Chronograph Testing

[DX]

So, I got this and finally had a *brief* window of opportunity to test it before a thunderstorm today. This is the only model of chronograph that I was able to find on the internet with water resistance. I'm still apprehensive about testing turbulent streams that break up a lot. Anyway, I took a string of 10 shots with an XP 150 and got:

69 FPS
37
49
55
54
56
67
57
52
42

This indicates to me that nozzle velocity testing will require multiple shots, you can't just shoot once and call that it. There also may be an optimal distance and angle from the chronograph to place the nozzle. I didn't take these shots from a set distance or angle, I was just excited to finally be able to measure this important stat. Those things will need to be standardized, because there is quite a difference between 69 and 37.

[the oncoming storm]

I would shoot level and at a range of 10' for all my tests,

[DX]

I'm actually thinking it would be best to shoot close to the first sensor, as close as possible. This is because the 150 stream, which I would consider to be well-laminated, still sprayed the sensors and I kind of want to avoid getting water in those areas if at all feasible. Streams give off less spray the closer you get to the nozzle.

[SSCBen]

Nice work. I've considered using my Chrony for this, but I was unsure whether the result would be accurate, and your concern about water damaging the electronics is valid.

The velocities seem to be quite reasonable. Based on the flow rate (output) and nozzle diameter of the XP 150 that I have in my spreadsheet, I estimated the nozzle velocity to be 46 ft/s. The average of your data is 54 ft/s. This fit is quite reasonable, especially after considering that the I believe the flow rates reported are too low. Dropoff in air pressure blasters makes the average flow rate lower than the initial flow rate. The chronograph should measure the initial flow rate.

[HBWW]

I think I'll try to set up video-based testing sometime, but I'll have to come up with a standard for my own testing. That is, if I find the time for it.

Your chrono results are all over the place though. Now I ain't no statistician or nuthin', but I'm pretty sure there's something odd going on when the lowest result is nearly half of that of the highest. Perhaps part of it has to do with it measuring the "last drop" correctly or incorrectly; maybe it won't detect a given droplet all the time?

What's the slowest speed this thing can clock?

[marauder]

That's pretty awesome. Is there any way you could take measurements in order to plot the change in stream velocity at different ranges? For instance, you could show initial velocity, then velocity at 10 ft, 20 ft, 30 ft, etc.

[DX]

These were tap shots, not full shots. With any soaker, but especially air pressure, the beginning is the most powerful part of the shot and that's what I wanted to capture. I can do initial, 10ft, 20ft, etc, but am very wary of approaching the end 3rd of a stream. The point at which the stream starts dropping is where it will go all over the sensors and possibly ruin my expensive toy. I'm not sure why there was such a huge gap in the beginning, I didn't change much when shooting those first two.

22ish FPS is the minimum threshold for this chrono. Anything less than that will not register (aka, the new definition of a bad soaker).

[soakinader]

I think that part of the problem is that it measures the time it takes for an object to leave the first sensor and hit the next one- and a variant length stream of water is probably going to cause some errors.

[DX]

I was under the impression that it measures the time from when the first sensor registers something, not when the sensor is completely cleared. Otherwise, (just mentioned by someone on NMW) how would it measure archery arrows? Those arrows are longer than the whole chrono.

[isoaker]

Neat "toy", though I also wonder whether the variation in FPS you've recorded are due to how well the sensors are detecting the stream passing through. This is where video-based stream speed measuring may yield more consistent values since one can see when one wants to begin counting as the stream passing a mark.

In terms of how to set up a video-based method, what I'd do would be similar to how I do video capture for determining output from pump-action blasters. I'd have the camera mounted perpendicular to the blaster on a tripod and have a darker, solid-color background with a ruler or markings on it. I'd then do a few shots while the video was rolling (at 30 to 60 frames per sec; higher if possible), then pull it into the software to step through frames to determine the video time the stream passed one point versus passing the next point. Granted, it'd be harder to measure stream speed at the tail end of a shot since there'd be more angle and more likelihood of hitting the camera.

[SSCBen]

Video-based velocity measurement probably is the most accurate. Right now I can think of 5 reasonably easy ways to measure velocity:

1) Empirical equation from the pressure differential.
2) Converting a flow rate (measured with a stop watch or video) to a velocity. This could be misleading due to dropoff.
3) Video-based measurement.
4) Chronograph.
5) Measuring recoil and converting that into a flow rate, and then converting the flow rate into a velocity.

