Accelerometer?

[bob4432]

i am not sure where to put this so i will put it here, if a mod sees that it should be in a better area, please move it.

i would like to do some testing on different types of recoil reducing devices on certain firearms (long guns) and was thinking about attaching an accelerometer to the firearm so i can have a readout of what happens.

unfortunately i don't know anything about them - how much they cost, what type i would need or anything.

any "pointing me in the correct direction" would be greatly appreciated.

 

[JohnCU]

we use accelerometers at work to measure any possible seismic activity that may occur (to evaluate how strong the earthquake was and see if it was beyond our design limits of the site) and they are hooked up to a computer (standard desktop) that constantly monitors the meters. if the thing starts shaking above .1g (i think?) it starts recording and analyzing. gives you all these neat graphs and stuff. the company who makes it is kinemetrics, our system was really expensive but maybe you can find a smaller version that comes with software, etc.

 

[PottedMeat]

I've used Analog Devices accelerometers before, although those were bare ICs soldered onto custom boards. Their products range from +/- 1g to 10g to 100g. From digikey you can get sample boards for ~50 bucks, those I believe can give you an analog output ( 0 - 5 V or 3.3V proportional to the force ) or a PWM output, where g is proportional to the time between the rising edge to the falling edge of an output clock. There are nicer ones with a digital output, but you need a little microcontroller experience to use them. The accelerometers themselves are ~5-10 bucks a piece and are a challenge to solder unless you've done SMD work before.

Before you buy an accelerometer you should know the bandwidth of the movement you're recording, and obviously the range of forces you want to measure. How are you going to collect the data?

 

[bob4432]

thanks for the info. i really don't know how i am going to do it. i was kind of wondering how to actually go about doing it. as far as how much data, again, not really sure the amount in bytes, but for just very short periods of time. maybe something like if i were to load a mag in the firearm, i could just shoot as soon as i re-stabalized and squeeze of another round. the a laptop could just keep tracking the data but i could do 20rnds in 3-5mins easy - don't need accuracy just need to get the acceleration of the weapon.

as far as smds, i have done a bit w/ them in the past but very few.

anybody know of a schematic or board layout that i could possibly have a friend make a pcb for? if that is the way to go. again, i am pretty lost w/ this whole setup so any direction would be welcome.

thanks in advance,
bob

 

[Cattlegod]

Whatever setup you end up going with, use temporary dental cement to secure the accelerometer to the pistol. When I was in vehicle development, I used to use accelerometers all day so I know a little about them. You will probably be better off renting some equipment, the stuff can get very expensive if you want good results. Remember, you will have to calibrate the accelerometer prior to taking your measurements too. Also, to ensure you get repeatable test results, you may not want to fire the gun yourself, but make a fixture to mount the gun.

 

[bob4432]

Originally posted by: Cattlegod
Whatever setup you end up going with, use temporary dental cement to secure the accelerometer to the pistol. When I was in vehicle development, I used to use accelerometers all day so I know a little about them. You will probably be better off renting some equipment, the stuff can get very expensive if you want good results. Remember, you will have to calibrate the accelerometer prior to taking your measurements too. Also, to ensure you get repeatable test results, you may not want to fire the gun yourself, but make a fixture to mount the gun.

where would one even rent this type of stuff at? a fixture i don't think would be that big of a deal, should be able to get one of those or make one - we can load stuff one at a time if we need to. we were planning on making some type of holder (that will be the easy part) for the electronics to go on a 1913 rail if we made the stuff ourselves, but don't know about how to do it for renting...

what would i ask for regarding rental?

 

[Cattlegod]

Originally posted by: bob4432

Originally posted by: Cattlegod
Whatever setup you end up going with, use temporary dental cement to secure the accelerometer to the pistol. When I was in vehicle development, I used to use accelerometers all day so I know a little about them. You will probably be better off renting some equipment, the stuff can get very expensive if you want good results. Remember, you will have to calibrate the accelerometer prior to taking your measurements too. Also, to ensure you get repeatable test results, you may not want to fire the gun yourself, but make a fixture to mount the gun.

where would one even rent this type of stuff at? a fixture i don't think would be that big of a deal, should be able to get one of those or make one - we can load stuff one at a time if we need to. we were planning on making some type of holder (that will be the easy part) for the electronics to go on a 1913 rail if we made the stuff ourselves, but don't know about how to do it for renting...

what would i ask for regarding rental?

Not sure what your application is for, but you may be able to contact a local university's engineering department and let them know what you plan on doing and offer to share the results. They likely have some if they do anything with NVH. Otherwise open your phone book and look for some acoustical companies.

