Strongest Bolt? Grades Explained & Dyno Tested For Science

Torque of TTC is working for Astro Tools who also make impact wrenches (air). TTC is not the only testing out there, always consider multiple sources when looking at a tool!

0:00 What we're testing
2:06 Grade 1
5:11 Grade 2, 3
5:56 Grade 5/Class 8.8
8:14 Grade 8/Class 10.9
10:44 Grade 9/Class 12.9
13:37 Bowmalloy
15:30 Stainless
17:54 Hardness

[MattLitkeRacing]

I’d like to to see the same test with ARP head studs. And compare to other head bolts or studs

[tylerw]

This channel is nothing but straight information with zero fluff or filler. Don't ever change.

[tedjordan]

Interesting video. I’m a metallurgist and used to operate an accredited fastner testing laboratory. Nuts and bolts are a complicated subject. The way they are installed plays a huge role in how a threaded connection performs.

[ADY-dljw]

Fantastic video. Very interesting.
I have a few comments from the perspective on an engineer.

1:50 - 18/8 is a stainless steel alloy, not a grade of bolt. It also called 304, or A2. It comes in its own three different strength classes - A2-50, A2-70 and A2-80. On metric fasteners this info is often stamped on the head. This steel can't be heat treated, as it doesn't have any carbon content (other than as impurities). The different strength classes are achieved through cold working the steel in this case.

8:36 - Grade 8 / 10.9 is always made from medium carbon (~0.4% C) alloy steel, quenched and tempered, never plain carbon steel. This is a requirement of the SAE/ASTM/ISO/DIN fastener standards. Plain medium carbon steel can't achieve these strength levels. Chromoly 4140 steel and 8640, 8740 are alloys often used for this grade for smaller diameters. For large diameters, 4340 is common.

11:10 - It is worth noting that metric 12.9 and ASTM A574 are standardized classes, but the so called equivalent "grade 9", as well as the bowmalloy are not (they are a product of some manufacturer that does not have to meet any legal criteria behind the manufacturer's promises). Therefore the last two will not be found in any product that needs to meet certifications.

17:45 - Regarding your question about the recommended torque specs, these torque figures are not really related to the maximum that a bolt can live with. Instead, they are calculated to generate a bolt tension/clamping force/preload that is equal to 75% of the bolt's proof strength. This guarantees reusability of the bolt after disassembly. The reason for it is that torque is a pretty poor proxy for preload - even among seemingly identical bolts torqued to the same spec, there is a +-30% variance in the generated preload. To make sure that a bolt does not yield, it has to stay below its proof strength. Hence if we calculate torque for 75% proof, and the natural bolt variance can mean that for some fasteners this will actually be 30% higher, we then get 0.75x1.3 = 98% proof strength, so even outliers won't deform and the bolts can all be reused.
In a permanent assembly, where fastener reusability is not a factor, the guideline is to use a torque calculated for 90% proof strength. Most bolts will not deform under this preload, but some will and will have to be replaced if it's ever disassembled.
If the engineer wants to use even more of the bolt's strength, he has to rely on more accurate methods than a torque figure, such as a specified degrees of turn from an initial snug position (you may recognize that from engine head bolts), a preload indicating washer, a yield sensing robotic wrench, or specified bolt elongation.
Regarding the failure point from torquing the bolt, it's not really something that can be usefully specified, both due to the aforementioned huge variability in torque, but mostly because the torque-preload relationship ceases to be linear once the bolt begins to yield.

Finally, in this specific video (not your usual impact tests), a better methodology would be to provide the dyno results as the gauge's direct reading of clamping force, instead of calibrating for torque, because
A. bolts are not chosen by the engineer to meet a torque figure, they are chosen to create some clamping force and the torque is calculated to generate this force. This torque is very different if the bolt is dry, oiled, greased, cadmium plated or used with anti seize, but the clamping force is the same.
B. You have to calibrate the dyno for every single bolt you test even if they're all the same size, due to the torque variability especially with unlubricated bolts.
C. Once a bolt starts to yield the multiplier constant in your digital gauge no longer converts the measured force to torque correctly, as torque no longer varies linearly with more tension.

Once again, this was a very interesting video to watch. If you ever do it again with a measurement of the force it would be very interesting to compare the ultimate force you get with the grade specifications. Should be essentially the same as a tensile test, and the forces should match or slightly exceed the specs.

[andycanfixit]

I've used tons of stainless bolts for bolting together RF transmission lines that have bronze flanges with o-rings that have to bolt together to make an air tight seal but the also not corrode or react in the open air environment for 20 plus years. Biggest issue with stainless is galling, but they are quite durable and you don't have to worry about a rusted nut not coming off years later. The bolts we use come with a wax coating so they actually have a bit of a grey hue to them instead of being super shiny as the wax helps prevent galling as you tighten them.

[abcdefg]

One could add, that the markings in the metric system actually make a lot of sense and tell you about the physical properties of the bolt. 10.9 for example: the first number (10) multiplied by 100 gives you the tensile strength of the bolt (1000 N/mm²). Multiply that by the second number (.9) and you get the yield strength of it (900 N/mm²). Simple as that.

[Christackleberry]

I used to work for Bowman. They are the company who originally created the Bowmalloy fasteners. They were bought by the Barnes Group and then MSC.
They kept the name due to industry recognition.
For a time, these were made in China and almost cost Bowmally their reputation.
I believe they have been brought back to domestic production now.

[jfellow]

Never apologize for a longer video. The more information out there, the better.

