Recoil testing of large caliber muzzle breaks

[choppero]

I am a gunsmith and among other things design, build and install muzzle breaks on large caliber rifles. I have designed a test sled to measure the differences in recoil with and without these breaks.
I would like some feedback on the design and ways to improve its' ability to measure recoil. I have attached three CAD sketches to illustrate.
The recoil assembly is made up of a cut down barrel from a Remington 11 shotgun, the magazine tube from the same gun, and the recoil spring. these are captured in the lower assembly. The upper assembly consists of frame made from preformed channel (the same for the lower assy) and the sides made from butcher block type nylon. This slides on the black Delrin rails.

The sled, under recoil, pushes the upper against the spring and the distance is measured by the upper assembly pushing a sliding rod (not pictured) that is attached to the lower. A tattle tail device if you will. The deviation of a three shot string is about .02". A 7mm Remington Magnum recoiled 2" with no break and .96".

I would like some thoughts of how to get better consistency sot to shot.
Would leveling the center line of the bore with the sled help?

Thank you for allowing me to participate in this site.

[RWOLFEJR]

You might want to look into a "linear slide rail" set-up. They are relatively inexpensive and have mounts for your feet and bolt holes in the sliding member to mount your moving rest. That would improve your repeatability. The spring is another area for variation... and also misleading information. A compression spring will skew your results on a curve. since you're looking for real world improvemnt created by your break you want to see actaul difference. A compression spring produces more force the more you compress it. You need a constant force spring... or a different way to keep the unit toned down for measurement within a reasonable distance.

Myself... I'd think simply adding weight to the moving member and eliminating the spring would give you a pretty darn accurate comparison of break vs. no break and not introduce the spring variable.

Slide rail with plate on top of slides that you mount your rest on... Hanging below plate on both sides are square tubes full of lead? their low center of gravity to the slide will also help counter the weight of the gun on top... sort of keeping your load centered to the slide bearing centerline. Play with the amount of weight depending on the caliber. A 22 wouldn't need any weight (no need for a muzzle break either...) where a 50 cal. would need a good bit.

[Kelly_Bramble]

I think a design where the slide and other mass features are in-line with the muzzle is the best approach. Reacting with a slide that is not co-axial with the muzzle could induce complex bending loads that cannot be accounted for in the difference muzzle-stock designs.

Keep it simple and eliminate false friction and reactionary loading.

[choppero]

I have tried a weighted carrage, 50 lbs with no spring and the results were not as consistant as the spring loaded version. I've also tried a hydraulic damping system using a one way actuator and a flow control valve. The near instant pressure blew the seals on the first shot of a 300 ultra mag. I do understand that the spring will increase tension as it compresses which is why I've tried the above mentioned methods. So far the spring system has been the best approach. If the linear slide rail would help by reducing the friction variable then I will look into that option. My thoughts about the spring is; although the resistance does increase over distance, it is fairly reliable. A 30-06 without a break travels X inches, a 300 magnum X+y inches. The 300 magnum with the break travels X inches. I can make a compairison and calculate a percent of change.

I have hung the sled from the front and added weight to try and graph a weight/distance. Is this the way to show anything useful?

[Kelly_Bramble]

If you could use a precision load cell in series with a mass or a spring recoil system you could measure and retain a lot more data for analysis.

[choppero]

load cells and data collection are expensive are they not? I did try an accelerometer circuit board from Analog Devices but the frequency response was not fast enough. It needs to capture a 6ms event.

[Kelly_Bramble]

load cells and data collection are expensive are they not?

Less than $200 all day long... What do you call expensive?

[choppero]

where can this all day special be had? This is a short budget project. $200 is ok if it includes the load cell and the data collection software or what is needed to store the data. A very nice piece of work in this area was done by Cal Poly San Luis Obispo in 2013 titled Rifle and Shotgun Recoil Test System. They determined "a sensor package that could attain 25 kHz to 50 kHz was deemed necessary". Also "The estimated peak force for the .460 Weatherby Magnum was assumed to be double to triple the peak force for the 300 Magnum. Thus, a sensor that could read over 3600lbs was deemed necessary, and with a safety margin, a 5000lb load cell was purchased".

[JAlberts]

What percentage of repeatability do you want to achieve? At this point, even at the short 0.92 in. recoil length you are only suffering a 2% variance.
My other issue regarding the percent of the problem potentially coming from your fixture is the question of what percent of the variance is being caused by the gun's actual firing discharge variances from shot to shot as opposed to what percent is being caused by your test fixture.
Also, I would like to know what the spring rate in lb/inch is for your current spring to get some idea of the amount of load variation you are actually getting within your .02 inch variance.

