Internal Ballistics Modellilng spreadsheet

[stephen m weiss]

Over the weekend, I worked up spreadsheet which models bullet motion within the barrel. I validated it with data from all my shooting. It works from 100% fill in 223 and 308 to 27% fill, assuing the powder is loaded back toward the primer. It matches data to within about 5% speeds.

The low fill low energy issues seem to have originated from slow powder heating. As soon as I flagged ignition to heat and to a lesser extent density, I got great data matching. I get very good prediction of speeds from 7 gr powder to 23 grains powder in a 223 with 50-75 grain bullets. There is some chance I have over sampled the data, ie just fit the data I had.

So, I would like to get some good data on different fills, calibers, bullet weights, etc.

Interestingly, I found that low fill detonations were very dependent on the factory crimp, ie the force required to push the bullet from brass. Since these crimps are set with no real great method of measuring them, and are entirely dependent on brass length, few people mention this data so most published data is nearly useless.

Also, I am hesitant to just publish the spreadsheet generally. I figure there is a good reason the existing published prediction methods are so weak. I expect the us govt would be less than excited to have the best ballistics analysis in enemy hands.. Not that I am saying mine is the best, but I dont want to get hung up in that issue.

Is anyone willing to work with me on this, providing good data in exchange for my modelling spreadsheet? I suppose also, I should check my spreadsheet against the simple cases that are represented in commercial work. So, tell me the programs that people use, and I will cross check results.


smw, msme pe

ps i give my creds so hopefully fewer sophmoric attempts to say stick with the easy and done will be posted. no bragging, it's not even a phd, not even Harvard.. so just chill....

[C. Latch]

Are you familiar with Quickload?

[stephen m weiss]

Just checked it out. Norton said it was bad. Given that, and that they offered nothing about what it does, I am all ears but not buying it unless it has good reviews here.

smw

[C. Latch]

Norton who?

[stephen m weiss]

Norton antivirus.. That doesnt mean it is bad. just got flagged bad in their data base I expect.

smw

[725]

Norton was the guy on the Jackie Gleason show. Everybody knows that!

[C. Latch]

You really ought to look into Quickload.

[gmsharps]

Give Quickload a good look. It's got all of the work already done. Just put the parameters in for the results you want like pressure, type and amount of powder, barrel length and a host of others. It has saved me loads of time and expensive components and gives a place to start. I always double check with other paper sources just to be on the safe side. I never trust just one source, thats just me. Some people don't like it but for me it helps a lot.

gmsharps

[dromia]

Quickload is the best predictive software out there bar none.

If you put good data in then you will get surprisingly good and accurate information out.

You must study the manual to get the best out of it.

[Digital Dan]

Surprised they let a person tinker with this stuff on Long-uh Island.

[stephen m weiss]

Surprised they let a person tinker with this stuff on Long-uh Island.

Heh, well I just build the stuff at home. I shoot upstate on my 120 ac wooded hillside. That way I can collect the range scrap, shoot milk jugs, and generally do whatever I want when my wife is not there... There is certainly very little freedom to be found on LI. There is an outdoor range some 40 min away but I prefer to save my tests and conduct them at my leisure where I can hunt, fell trees, mill lumber, hike, and laugh my butt off when the chipmunks get into the coffee grounds I throw out....hahaha makes em so hyper! They get into really loud squabbles with each other over the stuff.

Ok I will check out Quickload. Is it always $150 or is there a trial offer of something?

smw

[stephen m weiss]

I found the demo. And especially the 'reflections on interior ballistics' section of the user manual appendix. Good stuff. So far they have validated many of my observations on the modellilng. I am still looking for their pressure detonation exponent justification, and which control parameter does it. But their fast rifle powder model seems to kick out the same results as my model, so thats good.

Thx again guys.

