Motivated by my earlier testing with Blue Dot, where I was running into manageable but somewhat performance-limiting pressure issues, I decided to look into Lil'Gun. Lil'Gun is said to produce higher muzzle velocities with lower peak pressures compared to other powders. It also meets my safety requirement of filling more than 80% of the case at the loads I am interested in, to more easily detect double charges and reduce the likelihood of positional performance problems and air-gap-induced detonation.
Lil'Gun for the 5.7x28mm has been previously discussed. Some comments include:
All comments I found were at least a decade old. None of them presented any data, except in reference to a test done by Guns and Ammo magazine, which I can't find online, and which appears to have been related to developing a load for a 60gr projectile, which is much heavier than I am interested in."[someone else said] they [Guns and Ammo] used a custom T/C Contender barrel, and that many of those loads were unsafe for use in any of the FN semi-autos"
"very old info I would not trust it"
"[someone else said] the pressure curve is off, and it is not very efficient"
"Where did you hear lil gun was a good powder for 5.7x28? Unless you are shooting 5.7x28 out of a T/C, I wouldnt go near it."
"DO NOT USE Lil Gun for your FSN. You might be able to get away with it for the T/C but certainly not any of the semi-automatic guns. Guns and Ammo did an article a while back using a variety of powders (including lil gun) and they used T/C's. Most of us have concluded that many of those loads are not safe to use in FN weapons....but out of the T/C you should be ok....(SHOULD)."
"Lil' Gun: Too slow burning for blow-back actions."
I also looked elsewhere to find information on its use in .410 shotshells and especially the .22 Hornet, where it appears to have achieved legendary status for safety, predictability, velocity, and accuracy. Additionally, I did some modeling with QuickLoad (more on this in a minute), which seemed to bear out the claims of moderate peak pressures at high velocities. I satisfied myself that some carefully thought-out testing was within the risk posture I personally find acceptable.
Throughout this testing, I used unfired Federal American Eagle cases. I recently switched from once-fired to unfired. Pulling the bullets from loaded ammo with a kinetic puller goes fairly quickly with my wife assisting. I also saved some time in not having to deprime and reprime, but I still had to resize and chamfer to get the case mouths and necks in shape. I Dremeled off the decapping pin on one of my Lee full-length sizing dies to be able to full-length size a primed case.
I've determined through testing that the AE OEM primers show pressure signs in much the same way as the CCi 400 small rifle primers. That is to say, they begin extruding around the firing pin at significantly lower pressures than the CCi BR-4 primers, which have a 25% thicker cup. The OEM primers are a mitigation to help me identify excessive pressure early, though if the loads prove unpredictable, I know that might not help me.
My first loads to look for pressure signs were based on the Hammer Hunter 35gr bullet, seated to 1.495" COAL. I've not had amazing results with that bullet in 5.7x28mm, but recent testing and better statistical analysis indicates it is likely somewhat more accurate than the AE round. I would have preferred to do my testing with the Cutting Edge Bullets MTH 32gr, which so far has given me the best accuracy by a noticeable amount, but I'm still waiting for CEB to put that bullet back into production.
I'm adding Tubb Dust to the Lil'Gun.
I also used my MagnetoSpeed Sporter. The Sporter is the cheaper one where the sensing element has to be less than 0.375" away from the bullet as it flies past. You can't do that with the LC Carbine foregrip in place, since it puts the sensing element too far from the bullet's path and makes it impossible to mount the MS, unless you design and 3D print a special mount, which I did, and also use the magnet trick. BTW, regarding the MagnetoSpeed magnet trick, nobody seems to have realized that changing the magnetic flux field is likely to have an effect on accuracy, and it depends on how the magnets are positioned, so for the record I'll just say that I used 6mmx3mm disc magnets, centered by feel on each sensor, and the velocity measurement was ~2% faster (~50 fps at these speeds) than a properly-mounted Sporter that didn't need the magnet trick.
I ran the following loads. Charge weights are as reported by my Lyman Gen6, and are expected to be within +-0.05gr. Velocities are corrected for the 2% error.
