1. Attention: We have put together a thread with tips and a tutorial video to help with using the new software. Please take a moment to check out the thread here: Trapshooters.com Tutorial & Help Video.
    Dismiss Notice

A test of fast vs slower powders (Winston)

Discussion in 'Shooting Related Threads' started by Neil Winston, Jul 25, 2009.

Thread Status:
Not open for further replies.
  1. Neil Winston

    Neil Winston Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    7,856
    The subject of the relationship - if any - between powder burning-speed and recoil has hardly ever been more popular here and in the last few weeks. The cited thread above and another, “Softest shooting powder,”

    http://www.trapshooters.com/noframes/cfpages/thread.cfm?threadid=191782

    have between them corralled a hundred or more posts. There’s even been a test with a highly statistically-significant result. Rather than add to either one of those threads, I’ve decided to start yet another, justified by the fact that it’s focus will be not which or which idea is right, but rather, how you could find out. And have a high probability of being right.

    The latter thread, the softest shooting powder one, has the usual mix of contradictory opinion backed by the utter certainly of the writer. One of the points made several times is that there’s no way to for one person to know about the “inner” experience of another and so the “experiencer” is right (no matter what.)

    The idea that there no way to get inside the subject and find out what really going on has been out of date for 149 years now, ever since Gustav Fechner’s publication of "Elements of Psychophysics" in 1860. The science it spawned, psychophysics, has been going great guns ever since then and does exactly the sort of “looking inside” that we have to do if we are to convincingly answer the question posed by both these threads.

    A fundamental concept in psychophysics is the “just noticeable difference” or jnd. If two stimuli differ by a jnd, a subject can tell them apart 50% of the time. Of course, the magnitude of a jnd may vary depending on who is being tested, but the idea that there are differences which can be discriminated and differences which can’t be is obvious but important.

    In general, the larger the magnitude of a stimulus, the greater is the jnd. For example, if you can tell the difference between 10 and 11 ounces, then you can’t between 20 and 21, but can between 20 and 22. Said another way, the ratio between the magnitude of the stimulus and the size of a jnd is a constant.

    It’s time to define the task, not in the broad generalities used in both threads, but rather in terms of what the elements of the problem are.

    I thought I’d done all this before, I must say with a bit of dissatisfaction. I thought that if it were proven experimentally that the speed of the gun at shot exit is unrelated to the pressure in the shell and related only to the speed of the shot, the issue would be settled. Here it is:

    http://www.claytargettesting.com/study2/Study2.4.pdf

    Of course it could not have come out any other way and I knew that, but still it is a result that can’t be argued with and I thought it might have some effect. I should have known better.

    “OK,” I heard, “The speed may be the same, but the _acceleration_ is different, and it’s that which is giving the information about the powder being burned.” In counter, Pat and I go on and on about the basic similarity of all these pressure curves (and goatskin has, in a PM, given me precise information about the slope of the curves) but there’s been nowhere to go with that, beyond sitting back and reflecting about what really fast things and small distances we are dealing with.

    The task, then, from a psychophysics perspective, is evaluate the rate of acceleration of a gun from zero to about 12 feet per second when that acceleration occurs in about 0.0035 seconds and over a distance of 3/8 of an inch and to compare that to another rate of acceleration over the same time-period and distance and decide if they are the same or different.

    All of a sudden it sounds a lot harder than “telling Red Dot from PB” but it’s what has to be done (if acceleration is the cue.) The difference in rates must be larger than the jnd (just noticeable difference) I discussed above and the relationship between stimulus intensity and the size of a jnd begins to look like a real problem. Remember, the bigger the stimulus, the bigger must be differences to tell them apart. And the acceleration we are dealing with is big. Really big. I get nearly 800 g’s but I could be wrong, Whatever it is, I think we can agree it’s big and so the size of a jnd is going to have to be big too. And that’s what Pat and I have been talking about: the pressures leading to these accelerations are different, but not much. Can it really be possible to tell them apart when it all has to be compressed into a small distance and practically no time at all? Well that’s why Fechner invented psychophysics – to answer just that question.

