I've had this excellent document for some months now.  I wanted to share it with all of you.  Sorry it took so long, but I've been too busy to be around much lately.  I forgot where I attained it, so I am hosting it myself.  Please right-click and "Save as" or I may have to take it down.

Department of Civil and Mechanical Engineering, United States Military Academy, West Point, NY - M16 Bolt Failure Analysis

This document clearly shows that A LOT goes into making a bolt properly.  Not just HPT/MPI.  If it is not heat treated and case hardened properly/uniformly, the finish can wear exposing base metal.   This allows corrosion and pitting to occur, which in high stress areas will allow a crack to form and propogate from this point.

I've done the how long can it go without cleaning thing on my "experiments" rifle.  But on a rifle that you trust your life too, you really should clean and relube (atleast the bolt) as soon as possible.  On occation closely inspect the high stress areas of the bolt (especially where the extractor rubs near the rear of the lugs on either side) for exposed base metal, corrosion and pitting (use a magnifying glass).  This may or may not be enough to detect it, but it is worth a try.

Lastly, perhaps this will data will give rise to new rust+wear-resistant materials and/or coatings for bolts?  Or perhaps the remergence in popularity of old coatings (hard chrome)?  I have heard through the grapevine that manufacturers of bolts for the .mil were put on notice to pay extra close attention to the heat treat/case hardening of their bolts and that this process was revamped with closer/extra QC (this was some years ago, after a similar study) to ensure uniformity of proper hardness.  All current bolts (and those from the past few years) from .mil manufacturers should have nearly perfect case hardening.

That's an excellent read. Thanks for posting it

bump for those who missed it

Quoted: That's an excellent read. Thanks for posting it

+1

Thanks!

for what ever reason i cant open it

Quoted: for what ever reason i cant open it

Did you unzip it first?

Article content from document graciously made available by: wyv3rn




Failure Analysis of the M16 Rifle Bolt
V.Y. Yu, J.G. Kohl, R.A. Crapanzano, M.W. Davies, A.G. Elam, M.K. Veach
Department of Civil and Mechanical Engineering
United States Military Academy
West Point, NY  10996, USA


Abstract
Recently, there have been several occurrences of failure in the bolt of the M16 rifle at a United States Army installation.  Near the failure location, the bolt was subjected to repeated loading as the M16 was fired.  In order to determine the stress distribution of the bolt due to the firing process, a geometric element analysis was performed using ProMechanica®.  The fracture surface was examined using both an optical stereomicroscope and a scanning electron microscope in order to determine failure initiation and failure mode.  It was discovered that the fracture initiated at a localized corrosion pit and propagated by fatigue.  A controlled experiment, consisting of firing 1800 rounds using new bolts, showed that a region of wear developed near the site where fracture occurred in the failed bolt.  This suggests that exposure of the base metal may have facilitated the formation of corrosion pits.  In addition, Vickers microhardness profiles were taken on cross-sectional areas near the fillet region and 10 mm away from the failed locking lug.  Disparities between microhardness profiles near the fillet region and 10 mm away from this region revealed that the bolt may not have been uniformly case hardened.
Keywords:  failure analysis, abrasive wear, corrosion, geometric element analysis

1.  Introduction
The M16 rifle was fielded into the U.S. Army in 1968 during the Vietnam War.  The rifle has since been the primary assault rifle used by U.S. soldiers.  The M16 has been through several modifications in its more than 30 years of service in the military.  In July of 2003, an increasing trend in the amount M16A2 bolt failures was observed at a U.S. Army installation.  Figure 1 displays the data of bolt failures observed over a five year span.  These rifles were used over the past nine years during summer military training.  As a result, this paper investigates the leading cause of catastrophic fracture of the bolt under firing conditions.

This study used both a geometric element analysis and a metallurgical analysis of the bolt.  The goals of the methodology used are that 1) the geometric element analysis would reveal whether any elevated stresses existed in the bolt which would facilitate crack initiation and propagation; and 2) the metallurgical analysis would determine the fracture origin and failure mechanism.  The metallurgical analysis would also determine whether mechanical properties of the material were insufficient for the designed operation of the bolt.

A controlled experiment was conducted which consisted of firing 1800 rounds using new bolts.  After 1800 rounds, it was observed that there existed wear patterns which exposed new base metal to the environment at the same location as the failure initiation site on the fractured bolt.  This exposed base metal may therefore serve as a site for corrosion pitting.

