GeckoSub Mirage Evo - And Adventures in 3D Printing Speargun Parts

[Diving Gecko]

How deep the power regulator O-ring pass by the edge of the entrance for pumping mode? Try to make it minimal, just sufficient for closing the entrance.. You already know this...

Aha, you mean I should make sure the power regulator is minimally inserted into the bore - in order to reduce for how long a distance the force can act on it when opening the valve and thus reduce how much speed it will gain and as a result, the energy of the impact when it opens?

Sounds obvious now that you say it, but honestly I didn't think about it at all...
On the contrary, it's inserted quite a few mm more than it needs to be - to make sure it wasn't inserted too little. But I can likely screw the bushing further down the regulator rod and still have it seal.

If I can make my new valve idea work, it wont matter but it certainly sounds like a very worthwhile adjustment on a tradtitional valve like this, especially one with a larger diameter than the original Mirage had. Thanks Tomi

 

[Diving Gecko]

The sliding friction force does not depend on the contact area! It depends on the friction coefficient of the rubbing materials and the pressing force! The static friction force for rubber with metal depends on the contact area and is 10 times greater than the sliding friction force! The large barrel diameter allows for more efficient use of the pneumatic vacuum muzzle!

Also, in a Mirage system a larger shooting barrel allows for a bigger ratio between it and the pumping barrel - in effect making the pump/surcompressor effect more powerful
Another nice side benefit is that I can lower the overall pressure with a larger shooting piston and still have the same power. This could be helpful since I also have a dream of making my own cuttlefish (C4ish) style carbon reservoirs at some point.

 

[Diving Gecko]

New Bulkhead In The Making - Part 2
I took yesterday off, but back in the workshop today. Last time around, at the end of Op 1 which are all the features I can reach from the front side, I hadn't programmed two of the chamfers perfectly. So, they stopped short when they were close to the filleted sealing lip for the power regulator valve.

I considered just finishing those chamfers in hand but I had a go at re-indicating the part in the vise and it went well

The chamfers look almost perfect, just one little weirdness on the shooting barrel bore edge where it meets the sealing lip:

You can kinde see how it's ever so slightly higher, so maybe the Z-offset was not exactly the same. Anyhow, that little blemish can be ground and polished and overall, I think the finish looks great. Especially the inside bore walls come out very smooth.

I also engraved a version number on it so I can keep track of it down the line. But I got lazy and just used a 45d chamfer tool to do it. I have some tiny engraving tools, I have used to greater effect before.

OK, on to Op 2 which is all the remaining rear features.

 

[Diving Gecko]

New Bulkhead In The Making - Part 3
I didn't take as many pics this time around, but first let's start with the final product

No big dramas on the rear features. But I also spent quite a long time programming and checking the tool paths in the sim beforehand. It was mostly a case of remembering to have enough stickout on the tools so I wouldn't ram the spindle into the part:

And keeping track of the tools and programs. I have to keep each tool on their own separate programs on this silly low-end controller, otherwise I can't get the tool setter to work. That said, a precision tool setter is an amazing thing to have. I just need to get on with migrating my machine to LinuxCNC so I don't have these controller limitations. Then I can also use my new, yet unboxed 3D probe and not have to use the manual wiggler tool. The 3D probe should be more precise and way faster to use.

The only challenge was that I was intent on having a fillet on the bottom edge of the PR bore, like this:

But I didn't have a bullnose end mill in the right size so I ended up doing it with a regular flat end mill with a lot of tiny step overs. Better than not having it at all I guess. I could also just have chamfered it but in theory, air supposedly prefer to flow over filleted corners rather than chamfers.

Finally, since I also superglued the part into the fixture this time I had to soften the glue in order to get the bulkhead out:

I did scratch the outside a bit as I pushed it out - but for a prototype that's alright. If I was making more of these, I guess a better fixture would be in order.

I was curious about the weight compared to a few others:

It came in at 25.5g (which is ~0.1g lighter than the CAD calculated, haha). The original Mirage bulkhead (not the shorter V1) is ~20g and the Predathor is ~33g (all with a few valve parts missing). So, actually my version is not that heavy. If I really wanted to I could still shave a few grams off of this design.