So, does anyone else have fun doing their own dynamics modelling rather than just using the canned stuff?

smw

[stephen m weiss]

I finished their model justification just as they claim, their code doesnt work very well at low amounts of powder, like 30% case fill. The reason is twofold. They don't model the primer energy or pre-bullet motion at all, but rather just assume a starting pressure after ignition and bullet motion beginning. But, a big part of the burn for a low fill occurs before bullet motion. So their model predicts up to 100% bullet energy too high for a 30% case fill, like you would want for a 1500 fps bare base bullet. Also, they show nearly double the pressure than the bullet actually sees. While that may be safe, and perhaps relevant for barrel safety, it doesnt really work when considering bullet base yield and cocking in the barrel. For that, you need the real pressure pushing on the bullet, that really causes acceleration and bullet stress. My model does that very well.

The second issue is that they use a heat effect exponent of 1. This ignores what happens when very hot particles are very far from one another, which happens with low case fill as the
explosion blasts the powder down the barrel chasing the bullet. I found that a pressure exponent on burn rate of 1.22 worked great for even low fill rates.

I need to test how well it works at lower than 30% fill. My model may even break at those low densities.

smw

[Ricochet]

Hey, what do you mean about S.E.E. being related to factory crimp? S.E.E. has long been an interest of mine and I've recently been searching and reading about it online. My understanding is that ideas about the cause are still quite hypothetical. You're certainly right that we do a lot of things that don't fit the suppositions built into the old Powley Calculator (still a useful tool) and QuickLoad. But how are you improving on that? Are you recording dynamic pressure/time curves as well as velocities? While you're looking at reduced loads, I'd love to see hard data on how different fillers affect the pressure curve. And different bullet starting resistances. As well as different primers' effects on early pressures.

[stephen m weiss]

Hey, what do you mean about S.E.E. being related to factory crimp? S.E.E. has long been an interest of mine and I've recently been searching and reading about it online. My understanding is that ideas about the cause are still quite hypothetical. You're certainly right that we do a lot of things that don't fit the suppositions built into the old Powley Calculator (still a useful tool) and QuickLoad. But how are you improving on that? Are you recording dynamic pressure/time curves as well as velocities? While you're looking at reduced loads, I'd love to see hard data on how different fillers affect the pressure curve. And different bullet starting resistances. As well as different primers' effects on early pressures.

MM.. I wasn't talking about s.e.e. My reading about that about matches what you said, and I don't suppose my model will really predict that, although you could model a stuck bullet pretty well. What it does show is that much of the powder has not yet burned by the time the bullet leaves the barrel and the pressure goes away. Then the powder probably just goes out since the steel barrel will steal it's heat. If that powder were to build up, then you could easily get an overpressure. What I was saying was that in my model, using a hard factory crimp vastly increases the amount of early pressure build and so better ignition and more efficient burns. I think really low speed loads will be more subject to powder build up since even soft lead will not obdurate and scrub the old powder out.

If anyone is interested, I can outline my assumptions and approach. I am sure it can be improved. I would love to discuss it.

smw

I am not recording actual dynamic pressure/time curves. I just am doing mathematical modelling from first principles as we learned in college. Since my focus was bare base cast bullets and I use 223 and 308 with typical powders, I have low fills. So my data reflects this. So my model does a good job at this, while the Quickload model does a good job at production style loads, 90% fill or more.

Let me be clear though. I began last weekend and got some good results. Since reviewing the Quickload software and suppositions, I have found several things I was doing sort of crappy. I wasn't using true powder density, just the dipper density. Also, I didnt even realize a 22 inch barrel was really that less the case length plus the bullet seating...doh.

My model would make it pretty easy to predict the effect of fillers on combustion. I have just guessed the energy in typical primers based on silly things people do with primers, like shooting hot melt glue bullets and such. I plan to test my primers by weighing before and after ignition. At least I can know the amount of gas released, as a function of primer design. That will give some idea of the relative energy of small rifle vs large rifle vs large rifle mag, vs military. Another boundary on primer energy is that I accidentally didnt put powder in a few test rounds using cci primers. I was really surprised when the round didnt go off, so kept the dud and inspected everything after taking it apart. The primers had gone off but were unable to overcome the factory crimp I was using. Kinda nice since I didnt have to remove a stuck bullet. But, it shows that the pressure was less than the 3000 psi or so to get the bullet to move.