2.  Geometric Element Analysis
2a.  Procedure.  In order to analyze the stresses that the bolt experienced while firing, a three-dimensional model of the bolt generated in Pro-Engineer® was used [1].  Figure 2 displays the three-dimensional model of the bolt.  Subsequently, Pro-Mechanica® was used to post-process the model in order to calculate the von-Mises stresses in the bolt.  Pro-Mechanica® differs from traditional finite element packages in that it does not use linear shape functions.  Instead, Pro-Mechanica® fits polynomials up to 9th order as the shape function and is termed geometric element analysis.  Thus, geometric element analysis offers accurate computational results even if the mesh is coarse, since the polynomials offer better convergence to the shape functions.  As a result, the generated model of the bolt does not use purely linear shape functions since it can incorporate complex polynomials to fit the shape function.   Furthermore, since a coarser mesh can be used to generate an accurate approximation of the model, bi-linear quadrilaterals were not solely used; instead, a mix of triangular elements and bi-linear quadrilaterals were incorporated into the model as shown in Figures 3a and 3b.

From historical data at the U.S. Army’s Testing and Armament Command (TACOM) and the Army Research Laboratory, a stress of 414 MPa was used to model the instantaneous force of the propellant combustion of the 5.56 mm round in the M16 rifle on the face of the M16 bolt.  The conventional method for converting this type of dynamic process to a static analysis assumes that during the actual firing of the weapon, the pressure in the cartridge after combustion is dissipated by the rearward motion of the bolt.  In order to conduct a static analysis of the bolt, half of the cartridge pressure was used to model this dissipation of energy [2].  Therefore, a stress of 207 MPa was used as a distributed load on the face of the bolt in the model.  This force modeled the impact of the propellant igniting and exploding within the combustion chamber without incorporating the effects of recoil and the buffer assembly in the rifle.  In addition, boundary constraints were placed on the bolt which allowed for minute deformations that the bolt would experience when the cartridge exploded in the combustion chamber.

2b.  Results and Discussion.  
The von-Mises stress distribution in the bolt showed high stress concentrations present at the fillet of the locking lugs as shown in Figures 4a and 4b.  In particular, higher stress concentrations were present in the locking lugs which were immediately adjacent to the round extractor.  These two specific locking lugs experienced stresses on the magnitude of approximately 1070 MPa as shown in Figure 4b.  All of the five fractured bolts analyzed at the Army installation had fractured at these specific locking lugs.  Figure 5a shows a picture of a fractured bolt specimen and Figure 5b shows a picture of the fractured specimen at higher magnification.  In addition, these extremely high stress concentrations contributed to the crack initiation which is evidenced by the picture of a crack growing from the locking lug next to the round extractor, as shown in Figure 6.

3.  Metallurgical Analysis
3a.  Procedure.  The M16 bolt was also analyzed from a metallurgical viewpoint.  This analysis determined whether additional factors other than stress concentrations contributed to the bolt failure.  A chemical analysis of the bolt was conducted to determine if the material specifications were met, as displayed in Table 1.  A stereomicroscope and a SEM were used to locate the fracture origin and to evaluate the fracture surface.

Vickers microhardness indentation was performed on a cross-sectional area near the fillet of the lug and at approximately 10 mm away from the lug on the bolt.  Indentation profiles, consisting of five indents for each location, were taken which started 0.5 mm from the surface of the bolt and proceeded inward every 0.5 mm.  Hardness readings are shown in Table 2.

In addition, a controlled experiment was conducted where three new bolts were subjected to the firing of a total of 1800 rounds.  The experiment entailed firing the bolts in 300 round increments and subsequently cleaned with Royco 634 cleaner, lubricant, and preservative (MIL-PRF-63460D AM6) after each iteration.  After the 1800 rounds were fired, the surface of each bolt was then examined using a stereomicroscope to detect any surface anomalies which might have occurred.

