Thought some of you might be interested in these.

Here's a trace, Rem M700 22" 30-06, 58.7 gr. IMR 4350, Speer 165 gr., 3.390" COL, Fed 210 primer. Temperature was in the 30's. Velocity was 2700 fps, though it will break 2800 in warm weather. Judging by the velocities, I'd wager the trace is actually reading about 5000 psi high, but I elected to not adjust the calibration.


Here's another, same gun, 54.1 gr. Big Game, 182 gr. cast bullet, 3.133" COL. I have also tried 4350 and it does the same thing.


A 165 gr. Barnes X and 4350 had absolutely no secondary spike. This is all consistent with the prevailing "catch up" theory of secondary spikes. The slippery cast bullets do not provide enough resistance to make the slow powders burn correctly.

For more details, see mountain molds forum.

This is contrary to the conventional wisdom developed by NRA staffer Col. E. H. Harrison, who preached that slow powders like 4831 should be used for heavy cast bullet loads in the 30-06.

This is interesting, especially since the first load is a very common, ordinary load. I haven't seen any high secondary pressure spikes yet on my PT system. The second one doesn't suprise me at all, I would have used a fast powder with a cast bullet anyway.

The secondary spike is much more likely from another cause.

By the time the bullet gets to the muzzle, all the oxidizer in the powder is gone. The gas contains quite a bit of hydrogen, and is hot enough to glow. As the gas hits the air, it now has oxygen again, and that allows the hydrogen to explode.

Nobody knows for sure.

The re-ignition thing is well known in artillery. Since muzzle flash will gave away your position, they use additives to suppress this phenomonon.

The traces in the second figure show PSI rising to 107K plus. Wouldn't PSIs like this destroy the distal barrel?

It would be easy enough to determine if the spike occurs before or after the bullet leaves the barrel, simply by plotting the trace against a quickload simulation.

I'm just not buying the oxygen-detonation issue. Oxygen is not going to rush into a barrel when the gas is jetting out at 6,000 fps. Secondly, there is no significant difference in chemical composition between slow and fast powders -- most all the difference is in the coating. For another reason, I believe it was Charlie Sisk that managed to blow the last two inches off a barrel by jiggling things around "just so" on a 300 mag with 180 gr. bullets. A theory expounded based on that experiment was that the bullet was running so fast it ran into too much resistance for the barrel to handle. I'm not standing in line to buy that theory, either.

Intuitively, the improper ignition theory is attractive.

Denton, we should be able to simulate this with the HV bullets, as they are some of the most easily engraved around. Sounds like an experiment coming on, to me. FWIW, Dutch.

Dutch...

Unfortunately, yes, it will become an experiment. And HV's are far too precious to use for this purpose!!

There are problems with every theory that has been advanced, including re-ignition.

Sisk did indeed blow the last two inches off a barrel that had been drilled for muzzle braking. He ignored the secondary spike, and kept increasing his load, until the end of his barrel blew off. He had it recrowned, and redrilled, and blew it off again. The spike is real pressure.

One other theory put forward is that it is some kind of electromagmetic inteference. This does not match the known data.

The spike is more pronounced with lighter bullets, and, I think, when the gun is hotter. There is a clever trick for suppressing re-ignition, and as soon as the weather gets warmer, so the effect is easily detected, I'm going to give it a try...probably using 123 grain SKS bullets in my Finnish M39. I know that produces the spike.

If you give PT your muzzle speed, it will mark your approximate bullet exit point on the trace.

Denton, nah, even HV's are cheap, compared to ignorance.... I've got a few extra's, like I mentioned before.

Here is the thing I can't make fit. Usually, one presumes a pressure gradient, with the highest pressure at the chamber end, and the lowest pressure behind the bullet.

In this case, there is indication that the pressure at the bullet side is higher, since that's the part Charlie blew up. That's "incongruous" with what we assumed to know.