I plan on doing some tests on removal force of bullets with different crimps. I have some really big spring scales and vice grips grab bullet noses pretty well.

[stephen m weiss]

Inputs:
Primer energy, powder weight, powder solid density, bullet weight, bullet frontal area, bullet-bore friction force, bullet crimp removal force, nominal burn time, exponent on pressure for burn rate adjustment, powder energy per grain, barrel length, bullet seating depth, case length.
Approach:
Step through time using about 1/200 burn time as an increment. Begin at bullet position 0. Burn a little bit of powder in proportion to the time step and in proportion to pressure raised to the burn rate exponent. Keep track of the amount burned as % of powder charge. Calculate the gas volume based on case volume, bullet position, and remaining powder solid volume. Sum the primer pressure energy and burned powder energy. Calculate the resulting chamber pressure by ideal gas law. Sum the forces on the bullet: friction force and pressure force. If the resulting force is above the seating force then allow acceleration by a=f/m. The mass is the mass of the bullet plus the mass of the powder since the powder gets blasted down the barrel chasing the bullet. Integrate acceleration to speed and in turn to bullet position. All that was one time step, repeat with a new little bit of time. When the bullet position exceeds the effective barrel length, set pressure to zero. Check the whole speed and pressure column for maximum values and output these as well as the muzzle energy.
Calibration:
I am able to shape the burn curve to match test results by adjusting the nominal burn time and burn rate exponent. I use speed data at about 30%, 60% and 90% fills. I check interpolation accuracy at other oddball values with different bullet weights to make sure results are believable.
Approach weaknesses:
The burn rates are matched to a specific powder and chamber. Clearly, if you made a chamber 100x as long but kept the diameter the same, it would tend to act like a fuse and have a mass burn rate limited by its cross section. My model ignores this currently since I am using just two cases. I need data on several more cases to fit that reasonably.
Another issue is that real powder does not begin detonation at all until it gets to some minimum temperature-pressure threshold. That means zero combustion at room temperature and even up to 150F or more. My model always burns some powder, even at very low pressure-temperature conditions. On one hand, this matches powder chunks that have been lit and are burning, but really does not match powder that may never have been lit or has gone out due to proximity with cold metal surfaces or high velocity cold air. I should at least be aware of these values to assure that I dont extrapolate my model to inaccurately predict behavior, and thereby manage to get a stuck bullet or unburnt powder build up in the barrel and potential S.E.E.
A potentially big issue is that my model sort of assumes that powder bits act like one another. I have one pressure exponent instead of a series with combination coefficients. My model will do poorly with a combination of very differently sized powders or significant surface coatings.

[Wayne Smith]

Just a note - if by "factory crimp" you are referring to the Lee Factory Crimp die, that is a variable, not a constant. It varies by the amount of pressure used to make the crimp.

[stephen m weiss]

Yes, I am referring to the sum of the effect of the Lee Factory Crimp and the bullet to case neck press fit. I just built some simple jigs for testing this. The first was grabbing the bullet nose with vice grips and using some wire wrapped around the handles to attach to a spring scale. The case was gripped in a vice. I got 80lbs extraction force on a 308 fjm using my crimp setting. Then i realized most people do not have a tension spring scale going up to several hundred pounds, so I redesigned the test to use a bathroom scale. I take a 308 case and drill a .25 dia hole thru the empty primer hole. No primer can be mistakenly pressed in now. Then I load the bullet less primer and powder. Then i insert the greasy .25 drill bit point first straight at the bullet base. Then I drill a 3/8 hole in a 2x4 bit and slide the bullet nose first into the hole. I put my weight on the 2x4 pressing teh drill bit base into a bathroom scale. I use a bit of 2x4 under the drill base to prevent denting the scale top. By this method I could quickly measure 75 lbs for a case lubed bullet and 100 lbs for an fmj. The bullet was not even badly damaged.