Code: Select all
HH 35gr
gr Lil'Gun MV fps
7.9 2293
8.0 2375
8.1 2407
8.2 2320
8.3 2391
8.4 2509
8.5 2424
8.6 2417
8.7 2448
8.8 2442
8.5 2382
8.5 2440
8.5 2367
8.5 2399
8.5 2380
8.5gr average (6 rounds): 2399 fps
8.5gr standard deviation: 28fps
Throughout all of these rounds, including the 8.4gr outlier, I experienced *no* pressure signs. All rounds cycled and ejected without any issues.
I had previously documented that the Hammer Hunter bullets produce velocity spreads significantly greater than the other bullets I am using. I believe this has to do with the way the bullet seats into the neck. I use an LE Wilson inline bullet seater and a K+M Precision Shooting arbor press, which gives me good feel (literally) during the seating process. LE Wilson doesn't make an inline seater for 5.7x28mm, but I designed and 3D printed an adapter that allows me to use one of their 22-250 seaters for 5.7x28mm. Whereas other bullets have smooth shanks and therefore press in smoothly, the Hammer Hunters have multiple "drive bands" that make the bullet "crunch" into place as the bands click into the neck. I think I am not getting consistent neck friction with the Hammer Hunters. Furthermore, I am finding that to model my real-world results in QuickLoad, I have to enter a signficantly higher "Shot Start Pressure" than the default. So I think I'm getting the equivalent of a lot of neck tension, and it varies a lot. To eliminate this as a possible source of variation, for my next test I switched to Nosler BTLF 35gr bullets, which have smooth shanks and seat smoothly. The Noslers are also one of the more accurate bullets in this rifle. I am not crimping.
Emboldened (maybe that's an inappropriate word) that I hadn't blown the gun up yet, and having seen zero signs of pressure, I expanded my range of loads and got the following results:
Code: Select all
Nosler BTLF 35gr
gr Lil'Gun MV fps
7.0 2215
7.5 2275
8.0 2359
8.2 2329
8.3 2372
8.4 2397
8.5 2457
8.6 2420
8.7 2431
8.8 no reading
8.9 no reading
9.0 no reading
9.1 no reading
9.2 no reading
9.3 2591
Code: Select all
8.0gr 2469 (!)
8.5gr 2605 (!)
Again, I had *no* pressure signs.
I had intended to shoot a few more rounds, but now several things were really nagging me:
- Although velocity hadn't actually flattened out from 8.7gr to 9.3gr, I was missing velocities from 8.8gr to 9.2gr, so I didn't really know what was going on, and that made me uncomfortable.
- The velocities were substantially off from what QuickLoad had been predicting. For example, 7.0gr was predicted to be 1994fps instead of 2215fps. 9.3gr was predicted to be 2686fps instead of 2591fps.
- The 8.5gr load with the Hammer Hunter was 100fps faster than the 8.4gr load from the previous test, which was already 100fps faster than expected, so now I was 200fps off the chart using the HHs. The 8.0gr load was also substantially faster than predicted.
That's when I decided to call it quits for the day. I then spent several hours at home trying to make sense of my results.
I did another linear regression on the Nosler-based data and again got a nice linear fit of charge weight vs muzzle velocity, even across the expanded range from 7.0gr to 9.3gr. The velocity spread had also reduced considerably from the HH-based data.
Still, I couldn't find a way to reconcile QuickLoad with my real-world results. The velocities just weren't lining up with the charge weights across the full range that I had tested. Now, you should never use QuickLoad, on its own, to determine a loading schedule. QuickLoad can become useful once you have dialed in bullet seating dimensions to match a given load and you have captured some real-world velocity data to see how far off QuickLoad is. If QuickLoad is pretty close to your results for a given load, then QuickLoad can predict more results if you don't stray too far from that load.
If QuickLoad isn't agreeing with your real-world numbers, you can experiment with adjusting the propellant parameters. QuickLoad permits this, though they don't exactly encourage it. It appears most people who experiment with this tweak the "Burning Rate Factor", Ba. Increasing Ba will cause QuickLoad to predict a higher muzzle velocity (and a higher peak pressure), everything else being equal. The problem is this: you can dial in Ba to exactly match whatever real-world muzzle velocity you measure for a single charge weight, but QuickLoad might then not match real-world for other charge weights, with the error increasing the further away you get from the point you dialed in. In my case, if I dialed in Ba so QuickLoad matched my 9.3gr result, it would dramatically underpredict muzzle velocity at 7.0gr. And if I dialed in Ba so it matched 7.0gr correctly, it would dramatically overpredict muzzle velocity at 9.3gr. There wasn't even an acceptable middle ground.