[AlexanderBurgers]

Metric stainless fasteners are marked A2 or A4 for the material type (with A4 being higher corrosion resistant in salt water and chlorinated environments) and a number (50, 70, 80) that indicates its strength class, with 80 being equivalent to 8.8

As for the relationship between clamping force, torque, load, and so forth, there's whole engineering books about it, so the information exists, it might just not be explained in simple form anywhere easy to find. Odd. :) It would explain the amount of bolt abuse you see people do.
In short, the bolt fastening torque is mostly set to optimize clamping force. Clamping force makes sure that shear loads are absorbed by friction between the clamped faces, and that cyclical loading doesn't fatigue the bolt or joint.
A loose bolt would shear off or fatigue, or rattle around and damage the hole, where a properly tightened one would not.
An over tightened one could snap from being over tightened alone, or could damage whatever its clamping.

[FishMan]

Tightening torque is based on 66%-75% of the tensile loaded yield strength, and depending on who is doing the calculating for that factor of safety. It has to do with avoiding fatigue, which is based on yield. Adding an axial load onto the bolt adds some tension and reduces some of the clamping load until the clamped pieces separate. For most bolts, this ratio is 0.33 * load applied added to the bolt, or less. I believe steel is usually .2-.25. This ratio of force applied is part of what makes proper torque avoid fatigue. If the total load of the bolt varies by 5% and doesn't yield instead of varying by 50% or more, that bolt will stand up better to cyclical loading.

As much as I like your videos, this particular test doesn't mean much. Torque is heavily influenced by the lubricant you use. Molykote P-37 generally has the most consistent results across materials. Unlubricated threads vary widely between material and coatings, not grade 5 to 8 but steel to stainless or Monel or K-Monel.

And a comment on grade 8 and harder alloy steels (>34 HRC or >150 ksi UTS). You should avoid any kind of zinc plating because it actually increases susceptibility to hydrogen embrittlement, which is a science term that means a bolt can break at 30% the rated load from environmental conditions. There is some debate where the exact cutoff is, but the toughest alloy bolts can easily lose a lot of their strength from a coating, sometimes before it is ever installed if it was done wrong (See San Francisco Bay Bridge failure where hot dipped galvanized bolts broke throughout)


TL;DR. Bolt tension and reliability has a lot of factors and unlubricated torque runs til failure look nice but unfortunately don't really tell you much from a scientific point of view without concentrating on where they yield and controlling other factors by using a high quality lubricant, like Molykote P-37 or similar.

[speedyham]

Great video. I am an engineer and have known what the outcomes should have been for a long time, but I rarely get to see heads up testing like this. Always neat to see the real world line up with the specifications that I use day to day. When they don't line up, trouble happens. :(

[zjeepgozweeln]

Perhaps a test of thread lube? Grab a bunch of those grade 8 bolts and slather them in various thread treatments... ARP ultra, motor oil, a few greases, anti-sieze, threadlock, etc. You could test torque consistency and overall torque before failure.

[Fuddleton]

There are tougher bolts out there.
NAS8906 go to 220kpsi and NAS8907 go to 260kpsi. They're used in aerospace framing and are single point turned rather than rolled on a screw machine, and have a tighter class of fit than normal threads. You have to go to a dedicated screw house for those, and are the last stop before custom designed solutions, like special alloys and ground threads.
Most applications care more about performance at temperature and corrosion rather than just straight performance though, which is why A286 and inconel screws are also a thing.

[jozsefizsak]

I'm tremendously grateful for this very informative video. It's most definitely not too long.

[ultimatehandyman]

I used to work as a maintenance engineer in two local chemical plants, we used some large steel “bogies”, which weighed several tonnes each. The shelves were held into the bogies using ¼ unc bolts at one of the plants and M6 cap head bolts at the other plant. The cap head bolts (12.9) used to shear off for fun, yet the regular ¼ unc bolts never did.

As for stainless bolts- they are a real pain in the backside, especially when they gall up and cannot be cut off with the gas axe!

[scotttod]

Years back I had to replace 288 1" bolts that were holding a ski lift together because they were counterfeit. They had been in use for 35 years and none of them had ever failed and none of them broke on removal even with about 10 coats of paint on them and rust. A325 bolts and A563 nuts were all torqued to spec in a 3 step process 50-80 feet up the towers.

[kenchilton]

Grade 1 and 2 are fine for most wood fastening applications, since the wood will fail before the bolt. The carriage bolts you are testing are a good example - if building a wood carriage, just use grade 1.

Many people assume a higher grade bolt it better in all applications. This is incorrect. Some bolts are intended to fail to prevent more severe damage. Some bolts are better at shock loads than others. For instance, bolts on suspensions are intentionally not high grade, because higher hardness is usually more brittle, and the impacts in the suspension mean that high grade fasteners may snap during use. Lower grades will stretch, and that stretching and loosening of the fastener can be accommodated in the design, since the major loads can be designed in shear and not in tension.

The torque applied to the fasteners is also sometimes more critical than people assume, and too tight can often be a very big compromise to the fastener. Preloading the tension or creating additional stretch reduces the capability of the fastener in tension and in shear. Each application presents a window of proper torque for the fastener, including whether clamping force is required or just sufficient force to prevent the fastener from backing off. These windows also depend a lot on the fastener grade and thread configuration.

[needleonthevinyl]

14:04 FYI, that's just the Bowman Distribution company logo (now owned by Barnes) and doesn't necessarily mean it's their proprietary high strength alloy. I've got an entire bolt bin of old grade 5 bolts with the Bowman logo on the head. Great video as always!

[RustyShackleford]

One thing I would like you to try.. apply some anti seize to the stainless before it's run. reason being stainless binds up when heavily impacted. the antiseize keeps it threading freely.

[TesserId]

Very worth the watch. Where else are you going to see this kind of demonstration?