After lunch I will adjust the Lee Crimp to a higher crimp. I dont think it will change the value much because as long as you have more crimp than the elastic limit, it's that last bit to let the big bullet slide thru that takes the most force. We will see. More importantly is the friction force which varies like heck between a lubed bullet and a dry fmj. I examined my 223 case. It has the same thickness but only 72% of the circumference so the extraction force should be about 3/4. I gotta make a jig for that too. Same design, even same 2x4 bits, just a .219 drill bit.

The thing is, since it makes a really big difference in powder efficiency for low fill rates like cast bullets need, ya really need a way to measure and control the value. Unfortuneately, its the brass neck yield strenght and thickness that really dominates the value, and that varies from case to case and every time you fire it. They keep getting harder and harder.

Now that I have the technique down, I will test several different cases this way to get a feel for typical variations.

[rsrocket1]

I've found that Quickload is excellent for what it is meant to do and that is give predictions useful for starting points of mid range loads using a wide range of powders without actually empirically testing each one. It does break down at the edges, mostly at the low end where powder position of an undercharged cartridge can be a big factor. It also misses on muzzle velocity of extremely short (2" snub nose) barrels.

Otherwise, it is very reliable and accurate. Usually less than 10% difference between predicted MV and measured MV through a chrony, often less than 5% and often less than the standard deviation of the string of shots I am testing. So far I've compared my measured shots with Quickload using about a dozen powders in rifle and handgun loads.

As for the 30% statement.
If you are trying to load light recoil loads by using 30% capacity of rifle powders, I think you are not going about it the right way. You should be using fast shotgun/handgun powder in the 10 grain ballpark. You want the peak pressure above 10,000 psi for a good reliable burn. Quickload will show you that with the pressure/velocity vs barrel length graph. You want to see a pressure curve with a pronounced hump at the beginning rather than a slow rise to a low Pmax.

As for S.E.E. I don't buy that theory. I do buy the theory that the primer jams the bullet into the rifling, then the powder goes off (still burns and does not "detonate") and the spike causes the KB. I heard that there was a guy who could cause this pretty reliably with under charged cases with slow rifle power. I'll bet he did this by tilting the barrel down before firing. In either case, the result is a destroyed gun so I don't want to try that out myself.

[stephen m weiss]

Ah, when i say detonate instead of burn, it's because of the lack of oxidizer required and the large dependance on pressure. My testing seems to indicate that it's all controlled by smooth differential equations after ignition, so there is no sudden 'burn' of all the powder at once. I will avoid saying detonate in the future.

Stuck bullet S.E.E. That should show up on my model if I simply raise the crimp force to some large value, like 1000 lbs.

hmmm.. lemme see....

Yes, by reading the Quickload appendix, it's clear to me that they simply stopped complicating the model when it did what they wanted.

I will try to get some fast pistol or shotgun powder. After 3 kegs of powder, it seemed I should be able to work with what I have. It's kind of expensive to buy a lot, and a really bad deal to buy a little....And about frustrating to try to guess how the stuff will behave from published data. Of course for the cost of another keg, Quickload will give me a better idea, albeit not accurate at low pressures.

[stephen m weiss]

Ok, I put the bullet crimp at 4000 lbs, above that required to just shear off the sides of a hard bullet, and the barrel pressure peaked at 37ksi for a 223 at 7 grains of Alliant 1200R. The low fill just left too much empty space to create enough pressure to blow the barrel, even at full combustion. Either powder finds a whole lot more energy somewhere at very high pressures (I doubt it) or S.E.E. requires lots more powder from somewhere, like stuck to the sides of the barrel but not normally combusting due to conductive cooling from the steel, or a large fill of a dense fast powder. (or of course, my model is just **** and accidentally creates great low pressure results)

I know it is possible for powder to survive without combusting in a low pressure event because I was pouring 2495 from no filler low fill 308 rounds after firing them. All it would take to build up in the barrel would be a sticky barrel grease with flake powder and non-obdurating bullets. Eventually some would catch as the barrel warmed and there could be a LOT stashed in the grooves.

