GTA
All Springer/NP/PCP Air Gun Discussion General => Machine Shop Talk & AG Parts Machining => Share Your Simple Home Projects (TRICKS-N-TIPS) => Topic started by: Jay_sparky on April 09, 2020, 04:37:13 AM
-
Hello Fellow Enthusiasts!
We're under Covid-19 lock down here in New Zealand, so I'm not going batty by designing an air pistol.
These are the unrealistic things I want out of it...
1. More or comparable power to a .22 powder burner.
2. To be compact enough to put in a small pack while walking over the farm.
3. To get at least 2 or 3 regulated shots.
I want to use it for goats and maybe opposums (NZ) at short range, as it is a pistol...
I don't own a PCP yet, but I am on the waiting list for the Umarex Hammer and it's something I would like to get into... pistols are basically illegal in NZ but PCPs are not. I've brought a mould and am casting my own slugs for the Hammer when it comes out.
These were my design goals:
Long as possible barrel.
Electronic valve regulation, as it will be easier to tune and I can make different modes.
Maximum internal air capacity.
Water resistant.
Few as possible sealing faces
Design for 5000psi Working 3000 would be nice
My plan was to route it out of a solid block of aluminium (7075-T6) and my calculations came out with a wall thickness of 1.19mm at 5000 PSI - but I'm designing it with a minimum of 3 mm. Then I was going to build my own circuit board with an Arduino micro-controller and a HX711 load cell chip with my own pressure sensor; a small display to show pressure, mode, and battery level. I also wanted to mount 2 lasers to the front, that cross at the sighted in distance, as distance is very important with such a slow bullet.
Ill link to a google drive when I get enough reply's so people can build there own I am waiting on filament so I can 3D print my own prototypes
Now from my limited research, it looks like the air valve is the most important aspect of a PCP -
so that's where I need help... I've designed for an electronic valve that uses a small amount of air to control a large "ram" - see attached documents for more.... so please help/comment as I'm new to everything.
-
Can't really see how those features can be machined inside a solid block. Where is the air reservoir connected? You want over 100 FPE, in what caliber?
-
So there are 3 main Air chambers 2 in the handle one under the barrel there are 2 4mm chambers connecting the under barrel chamber to back 2 (not shown in cut view as they are not central in the gun)
so it has a internal air capacity of 88.5 cc where as the Rex pistol has 91cc and they make that in .45 where I was going to go for 9mm or larger if i get more than 5 shots Iv tried to plan it so I route all the inner chambers from the outside then thread and bung the ends the fill port is on the end of the handle which ill 3d print o cover for
-
Design for 5000psi Working 3000 would be nice
This is not the usual engineering practice.... A 3:1 minimum safety margin is typical.... The ASME requires 3.5:1 minimum to burst for a pressure vessel.... If you want a 3000 psi working pressure, you should be designing for about 10000 psi before failure.... Additionally, as understand it, you can't treat a flat-sided or rectangular reservoir the same as a tube, as the stress will be greater on the thin sides.... However, I do not have the math to understand that relationship fully....
Bob
-
Thanks for that Bob I am a novice so thats good to know I thought it sounded thin but even at 45000psi * 3.5 = 3.75mm wall so thats not much of a change from my 3mm..... (still pipe calcs though) do you think the Pilot/Main valve will operate fast enough?
-
Sorry, I don't know anything about piloted valves.... Maybe PM Gippeto….
Bob
-
So another day in lockdown another re-draw this time walls are 4mm thick and I added a 3 shot mag
-
Update ... not much has happened.... but ill give a run down on the electronics side of things... so originally i was planning on using a half bridge restive "loadcell" sensors on a squishy substrate as my pressure sensor as I couldn't find a sensor cheap and small enough.... but now iv found one! its a pezio restive element! (which i think must be a semi new technology as I was unaware of it) a 12.5mm D * 6mm high disk
So a pezio restive element (BX-PSC14FC 400bar) connected to a Hx711 chip controlled by a ATMega32U4 (running the Arduino IDE) with a 2 stage electronic trigger and 0.96 inch Oled 128X64 I2C display
this will control the length of time the solenoid is open / the power of the shot so I can regulate the speed / number of shots
More info and 3d print files in my Google ...... sorry still not aloud links email me if you want them
-
If you make the barrel as long as possible, than it's not a "pistol".