3b.  Results and Discussion.  
Chemical analysis of the bolt composition revealed no significant differences between the failed bolt and Carpenter Steel 158 specifications [3], as shown in Table 1. Micrographs from a SEM revealed that the M16 bolt experienced corrosion in the form of localized pitting, as shown in Figure 7, near the locking lugs adjacent to the round extractor.  From the SEM micrographs, the circumference of the fracture surface of the ruptured locking lug possessed shear lips.  The existence of the shear lips signified ductile failure at the surface.  However, the region at the corrosion pit did not have this characteristic shear lip.  The absence of the shear lip at this location indicates that the bolt material was discontinuous at the surface.  This discontinuity suggests that the corrosion pit is where failure initiated.  The corrosion pit provides an additional stress concentration which aids in the initiation of the crack.  In addition, the SEM micrograph displayed the presence of chevrons as seen in Figure 7.  The chevron markings point back to the localized pit which further confirmed that the pit was the site for crack initiation.

Near the initiation site, the fracture surface was transgranular with faint fatigue striations indicating fatigue crack growth, as shown in Figure 7 and 8.  Approximately 2.5 mm from the crack initiation site, the fracture surface transitioned from a smooth surface to a dimpled surface.  This dimpled surface signified that the crack experienced unstable crack growth, or ductile failure, in this region.
The Vickers microhardness indentations taken at both locations show that the hardness reading is higher at the surface than towards the center of the bolt.  This demonstrates that the surface was case hardened.  However, the Vickers microhardness at the surface near the lug’s fillet was 100 units less than the hardness readings 10 mm from the fillet region.  This indicates that the bolt was not uniformly case hardened.  Thus, the softer region near the locking lugs is more susceptible to wear.

After 1800 rounds were fired using the new bolts, wear was observed which exposed the Carpenter steel 158 base metal to the environment, as shown in Figure 9.  This area of observed wear on the surface of the bolt was in the same location as the crack initiation site on the fractured bolt, namely in the fillet region of the locking lugs adjacent to the round extractor.  The base metal exposed due to the wear makes this specific area highly susceptible to corrosion pitting.

4.  Conclusions
The fracture of the M16 bolt resulted from a cumulative effect of high stress concentrations at the fillet radius and the additional stress concentration imposed by the presence of localized pitting at the surface.  The bolt possesses many   fillet regions which impose numerous areas of high stress concentration.  In particular, two fillets experienced higher stress immediately adjacent to the round extractor due to the non-contiguous feature of the bolt.  These two specific areas of high stress concentration also corresponded to the same location where failure of the bolt occurred in all fractured bolt specimens.  Micrographs obtained from the scanning electron microscope of the fractured surface showed localized pitting at the failure initiation site. In addition, transgranular crack propagation near the pit formations in the fillet regions was observed.  The localized pits formed near the locking lugs also served as high stress concentration points.  The presence of pits in the material amplified the stresses of the bolt in the locking lug region which already had a high stress concentration due to the irregular geometry of the bolt.  This cumulative stress concentration provides a good indicator why the crack initiated and propagated from this region.

The wear observed in the controlled experiment indicates the mechanism of why the corrosion pits formed near the locking lug fillet by exposing the Carpenter Steel 158 base metal to the environment.  Vickers microhardness readings near the fillet region show that the bolt was not uniformly case hardened.  Comparison of the microhardness readings near the fillet region and 10 mm from this region show a disparity of approximately 100 units.  The softer, less carburized region near the fillet contributes to the formation of a wear area after firing just 1800 rounds.

5. Acknowledgements
The authors would like to thank Mr. Victor K. Champagne, Jr. and the Materials Analysis Group at the Army Research Laboratory in Aberdeen Proving Grounds, MD for helpful discussions and for performing SEM work.

6. References

[1]  Three-dimensional Pro-Engineer® model of M16 bolt from U.S. Army Testing and Armament Command, Rock Island, IL.

[2]  Individual Weapon Systems & 3-D Technical Data Development Team, U.S. Army Testing and Armament Command, Rock Island, IL.

[3]  Alloy data Carpenter No. 158® Alloy, Carpenter Technology Corporation, 1981.

_DR:

I'm do not believe it is proper to post their work, names, e-mails and phone numbers on this forum in such a manner without permission (spam/data miners, etc.).  At the same time the document shouldn't be posted w/o credits.  Also your post is missing all the excellent figures/pictures that really visalize everything for the reader.  I know your intentions are only good, but for the reasons stated you should probably delete the reprinting of the material in your previous post and just let everyone download the work in its original form, read it and see the figs/pics that go with it.

Windows XP has a zip/unzip utility built in.  For those with previous versions, free zip/unzip tools can be downloaded on the internet from Download.com, or found by google.com.