[popper]

I did the in-the-barrel sheet but not the chamber & powder part. I'm interested in your work. Powley doesn't have but a few powders & none at light loads.

[stephen m weiss]

I did the in-the-barrel sheet but not the chamber & powder part. I'm interested in your work. Powley doesn't have but a few powders & none at light loads.

Cool Popper. I know you have a ton of data. Is there some way to get the file to ya?

At this point it's only half a meg, and in excell 2000, so just about anything will read it. I havent done graphs because they take only a second and use up viewing area. I have never figured out a truly elegant way to do adaptive step sizing in excell so it has 200 evenly sized steps. It's easy enough to change that number.. but it runs off the screen a long ways. I expect making the number of step sizes an input will be something to be done soon, and then just haveing 1000's of rows that are not normally viewed. Prolly having inputs and outputs on one page and the calcs on a different page would be nice.. hehe, but harder to work on. At this point, each different caliber is on a different page, although really it is each caliber-powder combination since the burn rate and exponent change from powder to powder, case to case.

smw

[stephen m weiss]

Oh, I checked a 308 factory crimp extraction force on the same case as gave 100 lbs at a normal value. This time I cranked up the crimp 1/2 turn so that it put deep teethmarks in the brass, ie maxed out. That gave 110 lbs. The 10% increase could have easily been soley due to strain hardening, so this is some limited validation that extraction force is not very sensitive to crimp setting.

[popper]

There is a free equal to Matlab that will allow you to do partials & iteratives, don't remember if it will plug into cells of Excel. It is very good and has a graphic interface. I have matlab but never renewed the rental lic fee. My question of filler was serious, how much does it change the burn characteristics. Also question boolit oversize & burn rate. Try a roll crimp.

[stephen m weiss]

Ok, I tested lee loader crimps and got about 100 lbs for both on fmj.

Yes the filler should have a big effect. I didnt see a direct filler question. How much solid volume of filler for what fill 308 would you like me to test? I just did 24.5 gr of Accurate 2495 and .25" in the case of solid filler. Peak pressure went from 8496 psi to 12812 psi. Seating force was at 100lbs and Seating depth .41 inch. Muzzle Vel went from 1400 to 1650 fps.

Now knowing the solid volume of filler is not a trivial matter. For dacron, you would use published density and weight. For toilet paper, its wood so I use weight and wood density. For funny plastic particles, depends on the shape and size distribution.

[stephen m weiss]

Oh, that was 100 lbs for both 223 and 308 for bullet extraction force using a lee loader.

Does this site have a way to post files for others to download? Or is that by 3rd party sites. Really I shouldn't post it for general consumption, but if you can use matlab, you got a head on your shoulders.

I haven't used Matlab for 25 years since robotics grad school. I will look into that free version. I am sure if I think about it long enough I can develop progressive step sizing for spreadsheets lol.

Say, whats a roll crimp? Is that what happens to a rolled a...cigarette in a wallet? :)

Boolit oversize: as in pushing a very large bullet into the threading? I would just substitute that force for the bullet extraction force and adjust the case volume for the headroom. Or I could add a second check step just for that case.. That would require energy to push into the threads checking if it could stop the bullet.

[sackot]

There is a free equal to Matlab that will allow you to do partials & iteratives, don't remember if it will plug into cells of Excel. It is very good and has a graphic interface. I have matlab but never renewed the rental lic fee. My question of filler was serious, how much does it change the burn characteristics. Also question boolit oversize & burn rate. Try a roll crimp.

I think the Matlab equivalent will be "Octave", I used to use it with Linux.

[popper]

SciLab. Yes, pushing a 314 into a 308 bore or something like that. Roll crimp bends the case mouth into a crimp groove. Probably 30-50% powder fill for 308, dacron. Wonder if a non-burning fine particle would be a math equal?