A standard velocity 22lr 40gr is about 100fpe. If you make the barrel maximum pistol length (38cm or 15"), you might get close to that with 5000psi on deck. Not a very efficient way to get "comparable power to a .22 powder burner". In order to get 2 or 3 shots, you'll need lots of air.
It might be easier to do a 100fpe air pistol with a larger caliber.
I could imagine a 100-120gr 9mm, with a 6" to 8" barrel and 2 or 3 shots at 100-200fpe.
Based on the drawing shown, I can't quite figure out the operation of the valve/powerplant.
-
Yes aim high shoot low... I have lots more details and calculators for others to look at (very rough) but i am unable to link them ..... ill try upload some info now
I am planning a 8" / 200mm barrel (long as possible barrel meant to have the valve/mag as far to the back as possible) and my calcs seem off? but iv come to over 100cc internal H.P.A
1. Internal air chambers (20mm)
2. 4mm small off-center chambers connecting rear and front air chambers
3. PEEK air ram force on the back face (20mm) is greater than on the front (10mm)
and it holds itself shut until the solenoid vents the air behind the ram and it blows open...until the solenoid closes and the force is greater on the rear closing itself
4. Rear bung/ acces to Ram (3)
5. Mag – shoot through
6. Solinoid (see desription)
7. Pressure sensor
8. Arduino micro processor, pressure readout and electronic trigger
9. Air Inlet can be used on a lanyard
10. Burst disk and bung to access the pressure sensor (will be fine thread just coarse for 3D print prototype)
11. 3D printed inlet latch
12. 3D printed inlet cover
So the high pressure air passes the actuator and keeps the pressure on the rear of the ram. Then when the solenoid is energised the actuator is pulled back and air is vented from the rear chamber, also stopping the main chamber from refilling it.
When the rear chamber is vented, the pressure on the front face pushes the ram back, venting to the barrel. The safety is basicly a "tap" that stops the back of the ram from venting
-
This is all very interesting, but how are you going to machine the air reservoirs in the handle? There is no visible end plug in your drawing?
Where does the solenoid assembly go as I have a hard time finding it in the frame picture?
Don't guess on any of the design criteria on high pressure application, do your homework and make the calculations based on the material and design. One wrong guess and its your head, or in worst case the head of some one else!
Marko
-
I have not drawn the end plug threads as its a 3d print prototype, I was hoping to Pressure test it "virtually" on Solidworks to find the weak points
as for an explanation on how it works here is some more drawings hopefully some what self explanatory (drawings not shown on phones?)
on the basic drawings the purple part is magnetic and pulled to the solenoid venting the pressure on the back of the ram allowing it to move back venting the front to the barrel
-
I hope you get to the milling and lathe machining part soon. there is no point in playing with solid works any longer. calculate your max pressure and machine 10% thicker.
i dont see 14 inch barrels in .223 making 120fpe with 3000psi. maybe other bigger calibers.
i i am guessing this will take a long while that you still playing with your computer...been almost a month now. you should be making small parts by now if not the frame.
-
I'll just throw a few numbers out here for you to consider....
"Lofty goal" for an 8" barrel, 0.357 cal, at 3000 psi is (0.357 x 0.357 x PI/4) x 3000 x (8 / 24) = 100 FPE.... If you can manage to achieve that, you have done a great job, I have never achieved more than 90% of a "lofty goal" calculation....
To produce that power, the barrel will contain 0.8 CI (13.1 cc) of air at 3000 psi.... Three shots at that power will therefore require (13.1 x 3) = 39.3 cc of air at 3000 psi.... or the equivalent mass of air at a higher pressure.... roughly 9.4 grams of air.... That assumes you can close the valve at exactly the moment the bullet leaves the barrel.... otherwise you will be wasting air....
You have mentioned a total reservoir volume of 100 cc.... In order to have 3000 psi available after the 3rd shot, you will have to start at a much higher pressure, and carefully control the volume of air released so that all 3 shots will have the same (or nearly the same) power.... You will need to add 9.4 grams more air to the reservoir above what is required to fill it to 3000 psi (100 x .239) = 23.9 grams.... The total air mass when filled will therefore be (23.9 + 9.4) = 33.3 grams, or a density of 0.333 g/cc, which means a fill pressure of 4550 psi....