To finish this bulkhead, I need to machine the parts for the power regulator valve - it's already drawn up. Workholding might be a bit of a challenge for the valve cap as it's a funky shape and not very big. Oh, I also need to make the pumping barrel check valve. I haven't fully decided on a design for this one, yet - but since it just drops into a smooth, "oversized" bore I have flexibility on that. I am thinking to try out Zahar's o-ring idea for the pumping barrel vent valve but I will probably stick to a simple ball valve for the check valve - the mushroom type may come later.

So, a few more days of work on this bulkhead before I can test it in real life.

 

[Diving Gecko]

New Bulkhead In The Making - Part 4 - The PR Valve Cap
I did manage to machine the valve cap for the power regulator today - actually two of them.
The workholding turned out easier than feared. Only a few small issues. E.g. I would ideally have made this in brass but didn't have stock big enough. So, it's some old mystery alu - prolly 6061 from years back.

I held it in the vise between two V-blocks:

The shape, as mentioned is kinda "organic", almost bean shaped with convex and concave curves all over, but it was pretty easy to program in the end. First, just some roughing:

Also, as @tromic mentioned the other day, I should start using more shallow angles for the o-ring bore chamfers, but I only have 90d chamfer tools (45d chamfers) so this time, I machined a 20d chamfer with a bull nose end mill and it it seems to have turned out really smooth.

BTW, Parker's O-ring Handbook agrees with Tomi and propose chamfers as shallow as 10-20d for piston type applications.

I used the same type of tool path to make the outside curves smooth. Many, many tiny stepovers. In face, this op took +7mins, I think and the valve cap is just 5mm tall. But it did turn out quite smooth:

You can see the "bean shape" clearly here:

I parted it off on the lathe:

And then played around with it on the bulkhead:

It's supposed to be a tight fit between the inner wall of the reservoir and the shooting barrel - this will help keep it in place and not rotate:

As mentioned, I made two of them:

The first one above has an inner bore of 3.5mm, while the second is 4mm and it's also a bit taller. The idea with the smaller bore is that there will be less force on it, so it's easier to open. Remember, this is a ScubaPro/Salvimar type valve that operates opposite of the Mares/Cressi valves. You have to push forward against the pressure to open it.
The second valve got a bit messed up when I parted it off on the mini lathe - I really need to overall that machine and tighten up the gibs, etc.

And yes, there's no reason for this "fancy" shape other than aesthetics and wanting to learn to machine parts like these

Next up, how does the valve actually work...?

 

[Diving Gecko]

New Bulkhead In The Making - Part 5 - How Does The Valve Work?
I haven't seen a valve like this before - maybe it's out there as airguns have been around for ~70 years by now - but I hope it will work as I intend it to. That said, maybe I got something wrong, haha.

Let me try to explain how I hope it will work...

In Pos 1 for Loading I initially had a compression spring between the front "cap head" and the valve cap to force the valve cap closed against the lip. The plan was for the selector knob to be locked into the rearmost gate and then the spring would push the valve cap shut. But now I think it may not be needed, because I can use the pressure to my advantage here.
If the PR rod is 4mm in diameter, then at 30bar I'll have ~4kgf pushing the rod rearwards "out of the handle". So, instead of locking the knob into a gate, I will keep it in an open axial slot with extra space - so, it's kinda floating and can move rearwards as much as it wants until the valve head pulls the valve cap against the lip - which then seals it. Now, there will be 4kgf on that seal.

Up until this point the surface area of the valve cap itself doesn't matter - only the area of the rod. But as soon as I start pumping and building a pressure differential between the front side of the valve cap and the rear side, the cap's area now governs how much force is keeping the cap sealed against the lip. With one pump, the pressure differential will go from 0 to maybe around 7bar which results in a force on the cap of ~5kgf (I guess plus whatever pressure is now acting on the rod). After three pumps, the force could be 15kgf and after 5 pumps it's close to 20kgf (not 100% on the numbers, but they should be in the ball park).

