Hacking The Benjamin Titan NP Break Barrel

[WhatUPSbox?]

I set out to get a feel for the repeatability of the shot cycle and measurement. I used the setup as described before but added an extra roller under the stock (image 1). Previously the timing gear was supporting most of the weight of the Titan. It was OK but the lateral motion by hand was not smooth due to the weight on the printed gears. With the extra roller, it carries the weight and the timing roller is just closing the gap in the gears.

I took three shots. Image 2 shows an overlay of the first and third shot cycle as measured with the accel aligned with boresight. I overlayed them in a graphics program and aligned the trigger point (pellet at muzzle).

The third image is a  plot of the encoder distance (count) measurements for all three shots.

I'm good with the repeatability. I was going to try a few shots with deep seated pellets. Does anyone know the length or volume of the transfer port as baseline?

[George Schmermund]

Since the gas spring and piston are acting as a damped oscillator I don't know why your displacement measurements aren't indicating a reciprocating signal somewhere in the stroke. We're both getting about the same total displacement numbers. What am I missing?

[WhatUPSbox?]

My optical encoder does not provide direction information. I am fairly confident that the motion up until the piston impact is single direction. The impact brings it to a near stop but the post impact motion could be in either direction. The vertical axis in my plot should be distance traveled. When I had the gear supporting the weight of the rifle, there was more friction and the sharpie traces looked one directional. With the weight supported by the extra roller, there is little friction and some of the traces look like they reverse. I am also pinching the trigger with my fingers. This contact with the rifle is overwhelmed by the 180 lb gas ram force so I don't think it affects the pre-impact displacement curve, but could affect the settling point of the rifle some milliseconds after the cycle.
I am still thinking about how to implement a potentiometer. I would still like to use the gear approach to get some mechanical gain. I made the clamp on the roller with the thought of being able to switch out the encoder plate with one that has a gear section on it. I'm not sure how fast you can move a pot wiper and maintain contact

[George Schmermund]

How about blocking the middle slot on the encoder. Then you could tell each time it passed.

[WhatUPSbox?]

Maybe, the big change in recoil velocity after impact make detecting a different spacing a little harder.

[George Schmermund]

This image shows a response to a modal hammer tap on the muzzle. A couple of details in it separate out well. The DSO trigger is on Ch #1 which is the signal from the forward looking accel. Ch #2 the rear facing accel and is set to read displacement. Ch #3 is the force transducer (8200) and it's poking into the butt plate.

Ch #1 clearly shows how the displacement signal is modulating the acceleration signal. Ch #2 confirms the timing of the modulation. There's probably a Doppler effect in there somewhere.

Ch #3 shows the compliance recovery time of the butt plate cushion. The cushion looks like it's damping out the displacement motion.

Now that the testing looks good with the hammer It will be interesting to see how it compares to an actual shot.

[WhatUPSbox?]

Looking forward to your results.

I think I'll try using a limit switch to detect zero crossing for my encoder as a temporary mod until I try the potentiometer or equivalent.

[George Schmermund]

If you have a piece of fairly stiff foam about an inch or so thick it could be placed between the butt plate and something heavy to block it with. The muzzle can then be tapped sufficiently hard to assure a rebound. It works a lot faster than having to load and fire the gun. The black open cell foam that I used for my proxy shoulder tests worked really well.

[George Schmermund]

I changed the proxy shoulder to something possibly more realistic. The new foam and metal plate up against the butt pad now resemble more of a shoulder with a bone in it. This may actually have an effect on the recoil. we'll see. At any rate a butt pad against a shoulder is part of any real world shooting condition.

Hammer tapping the muzzle only gives the rifle a blow in one direction as can be seen in the previous post. There's a small signal in the opposite direction, but not much. The more time I spend staring at these DSO images the more pieces of a larger mosaic seem to be coming together. The hammer taps really do run the action in a form of slow motion. It also provides very fast feedback on any changes made to the experiments. The taps can be repeated as fast as the DSO can reset it's trigger.

At some point the modal hammer's signal could be run to ch#4 and a math function could then be used to standardize the other three channels. As always thought experiments easily outpace hardware experiments!

[WhatUPSbox?]

Interesting, though I'm not sure the muzzle taps are exercising the action. Even at end of stroke, the gas ram has 140 lb preload (Michael's spreadsheet), so the internal parts are not moving. I think at the low frequencies the taps are exercising a rigid body suspension mode. For the accel, I think the high frequency content is the barrel harmonics and the tube/scope mounts. I still think the shot data is the way to go.