The reason I am suggesting not doing this yourself is because you will need some extremely sensitive equipment. A gunshot is over in milliseconds and it will have a VERY high instantaneous acceleration. It would also be nice to have access to any of the processing software that you won't be able to provide yourself to see what frequencies you are dealing with (i'd imagine the bullet exiting the chamber would excite some interesting vibration modes in the weapon). In addition, signal processing software will allow you to see how smooth of a reduction your different coils provide over time.

 

[bob4432]

thanks for the information. will have a look around. first setup will be on a run of the mill ar style weapon since they can be had in many different calibers but still use just a couple of recoil assemblies based on which stock you use, either full length or one of the adjustable ones.

there is talk of barrel whip/vibrations, but that stuff is for the long range shooters

 

[OS]

http://www.endevco.com/

you can start with the 2222c part, i think i remember fixing them with just superglue, you need some adapter but you can get the output to labview on a PC.

keep in mind the basic accelerometers only measure force on one axis, if you care about motion in more than one direction you'll need a 3 axis model.

 

[Modelworks]

Pick up this months 1/08 issue of nuts&volts magazine.
It has an article in there about accelerometers and how to read the data from them.
They are reading data from a rocket, but it should help give you some ideas.

There is an even better article in the Sep 07 issue.
http://www.nutsvolts.com/preview.php?issue=61

This is the device they are using:
http://www.analog.com/en/prod/0,2877,ADXL321,00.html

This shows how simple the program to use the data really is.
http://www.nutsvolts.com/downloads.php

Pick the datalogger.zip

 

[Throckmorton]

Originally posted by: Modelworks
Pick up this months 1/08 issue of nuts&volts magazine.
It has an article in there about accelerometers and how to read the data from them.
They are reading data from a rocket, but it should help give you some ideas.

There is an even better article in the Sep 07 issue.
http://www.nutsvolts.com/preview.php?issue=61

This is the device they are using:
http://www.analog.com/en/prod/0,2877,ADXL321,00.html

This shows how simple the program to use the data really is.
http://www.nutsvolts.com/downloads.php

Pick the datalogger.zip

Won't he need more than 18g?

 

[Modelworks]

Originally posted by: Throckmorton


Won't he need more than 18g?

I seriously doubt it.
As he is measuring recoil on the gun itself , the part a person would actually be holding.
18+g recoil on the stock of a gun would be one hell of a kick from a rifle.

Still if needed the http://www.analog.com/en/prod/..._800_ADXL193%2C00.html
Can handle 250g and has the same data format.

 

[bob4432]

now would these items have the speed and repeatability to record the differences in the speed of the recoil on the 3axis? during the tests, the firearm will be nearly the same weight (+/- 5oz or so), but some of the internal moving mass will be different as to slow down the recoil.

it has been a long time since i have messed around w/ electronics like this, so i need to clear out the cob webbs, so if i ask a dumb question, please bear w/ me

thanks for your assitance

also, the firearms that ideally will be tested are all "consumer" level, not big game rifles or anything like that. there may be some f/a weapons as i have friends that have a couple of those, but mostly it will be 1 shot at a time for 10-20 rounds, change a part, then another set of shots, and contiune.

 

[canis]

Originally posted by: bob4432
i am not sure where to put this so i will put it here, if a mod sees that it should be in a better area, please move it.

i would like to do some testing on different types of recoil reducing devices on certain firearms (long guns) and was thinking about attaching an accelerometer to the firearm so i can have a readout of what happens.

unfortunately i don't know anything about them - how much they cost, what type i would need or anything.

any "pointing me in the correct direction" would be greatly appreciated.

Can't you just feel the recoil and see if it is reduced? Isn't that the point of reducing recoil in a firearm?

 

[bob4432]

Originally posted by: canis

Originally posted by: bob4432
i am not sure where to put this so i will put it here, if a mod sees that it should be in a better area, please move it.

i would like to do some testing on different types of recoil reducing devices on certain firearms (long guns) and was thinking about attaching an accelerometer to the firearm so i can have a readout of what happens.

unfortunately i don't know anything about them - how much they cost, what type i would need or anything.

any "pointing me in the correct direction" would be greatly appreciated.

Can't you just feel the recoil and see if it is reduced? Isn't that the point of reducing recoil in a firearm?

everybody is different - some may feel it others may not, but i would like to see how the whole thing happens from initial ignition, to bolt slamming back, to the bolt slamming foward, etc.

also, if you make a part w/ f/a in mind but dont have the $$$ (take ar15/m16 for example - ar15 - $600-$800, pre-86 m16 - $15K-$20K but they are basically the same minus a couple key parts) to purchase one, then you can see how such item would slow down the cycling rate of the weapon.