This is how I am hoping the loading part will work - that even before pumping, the pressure in the gun will keep the valve shut tight. And as soon as I start pumping, the valve will be pushed even harder against the lip.

And then after loading, we push the knob forward to the Equalizing position and the air will run from the forward chamber through the internal bore in the rod to the rear chamber until there is no pressure differential. How long this will take, I don't know. Maybe I need a gate to hold the knob in this position or maybe I can just hold it open against the pressure of the gun. With e.g. 35bar in the forward chamber the force to open a 4mm rod would be 4.5kgf. Whereas if I go to a 3.5mm rod, it would drop to 3.4kgf.
Also, in theory this position could be used for a throttled shot but I think the fully closed Low Power shot is a better option.

Once the pressure has equalized, the main valve can be pushed forward to expose the full area of the bore ahead of a Full Power shot. The force to do this would just be against the pressure acting on the 4mm rod as the cap area is now out of the equation again (same pressure on both sides of the cap now). That is about 4kgf @ 30bar and it would need a gate to hold it in that position. So, I think I only need two gates for the knob, maybe only one if I don't mind holding the knob foward whilst the pressure equalizes.

The main advantage to this valve - if it works - is that the forceful slamming when opening the valve after pumping would be eliminated. And another real world benefit is that going from High Power to Low Power would be very easy as you just need to flick the power regulator out of the High Power gate and the valve will close automatically and fairly gently.

Let me know if I got all this wrong

 

[Zahar]

so

 

[Diving Gecko]

so

Hi Zahar,
I am not sure what you are asking... If you're asking about the hole diameter for the venting, it's 2mm.

Remember, this is still a Mirage system. Which means:

You move air from the rear chamber to the front with the use of the pumping barrel (which is not pictured in that sketch). This lowers the pressure in the shooting barrel - but of course increases it in the forward chamber. Once the "moving of air" is done, then you load the shaft into the shooting barrel with ease and then finally at the end you open the power regulator valve.

That's the constraint of this system, not looking to change that yet - The question is whether my valve will work in that system.

 

[tromic]

New Bulkhead In The Making - Part 5 - How Does The Valve Work?
I haven't seen a valve like this before - maybe it's out there as airguns have been around for ~70 years by now - but I hope it will work as I intend it to. That said, maybe I got something wrong, haha.

View attachment 60878

Let me try to explain how I hope it will work...

In Pos 1 for Loading I initially had a compression spring between the front "cap head" and the valve cap to force the valve cap closed against the lip. The plan was for the selector knob to be locked into the rearmost gate and then the spring would push the valve cap shut. But now I think it may not be needed, because I can use the pressure to my advantage here.
If the PR rod is 4mm in diameter, then at 30bar I'll have ~4kgf pushing the rod rearwards "out of the handle". So, instead of locking the knob into a gate, I will keep it in an open axial slot with extra space - so, it's kinda floating and can move rearwards as much as it wants until the valve head pulls the valve cap against the lip - which then seals it. Now, there will be 4kgf on that seal.

So, far the surface area of the valve cap itself doesn't matter - only the area of the rod. But as soon as I start pumping and building a pressure differential between the front side of the valve cap and the rear side, the cap's area now governs how much force is keeping the cap sealed against the lip. With one pump, the pressure differential will go from 0 to maybe around 7bar which results in a force on the cap of ~5kgf (I guess plus whatever pressure is now acting on the rod). After three pumps, the force could be 15kgf and after 5 pumps it's close to 20kgf (not 100% on the numbers, but they should be in the ball park).

This is how I am hoping the loading part will work - that even before pumping, the pressure in the gun will keep the valve shut tight. And as soon as I start pumping, the valve will be pushed even harder against the lip.

And then after loading, we push the knob forward to the Equalizing position and the air will run from the forward chamber through the internal bore in the rod to the rear chamber until there is no pressure differential. How long this will take, I don't know. Maybe I need a gate to hold the knob in this position or maybe I can just hold it open against the pressure of the gun. With e.g. 35bar forward chamber in the the force to open a 4mm rod would be 4.5kgf. Whereas if I go to a 3.5mm rod, it would drop to 3.4kgf.
Also, in theory this position could be used for a throttled shot but I think the Low Power shot is a better option.