One other thing we have not thrown into the mix is that I have seen it mentioned that there is a "pressure wave" within the combustion chamber. Could this be some bizarre form of wave amplification? Perhaps increasing pressure locally by greatly increasing the burn rate of powder?

Two pressure wave pulses; one bouncing off the bolt face, one off the base of the bullet, meeting so they amplify, rather than cancel out? FWIW, Dutch.

The pressure wave thing is a possibility, but I see more problems with that that with re-ignition.

Heck, if you look at the CO2gas coming out of a jet engine, you have the conditions for a laser. It's called a gas dynamic laser, and, fundamentally, all you do is slap mirrors on either side of the stream, to extract energy. You get a huge amount of power out of it, megawatts. There was a lot about that in the open literature a few decades ago, and suddenly, it got very quiet. A lot of it may have become classified, I don't know.

If that's possible, you might do the same with the gas coming out of a rifle.

If thats possible, who knows what else might be?

My present guess is that the gas ignites outside the rifle, and that what we are seeing is concussion from the explosion. The fast risetime of the spike is probably even faster than what we are seeing on the trace. I suspect that all we are seeing is the risetime of the PT.

Couple of more shots at your "out of the barrel" hypothesis.

1) It would be obvious to the shooter, as the pressure (i.e. sound) would go from 10,000 PSI on a normal shot) to ten times that. Someone would notice, and shooters do not report anything out of the ordinary

2) How can you blow then end off a barrel when the pressure outside the barrel is the same as inside? If the detonation takes place outside the barrel, then the pressure upon the outside of the barrel is the same as the pressure upon the inside of the barrel! Compression isn't going to ring a barrel, or blow it off.

Still, a quickload trace should be able to tell us if it is "in or out", and so will an experiment. Dutch.

I highly recommend all of you call IMR/Hodgdon Powder company and talk with their ballistic engineers (call 913-362-9455). They have been very friendly and answered all my questions, when I've called. Calling them will not cause you harm, although you may not believe what they tell you.

They work in the a world were a MISTAKE = BODILY INJURY = LAW SUIT = GONE OUT-OF-BUSINESS. Yes, very serious - their livihood depends upon their expertise and mistake-free load recommendations.

They use the "best system available" (sic) for measuring internal chamber pressures - that is, the Oehler system 83, which is a piezo electric transducer system measuring internal chamber pressures. They don't use the Oehler system 43. Why? It's an indirect measure of internal chamber pressure, which they feel is much less accurate than the internal piezo electric system, especially at high and low PSI. In fact, they believe the Oehler system 43 is just a "crude estimate" of true chamber pressure, which doesn't correlate well with the system 83. They also say the NECO Internal ballistics program is a crude estimate - "but, it won't get you killed." (By the way, with the Oehler System 83, they don't see these peculiar secondary spikes, which is some artifact of the "stretchy" system 43. There are serious calibration/standarization issues with the system 43, which are obviated by the system 83.)

The cost of the Oehler System 83 is roughly $12,000 start-up with one barrel and universal receiver, and then $1400 for each additional barrel setup with the transducer. For those of you with money, these are not prohibitive costs.

For those of us, who don't have the System 83, they believe in monitoring pressure signs - one of the most reliable being extracter marks and loose primer pockets occurring at roughly 65,000 psi. At 70,000 psi you'll start popping primers with vast majority of primer pockets becoming toast in a new case with one firing at roughly 80,000 psi. Above ~80,000 psi you risk case-head separation and destruction of your gun by the massive release of gases back into the action - don't go there.

The say whatever system you're using to measure pressure always watch the physical signs of pressure developing in your cases as you increase powder charge, change bullets, shoot at high temperature, change primers, change powder lots, etc. The case is in the chamber! Use nothing but first-rate equipment and components. They also like using the "slowest burning powder possible" approach to achieve highest possible performance , while staying within SAAMI recommendations.