[stephen m weiss]

I cant predict the force for pusing a 314 into a 308 bore because it is heavily dependent on feed ramp geometry, lubrication, and material hardness. The easiest way would be to do it by hand using the bathroom scale and a pushrod.

I dont have a roll crimp tool, but if values were given, I could easily run the analysis.

I havent used dacron, though the solid density is about .04 lb/in^2, slightly more than water. Since its so fluffy, a volume measurement would be about useless, depending entirely on the packing. I suspect as used it's similar to crumpled toilet paper, which I use a 2"x2"x.001 thk sht for a 308 case. Thats .004 in^3, or about 1/80th the total fill volume.

For 30% powder fill, 16 gr, 2495, 150 gr bullet, no TP : 831fps at 3220 psi, with .004 tp 836 fps at 3273 psi. If I put in 25 gr of rice at .04 lb/in^3 on top of the paper, .09in^3, then its 1291 fps at 8783 psi.

for 50% powder fill, 23 gr, 2495, 150gr bullet, no TP: 1221fps at 6622 psi, with tp 1240fps at 7164 psi, with the rice 2018 fps at 25104 psi.

So clearly using a judicious amount of non-powder filler after making sure the powder is restrained back at the primer can give more bang for the buck. If the stuff has low dipper density like rice, you are probably unable to get an overpressure event even at maximum filler using only 50% powder charge. I would have to take some careful density measurements to be sure, but it seems reaonable.

But, you really have to have a separator so the powder and rice cannot mix, because ignition would get all weird and not predictable. But that separator would have to let pressure past and not be a perfect seal, otherwise pressure would be much greater, and potentially unsafe.

[stephen m weiss]

[QUOTE=stephen m weiss;2970703]
But, you really have to have a separator so the powder and rice cannot mix, because ignition would get all weird and not predictable. /QUOTE]

oops, no ya dont need a separator. By using over 100% total fill with the rice, the compression force just keeps cracking the rice till the powder is pressed and log jammed and cannot shake or move at all by pressing in teh bullet. I just made 5 at 17.4 gr of 2495, 26 gr rice, 155 grain cast bullet lubed with alox and primed with a mag rifle primer.

My spreadsheet predicts 1500 fps and 16ksi peak pressure. I skipped the tp packing because with powder ignition and quick 3000 psi plus load, the rice will compact to near solid at teh base of the bullet and make a perfect packing, that scrubs the barrel gently as it protects the bare lead from gas heat.

haha, testing to follow, let's see how close I got!!!! I may pull the trigger on the first shot from behind a tree again.. :) I am not really worried, but at $.15 a shot I can afford to be safe!

the chipmonks are gonna love me even more now! hahahah

[Doc Highwall]

Stephen, how are you going to measure the 16,000 psi peak pressure.

[stephen m weiss]

Stephen, how are you going to measure the 16,000 psi peak pressure.

The muzzle velocity will be the only thing I can measure. Sure there are different pressure profiles that can produce the same muzzle velocity, but it is very unlikely to get velocity matching over a wide range of loads without being fairly close to simulating the majority of the process.

The same is true in Finite Element Analysis, which I ran for over a decade. You frequently only test failure point, but the model predicts stress distribution, stress concentrations, and time histories for some models. That's one of the beauties of mathematical modelling, you learn so much more about what is going on.

[stephen m weiss]

I should say that like most reloaders, I look at other qualitative factors that show if something is substantially out of whack: relative sount intensity, primer condition, felt recoil, bullet accuracy, powder grime pattern around the case neck and whatever else may be going on.

At some point, I am going to stick an accelerometer on my barrel. I have some 3 axis analog output accelerometers that i read using a USB oscilloscope on my notebook pc. The oscy has fast enough input to get tons of data on the event, so the recoil time trace should have the same shape as the pressure curve. Heh, I can even substantiate that suspect barrel whip which kills my accuracy at full power on my savage axis 308 sporter barrel. I should note that the usb oscilloscope cost less than Quickload, and is much more versatile....I used oscilloscopes and accelerometers for years in vibe testing, so this shouldnt be too hard.