In fact, I searched the entire QuickLoad propellant database to see if there was any powder that could match my testing. None of them did.
I believe this is an indication of a significant limitation in QuickLoad's model. I don't mean that QuickLoad has an error. I mean that I think QuickLoad's powder burn model does not accurately reflect some powders when you are trying to get results across a broad range of charge weights. No model is perfect, you just need to understand the limitations of the model, but I can't find any information in the QuickLoad documentation that explains the model with enough detail to understand why it's not working for Lil'Gun, or any other powder, in my testing. I'm guessing, and I believe I am not the only one that has guessed at this, that the QuickLoad Ba parameter is measured for a given powder in a laboratory bomb apparatus that is filled to 100%. If so, then it's less valid the further away you get from 100% fill, at least for some powders. This implies that, without manually varying Ba as a function of fill % based on real-world test data, QuickLoad may only be suitable for fill ranges near 100%, again at least for some powders.
I did find a forum post from someone who presented his own real-world data of using Blue Dot in a .44 Magnum. He observed that velocities increased with increasing charge weight, until they leveled out and then decreased once the charge became compressed.
Then I found a different forum post from another individual working from his own data using Lil'Gun that suggested the Ba parameter in QuickLoad needed to be adjusted in order to better fit a wider range of charge weights, because it appeared the QuickLoad default value was only valid for charges near 100% fill. I tried his suggested value for Ba, but it did not work for my results.
So, out of frustration more than anything else, I decided to figure out what value for Ba was necessary in order to get QuickLoad to agree with my results at each of the charge weights I had tried. To minimize the noise in the signal, I used the linear regression model I had developed for the Nosler. I found I needed the following (not all charge weights shown):
Code: Select all
7.0gr: Ba = 1.1
8.4gr: Ba = 0.96
8.7gr: Ba = 0.92
9.3gr: Ba = 0.885
That is, they didn't match until it dawned on me to plot Ba as a function of "Fill %", "Fill %" being the percentage of Usable Case Capacity occupied by the powder. The Usable Case Capacity with the Hammer Hunter 35gr bullet is different than with the Nosler BTLF 35gr. When normalized on the basis of Fill %, lo and behold, I got a straight-line linear relationship between my Ba numbers (which matched the real world) and Fill %. Moreover, that relationship (refined by linear regression) now matched my Hammer Hunter results, with the exception of the outliers, *and* it matched fairly closely the default Ba number from QuickLoad at 100% fill.
For illustration purposes only (don't use these numbers yourself), the table looks like:
Code: Select all
Fill % Ba
70 1.155
80 1.059
90 0.963
100 0.867
Thinking of Lil'Gun (and Blue Dot) in the shotshell and .22 Hornet contexts, the following occurred to me:
- Plastic shotshells are always loaded to 100% fill. In other words, there is no air gap, like you would have in a brass rifle or pistol case with less than 100% fill. The charges might not be very "compressed" per se, but they are constrained, they can't slosh around, and there is no significant air gap.
- The best Lil'Gun loads for .22 Hornet are reportedly all ~100% fill or even compressed. Some have even gone so far as to opine that it may be physically impossible to cram too much Lil'Gun into a .22 Hornet case. The decrease in Ba as Fill % increases suggests that the powder self-moderates, at least to some extent, as the charge becomes constrained and packed.
Encouraged that I now had a model that was matching reality (albeit on not a ton of data), I decided to try a new round of loads. For this test, I used the Lehigh Defense Controlled Chaos 32gr round loaded to COAL=1.580". This bullet had produced accuracy on par with the Noslers in previous testing, it seats smoothly, and the lower weight produces higher velocities while mitigating risk from heavier powder charges.
I tried these loads: 7.7gr, 8.2gr, 8.6gr, 9.0gr, 9.3gr, 9.4gr, 9.5gr, 9.6gr, 9.7gr, 9.8gr, and then 15 rounds of 9.9gr (96% fill) (this time using my Brifit electronic scale to confirm charge weights were within [+0.00g, -0.02gr]). I did not measure muzzle velocity. By this point I had already reassembled my foregrip with the barrel dampener back in place, and I did not have time at the range to wrestle it off and back on again. All I wanted to do was document pressure signs, and if no pressure signs were evident, then I wanted to shoot for accuracy, which required the barrel dampener.