Denton,
I heard, through the grapevine, about charlie's barrel issue.... Gun cranks in houston are a fairly small frat!!

that's one of the reasons I bought the RSI.


much nicer to know what's what that guessing at primers and cases.. though that's still fairly valid!!
jeffe

Jef, extracter marks and loose primer pockets are not hard to recognize consistently and reliably.

quote:

Originally posted by Ackley Improved User:
Jef, extracter marks and loose primer pockets are not hard to recognize consistently and reliably.

and are at least 15% higher than I will ever MEAN to go

a blown up rifle is clear too.. and not something I want

jeffe

quote:

By the way, with the Oehler System 83, they don't see these peculiar secondary spikes,

Odd, since they blew the end off Charlie Sisk's rifle....twice. Must be an electronic artifact, that isn't really there.

It seems that the spike occcurs after the bullet leaves. At that point the rifle hasn't moved much (0.1 in ?). It could be signal caused by the wires moving around (tribo-electric effects). Some people have reported 10,000 psi due to the firing pin falling.
Try tying the wire down better. There is such a thing as "low noise wire", but it's tough to use. You can verify what I have said with an oscilloscope, just watch while you wack the wire.
Good luck1
Good

hawkins...

What you have said is absolutely correct, and is one possibility. It doesn't even depend on triboelectrics... take a pair of wires connected to a high impedance device like an old VTVM, and quickly flip them apart. The capacitance between the two wires drops dramatically, and because Q=CV, the voltage shoots way up as C goes down, and Q has no place to go. As you say, the same thing happens on a scope.

All theories I have seen have one or more problems, so I'm not trashing your statement, which is correct... just trying to reason through.

The place this theory runs into trouble is that we can make the spikes come and go, without changing the type of wire. Switch to a heavy bullet, and it's gone. Switch back to a light bullet, and it's there again, with no change in wiring.

It's more of a factor in longer barrels, and with lighter bullets. My 27" Finn generates some real beauties with 123 grain SKS bullets.

quote:

Some people have reported 10,000 psi due to the firing pin falling.

It does happen, just as you say. Some kind of vibration is a possibility.

Hard to blame the system, since you can make it appear with a change of bullets, and go away with a heavy crimp.

It's also hard to blame the wires, since by simply switching to a faster powder, the symptom goes away.

It's also kind of silly to categorically state that they have not seen the spike with the Oehler 83. Have they shot cast bullets at high pressure with slow powders much lately.

and then there is that shot-off barrel tip, again. FWIW, Dutch.

SEE ? =

Secondary explosion effect.

Slow, single base powder / reduced load is the usual scenario, some powders more prone to this than others.

Lloyd Brownell explores amongst other the theory of pressure wave amplification down the barrel.

This based on erratic and or incomplete combustion of propellant ( due to low temperatures and relatively low pressure)

Thus setting off a series of presssure waves and when two superimpose down the barrel you get a secondary pressure impulse, he talks of an in barrel sonic boom so to speak.

Evidence of something like that is as plain as can be in Brownell's data, though he didn't exploit it as fully as he could in arguing with people who insisted that reduced loads of 4831 couldn't blow up a gun.

If you look at his graph of pressure vs. charge, there is a dramatic increase in variation of pressure below about 30 KPSI, including one point that he put a "?" by.

To get that kind of amplification, the wave would have to pass through a region where it could coherently pick up energy, as photons do in a laser. I haven't been able find a way to model that.... doesn't mean it isn't so, it just means I haven't figured it out yet.

ALF, just my opinion, but I think the slow powder/small load/diassemble your gun thing is probably different from this. The 4831 thing is characterized by random spikes all over the pressure curve. That's not what we're seeing.

quote:

Originally posted by Ackley Improved User:
...IMR/Hodgdon Powder company ...ballistic engineers ....They don't use the Oehler system 43. Why? It's an indirect measure of internal chamber pressure, which they feel is much less accurate than the internal piezo electric system, especially at high and low PSI. In fact, they believe the Oehler system 43 is just a "crude estimate" of true chamber pressure, .... the "stretchy" system 43. There are serious calibration/standarization issues with the system 43, ....

