Adaptors and fittings for full pressure flipper

Fellow roboteers,

After 5 years doing spinners and some wedges in Brazil, I decided, since moving to the UK, to do a pneumatic featherweight.

Almost all components are defined and some were already bought, but I am heaving trouble defining the adaptors and fittings. What do people normally use for full pressure? I thought about using hydraulic adaptors from Burnett & Hillman rated for 350bar on a Aluminium manifold.

The system will be a custom, 65mm bore ram, with integrated valve. So the fittings for the Buffer tank will be 5/8~3/4 BSPP and for the main cylinder, to avoid icing, I thought about G1/8 or 1/4.

What are the experienced builders opinions on these matters?

Also I would like to take this opportunity to thank Woody for his resourceful website and all the forum members for the shared knowledge.

Regards,
Eduardo

Yes, you really can only look at hydraulic fittings to be able to cope with the pressure. I also do use custom Aluminium manifolds, a good grade of ally will help with getting strong threads. Personally I would use 1/4 fittings in the rest of the system, don't really know why probably got more that 1/8 at home to save spending money. Either way you should get the flow to the buffer ok.

hi again :)

What sort of setup is it? Buffer and valve as part of the ram? If thats the case i dont think you will need big flow pipes, even on the vent side of the valve you could get away with a Q.E.V straight on the vent port then small pipe to your firing valve..

If i was going to make a FP featherweight setup id go for the same pipe used in paintball guns.. its like low pressure plasticc push in pipe but its rated to 3000psi.

Ive only ever done heavyweight setups so weight was never much of an issue.

Id suggest if possible you stick to BSP fittings over here rather then NPT as you'd use in the states. just for ease of getting components.

Stick up a sketch of the system your planning so its abit easyer to help.

...Is that a freakazoid costume your wearing in your DP? lol

Thanks for the replies, I thought about hydraulic adaptors due to the pressure, nice to know that I am on the right track.

Ed, I was thinking about using 7050-T7651 or 7075-T6 due to their high tensile strength, what do you use?

David, The problem to use plastic tubing is the layout. On a small manifold it will probably fit the bottle, SPRV, Solenoid Valve, the RAM and, for the RFL, a gauge (:angry:), also it will be shock mounted. I will probably use those plastic hoses from the buffer to the dump valve, just need to find a place to put it! LOL! The costume is a custom made Freakazoid! My brothers always said that Dexter resembled me, so when the opportunity came, Freakazoid was reborn. Probably in July I will compete in Brazil dressed as such, the organizers normally wear costumes.

Back to the robot, the RAM exhaust is will be handled by the main valve, it is a really unusual design, its triples as admission, exhaust and check, depending on the position. It is similar to a QEV spud valve, mixed with Trent Elis' design, although different enough to be original ;) Isn't it weird how ideas come to different people at the same time? I thought about a QEV integrated RAM just when I decided to move to the UK, but since my knowledge of high pressure pneumatics were really limited, it took a long time to perfect the idea.

Definitely all the fittings will be BSPP, just the paintball gauge that will be NPT - I think - but it will use a BSPP-NPT adaptor. The firing valve will probably be a FAS 04-211-101-40 12V 10W, triggered at 19.8~23.1V by a RC Switch. I still have to calculate the time needed to empty the main valve chamber through this firing valve and do some flow analysis for the whole system.

I found these CGA320 adaptors on eBay, does anyone knows were to buy something similar in the UK? Just in case I buy this, is it possible to chop the NPT thread and machine a new BSP G1/4 on it? I came across some CGA320 - BSP G1/2 adaptors from Air Liquide, proper to use with CO2, but they are not the swivel type, any suggestions?

Cheers,
Eduardo

is it possible to chop the NPT thread and machine a new BSP G1/4 on it?

Yes to both the 1/8 and 1/4 ....simply run a bsp die over the npt thread ...not a perfect solution but ...IMO ..a perfectly safe solution.

http://www.teamonslaught.fsnet.co.uk/threads.htm

Regards Woody

I mostly use Alumec 89 from Blackburns metals, simliar tensile and hardness properties to mild steel, very good stuff.
http://www.blackburnsmetals.com/
But the grades you suggested are more than good enough.