    Let’s agree on three possible outcomes of an experiment to test the proposition that the differentiation can be done.

    1. The subject can’t tell one powder from another. This does not mean it can’t be done – the test procedure might not be sensitive enough, the subject a klutz, but it_is_ a failure to demonstrate that it can be done, brings the theory into question, and puts pressure on the supporters of the theory to replicate or improve the test to show that the theory is right or accept that it isn’t.

    2. The subject can tell the difference a statistically-significant percentage of the time. This, unfortunately, does not prove the recoil is different since other cues may be controlling the decision – noise, smoke, vibration, anything. To prove it’s recoil will take a lot more work, but at least there’s “something there” and just how it works is probably not important at this stage.

    3. The subject is right most of the time, but not so often that we have to reject the idea that his judgments are just chance. This would suggest that it might be worth a bit more work; there may be something there.

    The shells:

    First we need the shells – same speed, different pressures. Measured, since otherwise we won’t know what the stimuli are and the results won’t mean a thing. We can’t use a loading guide; none is close enough.

    Here are the Red Dot loads:

    <a href="http://photobucket.com" target="_blank">[​IMG]</a>

    The average speed is down near the lower left corner, it’s 1196 fps and the SD is only 6 (For extra credit, spend some time on the pressure rise data, the time in microseconds the powder took to raise the pressure from 25% to 75% of its maximum. What should surprise us? And what else about the pressures is unexpected, though unimportant to this test.)

    The Green Dot shells:

    <a href="http://photobucket.com" target="_blank">[​IMG]</a>


    And the PB shells:

    <a href="http://photobucket.com" target="_blank">[​IMG]</a>

    (For extra credit, spend some time on the pressure rise data, the time in microseconds the powder took to raise the pressure from 25% to 75% of its maximum. What should surprise us? And what else about the pressures is unexpected, though unimportant to this test.)

    All the SD’s were similar; these are as good shells for this test as can be produced.

    The test:

    The test should be as “blind” as practical. Certainly, the subject can’t be told what he’s shooting and it’s better if the experimenter doesn’t know either. Presentation should be random, but what’s the task? Going back to first principals, it’s no more than to tell if two successive shots are different or the same. If the subject wants to add more that’s fine, but the dependent variable is the verbal response “Same” or “Different.”

    There will be ten comparisons; in five the powders will be different (sometimes one first, sometimes the other) and in five comparisons the powders will be the same. Thus he has five chances on being right when they are the same, five when they are different. If he just says “same” all the time he will be right on half the trials, same for “different”, same for calling “heads” every time a coin is flipped. So by chance alone the most likely outcome is about five right. Though four to six right is clearly no better than guessing, seven is a hint of skill, and eight looks pretty convincing, occurring by chance (if there’s no “real” difference) only about 4 times in 100 tests as I remember.

    Here’s how the order of presentation is controlled. I made a list of pseudo-random pairings like this

    1 Red Dot/Green Dot

    2 Green Dot/Green Dot

    3 Green Dot/Red Dot

    and so on.

    The shells were then loaded into a rack which kept them in order:

    <a href="http://photobucket.com" target="_blank">[​IMG]</a>

    The subject, Roman P, took the top left, shot it, then the next one to the right, shot it, reported “same” or “different” and maybe “the second kicked more” (though, of course, he doesn't know what's in the shell.) He then shot the next pair down and so on. A pair, a report, a pair, a report, and so on.

    <a href="http://photobucket.com" target="_blank">[​IMG]</a>

    Bob L pulled the targets, all straightaways from 3, and Brian A witnessed to add another voice to the results.

    Results:

    The first comparisons were between Red Dot and Green Dot loads, at 1196 and 1194 fps respectively.In the table below, “R” means Red Dot; “G” means Green Dot.

    <a href="http://photobucket.com" target="_blank">[​IMG]</a>

    Roman got nowhere on this challenge, hitting chance right on the nose. He tried hard, took it seriously, but couldn’t do it.

    Conclusion:

    For this subject, the difference is shells was smaller than a just noticeable difference (jnd); he could not tell the two apart.