Must admit it does look like the Experts know what they are talking about.

But, to help simplify it for the regulars of this Board:

A HSGS not Calibrated to a Known Standard = Fool's Gold (aka Reloader's Pyrite)

Don't you just love it when the Experts verify what you have said all along.

Thanks to everyone for the intelligent and thoughtful responses.

If you look at the first graph, it does have the "+" marks indicating the calculated point at which the bullet exited the barrel -- after the spike, not before.

PT refused to show the exit points for the cast bullet load, but Quickload predicted 1.32 ms. BTW, the actual velocity averaged 2569, QL predicted 2597 fps @ 47,800 psi. That's far different than what Western Powder suggests for Big Game.

I'd encourage anyone else who has a trace unit and a rifle that they don't mind abusing to repeat my experiment. Try a Barnes X, a jacketed, and a cast all with the same powder charge. Let us know what you find.

We all know that slow smokeless powders need pressure to burn efficiently. Some powders just won't burn right below a certain pressure level. When the bullet first begins to move, the crimp and engraving forces may provide the required resistance. Then inertia forces come into play. But after the bullet has traveled 10 inches or so, the inertia forces have dropped off, and friction, or the lack thereof, may be an important factor in maintaining an efficient powder burn. I suspect that is what we are seeing with the Barnes vs. jacketed vs. cast comparison. The cast bullet doesn't have enough friction to maintain the critical pressure, so the flame goes out, so to speak, only to reignite near the muzzle.

We should not fail to consider that differences in initial engraving pressure could explain the differences in the secondary spikes, but so far I have not seen any evidence to support that, i.e., whether the bullet is seated to engrave hard or whether it has to jump to reach the rifling doesn't seem to have an effect on the secondary spike. Perhaps more experiments with crimps, magnum primers, and seating depth are in order, just to be sure.

Hot Core, I am not going to fully engage in the husband-and-wife argument concerning the accuracy of the strain gage system. So far, my impression is that the Pressure Trace provides good resolution but the accuracy is pretty iffy, with the adhesive being the weakest link. It can be challenging to get the strain gage to adhere correctly, and if there is any failure with the adhesive, it can cause the indicated pressure to be off -- usually low, but sometimes high. Let's not quibble over a possible 5000 psi error in the calibration, because a 5000 psi error can't explain the huge secondary spikes.

hell hotcore,
just change your tag line to "anyone that looks behind the curtain with anything other than a $100,000 piece of lab equipment is a fool.. and that applies to CHronos TOO"

or, HC, do you "calibrate" your chrono everytime?

jeffe

popenmann...

You're right. I hadn't noticed that the trace was tagged with an exit point. That sort of does make it difficult for the re-ignition theory, doesn't it?

Usually, when I get the spikes, the exit point is just as the spike is starting up. I guess I saw, or failed to see, what I expected to see. I had simply been assuming that the exit point calculation was a little off. It would have to be a LOT off to be consistent with what you are showing here. That's harder to believe.

One knowledgeable person thought it was barrel vibrations. He designed his system with reduced bandwidth, to suppress them.

I've wondered about torsional vibration... the bullet probably puts quite a bit of torque on the barrel. There are ways to measure that.

So far, on this issue, my score at totally satisfactory explanations = 0.

One theory that I drempt up about 'secondary spikes' (and I am still in the "its an artifact" crowd because a 107kPSI spike like was shown will show effects to brass and rifle pretty quick) is thus:

Compressed or near compressed loads do not allow the primer explosion/flame to reach the front of the powder column (shaky assumption, I know, but IF its true???)...so the front portion of the powder is pushed rather violently into the bore unburned, getting cracked and crushed along the way...as the flame from the rear burning powder reaches the crushed powder, it ignites VERY explosively as it is probably near critical combustion temperature by now anyways. A very small explosion at/near the rear of the bullet could cause a pressure wave to resonate back through the gases trapped in the bore.