Once the pressure has equalized, the main valve will open with just the force acting on the rod. So, it would need a gate to hold it in the open, Full Power position.

The main advantage to this valve - if it works - is that the forceful slamming when opening the valve after pumping would be eliminated. And another real world benefit is that going from High Power to Low Power would be very easy as you just need to flick the power regulator out of the High Power gate and the valve will close automatically and fairly gently.

Let me know if I got all this wrong

Very interesting design David! I believe it would work as you expect...

 

[Diving Gecko]

Very interesting design David! I believe it would work as you expect...

Happy to hear you think so and thanks

For quite a few years, I have been thinking about how to eliminate the slamming, have a valve that's easy to open, doesn't restrict the flow and doesn't take up too much space - but only now do I have the machines to make it happen

Before the pandemic, I had another two-stage valve drawn up that opened forwards like a regular Mares/Cressi valve but still with a smaller valve inside it like this one. But then I got stuck on this island and borrowed an old friend's Predathor and realized how that design has some features that might work well for this use case so I just merged my own idea with the old Scubapro design which I think is basically what the Predathor uses.

Another option would have been a ball-valve, like the "real" kind with a rotating, perforated ball. It would likely be the easiest to open and for sure have zero slamming but it takes up quite a lot of space relative to its bore size. But on a gun with more bulkhead space than a classic Italian gun, it could come into play.

 

[Diving Gecko]

New Bulkhead In The Making - Part 6 - Valve Gaskets
Today, I cut the gaskets for the valve cap on the CNC. I could have glue it to the valve cap and cut around it with a knife, but I wanted to try cutting it on the CNC and it went surprisingly well.
The gasket material is some old 1.3mm rubber that was in some of the boxes I recently had sent here from my storage unit in China. It's definitely not the best quality but I have some new stuff coming in that will hopefully be better.

I started by sticking it down to an alu blank with some double-sided tape:

And then cut it out on the CNC with a 1-flute end mill meant for plastics and some feeds & speeds ChatGPT proposed. I think the ideal cutter would be a down cutter as is often used for plywood as the the rubber was "flashing" a bit but from the top edge. Still, after a few tries, it cut pretty good as you can tell by the last one to the right:

Then I glued it onto the valve cap with some thickened, toughened superglue that's meant for rubber and has a little bit of flexibility:

Not sure how it will stand up to the oil, but if it fails, I think 3M5200 would be a good candidate.

This is how it looks on the bulkhead:

And then the productivity went downhill as I changed to the lathe to cut the power regulator rod with the internal venting bore. I messed up the first two tries despite just using easy-to-machine brass. I really need to take a time out and give the lathe some TLC - it's so "loose" it will just dig into the part or jump out of a cut. I know I can get it up and running well again - it's just tedious work and the lathe isn't exactly easy or to work on for these types of things.

Anyhow, I gave up on making a power regulator rod today but did cut one quick o-ring groove in a piece of scrap Ø4mm rod to get a feel for how well it would slide into the valve cap:

I am happy to say that both the regular NBR ring and the harder green o-ring slides in easily, yet still with what feels like good compression at the end. This is with 20d of chamfer angle so I will never use 45d again if I can avoid it, haha.
And that 20d chamfer was cut with a bull nose end mill, not even with a real chamfer tool. Pretty cool to know it turned out smooth enough - I will surely use that trick again.

Not sure when I will finish this valve. I mean, I really want to finish it ASAP, but I have another big project, one that can potentially make some money, that I really need to get stuck into. I tend to get carried away with whatever project catches my fancy, but I will have to prioritize the bigger project and then perhaps "schedule" one day a week to play with other stuff

 

[Zahar]

All rubber seals must be symmetrical and concentric with valve movement! Otherwise, the asymmetrical load of the valve body will cause the rubber thickness to change and its asymmetrical deformation! If there is even the slightest play, the valve will not seal! This valve can work 10 times and stop sealing! A very bad idea for sealing!