 

[thebeyonder]

however it's attached, make sure it's bolted on tightly or you're wasting your time, the results won't be as accurate as the instrument is capable of.

there are 3 different things you can examine with your testing.

1. acceleration. measured by the accelerometer. may be useful to you , since you've already got your ideas.

2. impulse. probably what you should look at, probably what the engineers actually consider instead of acceleration, and it can be figured out with the accelerometer readings, if not already displayed as a function of the readout. impulse is the 1st derivative of acceleration with regard to time.

you are thinking acceleration is important, because you're attaching some mass to that acceleration, so you're actually considering force. F=MA. but instead you should look at how much impulse (I) the rifle gives. again, it's the first derivative, and so the equation is I=MV. that is what pushes your shoulder back. a mass with some velocity.

3. energy. that's the raw number people refer to when trying to figure out how much kick a rifle has (or the bullet on the other end). it's the 2nd derivative of acceleration, and so you see the equation is E=MX. ...foot-pounds.

anyways, the recoil from AR type rifles is not negligible, but, shall we say, inconsequential. you would do better to develop a recoil reducer for larger caliber rifles (with more impulse) or make one that eliminates muzzle climb, the testing of which would require a whole different platform that would be capable of measuring torque seperately from the pure negative thrust vector.

 

[thebeyonder]

btw, if you're really interested in how muzzle brakes, cyclic rate and recoil relate together, check out the design of the Nikonov AN94. the russians are decades ahead of us.

http://www.enemyforces.com/firearms/abakan_2.jpg

haven't been able to figure out that muzzle brake yet.

 

[bob4432]

Originally posted by: thebeyonder
however it's attached, make sure it's bolted on tightly or you're wasting your time, the results won't be as accurate as the instrument is capable of.

there are 3 different things you can examine with your testing.

1. acceleration. measured by the accelerometer. may be useful to you , since you've already got your ideas.

2. impulse. probably what you should look at, probably what the engineers actually consider instead of acceleration, and it can be figured out with the accelerometer readings, if not already displayed as a function of the readout. impulse is the 1st derivative of acceleration with regard to time.

you are thinking acceleration is important, because you're attaching some mass to that acceleration, so you're actually considering force. F=MA. but instead you should look at how much impulse (I) the rifle gives. again, it's the first derivative, and so the equation is I=MV. that is what pushes your shoulder back. a mass with some velocity.

3. energy. that's the raw number people refer to when trying to figure out how much kick a rifle has (or the bullet on the other end). it's the 2nd derivative of acceleration, and so you see the equation is E=MX. ...foot-pounds.

anyways, the recoil from AR type rifles is not negligible, but, shall we say, inconsequential. you would do better to develop a recoil reducer for larger caliber rifles (with more impulse) or make one that eliminates muzzle climb, the testing of which would require a whole different platform that would be capable of measuring torque seperately from the pure negative thrust vector.

thanks for the info. the housing will be no problem and attached to a ff rail - no play, it will be solid.

as far as the ar platform, recoil is relative - years ago i shot all types of firearms w/out so much as a thought regariding recoil (i would should 3" 12ga magnum shells all day like a 22 - no isues), enter in a couple neurological diseases/conditions and the ar now has a lot of recoil perceived by me.

plus when you mix in the fact that there are so many different brakes and buffers for that platform i wouldn't have to create or have created each piece to test. buffers alone, w/out even thinking about it there are at least 5+, and brakes, again, 5-8+ w/out even looking them up. honestly if i didn't have my medical issues i wouldn't care, but unfortunately that is not the case.

my main concern is still the speed of the sensors...i need to read more on them. the model rocketry unit is done in mS, but i think i need resolution much lower than that.

i have a great deal of respect for the russians and some of their designs - have always been impressed w/ them in that realm. that is on interesting muzzle brake

 

[thebeyonder]

how about zero response time? no electronics. like, put a pen at the end of the buttstock, and set it on a sheet of graph paper coming off a roller, going perpendicular to the motion of the buttstock. you can adjust the graph speed and whatever motion damping system you rig up to suit your resolution requirements.

things to take into account:

-mass of rifle, magazine, the spent cartridge, carrier assembly

-coefficient of friction between carrier and platform

-characteristics of whatever motion damping system you come up with. if you use a spring, for example, figure out the spring constant. or a plunger going into sand. or a peice of foam rubber. anything with reproducible effects.

that would be more work than just bolting on an accelerometer, but by the time you're done you would have a clear idea of how everything works, including the math.