As for blowing off the end of a rifle, well if it is drilled/threaded for a brake, it is much weakened anyway, so maybe that was the weakest link? I also don't criticize anyones data, but a sample of 1 is almost useless...

AND, as for Oehler 43 vs. 83 showing vs. not showing a secondary spike, well if the piezo system has to phsically move a rod to 'shake' the crystal, that mass has a definate dampening effect, so a VERY short duration event could easily be missed or drastically reduced...

BUT I still think the barrel hits a harmonic with some loads where the stretching and contracting from the gas pressure just causes the reading to spike. But that is just a semi-educated theory...

Just something to chew on. I have too much time to think on my way to/from work these days!

Is really "pressure" that instrument is reading with these pressure spikes. I ask for two reasons.

Firstly, the use of of powders that are too slow has been going for a long time, especially with double rifle users. I have done a lot of it myself with two previously owned 460 Wbys and with 400 grain bullets. I think some of the double rifle users have cautioned against using faster powders and prefer the low pressure loads of 4831.

Secondly, my limited understanding of how you are measuring pressure is based on the expansion of the chamber. If this is the case then one would think that readings of 100,000 PSI would be reflected on the brass.

Mike

popenmann,

What is the trace like if you place the strain gage 4" rearward of the muzzle? How about 6" back from the muzzle, then try 8" back? The results may be surprising.

ASS_CLOWN

quote:

but a sample of 1 is almost useless...

Well, he actually blew the end off the same rifle twice, so it's technically only half useless....

Yes, being drilled for a brake certainly makes it the weak point. It might also make it a venturi... maybe.

If anyone wants to test the re-ignition theory, I think this experiment might answer some questions:

Fire and record several shots, to make sure that secondary spikes are being generated. Then put a large chunk of dry ice in a garbage bag, and let it inflate, driving out all the oxygen. Poke a little hole in the bag, and put the muzzle through the hole. Fire a few more shots, so that the bullets travel through CO2 for the first foot or so. This deprives the exiting gasses of oxygen. If the spikes disappear, then I think that means that it was re-ignition. If they continue, then I think re-ignition bites the dust.

I can't argue with people who say that pressure of that magnitude should leave evidence on the brass...

denton

On the muzzle brake issue a mate of mine had several rifles magna ported in Australia with what we called double magna porting. There were two extra slots. This was done on calibres in 7mm and 30 and on barrels that were .72" at the muzzle on 26".

He puchased one of the Made in Japan Weatherby in 378 which had extremely thin barrels. If you have not see one of them they almost look like a.410 shotgun at the muzzle Initially the people doing the magna porting would not do the 378. However, they changed their mind but would not do the extra two slots.

On the surface your pressure spike readings seem like a car on a slippery road and the speedo is reading 30 mph but the car is only doing 10 mph....or perhaps the duration of the pressure needs to be longer for it to register on the brass.

Mike

Charlie Sisk while working with an 8x57 found a dirty barrel didn't produce secondary spikes,after cleaning it did.

quote:

Charlie Sisk while working with an 8x57 found a dirty barrel didn't produce secondary spikes,after cleaning it did.

Now that's interesting and useful data.... Dang, I wish I knew what it it means!!

I don't see how it could "reignite" before the
bullet leaves the muzzle. Another thought to muddy the water; There is a site that claims the
strain waves in the barrel constrict the bore.
They attribute "sweet loads" to this effect.
Could have some effect here, explain the dirty/clean barrel?.
Good luck!

Forgot address. www.speakeasy.org
Take Care1

quote:

Originally posted by popenmann:
We all know that slow smokeless powders need pressure to burn efficiently. Some powders just won't burn right below a certain pressure level. When the bullet first begins to move, the crimp and engraving forces may provide the required resistance. Then inertia forces come into play. But after the bullet has traveled 10 inches or so, the inertia forces have dropped off, and friction, or the lack thereof, may be an important factor in maintaining an efficient powder burn. I suspect that is what we are seeing with the Barnes vs. jacketed vs. cast comparison. The cast bullet doesn't have enough friction to maintain the critical pressure, so the flame goes out, so to speak, only to reignite near the muzzle.