@Woody,

Thanks once again! Your site and spreadsheet are really usefull.

@geeza,

This Alumec 89 have really nice specs! Added to my list, thanks again!

@All,

Does someone know a good BSP G1/8 ball valve to use as a high pressure dump valve? I searched Google but the choices are bulky or expensive. I was looking for something like the one on the below picture:

http://www.fightingrobots.co.uk/foru...fce39162a5a4e5

Also, can anyone confirm if Wizard of NOS Self Release Pressure Valves (SPRV) complies with FRA current rules? I managed to get their support and receive three of these valves. :D

Cheers,
Eduardo

http://www.palmer-pursuit.com/cart/i...roducts_id=116

The relief valve looks like a good find ...but setting the pressure and certifying that you have set it correctly maybe a problem.

Cheers Woody

I use a G 1/4 ball valve as a dump on the heavy, got it from technobots years ago. It works although takes a while to dump all the buffer, would really like it to be a bit quicker. But don't see a problem with G1/8 for a feather.

Finding an 1/8 ball valve rated for 1000psi is ...I believe ..the problem.

Suppose never thought of that.

@Woody,

Thanks for the link, that is the only small manual valve I've seen to 1000PSI! I guess the robot will have one NPT more... I also found some Parker compression fitting valves rated to 1000PSI, namely the MB Series, although nowhere sells them.

The SPRV comes factory setted to 1000PSI (+-50), I think they can send a certificate stating that, will that be sufficient?

@geeza,

I found several valves G1/4, the problem is their size.

@all,

Does any one knows if there exist an 5/8-18 UNF to G1/2 BSP adaptor?

Since the arm and ram are bolted to the mechanism bars, does a 4 four bar mechanism need a tether?

Cheers,
Eduardo

I am really thinking about reverting all the threads of the project to NPTF. The CGA320 adapter is 1/4 NPTF, the Wizard of NOS SPRV is 1/8 NPTF, the gauge that I will have to use on RFL events is 1/8 NPTF and the dump valve looks like is going to be 1/8 or 1/4 NPTF also... :(

The Dump valve will probably be savaged from a Paintball refilling station that I bought last year, the valve is pretty small 51 x 28 x 48 and can handle 65BAR, one alternative is to use a NPTF 1/4 Male to G1/8 female adaptor from Burnet and Hillman to convert it.

The other dump valves options are:

-Parker MB or B PR Series (with compression fitting tubes):
http://www.fluidcontrols.co.uk/valves/ball-valves
-Aerohydraulics 1/8BSPP High Pressure (I don't have any information about size or weight, but there is G1/8 option):
http://www.aeroindustrialproducts.co..._p-99-438.html
-Legris 4730 - 2way Female BSPP Thread G1/8 100BAR - 4730 04 10 22 (A bit heavy... 350g):
http://legris-autoline.com/Legris_ec...hoisis=1_G1/8@
-Mars One Piece Reduced Bore Stainless steel Ball Valve G1/4 (Only 70g and as small as the aforementioned paintball refilling station):
http://www.valves-online.co.uk/acata...all-Valve.html

Today three CGA 320 to 1/4 NPTF adaptors, one anti-syphon tube (the main tank will be almost horizontal to keep height low) and two paintball on-off valves and the 1500 PSI dye gauge arrived. Just waiting for the two 16oz bottles from Pure Energy (75 x 205mm).

The paintball on-off valves from JustPaintBall.co.uk have a smaller bore than I thought, maybe this will be useful to prevent bottle freezing, although I will search for other alternatives, any suggestions?

Cheers,
Eduardo

Quote:

---

The SPRV comes factory setted to 1000PSI (+-50), I think they can send a certificate stating that, will that be sufficient?

---

IMO yes.

Quote:

---

Since the arm and ram are bolted to the mechanism bars, does a 4 four bar mechanism need a te

---

IMO no.

Quote:

---

Does any one knows if there exist an 5/8-18 UNF to G1/2 BSP adaptor?

---

Well you could consider drilling out the existing paintball valve to 8mm ... CGA 320 is very close to 1/2 bsp apart from the thread angle.

or
http://www.thinkauto.com/acatalog/On_li ... __162.html

MMS9-8-7
1/2BSP x 5/8UNF- steel
£2.57 (Excluding VAT at 17.5%)


ball valve ..

http://www.hydrasun.com/catalogue/page.aspx?i=9451

Woody,

After these tips, I will have to name the robot after you!!! Thanks a lot mate!