 

[tromic]

New Bulkhead In The Making - Part 6 - Valve Gaskets
Today, I cut the gaskets for the valve cap on the CNC. I could have glue it to the valve cap and cut around it with a knife, but I wanted to try cutting it on the CNC and it went surprisingly well.
The gasket material is some old 1.3mm rubber that was in some of the boxes I recently had sent here from my storage unit in China. It's definitely not the best quality but I have some new stuff coming in that will hopefully be better.

I started by sticking it down to an alu blank with some double-sided tape:
View attachment 60881

And then cut it out on the CNC with a 1-flute end mill meant for plastics and some feeds & speeds ChatGPT proposed. I think the ideal cutter would be a down cutter as is often used for plywood as the the rubber was "flashing" a bit but from the top edge. Still, after a few tries, it cut pretty good as you can tell by the last one to the right:
View attachment 60882

Then I glued it onto the valve cap with some thickened, toughened superglue that's meant for rubber and has a little bit of flexibility:
View attachment 60883
Not sure how it will stand up to the oil, but if it fails, I think 3M5200 would be a good candidate.

This is how it looks on the bulkhead:
View attachment 60884

And then the productivity went downhill as I changed to the lathe to cut the power regulator rod with the internal venting bore. I messed up the first two tries despite just using easy-to-machine brass. I really need to take a time out and give the lathe some TLC - it's so "loose" it will just dig into the part or jump out of a cut. I know I can get it up and running well again - it's just tedious work and the lathe isn't exactly easy or to work on for these types of things.

Anyhow, I gave up on making a power regulator rod today but did cut one quick o-ring groove in a piece of scrap Ø4mm rod to get a feel for how well it would slide into the valve cap:
View attachment 60885

I am happy to say that both the regular NBR ring and the harder green o-ring slides in easily, yet still with what feels like good compression at the end. This is with 20d of chamfer angle so I will never use 45d again if I can avoid it, haha.
And that 20d chamfer was cut with a bull nose end mill, not even with a real chamfer tool. Pretty cool to know it turned out smooth enough - I will surely use that trick again.

Not sure when I will finish this valve. I mean, I really want to finish it ASAP, but I have another big project, one that can potentially make some money, that I really need to get stuck into. I tend to get carried away with whatever project catches my fancy, but I will have to prioritize the bigger project and then perhaps "schedule" one day a week to play with other stuff

Yes, 20d is better choice than 30d... Iv read about it long time ago so I forgot the proposed optimal value..

 

[Diving Gecko]

Yes, 20d is better choice than 30d... Iv read about it long time ago so I forgot the proposed optimal value..

It does make sense that shallower is better. And in the case of the valve cap, I was really surprised how well it went in.

Here are two screen grabs from the "bible" on o-rings: "Parker O-ring Handbook":

 

[Diving Gecko]

All rubber seals must be symmetrical and concentric with valve movement! Otherwise, the asymmetrical load of the valve body will cause the rubber thickness to change and its asymmetrical deformation! If there is even the slightest play, the valve will not seal! This valve can work 10 times and stop sealing! A very bad idea for sealing!

While the cap is not symmetrical in shape, the center of effort is perfectly in line with the power regulator rod - meaning when it get's pushed open, the force from the rod is almost right on the center of effort of the pressure on the valve cap. And when the pressure pushes on it to close, it should also not be pushed out of whack.
The "target" on the pic below is the geometric center of the valve surface area which I assume is the same as the center of pressure/effort - and as you can tell it's almost centered with the bore for the power regulator rod.

I actually think that the banana-shaped ScubaPro/Salvimar valve does not have its load aligned with the rod and that design works well enough and have so for 40 years or more. What keeps that valve aligned and "landing on its seat" is that it rides against the inner wall of the reservoir so it can't rotate or get skewed.

