10 Tricks Engineers Need to Know About Fasteners

You’d expect an engineer to know everything about a basic subject like nuts and bolts right? Well fasteners are one of those topics that seem simple at first but turn out to have much more depth than you expected. What follows are 10 tricks on this ‘basic’ subject that might surprise you!

UPDATED MARCH 2017: It has come to my attention that there is testing data available that makes the counter argument to my first point in this article. In the interest of transparency and good scientific discussion, I’ve provided links to all sources mentioned at the end of point #1. When faced with contradictory test results, if we assume the methodology and integrity of all sources are without fault, it is reasonable to conclude that your results may vary. Many factors could significantly affect testing results including variances from manufacturing process, fastener & clamped materials, heat treating, surface finishes, ambient conditions, and bolt diameter.

1. Split washers have been experimentally proven to be ineffective locking devices and can even aid self loosening over time. And yet I see these things in use everywhere, so what gives?

split washers dont work

In theory split washers (aka lock washers or helical spring washers) are supposed to work by squishing flat between the nut and the mounting surface when you tighten them. At this point the sharp edges of the washer are supposed to dig into the nut and mounting surface to prevent counter-clockwise rotation.

In practice a split washer is unable to gain any purchase against hard surfaces and does not actually prevent rotation. The problem is that split washers make for poor springs and bottom out after only a small percentage (on order of 10%) of a bolt’s total clamping load.

The only time a split washer might prove useful would be for fastening onto soft easily deformed surfaces such as wood, where the washers springiness & sharp edges could actually work.

The evidence against split washers started stacking up in the 1960’s when a gentleman named Gerhard Junker published some of his lab experiments. He invented a machine specifically for testing the effect of vibrations on threaded fasteners. The first thing he discovered was that transverse vibration loads generate a much greater loosening effect than do axial vibrations. Good to know.

transverse axial loading

His second discovery was made by plotting the bolt tension vs vibration cycles to create a ‘preload decay chart’. When he compared the preload decay of a bolt & split washer combo to bolt by its lonesome, he found that the split washer caused the connection to loosen sooner, as seen below.*

preload decay chart

Not to worry, there are better locking options available. Chemical lockers like Loctite, deformed thread lock nuts, and Nyloc nuts should be your everyday go-to locking devices. If you have some money to burn then wedge lock (Nord-lock) washers &
Serrated flange nuts are probably the best way to go.

When lives are on the line you may want to employ a ‘positive locking device’ such as a castle nut or a slotted nut. No amount of vibration will break this kind of connection:

castle nutSince this one is sure to stir things up when you mention it the guys at the office, I‘ve provided my sources below.

A) Article 1 n boltscience.com and Article 2 on Boltscience.com and Article 3 on Boltscience.com, all condeming split washers
B) pdf file from hillcountryengineering.com condeming split washers
D) Awesome video showing actual testing and how preload decay charts are generated.
*E) Alternate Testing Video #1 making the counter argument in favor of split washers.
*F) Alternate Testing Video #2  making the counter argument in favor of split washers.


#2. Double Nutted joints with jam nuts are affected by order of clamping. While I’m talking about bolt locking techniques, I’ll share another interesting one: For double nut connections involving the use of a jam nut & a standard nut, it REALLY matters what order you install them in.
The jam nut should go on first! Otherwise the effectiveness of the nut pair is greatly reduced. Double nut Source.



Before I move on to the next one I need to clarify the difference between static loads and fatigue loads. Static loads do not change over time. If a bolt is rated to yield at 3,000 lbs of tension, any static load less than that will not have a permanent effect.

However, if you were to vary that applied load over time you can fatigue the bolt until it breaks using less than 3,000 lbs! In the same way that a small stream can carve out the Grand Canyon, fatigue loads gradually chip away at the structural integrity of fasteners over time.

static vs fatigue

#3. The relationship between fatigue load and the number of cycles until bolt failure occurs can be predicted using experimentation. It turns out that you can make reasonably accurate predictions of the cycle count at failure by performing as few as three experiments (though I would recommend doing at least 6 to attain some real accuracy). All it takes are a few data points and a regression line to create a high cycle fatigue life curve (aka an S-N curve).

fatigue life

We don’t put the ultimate static loading on the graph, but if we did it would be the highest point on the ‘applied load’ axis and the zero point on the ‘cycles’ axis.