    But how about the comparison between Red Dot and PB? The pressure traces are more different than were the Red Dot/Green Dot ones – does that make the task possible? In this table, R is still Red Dot and PB is, naturally, PB.

    <a href="http://photobucket.com" target="_blank">[​IMG]</a>

    Roman got nowhere on this challenge, hitting chance right on the nose. He tried hard, took it seriously, but couldn’t do it.


    Conclusion:

    Once again, for this subject, the difference is shells was smaller than a just noticeable difference (jnd); he could not tell the two apart.

    Discussion:

    I consider this a well-designed experiment, one which gave the differences, if any, a fair chance to show themselves. There was no evidence that the shells could be differentiated. The subject did no better than he would have if he hadn’t shot the shells at all and just flipped a coin.

    To those dissatisfied or disappointed with the results of this experiment. Consider carefully the “real” task, telling two accelerations apart which happen very fast and over a very short distance. Is it reasonable to think anyone could tell them apart?

    If you want to challenge these results do your own experiment, and I’d suggest the model here as one as good as any, better than most. See what you get and post it here on this thread.

    Thanks for your attention to this rather long and technical article; I hope you have enjoyed it and learned something.

    Neil Winston

    © 2009, Neil Winston
     
  2. Paladin

    Paladin Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    4,398
    Neil,

    Eluding to what DB wrote, the only thing written in stone from the results of your experiment is the fact that Roman could not tell the difference. No more, no less.
     
  3. 221

    221 Banned User Banned TS Supporters

    Joined:
    Jan 29, 1998
    Messages:
    4,736
    What test equipment, specifically, are you using to generate these pressure curves. It's doubtful that the equipment you have can accurately plot these graphs in us(microseconds)
     
  4. Neil Winston

    Neil Winston Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    7,856
    Actually, Paladin, that's all I claimed.

    Neil
     
  5. PerazziBigBore

    PerazziBigBore TS Member

    Joined:
    Jan 29, 1998
    Messages:
    3,859
    Our PB shooters are sitting back and laughing...
     
  6. Neil Winston

    Neil Winston Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    7,856
    221, I wonder how you can call it "doubtful" if you don't know what equipment I used. Or that you don't know the curves I used cover (I guess) about two milliseconds. The microsecond data concerns just the rising limb of the pressure curves and yes, my Oehler Model 84 can do this just fine.

    Neil
     
  7. 221

    221 Banned User Banned TS Supporters

    Joined:
    Jan 29, 1998
    Messages:
    4,736
    Your sure? Which way is it????? Milliseconds or Microseconds......It makes a big difference.
     
  8. Neil Winston

    Neil Winston Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    7,856
    DB BIll, I considered that - asking after every shot, but if the results are chance, and these are, more data won't help and this is a clean design (suggested by Pocatello, actually) that lends itself to uncomplicated binomial analysis which is the easiest and clearest possible statistical treatment.

    Your second observation is interesting because an old experiment did exactly that and found that while novice shooters thought slow powders kicked less, expert shooters thought the opposite, that fast powders were "softer." The trouble was, if you went carefully through the results, they all were doing little better than the subject in this experiment, that is, it was not possible to reasonably say the results stemmed from other than chance.

    Neil
     
  9. Neil Winston

    Neil Winston Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    7,856
    Yes I am sure, 221. I suggest you re-read what I wrote. The numerical data is quoted in microseconds, but the horizontal scale on the curves is about 2 milliseconds which is 2000 microseconds, which is the general level - microsecond - at which the equipment works. And it does not make any difference in this experiment, they are just posted as proof that I was using the shells I said I was to counter the inevitable claim that the slower-powder shells were going faster and that's why the results were as they were.

    Neil
     
  10. markdenis

    markdenis TS Member

    Joined:
    Jan 29, 1998
    Messages:
    1,123
    Neil

    I have a few questions regarding published pressures and velocities.

    1st. Have you tested several different loads other than red dot and green dot.

    2nd. When loading and reading a loading manual, I assume the published data is a close outcome of the actual shell reloaded. So, I just wondered if you adhere to the same assumptions I have ..that the information is basically correct, with general tolerance deviations expected.