It is obvious you have an excellent grasp about what is going on. Real glad to see it so you don't end up destroying a barrel or hurting yourself. No doubt those improperly or totally non-Calibrated HSGSs can certainly be misleading as someone has mentioned on this same Board numerous times.

In fact you even mentioned your HSGS was probably wrong when you said:

quote:

Judging by the velocities, I'd wager the trace is actually reading about 5000 psi high, but I elected to not adjust the calibration.

I totally agree with you that when a Calibration to a Known Standard has not been done, the results are as you implied, "misleading".

quote:

Hot Core, I am not going to fully engage in the husband-and-wife argument concerning the accuracy of the strain gage system. So far, my impression is that the Pressure Trace provides good resolution but the accuracy is pretty iffy, with the adhesive being the weakest link. It can be challenging to get the strain gage to adhere correctly, and if there is any failure with the adhesive, it can cause the indicated pressure to be off -- usually low, but sometimes high. Let's not quibble over a possible 5000 psi error in the calibration, because a 5000 psi error can't explain the huge secondary spikes.

Good. No need to argue about it at all.

It is quite obvious the IMR/Hodgdon Experts have as little faith in the (non-Calibrated to a Known Standard) HSGS as I do. In fact, it appears they have even less confidence in "Home Strain Gauge Systems" in general than I do.

But what do they know, they are just the Experts with plenty of time and money to verify the accuracy of all the various Systems.

I think the secondary spike is illusion of a technical nature. One way to explore that is to find a load that generates the circumstance and fire a few dozen or more. If it really is generating those pressures in the 1-1.3 millisecond time frame there will be evidence in the bore section that corresponds to that time line. Look for cracking with a bore scope, much as you see in the throat region of any high intensity chambering after a short shooting history. JMO

As to Charlie Sisk's experience, don't know, and won't speculate. I will say that it seems logical that more than a few people would have experienced such events regardless of the difficulty of replication. There's a lot of tinkerers out there a lot less caution than Charlie Sisk.

quote:

Originally posted by ASS_CLOWN:
popenmann,

What is the trace like if you place the strain gage 4" rearward of the muzzle? How about 6" back from the muzzle, then try 8" back? The results may be surprising.

ASS_CLOWN

I haven't tried it myself, but vaguely remember reading about someone trying it, and confirming the existance of the spike. I've got a bunch of gages on order, so maybe I will be able to try it later. Ideally, you'd want to be able to read all the gages at once and synchronize the output, which I am not equipped to do.

Digital Dan, I haven't measured my bore in the allegedly affected region, and I don't have a bore scope or air gage to measure a pressure ring if it does exist, but there have been reports of verified pressure rings in other barrels. I think one case is mentioned in one of the articles on the Pressure Trace website.

The NRA Cast Bullet Handbook, which dates back about 50 years, alludes to sporadic reports of erratic pressures and ringed barrels in bottleneck rifle cartridges, so it's not accurate to say that there is no history of pressure problems with cast rifle bullets.

I am not sure that a pressure spike would neccesarily have to harm the brass. The gun seems to have minimal headspace, and the resized brass is a fairly snug fit in the chamber, so there's nowhere for the brass to go. The flattened primers and obliterated headstamps that we associate with high pressure may be due to the initial movement of the brass and the bolt, when all the slack is taken up and the case head is slammed backwards. When the secondary spike hits, if it is real, the brass is already pressed tightly against the bolt, so it's not going to be hammered against the boltface like it was during the initial pressure rise. Likewise for the bolt bearing against the receiver. For those reasons, it is conceivable that "X" amount of pressure in a secondary spike is less damaging to the brass and the action than the same amount of pressure in the primary pressure peak. That would explain why there are so few reports of guns being blown up or cases seperated by secondary spikes. It seems to me that the main concern is ringing the barrel, and if barrels are getting ringed, the rings may be so slight that they are not being noticed.