For those that don't have a fully fledged copy of Microsoft Word, they can download the viewer here:
Microsoft Viewers

Quoted: _DR: I'm do not believe it is proper to post their work, names, e-mails and phone numbers on this forum in such a manner without permission (spam/data miners, etc.).  At the same time the document shouldn't be posted w/o credits.  Also your post is missing all the excellent figures/pictures that really visalize everything for the reader.  I know your intentions are only good, but for the reasons stated you should probably delete the reprinting of the material in your previous post and just let everyone download the work in its original form, read it and see the figs/pics that go with it. Windows XP has a zip/unzip utility built in.  For those with previous versions, free zip/unzip tools can be downloaded on the internet from Download.com, or found by google.com. For those that don't have a fully fledged copy of Microsoft Word, they can download the viewer here: Microsoft Viewers

What about now? edited names/numbers out, retained credit for institutions responsible. Added tables/figures in sequence. Just thought that since this is all funded by taxpayers dollars and is not classified that it was therefore public domain. maybe make it easier for those less computer savvy to see it. This would actually be an excellent article for the AR15.com articles section.

However I understand your concern and if you wish I will delete my entire post. Just say the word and it's gone.

_DR

It would be acceptable to me now if it was my work and now their phone, email, etc. is safe(r) from automated spam/data mining tools so I say leave it and see if anyone complains.  Thank you _DR.

ETA: Hmm, actually could you add their names back?  The more I think about it, that should be ok.  Thank you.

Quoted: It would be acceptable to me now if it was my work and now their phone, email, etc. is safe(r) from automated spam/data mining tools so I say leave it and see if anyone complains.  Thank you _DR.

Thank you for making it available.
Name credits restored per your request.

I appreciate all the thank you notes, but these guys did the real work.  I'm just making putting it up for download.  Besides, this is really just returning the favor for all the help I've received from all of you on the forums.  Keep the free flow of information going!  THANKS TO ALL OF YOU.

Many thanks for posting this very interesting and useful data.

It would be neat if we had a study of similar methodologies pertaining to bolt breakages at the cam pin hole.  I suspect the failure mode is often similar:

Base metal becomes exposed, corrosion pitting occurs, a crack forms and propogates from its origination point.

CLEAN (not too much!) AND RELUBE YOUR RIFLE!

w00t!

Good info.  Need to read and compehend it later.

Quoted: I cannot believe minor corrosion could cause such a catastrophic failure.  We have used these things in jungles for 40 years in monsoon conditions.  Something else is in the mix. As far as the failures per year at the installation graph, how much more traning is being conducted today vs 1999? More rounds fired=more failures.

Inproper, non-uniform case hardening.

I have inspected a bolt (Colt) with over 3k rounds on it in the area where the extractor is said to cause wear and there is none.  Why?  Because the case hardening is uniform and correct.  It is possible that for a small number of years in the late 90's some batches of bolts were not being properly, uniformly case hardened.  This allows the surface to wear and the base metal to be exposed, allowing corrosion pits to form if exposed to the elements.  Corrosion pitting in a high stress area that is already at the max in its' ability to hold up against said stresses as it is, can be expected to cause failures.  Notice the broken specimen was not uniformly case hardened.

Now if it is properly, uniformly case hardened, it might still wear, expose base metal, corrode and break but later (say 10-20k rounds) instead of showing wear at as little as 1,800 rounds.  I've seen many bolts with broken lugs on either side of the extractor and the path of the crack looks almost exactly the same.  The origination point is likely the same.  The failure mode is likely to be the same, but it just happened later in the bolt's life.

This coincides with the rumors I've heard about manufacturers being put on notice awhile back about case hardening procedures, which were supposedly revamped along with the QA processes that go along with it to ensure every bolt is properly, uniformly case hardened.  Basically it has been getting a little extra attention.  HPT/MPI is not the only QA a bolt goes through.

Likely case hardening would occur after shot-peening to provide the most robust surface finish possible. Interesting that case hardening is a process used on small arms since the mid-1800's that is still proving it's value in the space age.

I have always believe that an addtional persistent lubricant such Tetra grease applied in a thin layer to critical high-heat, friction bearing contact areas such as the cam pin and cam pin hole can make a difference when a bolt is pushed to perform under extreme use. This data does not seem to contradict the idea that stress and heat reducing lubricant may stave off failure at least for a time.

Now what would be very interesting is a similar study done with a gas-piston impinged system where the temperatures of the bolt group are typically far lower than with a direct-impinged system such as the M16, with extended operation.