[choppero]

Thanks JAlberts for the reply. I am wondering if hanging the sled from the front and then adding weight to the back and measuring the distance each weight pulls the sled would give you that answer. I have done that. I used cast lead inguts so the weight is not uniform

weight	distance
3.95	0.2
7.89375	0.58
14.63125	0.753
21.3125	0.815
28.0875	1.02
34.9625	1.25
40.9675	1.34
46.9728125	1.537
53.9540625	1.736
60.9165625	1.856
67.7103125	1.957

Cal Polys' experiment showed shot to shot variations using rather expensive load cell technology of around 2%, so maybe my unit is gathering reasonable data.

[choppero]

Would moving the recoiling system up and to the rear of the butt stock, in line with the bore have a better result? This would allow the use of the linear rail and a lower profile.

Kelly Bramble quote "I think a design where the slide and other mass features are in-line with the muzzle is the best approach. Reacting with a slide that is not co-axial with the muzzle could induce complex bending loads that cannot be accounted for in the difference muzzle-stock designs.

Keep it simple and eliminate false friction and reactionary loading."

[JAlberts]

choppero,

Below you are going to find what can only be described as a “long response” to your last post.

First, with respect your above design revision, moving the recoiling assembly up to be inline with the butt plate (and as much inline with the gun barrel centerline as possible) would definitely be an improvement. This is because the current vertical offset between the gun loading on the butt plate and the recoil spring assembly is what creates the major amount of friction loading on your slide assembly bearings.

Now that I have addressed that issue there is the result of my spring rate analysis using your test data. I was feeling pretty good about your last response with regard to the 2% repeatability variance until I analyzed the load vs travel data you gave me to determine your spring's spring rate.
In a friction free system the rate of the spring, its force vs. travel distance value, should be constant; however, in analyzing your data, I found that as your test loading increased so did the spring rate, far beyond any amount that would have been caused by the small errors in your test loading weights. Unfortunately, this indicates an outside effect that is most likely friction in some element(s) of your assembly. In that note, I do believe that static loading in each step used in your loading test tends to increase the effects of friction in your system beyond what is present during the dynamic motion of a firing test.

As you stated in your earlier post, as long as this effect would stay constant, it would still allow comparative testing; but, there is no guarantee that it will not change over time; and, it prevents you from being able to give a reliable maximum recoil load value from yout testing if should want that at some point.

From my view, are three potential sources for this friction loading:

The first, and I think the most significant, is loading on your guide bearing due to the offset between the butt plate and the recoil assembly as discussed above. In that respect, while you could reduce this effect on the guide assembly by going to a more sophisticated roller type guide as suggested above; my first choice would be to make your suggested move of the recoil system to the butt plate level regardless of the type of guide used.

The second potential friction source has to do with your spring and its tube enclosure. If your spring is very long then during compression it can become "unstable" meaning that at larger compression displacements it will tend to misalign in the center range of the spring, This misalignment can, or will, be restrained by the inside surface of a tubing enclosure but unfortunately this results in sliding friction between the spring coils and the inside tubing surface that is then added to the spring resisting force possibly causing some of the apparent spring rate increase I saw in my analysis of your loading test. At the same time, your spring would have to be relatively long and the drag force would need to be pretty extreme to cause the type of load rate increase I am seeing so I placed it second among my list of drag candidates.

The third potential source that has been raised by others is the sliding friction between the nylon and Delrin guides in your guide assembly. I have left this to last because in my experience I have found that this combination actually works pretty well as long as the travel velocities and repeated test cycles do not result in overheating of the materials. If there is any difference I would have chosen it would be to have used Delrin for both elements and Teflon filled Delrin, which is reasonably available, for at least one of the guide bearing components. All that being said, I think aligning your reaction spring assembly with the butt plate centerline would effectively eliminate the issue of side loading and friction in your current guide assembly.

Now that I have given you the above information that I hope will be helpful, I want to address one more issue with regard to your spring assembly (and this is something you may already aware of from your gun assembly experience). Much of my engineering background is in the design of spring loaded safety relief valves and in this industry there is a standard that requires a spring must be installed with a minimum preload deflection of 10% of the spring’s free (uninstalled) length. The purpose of requirement is to insure that the spring loading will be distributed evenly across all of the spring coils for maximum spring rate consistency during the spring’s operation. The reason I mention this is that my analysis of your data indicated the greatest change in the load vs. deflection at spring compressions below 1.00 in, which could be a result of low spring preloading.