In order to achieve your goal of 3 shots of 100 FPE, using a 9mm barrel that is 200mm long, it looks to me like you will need at least 100 cc of air at 4550 psi, and perfect control over each shot to insure that no air is wasted.... IMO that is a challenging goal, and I wish you luck with it....
Bob
-
Now I see it, well you need to make the main piston as light as you can and at the same time you need to make it long enough that it doesn't rub on the cylinder. You can consider atmosphere as vacuum when talking about high pressure.
The valve piston is sucked against the port where there is atmospheric pressure.
It will be forced against the cylinder wall and if its too short it will seize. Don't ask how I know.
Marko
-
Thanks for the input guys Im working on a 9mm ".357" bore I have over 100cc of air @ 4500 psi and I am using electronic valve con troll so i can switch to the micro second...... But... Im waiting on the guy that was going to supply the aluminum .... hes also an enthusiast and he sounded keen but hasn't got back to me.... and this was going to be a "long term project" as I have a young kid and other projects as well. Sorry :-[
-
dont be sorry man, i am just trying to push you into the production stage...
go online and buy the 7075-t6 and have it shipped to you. let no one control the fate of your projects, even kids...just kidding about the kids...but ditch that other supply guy
-
Thanks for the input guys Im working on a 9mm ".357" bore I have over 100cc of air @ 4500 psi and I am using electronic valve con troll so i can switch to the micro second...... But... Im waiting on the guy that was going to supply the aluminum .... hes also an enthusiast and he sounded keen but hasn't got back to me.... and this was going to be a "long term project" as I have a young kid and other projects as well. Sorry :-[
Now that looks to be a realistic goal. I like it and I like the valve concept.
-
I can understand how the valve opens.... but how are you going to control the closing of the valve within a microsecond?....
Bob
-
It's quite possible to time the pulse to within a microsecond using a simple Arduino.
It's harder to control the valve timing, as then one needs to consider, springs, mass, inertia, coil inductance, etc. The faster one tries to cut off the current in the coil, the higher the back EMF. At a minimum, one needs diodes across the coils, just to ensure not blowing out your drivers.
One still doesn't know (directly) if the valve is closed, except by measuring the report, or air consumption. It would be interesting to add some instrumentation, like a linear transducer to the poppet so one would know when it's truly closed. Might make tuning a lot easier. Of course the linear transducer will affect the timing, as there's no free lunch.
The first low mass linear transducer that I found is based on a coil with a moveable aluminum slug. The slug mass is 4.5 grams. The response time is 35us. However, that is the response time of the output, which is run through a small dedicated processor. Since one already might be using an Arduino, for timing, perhaps the transducer processing might be done with the Arduino as well? https://files.sentechlvdt.com/datasheets/DCFastar.pdf (https://files.sentechlvdt.com/datasheets/DCFastar.pdf) I'm sure there are other transducers, this was the first hit. Maybe it's possible to just monitor the firing coil's current or back EMF and infer the solenoid shaft position?
-
Analog hall effect sensor could be used for "ram" position measurement, only needs a small magnet and it would be a solid state measurement. I don't remember the model I used but it was TO-92 package and analog 0-5 v output, very cheap part. Using arduino is not ideal if high precision timing is required, problem is the language itself, of course it is possible to get close with clever use of interrupts and inline ASM, but I guess you already know all this.
-
Make the valve NC with a spring / pressure and all you need to control is open time. No need to over complicate the system for no reason.
Coil energizing time is the only variable then with pressure.
You dont need to control or monitor the closing, what there is to gain for it?
Marko
-
Analog hall effect sensor could be used for "ram" position measurement, only needs a small magnet and it would be a solid state measurement. I don't remember the model I used but it was TO-92 package and analog 0-5 v output, very cheap part. Using arduino is not ideal if high precision timing is required, problem is the language itself, of course it is possible to get close with clever use of interrupts and inline ASM, but I guess you already know all this.
What Marko said, so...