Why is this useful? Now that you know bolts can break from small fatigue loads, imagine trying to build a bridge using riveted or bolted connections. How could you trust that you ever had a big enough fastener? It turns out that fatigue loads below a certain threshold will never cause the fastener to break.

As a very general estimate, a bolt will require an infinite number of cycles to break if the fatigue load is around 30% (+/-15%) of the ultimate static load. You can expect the bolt to break in a few thousand cycles if the fatigue load is about 80% (+/-10%) of the ultimate static load. (Note: the exact percentage can vary dramatically based on material composition and ambient conditions.)


#4. (UPDATED) For maximum strength tighten bolts up to the yield point…For maximum durability, don’t! There is a common misunderstanding that a bolt within a securely fastened connection is impervious to outside forces if hey don’t exceed the clamped load of the connection.

That is, the myth says a bolt clamped to 500 lbs won’t experience additional stress unless outside forces applied to the clamp exceed 500 lbs.  This is not so! In fact ANY additional load, no matter how small, will add to the tension in the bolt. But not at a 1:1 rate.
Think of pulling on a fastened connection as if it were 2 stacked springs. Both springs stretch measurably, but the weaker one stretches more. Part of the external load is absorbed by the joint and part by the fastener.

pic 1

 To be crystal clear, as you tighten the nut the bolt will compress the two parts together. The bolt itself has an internal reaction force equal to the amplitude of the compression force, but the bolt itself is in tension.  If you were to graph the tension on the bolt while you tighten the nut, the plot would look like the graph below. To get the greatest clamping force out of the bolt we would to tighten it all the way up to the yield point. Any more force and the bolt will enter the plastic region and permanently deform.

force displacement curve

In practice engineers don’t design that way. Since any additional force will begin to yield the bolt, you want to give yourself some margin for error.  Engineers select a bolt tension that is somewhere between the calculable minimum functional clamping force and the yielding force…. while also accounting for error in the tension measurement method.

(Sources: Shingley’s Standard Handbook  & Article on Fastenal )

#5. It is actually quite difficult to determine the exact load the fastener sees during clamping. We now know how important it is to avoid over tightening a bolt, but how do we know when it is yielding?

   For everyday purposes the clamping force can be approximated by measuring the tightening torque. You can look up the recommended tightening torque for a given fastener size in my bolt sizing calculator or in a table like the one found here.  An alternate method is called the ‘turn of the nut’ wherein you tighten the bolt until it ‘feels snug’ before rotating it another 90 degrees to ensure adequate tightness.

Those methods work OK for most things, but some critical applications require you to be certain of the clamping force (think spacecraft or large weights above your head). The torque method has difficultly accounting for friction and lubrication, but at least the torque is mathematically correlated to the clamping force. On the other hand the turn of the nut method uses rotational displacement to bypass lubrication affects, but it doesn’t even consider forces at all.

There are better options though. Load indicating washers can accurately verify bolting loads by squishing open a paint sack after reaching a specific load. The drawback with these is that they only work once.  http://www.boltscience.com/pages/tighten.htm The other option comes from a company called smart bolts who came out with a fastener featuring a built-in tension indicator. This is by far the most accurate method of measuring bolt clamping load. On the other hand, a single box of these bolts can cost around 10 times as much as a standard fastener!

smart bolts

Neat, I just wish I could afford one.

How different tightening methods compare in terms of accuracy.


#6. If you’ve ever designed a part with a threaded hole, you may have wondered:How many threads do I need to make a strong connection?’ The answer is that it varies, but six at most.

Bolts actually stretch very slightly when force is applied, which causes the loading on each thread to be different. Because of this stretch, when you apply a tensile load on a threaded fastener the first thread at the point of connection sees the highest percentage of the load. The load on each successive thread decreases from there, as seen in the table below.

Additional threads beyond the sixth will not further distribute the load and will not make the connection any stronger. 

6 threads at most

So will a bolt break before the nut strips? Yes! Nuts typically have no less than three internal threads, but nut thickness standards have been selected on the basis that the bolt will always sustain tensile fracture before the nut will strip.

#7. Have you ever seen a fastener labeled with a 2A or 3B rating and wondered what that meant?That number-letter combo is used to indicate the thread class of the fastener. Thread classes include 1-4 (loose to tight), A (external), and B (internal). These ratings are clearance fits which indicates the level of interference during assembly.

  • Class 1 is a good choice when quick assembly and disassembly is a priority.
  • Class 2 is the most common thread class because it offers a good balance between price and quality.
  • Class 3 is best used in applications requiring close tolerances and a strong connection.
  • Class 4 is precision tight, typically used for lead screws and such.