However, my valve is much shorter - less elongated - than the ScubaPro/Predathor so the centering effect of it touching the reservoir is less pronounced. I wish I could have made it longer, but I can't. There's a pumping barrel in the way as seen in this pic:

So, the shape of the valve cap is a result of the following:
- A wish for the center of pressure/effort to be aligned with the power regulator rod
- The positions of the pumping barrel, the shooting barrel and the inner wall of the reservoir and finally:
- A machining constraint that governs how small a radius the corners of the valve bore - and thus the valve - can have

But in my effort to make a decently sized valve bore, I only have about 1mm from the outer edge of the gasket to the center of the lip, which does mean there's a risk that if it closes in just a slightly twisted/rotated position it will leak. But that's why I designed it to have only a 0.15mm gap to the inner wall of the reservoir and also to the pumping barrel and to the shooting barrel. Hopefully, the valve being butted up tight against the two barrels and the reservoir will keep it aligned.

Here's what I think/hope will happen. On the first use, if I make sure the valve is aligned properly, the gasket will take an imprint of the lip. A debossing which will then hopefully help align the valve every time it closes - the closing force is just 3-4kgf so with some oil in there, the valve may actually wiggle its way to the "home position" each time - even if it is ever so slightly rotated.

[Edit] I read your message again and I don't think you are talking about alignment of the rod or the valve twisting, but you're saying that one side will get pushed more than the other in an assymetric valve.
I guess I was thinking more in pressure and not force - so I guess you're right. But let's see how much of a real world issue it will become - one thing that might help me is that my valve cap and sealing lip are aluminium and wont flex, so it may help distribute the force enough that it will not be an issue. Time will tell.

Also, for my own handle which will come sometime later this year, I have more freedom when it comes to the shape of the valve as I decide the location of the power regulator rod - which means, I can make the valve less assymetrical.

 

[Zahar]

so

 

[Zahar]

so

 

[Diving Gecko]

so

Once again, I am not sure what you're saying?
These screenshots are from a source on how to design o-ring grooves. Granted from a commercial seller of o-rings but supposedly widely regarded as best practice.
The only point I was making was that a shallower angled chamfer, not surprisingly, really helps when inserting a piston type bushing into an bore.

I have used 45d previously as that's what my chamfer tools are. Tomi suggested something shallower so I looked it up.

 

[Zahar]

Once again, I am not sure what you're saying?
These screenshots are from a source on how to design o-ring grooves. Granted from a commercial seller of o-rings but supposedly widely regarded as best practice.
The only point I was making was that a shallower angled chamfer, not surprisingly, really helps when inserting a piston type bushing into an bore.

I have used 45d previously as that's what my chamfer tools are. Tomi suggested something shallower so I looked it up.

With a small angle d, there is a risk of the O-ring being blown off the groove or breaking! Tromik is Smart! But I rarely used valve guns! On valve guns, the O-ring was often blown off during the shot and I had to make special valves with mechanical retention of the O-ring or with a short chamfer at a large angle d!

 

[Diving Gecko]

With a small angle d, there is a risk of the O-ring being blown off the groove or breaking! Tromik is Smart! But I rarely used valve guns! On valve guns, the O-ring was often blown off during the shot and I had to make special valves with mechanical retention of the O-ring or with a short chamfer at a large angle d!

I don't think it will get blown off.

The chamfer is only for when closing the valve from open to closed position and there's no air rushing past in that movement. When the big rush of air happens the o-ring is not inside the bore at all. So, yes, it's out in the open exposed to the "storm" but at least it sits in a normal deep groove with just the tiniest possible edge break chamfer:

But yes, I also experienced an o-ring getting blown off on a power regulator - but that was when I used the original Mares bushing which is a bad design. It's an 8mm bushing (if I remember correctly) but the bore it sits in is 9mm, which means the walls of the groove were almost 0.5mm lower than they could and should have been. Not an issue on a normal Sten type gun but when I turned that handle into a Mirage the o-ring got blown off. I made my own bushing with a smaller gap between the bushing and the bore and the o-ring stayed put.

Actually, even on a Sten type gun, it's bad design to have such a big gap - and on top of that, there wasn't any free space in the groove for the o-ring to expand into when compressed...