    The reason I am asking these questions is definitely not to start a pissing match between us,(we have already had plenty of them) but to ask how you would interpret published reloads and how they react in my Remington 1100.

    Here are the two loads I have questions about that are published loads in their respective manuals.

    Once fired AA hull, Remington 209P primer, 17.0 grains Titewad powder, Windjammer wad, 1 1/8oz. of shot. (1200 fps & 11,100 psi).

    The other load is a Once fired AA hull, Winchester 209 primer, 21.0 grains Green Dot powder, Windjammer wad, 1 1/8oz. of shot. (1200 fps & 9,000 psi).

    Both loads have been weighed with reliable scales, and both loads are shot with the same gun.

    Out of a full box of shells, the first load will only operate the action and throw the shell out of my gun maybe 3 or 4 times out of 25 shells. With the few it does operate, they will fall at my feet. Most of them will remain in my gun.

    Out of a full box of shells, the second load will operate the action all 25 times and throw the shell ten to fifteen feet away from me.


    There is notable increase in felt recoil with the second load. Now I understand my old 1100 is not new and has shot many shells, but the actual results don't lie. You could strap my gun in a vice and see the actual difference the gun throws the hull with both loads.

    What's your thoughts on that?

    Mark Rounds
     
  11. Neil Winston

    Neil Winston Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    7,856
    Mike, they will really be laughing if they can do as carefully-controlled an experiment as this one and get a positive result. Until then, they are just typing.

    Neil
     
  12. 221

    221 Banned User Banned TS Supporters

    Joined:
    Jan 29, 1998
    Messages:
    4,736
    What Winston has suggested has about as much chance of a successful outcome as trying to square a circle.....Especially when he controls all the variables which are meaningless with regards to this "well designed experiment" Seems he has the deck stacked to his favor.

    I enjoy the spin being employed though....When you cannot hide behind Newton, Alliant, Logic, Science, and Math....you drag out psychobabble......You won't mind if I use your approach the next time one of our projects goes sour.
     
  13. Neil Winston

    Neil Winston Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    7,856
    As I said on that other thread, Mark, there's a lot about this pressure stuff we'd have to both know to answer your question. Why, for example, would my ported (and so enlarged gas-port) 303 eject light AA's and not heavies? I am forced to answer, "I don't know."

    My guess - and that's all it is - about the recoil difference you report (and I don't doubt at all) is that it's a result of the particular physical principle that semi-autos use that reduces recoil, one which no one here as ever gotten close to, as least as far as I remember. It's in Donald Butler's "The American Shotgun" and well worth spending the considerable time it takes to understand it if that's the kind of thing that interests you.

    I've tested dozens of loads and many powders and found few surprises. Most are pretty close to what I expected going by the reloading guide,some diverge to a greater or lesser extent, and there are lots of shells in those guides which have lousy ballistics, though you'd never know it on the field (or following the guide.)

    Neil
     
  14. Neil Winston

    Neil Winston Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    7,856
    221,I'll never get any better at this until you tell me where I've gone wrong, what I should have done differently, and what squaring the circle has to do with any of this. What, specifically, is wrong with the design?

    I'm especially having trouble with your "Especially when he controls all the variables which are meaningless with regards to this "well designed experiment." I can't figure out what you are getting at at all. If I have controlled al the meaningless variables, what are the meaningful ones? And did I control them or not? How? How not?

    Regarding your recently added last paragraph, I think if you equate psychophysics with "psychobabble" you ought to spend more time at the library.

    Neil
     
  15. 221

    221 Banned User Banned TS Supporters

    Joined:
    Jan 29, 1998
    Messages:
    4,736
    Neil.......My problem with all of this is what you are trying to accomplish here.This "experiment" of yours has no outcome....It's more of a mockery of the real question and, that is, is there a difference. Whether or not "YOU" or "I" can feel the difference is really up to you......I don't really care what you can feel or your test subjects can. To keep using all this fuzzy logic to avoid the real issue is beyond me.Why should you or Pat care that pressure affects recoil. It will not affect your status one way or the other. Technology changes so rapidly these days and everyday things change exponentially, It's difficult to keep up. That's why I question your test results, as I have a day or two experience with test equipment, and most times you don't have what you think you have especially if it is of any age. What would be necessary to properly do these tests, is not in the serious hobby range. IE....You cannot just tape a pressure transducer to a barrel and accurately test in us or ns.