[choppero]

While I am moving the recoil assembly behind the butt plate I will order a new remington 11 spring as this one is missing about 3 inches (cut down). Your help is very much appreciated. This unit is made from mostly stuff laying around the shop. here's a picture of the sled. The only purchased material is the frame.
It looks like a re-design is in order

[JAlberts]

choppero,

After seeing the actual unit I am really impressed by the design, regardless of its "shop components" basis. Although I shouldn't be considering how particular and precise gun smiths are about their work.
Looking at it makes me question that the firing centerline to guide rail height offset should be creating any significant friction or resulting drag in the unit's operation. As a result, I am becoming more suspicious about the source of the load test spring rate variance from the testing; and, I cannot see that your load test method should have created this problem.
With what appears to me to be the low likelihood of a guide friction problem I am becoming more focused on the reaction spring carrier assembly.

While there is no question that aligning the reaction spring assembly with the barrel centerline would be a the best arrangement from an engineering standpoint, what you are trying to achieve here is a functionally accurate usable assembly; and, your current design defifntely has the advantage of being compact and easily mounted.

As a result, before you start a major redesign, I would like to see pictures of your actual reaction spring assembly both assembled and broken down into its individual components so that I can see if I can see something that would indicate to me that this might be a source of the apparent friction indicated by your static load test.

[choppero]

Here are some pictures of the recoil spring assembly installed and outstalled.
I found that the rear guide block was binding and have opened the cuts. I'll have to pay attention to the level when installed as it may have a desire to bind on the front ring.

[JAlberts]

choppero,First off, let me say that even with all of the photos I am having a bit of a problem visualizing your spring assembly but One thing that I think I see is that there seem to be a number of components that are being used in addition to your basic fixture guide rails specifically to independently guide this assembly and that, as you stated in your above post, can potentially create friction and drag. As a result I am attaching the below rough drawings of a a couple of simpler spring assemblies for your consideration.

[JAlberts]

choppero,

Just as an added note to the above usinng your spring picture and comparing it to a Length/Diameter spring stability reference it appears that you should not need any center guide as long as your maximum spring deflection is no more than 50% of the spring's free length and the suggested simplest design without a center support rod should work for your application.

[choppero]

Yep, I see the simpler solution. In my top photo see that the two bases are bolted to the bottom frame. The barrel rides in a dovetail at the rear and and a ring soldered to the bottom of the barrel that slide over the tube with the spring in between. The second picture the square pin sticking up from the rear of the barrel is attached to the slide and that is what drives the rig (ref. CAD sketch). Now I see where all that friction comes from. Good news is the spring slides, and is cradled inside the frame. Sort of like an outside guide.

[choppero]

I've installed the spring as you suggested and the results are very consistant now. A 308 win without the break recoiled 1.096, 1.075, 1.104. With the break .689, .683, .692. I'm now thinking of using a full length spring to see if it has any change. Thanks for the great insights.

[JAlberts]

Glad I could help. You should be able to the lb/in spring rates from the supplier so that you can determine the relative forces as well.

With regard to a longer spring, remember that a key issue in the selection of a spring length is its unguided stability under compression (for the best accuracy you do not want the spring to come in contact with your guide rails at any point in its travel). Sliding contact = friction = inaccurate force readings.

Long story short: If you don't anticipate that the recoil force of the largest gun you intend to test will result in your spring going solid at its maximum travel; then the shorter the spring the better. If you were to install a longer lower rate spring then the measured differrential length between a gun with and without your muzzle brake would be greater but the amount of travel variations (repeatiblity) in your test would also increase. So it ends up being wash between the two springs. For example: If the travel difference between the two test conditions is increased by 20% then the measure variation lengths in multiple firings at each muzzle arrangement will also increase by 20% and the calculated % error will be same as you now have.

Generally, there are only two reasons that a longer spring is selected by a user; and, as I see it, if your assembly is working well now, neither of these should improve its performance or accuracy.

The two reasons for selecting a longer spring are:
One, the currently installed spring length results in the spring coils going solid under the operation stroke required by the mechanism in which the spring is installed.
Two, a lower spring rate with more spring travel at the same maimum spring load is required by the mechanism in which the spring is installed. For exmple: in a spring loaded relief valve a certain minimum opening height (and spring travel) is required to achieve a required flow discharge rate from the valve at full lift.

Of course, if you have any questions about any of this I will happy to discuss them with you.