The Sisk report about spiking in a clean barrel but not a fouled barrel is consistent with the theory that increased bullet friction helps to maintain a minimal pressure required for the powder to burn efficiently. I wonder if I could duplicate that by putting lapping compound on my cast bullets? Let me write this down -- I've got so many experiments going on that I can't keep track of them all.

Guys, I sure don't have all the answers. I just wanted to make people of aware of a possible risk and maybe inspire other experimenters to get out there and collect more data. Just be careful, and don't use your most prized barrel.

Silly question, possiable audio effect of the second spike.
Eveyrone that has been hunting much has heard the double WA-Womp of a 30-06 (as opposed to the single blast most others produce) that makes it (the 30-06) fairly recognisable in the distance. Could this be a side effect of the spiking?

popenmann, you show one trace wherein the psi exceeded 107,000 (massive over pressure) - the case should show signs or that excessive pressure. Did the case freeze in the chamber, did you have a blown primer, did you have a diffucult extraction, and/or did you find prominent extractor marks?

popenmann

I can't agree on the brass failing show pressure because the slack has been taken up. One reason being that cases that have a very tight to chamber because they have only been neck sized for a few shots will show pressure signs. Certainly flattening of primers is different with fireformed brass to brass with headspace.

To my mind the fact that your pressuring measuring equipment is really measuring expansion of the chamber but the 100,000 is not showing up on the brass means either:

1) The duration of the pressure is not long enough to register on the brass or

2) For some reason the instrument is giving a false reading.

Let's keep in mind that pressures of over 100,000 would mean more than a shiny spot on the case head.

As to ringing of the barrel with reduced loads all the reading I have seen on that has been in relation to the use of some fillers.

Mike

quote:

so that the bullets travel through CO2 for the first foot or so. This deprives the exiting gasses of oxygen.

Denton, interesting experiment, but your assumption of CO2 depriving the combustion gases of oxygen is shaky. I saw an experiment/demonstration in college whereby a wad of magnesium was ignited and then placed inside two pieces of dry ice hollowed out to just the same size. The metal continued to burn...completely.

At a high enough temperature CO2 goes to CO and releases oxygen. I do not know if the combustion gases are hot enough to do this, but before trying your experiment I would substitute a bottle of helium or nitrogen to inflate the trash bag.

For the rest...one day I am going to beg/borrow/steal a 4 channel digital oscilloscope, 4 strain gauges and place them equally spaced on the muzzle of a rifle. Power the gauges and freeze the traces on the screen. If there is a secondary spike, it should show up, and I expect to be able to see it transiting the length of the barrel, because I still think it is a mechanical harmonic of certain barrel/load combinations...hoop strain harminics if you will. Note that the traces showing it are ALWAYS right at the end of the bullet travel, and NEVER in the middle...so something with the exit of the bullet is causing it.

Another experiment that might shed some light...take a very long barrel with the second peak. Is it at the end of bullet travel also? Cut the barrel back. Is it still there, and still at the end of bullet travel?

CDH,
You make a good point concerning the use of CO2, but my chemistry is too rusty for me to remember much about the reaction you mention. This may well be a peculiar reaction with magnesium. Magnesium is one of those materials which when burning, you must not attempt to extinguish with water. Hot magnesium will strip the H from H2O, consume the O in combustion to further propagate the reaction, and leave the
H2 free to react with atmospheric O2. Remember the Hindenburg?

The reaction you mentioned with CO2 may be due to a similar effect of magnesium, rather than simply due to high temps. On the other hand, it may be due to high temps, and at the temperatures generated by 60,000 psi N2 may also become reactive with unburned H2 or other unburned combustables.

My point in all this rambling is only that I know of no reactive issues with argon or with helium, hence their use in high temperature applications such as MIG and Tig welding of steel, aluminum, and various Al-Mg alloys. Both of these gases are readily available from any welding supply house.