#8. All fasteners are available with either coarse or fine threads and each option has its own distinct advantages.
Finely threaded bolts have slightly larger cross-sectional areas than coarse bolts of the same diameter, so if you are limited on the bolt size due to dimensional constraints, choose a fine thread for greater strength. Fine threads are also a better choice when threading a thin walled member. When you don’t have much depth to work with, you want to utilize their greater number of threads per inch. Fine threads also permit greater adjustment accuracy by requiring more rotations to move linearly.

On the other hand, coarsely threaded bolts are less likely to be cross threaded during assembly. They also allow for quicker assembly and disassembly, so choose these when you will be reassembling a part often. If the threads will be exposed to harsh conditions or chemicals, a coarsely threaded fastener should be considered for its thicker plating/coating. Coarse ly threaded fasteners are much more commonly available in the united states.

#9. Would you expect a bolt to be stronger or weaker at very high temperatures? How about at cryogenic temperatures?

Most people answer ‘weaker’ to both question, but being weaker at both temperatures doesn’t even make sense when you think about it. Why would steel be strongest at whatever typical room temperature happens to be?  It’s not.

As a rule, metals are strong & brittle at low temperatures and soft & ductile at high temperatures, within their solid phase temperature range. Room temperature is just another non-extreme point on the curve.

test temperature chart

#10. You can make bolted connections more resistant to shear loads by using clever design instead of larger bolts. For maximum strength, try to use the correct thread length for the connection. In the image below you can see two connections which are identical except that the one on the right has a properly sized thread length. It exposes the bolt shank (rather than the threads) to the applied load at the connection seam.

All else remaining the same, the connection on the right will be stronger because the shank has a larger cross-sectional area and no stress concentrations.

pic 2

Another clever trick is to design connections so that the applied load will be on multiple sections of the bolt, as opposed to just one section. In the images below there are two connections. The one on the right is twice as strong as the one on the left because it would have to shear the bolt off in two places to become free. Also, the single shear configuration can also lead to bending loads on the fastener and loosening of the connection (see #1).

pic 3

#11. Have you ever cursed the day you were born because you just stripped out a Phillips head screw? While it is nice that Phillips screw drivers don’t slip off screws like flat heads do, it’s a real pain when the head can no longer be rotated because the screw head has melted into a hollow cone.


As frustrating as that is, it turns out that Phillips head screws are designed to strip out via the tapered point and rounded edges. The technical term is called a cam-out and every time it happens the relative surface motion wears out your screw.  Alternate screw heads like torx and pozidriv are specifically designed not to cam-out.


If fascinating facts about fasteners are your thing then you should check out  Carroll Smith’s “Nuts, Bolts, Fasteners, and Plumbing Handbook” aka ‘Screw to Win’. Smith’s Engineer To Win is another good one. Actually every book by Carroll Smith is pure gold.

carrol smith book

You might also enjoy reading about a custom MS Excel based EngineerDog Bolt Sizing Calculator here.



  1. I don’t think that is the case. I think the spring washer actually (at least in current usage) is to increase preload with a controlled deformation. When you have say a safety item with 10.9 grade steel to the preload it will start to deform the steel of say more than 980 MPA steel. This causes a preload that puts the energy back into the tightening torque and not the strain energy of the steel. Basically it controls your torque limit so you can keep it fastened (without torquing the nut) but still maintaining the preload to prevent loosening from NVH or light brushing torques. Basically, lower loosening and install torques but still same preloads and tolerancing.


  2. Hello,

    First, thank you for the article.
    Second, I found on youtube this video also from boltscience. From what I see the test shows now that helical spring washers do help. The amplitude in change but still.
    Can you please check out the video, what do you make of this?


  3. What about the general use of flat washers? My general practice is to always have a flat washer under the bolt head and under the nut, but I’ve ran into conflict over the years where some engineers dont feel it’s required under the bolt head in some instances.


    • In theory I can’t see why plain flat washers would hurt as long as they are harder than the bolt AND their ID is no larger than the joint through hole ID.
      Washers are used to provide a bearing surface for a nut or screw head, cover large clearance holes, and distribute fastener loads over a large area, particularly on soft materials. They are useful hardware for many reasons but are not intended as a reliable source for vibration protection.
      Also, I wouldn’t use a washer in addition to a serrated flange nut because it would negate the effect of the serrations.


What do you think?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s