I will order the UNF to BSPP adaptor right a way!

Just sent and email to Wizard of NOS asking for a certificate of compliance stating the factory setted pressure.

Cheers,
Eduardo

Happy to help Eduardo.

Just noticed this.....

Quote:

---

Just waiting for the two 16oz bottles from Pure Energy (75 x 205mm).

---

Pure Energy is American and EU laws require the bottles to have a pi mark ....they may however comply...worth checking though.

http://www.luxfercylinders.com/products ... tcylinder/

The Paintball tank valves that I've taken apart have very small drilled passageways 1/64 as I recall and a convoluted route.

You might be able to purchase a MEGA / SUPER / HYPER flow version ....however the problem is quantifying the hype.

Cheers Woody

feel free to email me direct should you wish.

Woody,

I've bought one Pure Energy 16oz from Justpaintball.co.uk, a London based webstore, so I guess they are clear to be used here. If not I can try to return them and get my money back or store credit. They have exactly the size that I wanted for the project, although the project allows bottles up to 210 x ˜81mm, allowing Catalina and DCEC Cylinders as well.

For the Valve, I wanted 3~4mm bore, straight ball valve or even flux control valve, that way I could fill the buffer quick enough without freezing the bottle. One idea is to adapt the bottle from 5/8-18 UNF to G1/4 and put a regular G1/4 ball valve, but there is the burst disc issue, the adaptor would need to be custom made, pressure tested... How I miss my Lathe and Lab!

The on-off valve that I've got have a passageway around 1~1.5mm and clearly two 90° bends, living if with a pitiful Cv. They were advertised as increased flow, but the comparison is probably a pin valve... Which make the statement true.

I prefer talking on the forums, since more members can have the same questions as mine, making the answers even more useful! But I am co-writing the next version of RioBotz Combot Tutorial, both in English and Portuguese. On the previous versions I collaborated with two chapters and contributions on several others, now one of my goals is to improve the pneumatic and hydraulic chapters. Therefore any help will be greatly appreciated and credited! Your website will definitely be mentioned, as some of this forum topics! With this I hope the Americans (as a continent) start to build machines other than spinners and wedges, bringing more fun to the audience... Hehehe.

Thanks again!

Cheers,
Eduardo

Use a liquid pick up in the main supply bottle and expand to gas IN the buffer ... it'll only take a second or so to refill even with the poor cv.

I've thought of that, the problem is the integrated shuttle valve design. Maybe I can change the design to allow liquid to the buffer or do a by pass with a macro line so the liguid doesn't have to go through the main valve. Later I will post a drawing of the system.

Thanks again!

Cheers,
Eduardo

Energy is required to turn the liquid into gas.
Freezing or the transfer of heat from the surroundings ...including the co2 liquid itself ... happens at the point of expansion from liquid to gas.

So a word of warning on expanding liquid co2 to gas into your shuttle valve ...it'll freeze!

I have had success expanding the liquid to gas thru the buffer.

I look forward to a sketch of the system.

Woody

Quote:

---

So a word of warning on expanding liquid co2 to gas into your shuttle valve ...it'll freeze!

---

The design is based on this spud gun valve: http://potatogunwarrior.tripod.com/valve2.htm. So the CO2 is fed to the buffer through the shuttle valve, this way I managed to come with a really compact system, as you can see in the attached PDF from my early design made in PowerPoint. For better viewing just set the PDF reader to presentation mode.

In order to prevent liquid from the tank into the valve, I thought about using an anti-syphon tube. But, thinking again, this might not suffice, still having a lot of risk to freeze the valve.

To solve this problem I decided to run a macro line or other pressure resistant hose from the main tank manifold to the buffer. And place the Auxiliary Valve directly on the Ram. With this the dump valve will be placed on the main manifold and the CO2 will flow from it to the buffer manifold. Time to change the design a bit, back to Solidworks!