No need to sense the position. If the time is adjustable and can be programmed, that's enough. The design has a pressure sensor. The time should be based on the pressure available. The time needed for any given pressure can be determined experimentally after the build is complete. That will be the tuning process. That time-vs-pressure curve gets mapped into the processor. As an example: 1st shot at 4500psi might require 0.0010s. 2nd shot at 4000psi might require 0.0015s. 3rd shot at 3500psi might require 0.0020s. As the pressure falls, velocity stays relatively constant for the three shots.
-
Yes I know it is useless to sense the position, I thought it would just be interesting to see whats happening.
-
Well, if you guys like to run blind, that's ok with me. Yeah, it isn't necessary, you could always do it the way you always have. That's not a way to progress, in my opinion.
One could, with adequate internal sensors have the peak air efficiency for a projectile, by measuring what's going on. Or dial in the exact FPE you want. The computer would adapt the dwell time to the existing air pressure to maintain consistent shots. Whatever.
Just because you kill the pulse to the solenoid, doesn't mean you know it's position. Postulating what happens is not the same as measuring and knowing what happens. My two cents. Spend them as you may ;)
-
Well, if you guys like to run blind, that's ok with me. Yeah, it isn't necessary, you could always do it the way you always have. That's not a way to progress, in my opinion.
One could, with adequate internal sensors have the peak air efficiency for a projectile, by measuring what's going on. Or dial in the exact FPE you want. The computer would adapt the dwell time to the existing air pressure to maintain consistent shots. Whatever.
Just because you kill the pulse to the solenoid, doesn't mean you know it's position. Postulating what happens is not the same as measuring and knowing what happens. My two cents. Spend them as you may ;)
I agree that even if the final model does not need position sensing, position sensing could serve as confirmation during development. But it would complicate the development prototype.
"...run blind"? - hardly, position can be calculated:
https://www.gatewaytoairguns.org/GTA/index.php?topic=170568.msg155946628#msg155946628 (https://www.gatewaytoairguns.org/GTA/index.php?topic=170568.msg155946628#msg155946628)
This thread should probably be in the "Engineering Research & Development" section as it is not "Simple...".
-
i would vote for some type of pressure sensor in the exhaust area between the pellet and the exhaust. if anyone know of such a sensor please post it here. it would be very helpful to any build
-
I must be missing something, I still don't understand what closes the pilot valve to allow repressurization of the main valve to close it....
Bob
-
I must be missing something, I still don't understand what closes the pilot valve to allow repressurization of the main valve to close it....
Bob
I'm guessing the armature return spring (spring force would be = 1/2 solenoid force). My feeling is that the difficulty will be to get a small enough solenoid that can operate the pilot valve.
-
In the drawing the "purple" piece that switches the flow is not shown attached to the armature of the solenoid, but "drawn" towards it because it is made of magnetic material.... I'm hoping the OP will explain, rather than us guessing....
Bob
-
In the drawing the "purple" piece that switches the flow is not shown attached to the armature of the solenoid, but "drawn" towards it because it is made of magnetic material.... I'm hoping the OP will explain, rather than us guessing....
Bob
I missed that part - "magnetic material". I'm not familiar with them but that could be a bistable latching solenoid with two opposing latched positions. Apply a pulse and it latches in 1st position. Apply an opposite polarity pulse and it latches in the opposite position.
-
Looking at the picture, that slide valve for the pilot will never work!
There is no way and no how to move the slide against the full reservoir pressure, when the pressure is pushing it sideways against the port wall where it is sealing the port, if you get it to seal in the first place.
You have pressure in the end of the magnetic piece and at the side.
With an air gap to the coil. No bueno.
Let's assume 4mm sealing solenoid plunger port and at 300bar you are seeing 376N of force on the sealing surface. So basically you are trying to move a 37kg with a tiny solenoid. Given my rough calculations are correct.
Marko
-
Looking at the picture, that slide valve for the pilot will never work!
There is no way and no how to move the slide against the full reservoir pressure, when the pressure is pushing it sideways against the port wall where it is sealing the port, if you get it to seal in the first place.
You have pressure in the end of the magnetic piece and at the side.
With an air gap to the coil. No bueno.
Let's assume 4mm sealing solenoid plunger port and at 300bar you are seeing 376N of force on the sealing surface. So basically you are trying to move a 37kg with a tiny solenoid. Given my rough calculations are correct.