I had some tests run on some failed 50AE bolts back in the late 90's and the results showed that the carburized depth (case hardening of the surface) was all over the place.  If the carburized depth was too deep, the lugs would break off.  If it was too shallow, the lugs would curl inward.  I did not see a chart showing the hardness at various depths around the bolt in this analysis.  When they analyzed my bolts, they sliced very thin cross sectional waffers through the bolt body and analyzed the hardness at various depths.  I am sure adding some pitting to the surface doesn't help things either.

The bolts are actually heat treated to the core first, as hard as Carpenter 158 can get.  Then they are carburized, to harden the surface.  But if you over do it, they will get harder, deeper into the core.  I know because they screwed up and over carburized 100ea of my bolts and scrapped the whole lot.  That sucked.

Tony Rumore
Tromix Corp

It is a good read, if you are into such thing, but one thing I notice is that the photographs apear to be of old worn M16A1 bolts rather then M16A2 bolts.

Another odd thing is the graph.  It does not include info on the age of the weapons, or the size of the pool.  First thought I had was that their rifles had reached the end of their service life, duh.  About as informative as the pirate graph chart with temperatures.

Hey, me again, still thinking out loud.  I got an explanation on the pictures of busted M16A1 bolt, maybe these are bolts out of M16A1 remanufactured as M16A2's.  The upgrade kits come less bolt groups.

Quoted: Anl like I said before, what about the change in optempo due to the global war on terror? How much more training has been going on per annum since 9/11?

Yes, we are on the same page, the info left out is important in understanding the large number of failures, and the pronounced increase in occurrences.  There is no mention that the bolt pictured is 40 to 23 year old.  Just the same the info on how they fail is very interesting.

This is quite an interesing post......tagged for further reading.

Fortuantely the bolts redundant locking lugs provide a generous safety factor.  Still the increase in bolt fractures indicates a QC problem.  No wonder these are like hen's teeth today.

Quoted: Fortuantely the bolts redundant locking lugs provide a generous safety factor.  Still the increase in bolt fractures indicates a QC problem.  No wonder these are like hen's teeth today.

Still, a broken bolt lug can be caught in the barrel extension's locking lug recess or down in the trigger assembly and jam things up good.

Ekie, excellent observations re: M16A1 bolts.  I can't believe that didn't register with me, I usually don't miss things that obvious.

Quoted: Ekie, excellent observations re: M16A1 bolts.  I can't believe that didn't register with me, I usually don't miss things that obvious.

Well, I have had that report for over 6 months, and it was not until your thread that I noticed that the failures were dealing with M16A2's.  Am now thinking, what, that ain't no M16A2 bolt.

What's the difference between M16-a1 and M16-a2 Bolts?

Quoted: What's the difference between M16-a1 and M16-a2 Bolts?

Bolt lugs have a radius cut on the corner.  1960's and 1970's bolts had the radius cut in such a manor that it left a 45 degree mark in the bolt body.  Later bolts lack these cuts in the bolt body.  The busted bolt in the picture is an old bolt with the cuts.  

We talked about this subject here:

www.ar15.com/forums/topic.html?b=2&f=29&t=146824

Need to take a closer look, but my first impression is that the busted bolt pictured has been refurbished.  Looks like it was rusty and pitted at one point, and then refinished.

Thanks Ekie.

Wow!  In some of the failed bolt pictures you can actually see how deep the casehardening is!  That's metalurgy that just about anyone can understand!  And the stress locations on the bolt lugs makes a lot of sense-and that's why EVERY TIME I shoot I clean, inspect and lube the bolt (along with the rest of the rifle-I have a flat spot on the back of my head from "reinforcement" for doing a good job at rifle cleaning!), paying particular attention to places like the lug roots and the recessesed bolt face.

Thanks wyv3rn for the wonderful document!

Quoted: Need to take a closer look, but my first impression is that the busted bolt pictured has been refurbished.  Looks like it was rusty and pitted at one point, and then refinished.

Well.. I'm trying but I don't see that.

Quoted: Quoted: Need to take a closer look, but my first impression is that the busted bolt pictured has been refurbished.  Looks like it was rusty and pitted at one point, and then refinished. Well.. I'm trying but I don't see that.

That looks like pitting, or maybe carbon?  I would need a magnifying glass to get that close a look at one of my bolts though.