Once more, I had no pressure signs, at least not signs like I was expecting. Two rounds extracted but failed to eject and stovepiped. I'm not going to read a lot into that just yet, because I also had a known-good mild Blue Dot round FTE on me. The rifle hasn't had a proper tear down, cleaning, and inspection for about 300 rounds. I also saw a few dented necks, presumably from slamming hard into the ejection port on the way out. One round had the case body dented so badly that it's now unusable. So you can call those pressure signs. Conservation of momentum says the bolt is going to move backwards at higher velocity due to a hotter round, maybe not initially due to the lower pressure peak with Lil'Gun, but it will experience higher average acceleration due to the higher average pressure as the bullet moves farther down the barrel. As far as the usual pressure signs go, the primers were fine, cases were not getting stuck, and there was no hint of neck cracking or neck separation or base separation. Bullets receded into the case about 0.0005" per previous round fired (half a thousandth).
As far as accuracy goes, it was one of the best groups I have shot, and that's across 15 rounds. SD Radius was 1.0 MOA, matching or beating all previous results except a couple 7-shot groups that beat it out by 0.1 or 0.2 MOA (which could easily be by chance). Furthermore, that was 1.0 MOA in the rain. Raindrops likely opened up the group.
As I mentioned, I didn't measure velocity, which is a significant shortcoming of this test, at least in terms of validating the improved QuickLoad model. At some point I'm going to have to get a MagnetoSpeed V3 (which has a longer sensing range) or a LabRadar. For now, I'm going to be content that my improved QuickLoad model is predicting just under 2700 fps and almost 520 ft-lbs for the Controlled Chaos 32gr+Lil'Gun 9.9gr at COAL=1.580". I'm also going to think about whether I want to back off the charge a tad. I'd be perfectly happy with a 2600 fps round, assuming it's no less accurate. In any event, I'm now shooting 300 fps faster with as good, maybe better accuracy and 100 ft-lbs more energy than my previous preferred load based on Blue Dot that was edging towards pressure issues. It more accurate than, 600+ fps faster than, and 150 ft-lbs more potent than the AE 40gr rounds I started with.
I'm getting more comfortable with the thought that the velocity variations I saw with the Hammer Hunter might be because of a) the neck friction issue I mentioned before, and b) the fact that those loads were down around 85% fill, not up near 100%. Although some of my recent results were looking more promising on accuracy, I think the neck friction issue will steer me clear of loading Hammer Hunters for this cartridge.
A final observation, about the Lil'Gun pressure curve, is that it's a somewhat lower peak but then a firmer push down the barrel after about 3.5", at least based on the QuickLoad pressure vs. distance curves. One commentor wrote: "Too slow burning for blow-back actions" without being more specific. Well, even my weakest loads cycled just fine. Compared to Blue Dot, the pressure peak occurs a scant 0.05" (five hundredths of an inch) further down the barrel. For the same energy delivered to the bullet, it's about 15% lower peak pressure. To duplicate the velocities I am now seeing with Lil'Gun by using Blue Dot, I would have to increase the charge up to a point where I know I would be punching primers. With my Lil'Gun 9.9gr load in the LC Carbine, I am getting 89% propellant burnt and 26.5% ballistic efficiency according to QuickLoad. That's pretty good for a rifle.
Maybe that's a last point I will make. I'm only interested in reloading for the Ruger LC Carbine. I'm uninterested reloading for my Ruger 57, and even if I did want to reload for that firearm, I'd get better results with a faster powder, with a whole lot less unburned powder flaming out the muzzle. In my opinion, I don't think the Lil'Gun load would be unsafe in the pistol, other than setting off forest fires, as the pressure peak is well below True Blue maxloads according to QuickLoad, the pressure-vs-distance profile inside the barrel is lower for the first 3.5 inches (by which point the pressure has subsided significantly), and the pressure peak happens just 0.13" further down the barrel. This depends a lot on random pressure spikes not showing up, of course, and I haven't yet shot hundreds of rounds to really confirm their absence. It also depends on the firearm surviving the recoil.
So don't go running out and start loading 9.9gr Lil'Gun because I said so. That would be a gross misinterpretation of this post. Rather, if I've helped to advance the discussion with new data and, more importantly, an explanation of how I went about things and why I came to my own (limited and error-prone) conclusions, that's great.