I'd be interested in seeing a comparison of all these different methods. I might do that myself.

[HBWW]

See if you can position the chrono in an inverted position, so that the droplets aren't falling on it.

Neat "toy"

And has far more to contribute to the hobby than anything Nerf has made.

In terms of how to set up a video-based method, what I'd do would be similar to how I do video capture for determining output from pump-action blasters. I'd have the camera mounted perpendicular to the blaster on a tripod and have a darker, solid-color background with a ruler or markings on it. I'd then do a few shots while the video was rolling (at 30 to 60 frames per sec; higher if possible), then pull it into the software to step through frames to determine the video time the stream passed one point versus passing the next point. Granted, it'd be harder to measure stream speed at the tail end of a shot since there'd be more angle and more likelihood of hitting the camera.

The setup I had in mind involved setting the camera a bit far out, facing the side and getting the whole shooting range in the frame. (Think Super Mario Bros. camera position) The blaster is shot flat at shoulder height, and I only had one or two wall markings in mind, although more would certainly be useful.

From there, I'd look at 2 measurements: front end and tail end of stream. I have to check the max frame rate of the GoPro (I think the HERO2 can go to 120 FPS at low res: 640x480), but obviously, I'm not going to try anything below 60 FPS. (Which I can run at 720p.) From going frame by frame when editing the Frozen Fury 2013 video, I've noticed that for each frame at 30 FPS, the initial stream seems to travel about 1-3ft. Of course, that doesn't really mean anything: 30-90 FPS is a useless range of numbers. However, doubling the framerate would double the accuracy of the measurement.

If velocity does indeed turn out to be a useful statistic to collect, we should establish measuring standards that we can all use reliably, and start collecting data. The only real reason I can see that it's less useful for water blasters is stream breakup to air resistance, but the data would still help us find that, at ideal nozzle conditions, what maximum muzzle velocity you can get without the stream breaking up to the point of uselessness. (Which is, of course, not an actual point, but more of a gradual change.)

[SSCBen]

If velocity does indeed turn out to be a useful statistic to collect, we should establish measuring standards that we can all use reliably, and start collecting data. The only real reason I can see that it's less useful for water blasters is stream breakup to air resistance, but the data would still help us find that, at ideal nozzle conditions, what maximum muzzle velocity you can get without the stream breaking up to the point of uselessness. (Which is, of course, not an actual point, but more of a gradual change.)

Nozzle velocity is definitely a useful number to have. Most of the correlations for breakup distance of a liquid jet that you can find in the engineering literature are written in terms of the Weber number, which is a function of the velocity (squared). Drag force is also a function of the velocity (squared, usually). In fact, I posted the recent thread on nozzle diameter so that I could convert from a flow rate to a velocity.

The measurement precision for video seems to be pretty bad based on what you've said, CA99. I've been meaning to get a high speed digital camera, and this might motivate me to finally do that.

[HBWW]

It could work out actually, if you're able to figure out what length the stream travels in each frame. That would give you a measurement method that doesn't depend on the framerate of the camera. For most cameras, it's a tradeoff between resolution and framerate. Higher res will allow for more precise numbers reading, while higher framerates are more likely to capture frames where the stream is closer to the measurement interval. Either way, it's trivial to multiply the length traveled between two frames by the framerate itself to get the muzzle velocity. The trouble is getting that measurement of length in the first place.

[SSCBen]

The trouble is getting that measurement of length in the first place.

Ah, I misunderstood. This won't be hard if you use the test method isoaker recommended. The setup would look something like this (from fire hose fog nozzle tests):



You can use the background to figure out a conversion factor from pixels to physical dimensions. I've done similar things before with good results and I recall I recommended a similar procedure for measuring nozzle diameters.

[SEAL]

This is cool. I've been meaning to do something like this but never got around to it. I think velocity is an important stat, though probably not as important as range and output. I think I'd prefer the chrono to some kind of camera setup because it's so much easier to use. Just turn it on and shoot a stream through it. It may be imprecise, but do it enough times and you should get a rough idea of how your blaster performs. I honestly don't think accuracy is all that important as long as you use the same measurement method for each blaster. I mean really, stats like these are only really useful for comparison purposes. In the real world, range and velocity vary greatly depending on conditions.