Marko
I also have some reservations about the valve but it's not as bad as you think. That 4mm in your example could be closer to 1mm.
-
We're all just guessing.... ::)
Bob
-
I was going to start with very short on times until the valve starts opening ....(the mass of the actuator and the back of the ram vent times will be reasonably long) then plot the muzzle velocity vs on time until the curve starts flattening off and this will be the when the bullet has left the barrel do this at 3000 or so PSI and this will tell me the max on time then see how long I need it on to get the same velocity at max pressure
I think my biggest problems will be 1: having enough energy to move the actuator off the vent port and 2: being able to close the ram fast enough (for efficiency)
So 1: I want to use 5v (very low I know) and 4amps will let me lift about 800grams with a m8 bolt more amps doesnt seem to mean more "lift" just more heat. Iv found .5mm wire the best so far.
It seems like to get more "lift" I need more iron m12 bolt will lift way more for the same number of turns&s (more of an electronics question I know) but does anyone know if "800grams" of force will move my 5mm˛ actuator off the vent port (4mm˛)
2: when the ram is open and the actuator closes the vent it will take time for the back of the ram to pressurize and I think it will close "slowly" will i need to but a partial block in the air tank between the ram port and the actuator port? which should help with a pressure difference when air is flowing
-
Last reply I did not realize there was another page of posts.... attached is a more detailed picture of pilot valve I was just going to scratch the wall of the valve to let air to the back of the actuator
And I KNOW I will not be able to make the actuator seal! so I was going to put the "safety" after the actuator which will seal it.
So when you're ready to fire it'll start a slow leak, but will only do so when the safety is off.
As far as actuator material...I was thinking about having a magnetic actuator slightly repelled by the solenoid core (also magnetic) When you fire you pulse the core and pull the actuator in and then you reverse the coil and push it closed again and the hysteresis should keep it slightly repelled... but I'm keeping it simple.
As far as the timing in the arduino goes I was hoping an interrupt would do it.
The pressure sensor I was going to use is very small - 12mmDx6mmH I could put one in the port but also I'm trying to keep it small and simple.
-
Your valve closing force at 300 bar is something like 147N or 14kg with 2.5mm port or ~5mm2 surface area. Not accounting surface friction in to that.
Your 800g force is too little by a large margin!
Not trying to put you down, but do some calculations before you make something out of metal.
Marko
-
Thanks Marko that is what I am here for Your knowledge! ;D that is sideways force not the force to slide it? but still to much force :'(
So here is another design that I am not happy with as I do not think it will seal
So the ram is completely balanced and contains a magnet that holds to the electromagnet keeping it shut... when fired the electromagnet is turned on and repels the ram....
then reversed and pulled shut again
But.. Iv spent a bit of time thinking about the next design..
-
So here it is! I think it looks good and I am keen to hear what you think
So you hit the pin with a hammer (not necessarily electronic :'( ) this vents the back of the ram through the tube with the sleeve covering the ports..... the ram moves back opening the H.P.A. to the barrel, then it reaches the back the travel in the screws runs out and pules the sleeve back opening the ports by now the hammer has backed off the pin resealing the tube and the air from the ports flow through the tube into the back of the ram, closing the ram with the ports open until nearly closed then the ram pushes the sleeve over the ports..... with only leakage keeping the pressure on the back
tuning would be done by adjusting the screws that allow travel of the sleeve or by adjusting the expanding air past the pin
-
If you are asking for opinions on a particular design I think it would be much better to draw it correctly first. Put all the seals and correct dimensions in there, often the devil is in the detail, it is easy to come up with craziest of ideas and only when you start making parts you realize that something does not fit or is too difficult to make.
-
I would agree, I don't understand the operation from that drawing and description.... If there is pressure inside the sleeve at any time, then there is a closing force on the "pin" which is driven by the hammer, because there is atmospheric pressure on the left in that drawing....
Bob
-
That last idea might be feasible but like Albert said there are complications bound to happen if you don't start to design the parts properly.
The design idea could work but the design in it self is overly complicated.
Think what the forces are, what holds the main valve close, and what you need to do to close it in the first place.
Two moving parts is all it needs, rest of it is just passages and or springs if you choose to use one.
Marko