I kind of liked your previous configuration with the rigidly mounted force transducer pushing on the relatively rigid rubber. This should have given you the ram preload for the first part of the piston motion and then a sharp fall off towards impact.

[George Schmermund]

I should point out that my modal hammer enthusiasm wasn't intended to suggest that the hammer would or could replace the actual shots. The hammer just gives a good impulse to test the entire measurement system without having to cock, load, and then reposition the rifle and wiring where you want it. It's quite annoying to take a shot and realize a DSO or preamp setting needed to be reset. The tapping can be done (with care) while the rifle is ready to go and before you take the real shot.

If you go back to page 2 of this thread and look at the slo-mo or even real time YouTube video @5:20 (now disabled) you can see how much energy the shooter's shoulder is absorbing during the shot. This leads me to think that real world shooting of a springer (or any rifle) includes the shooters shoulder in the recoil absorption system.

My simplified objective at this point is to evaluate the magnitude of the optical scope's impulse loading. I fell that it isn't fully defined without the shooter's shoulder being taken into account. The 8200's measurement is more curiosity than anything else. The real stuff (for me) is happening where the 2 accels are mounted.

[WhatUPSbox?]

Agreed, having an impulse for checking the instrumentation settings is a great idea. I've take a number of shots with the DSO not set right or a power supply not turned on.

So tonight I added a geared potentiometer to monitor the carriage block. This was to take the ambiguity out of the encoder readings by identifying any reversals. The rack and gear (just a scaled down version of the other set) are shown in image1. It is a 10K pot, that I'm using in the 2-3 Kohm region. There is a 1K resistor in series to provide a voltage divider.

Image 2 shows the overall trace out to 100 msec. The accel data (Ch-3) and timing is a repeat of the traces from the other day. The potentiometer data (Ch-2) does not clarify the position data. Based on both markings and measurement of the travel needed to get to the starting voltage, the difference between the start and end positions on the carriage is about 8.5 mm. The encoder (Ch-4) total count does not match that and the final position of the encoder is not 8.5 mm from the start. There appears to be some relative motion  between the encoder and the carriage.
Well, more work to be done.

[George Schmermund]

Can you move the rifle back and forth some distance by hand to confirm that the pot is reversing the signal?

[WhatUPSbox?]

Yes, I marked the carriage at the start and recorded the voltage. After the shot, I can roll back the carriage until I get the same starting voltage and I'm back to the marked starting position.

I think I need to add some constraints on how the rifle rests on the carriage. I've been depending on the low friction of the carriage bearings and the arch shape of the stock to maintain contact, but I think more needs to be done.

I should also start integrating the accel data as a comparison. The DSO only does single integration internally so I probably need to capture the dataset and do it externally.

[WhatUPSbox?]

My guess was right. The restraint on the carriage block was not good enough (even strapped down) and there was about 3.5mm slip (image). I will add some mechanical constraints.

[George Schmermund]

You got me curious about the drag on my system. I decided to do a static friction test to see what the break-free force would be. It's pretty consistent at ~ 2 1/4 lbs in both directions. I'll go back at some point and see what the 8200 comes up with for actual shot forces in both directions.

[WhatUPSbox?]

I'm making some progress with the potentiometer as a motion instrument approach. I added some blocks to control axial slips (image 1).
I also took some time to pick a portion of the potentiometer range to work in, calibrated the voltage to axial motion, and scaled the DSO accordingly.

I put it all together and took a shot. Image 2 shows the overview of the DSO trace. I took the timing data for the encoder and the voltage data for the potentiometer and converted them into motion (image 3). The plot also show a simple local (trailing data) slope calculation for the velocity, just to give a feel. The data later in the cycle is probably influenced by slow things like my hand but I'm starting to gain confidence in the shot cycle timing itself.

Some more testing to be done for sure, but it looks like there is some potential in that potentiometer.

[George Schmermund]

Very puny! It looks like some bidirectional action is starting to show up.

[George Schmermund]

Stan - Have you measured the cocking force on your Titan?

[WhatUPSbox?]

I have not. I don't have the dimensions to calculate the leverage and right now I have wires down the barrel so clamping on a scale mount will take a little thinking. For now I've been working from Michael's estimate for a gas ram of 180 lbs max. Though my accelerations were showing less.