six threads at the most” yet he says and the graph shows that it takes six threads to carry 100% of the load.
So six threads at least.
“Additional threads beyond the sixth will not further distribute the load and will not make the connection any stronger.”
Remember also NYLOK nuts aer an interesting beast.
Not only are they an insulator (nlylon sleeve) between the bolt and the outer nut, but choose the right colour – otherwise you won’t be removing them without a dremel or other way of sawing the head off !
The only lock washers that work in my experience (industrial machines) are the Nord-Lock™ type.
They bite hard and are reusable…every time I’ve disassembled anything secured by them, I swore the bolt was seized until they broke loose.
Split washers work … just as well as plain flat washers. Or no washers, for most applications. Their greatest help is acting as a spacer to increase the amount of screw that is available to stretch, which any spacer would do. They do not help retain a loosening fastener in the slightest.
Split ring washers can be worse than useless, they may speed loosening of the nut. This link says faster than just a nut with nothing!
https://www.boltscience.com/pages/helicalspringwashers.htm
Junker doesn’t test vibration. The machine forces relative movement between parts. In a properly designed joint there is never relative movement between parts. This can be enforced with shear pins or by making the clamping related shear force higher than any that is carried across the joint. If there is motion the parts will eventually wear and the joint will fail no matter what sales pitch the washer salesman gives.
“It turns out that fatigue loads below a certain threshold will never cause the fastener to break.” This only applies to steel.
Plastic deformation doesn’t make the fastener is no longer elastic. For small amounts of extension it makes the fastener a bit longer. Just because it yields doesn’t mean it no longer provides a preload.
It is rarely the case a Phillips head simply strips out. What does strip out is using a Phillips screwdriver on the Japanese JIS recess or vice versa or using the wrong size Phillips driver on a Phillips head.
In my experience Pozidriv and phillips mix ups are more usually the cause of screws getting buggered. They seem to fit just fine, but with the tip angle being just slightly off escpecially a Pozidriv screwdriver in a Phillips head will bugger it up in short order
What is wrong with you guys? Have you not read the 1945 engineers hand book? With Spanish and French translations…tell the young ‘uns today …..they don’t believe ya…and c+ was a mark you got in the exam room in them there days.
“your results may vary.”
My results do, in fact, vary. I worked in a company that bolted things (usually with 6-10mm sized fasteners) into cabinets that were then mounted onto earthmoving equipment. One of our engineers read the guff about split washers and decided to not use them resulting in me being the poor sucker who had to fly to the other side of the country to retrofit them.
The joint and screw was undersized for the load, then. Simple. But it is easier to blame the engineer and theory, and it makes for a nice story.
Split ring washers work, for certain values of work. Better alternatives are available. That is pretty much the whole issue. You CAN use split ring washers, the question is whether you should.