    Now...Velocity measured by a chronograph is just that, a measurement.....I know you already knew this....Just imagine if we were able to 2 stage a powder charge....Same velocity... Totally different acceleration rates.....Thus even recoil has acceleration and velocity .

    Unfortunately for me my pen is not as eloquent as yours
     
  16. Neil Winston

    Neil Winston Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    7,856
    I think I've had my say to you, 221. Thanks for reading the test.

    Well, no. The equipment is what factories would use if they could justify the cost (though many do) , at least $15,000 all set up, to say nothing of the time and effort to understand it and exploit it. All the supporting bits are per SAAMI standards - a PCB 167a pressure transducer in a genuine test barrel, Kistler electronics, the whole bit. All assembled, wired up right and extensively tested, which is the only way I could get the remarkable results I do.

    Neil
     
  17. Shooting Coach

    Shooting Coach Well-Known Member

    Joined:
    Sep 14, 2006
    Messages:
    8,358
    Location:
    Nashville Tn
    I wonder if the test would have been more objective if Neil had the subject shooting 100 rounds on different days with the different ammo?

    If I shoot 100 rounds of Caps with a bomb Green Dot load, then the next time out, use my favored Solo 1250 load at the same speed, I feel less fatigued after the last target with Solo. YMMV

    Not poking anyone with a pointed stick. :^)
     
  18. Neil Winston

    Neil Winston Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    7,856
    Coach, since the test was objective and couldn't be more so, I think you are wondering if the results would have been different if I'd done it the way you suggest. While I don't know, I wonder how I would analyze the data. If I did it for 200 total shots, all I'd have is one comparison, and you can't, in principle, differentiate that from chance.

    No poking inferred. It's a reasonable question.

    Neil
     
  19. 221

    221 Banned User Banned TS Supporters

    Joined:
    Jan 29, 1998
    Messages:
    4,736
    Neil you really need to make up your mind ....you ridicule me then you claim it's an oehler 84... now it's kistler and pressure guns You really need to get more fiber in your diet.....It's too bad that your status shields you from ....Just keep BSing the folks
    You are entertaining.
     
  20. crusha

    crusha TS Member

    Joined:
    Jan 29, 1998
    Messages:
    5,762
    Great job, Neil. These kinds of experiments, where the human and his perceptions are the output variable, are always fascinating to me. This is basically a nonparametric test with two response levels, and I really can't see how it could have been better-designed. If, as someone suggested, you shot 100 of one and then 100 of the other, as you've said you really only have one observation...plus you now have a "fatigue" variable that hasn't been blocked in the design.


    The only further control I can think of is to have someone else unload the gun for the shooter, away from him, and have some kind of fan blowing the smoke away to prevent telling any difference in the odor of the smoke...but even without such controls, the subject in this case still couldn't tell a s/s difference, so it really just strengthens the eventual conclusion.


    Of course, you realize the proponents of the pressure theory are going to feel like they're being violated here. They, in choosing "personal sense of feel" (as opposed to ballistic pendulum readings and such) as their output variable, believe they've insulated themselves against any attempt at scientific refutation. "I feel this...you feel that...there's really no way for anybody to judge or scientifically attack my opinion on this," they think. Aha, but you have gone and applied science to it _anyway_; you've done something foreign to them, you've turned them upside-down and "stuck it in the wrong hole" on them, so to speak - and they simply will not be able to abide this!


    Love it. They know they're had...but yet not one of them will do the work to prove otherwise. "I know what I feel, and you and Roman P. can't tell me what I feel, so there...Harrumph!" They'll just sit here and type, that's all.
     
Thread Status:
Not open for further replies.
Search tags for this page
neil winston green dot test
,

neil winston powder test

,
niel winston green dot test recoil test
,
niel winston recoil test green dot