[popper]

Be careful with the rice, it absorbs water & can make a hard plug. Wonder if L.G. has some dacron/ no filler data he would be kind enough to add for you. He does have pressure measurement equip. We know the pressure goes up as the fps does, we don't know how it effects really light loads, or I don't. I've used it in 17 gr 2400 for 308 & didn't blow anything up, 168 gr. LFN no GC. IIRC, ~1750 fps, maybe 50% fill. Acccurate but didn't cycle AR so I stopped testing.

[stephen m weiss]

Be careful with the rice, it absorbs water & can make a hard plug. Wonder if L.G. has some dacron/ no filler data he would be kind enough to add for you. He does have pressure measurement equip. We know the pressure goes up as the fps does, we don't know how it effects really light loads, or I don't. I've used it in 17 gr 2400 for 308 & didn't blow anything up, 168 gr. LFN no GC. IIRC, ~1750 fps, maybe 50% fill. Acccurate but didn't cycle AR so I stopped testing.

Ouchie on the hard plug.. I gotta think hard on that. I was already starting to conclude I should build low powder loads and work up to see if the model stays accurate. If anything weird starts happening, then I would not fire the higher powder loads. I just dont want to get really low speeds and a stuck bullet. I should do some models where I up the friction force, use a fast powder, for really low speeds. I bet I could get a bullet to skid to a stop, in the barrel, in my model.

btw, you got 1750 fps with a 168 gn in an ar length barrel? 20"? Just dacron as filler?

[popper]

119474
Yes, 16" carbine, didn't write vel on target but did chrony. . Got 1875 from 150 gr PB in 308ME 18gr of 2400 (slightly smaller case than 308W), 1825 from 145gr PB in 300BO, using 4227. MX did 1.6 MOA, 0.664 ATC @ 50. 308MX used dacron from used doggy toys. No filler in the BO, probably 80% fill.

[stephen m weiss]

Ah oh.. My model does not predict those values at all, it's saying 1100fps. Yikes. More testing to do.... I definately will make some low powder versions.

I can switch to barley from rice. It's more expensive, around $2 a lb, but has some of the shell left on, and wont cake together. Isn't dacron polyester basically about $5 a lb? I had to buy some 50lb sacks for some electronics cases the lovely chinese were injection molding for me. On the other hand, when little girls grow up, they leave about 50 lbs of the stuff lying around lol. I noticed coffee grounds are nearly the perfect size, but thats $10 a lb, approaching that of gunpowder. After you dry out used coffee grounds, it's really fluffy like sawdust, and wont give good pressure curves.

Thx!!!

[stephen m weiss]

Oh.. to put a factory crimp or bullet at rifle engagement condition into your model, try a line like

=IF(AND(I2<A$14,D2=0),0,I2/J2) for the bullet acceleration column

Where I2 is the force and A$14 is the bullet crimp force, D2 is bullet travel and J2 is the current free volume

I had to add the 'And' part because as pressure would fall towards the end of the barrel, acceleration would be set to zero as if teh bullet crimp were still working.

[stephen m weiss]

Ok Popper,

after much effort and a help call to my wifey who does numerical modelling, actually writing the code, for a living, I got adaptive step sizing going pretty well on my internal ballistics spreadsheet.

What I found was that the falling pressure zone is hard to model accurately, but we got it. And seriously, for the loads you are referring to, the solution seems to pop between 1500 fps and 1750 fps, depending on exactly how much primer or crimp or probably temperature or batch of powder. It's right on the edge of two dynamic modes.

That probably means that if I shoot at the top of my mountain at 45F or you shoot at sea level and 80F, we will get either of the two results. Of course this is not desirable in a load so pushing out of the unstable zone would be advisable. Either go high and waste some powder or go low and waste some powder, but ya gotta know what it will do day by day.

It kind of doesnt surprise me that you found a local maxima....

smw