On the main valve, I am thinking... Since the hole system will be pressure tested, I might use the core from a Parker MB valve and machine a body for it with 5/8-18 UNF Male on one site and G1/2 or ASA (CGA-320) Male on the other. The burst disc would be placed before the valve itself, just after the 5/8-18 UNF thread.

Also, the Pure Energy bottles have the PI Mark, just checked on the bottle I received yesterday. Sadly they sent by mistake an 20oz instead of an 16oz. :(

Cheers,
Eduardo

Ah yes ...the spud gun valve concept

http://www.robowars.org/forum/viewtopic ... 7348d37ae5

Quote:

---

The trick is to get the Bl00dy things to work for your application!

---

Yes... That is always the issue: Getting the bloody thing to work! Hehehe

The redesign is almost thought through: need to shorten the main tank manifold, connect a flexible line from it to the buffer tank manifold, place the SPRV and Dye Gauge on a T and the dump valve directly on the main manifold. The auxiliary valve can go on top of the main manifold, connected directly to the shuttle valve back chamber by two 90° Elbows or along the main tank with only one 90° adaptor.

The major issue, apart from the obvious ones, is weight. Right now with all the components for the pneumatic system, drive train, electronics and battery, I still have near 6kg to build the outer structure, armour and wedge. I might save a bit more since some of the weights are a bit overrated in my spreadsheet (eg. manual valve 300g).

I've read that Robowars Australia topic mote than 100 times, it is a pity that Trent never completed that robot. My goal is to get is ready for the UK Featherweight Championship in September, but I still need to check the cash flow and finish the design before starting the construction per se.

Thanks again for your attention and awesome material!

Cheers,
Eduardo

Here's my solution to the std paintball valve.

A 16 oz paintball tank and a rethreaded valve from an Aqua medic , JBL or similar bottle.

http://www.reptilica.de/images/products ... e-500g.jpg

Woody,

That is a really good idea! I definitely will need to re-think de components layout, since it was supposed to use a straight valve. But it is doable ;)

I've searched the last couple of hours for those valves, but only managed to find bottles with valves at a really expensive price, more than £40 :( Do you know any where that sells these valves without the bottle?

On this search, I came across with one valve that might be the one: Co2 & Mixed gas valves. It looks heavy, but might be available threaded to 5/8-18 UNF.

I've also found other interesting valve: http://www.rap4.com/paintball/os/rap...ff-air-adapter. It isn't suited for the application, but using a drilled trough Pin Valve as an adaptor and burst disc support might work. Just an idea...

Since from the tank to the buffer the CO2 will be liquid, the paintball on/off valve might work. I will disassemble one of the valves that I bought to see if their bore can be enlarged a little.

On the other hand, some things are decided and already arrived, such as the Wizard of NOS SPRV, threaded to G1/4 BSP, and the 5/8-18 UNF to G1/2BSP.

Cheers,
Eduardo

John from roaming robots uses Jackson Keay as his supplier.

http://www.jk-cylinders.com/?content=supplies

might be worth asking them.

Quote:

---

Originally Posted by evristow

@Woody,


http://www.fightingrobots.co.uk/foru...fce39162a5a4e5

---

That setup looks familiar.

Quote:

---

File comment: Wazio's Highroller pneumatic setup.

---

Quote:

---

Originally Posted by woody

Quote:

---

File comment: Wazio's Highroller pneumatic setup.

---

---

That's why it looks so familiar.

LOL! I just harvested some pictures from the forum and saved them for future reference. That was a really neat setup!

Since the day I saw The Judge, Ziggy and Rocket (Wazio's 60) live at the Robogames, I have the desire to build a high pressure pneumatic system, last February seeing the UK heavyweights just confirmed it.

Back on the robot, I am really thinking about what woody said: I have had success expanding the liquid to gas thru the buffer. Since liquid CO2 is a lot more dense than the gaseous form, the small flow from the paintball on/off valve is not a issue any more, right? Is it worth to use a dip tube?

Cheers,
Eduardo

Quote:

---

Since liquid CO2 is a lot more dense than the gaseous form, the small flow from the paintball on/off valve is not a issue any more, right? Is it worth to use a dip tube?

---

I estimated that liquid gave a 10 fold improvement....I'd use a tube that dipped IN the liquid IF deciding to go with a liquid feed.

Since CO2 as a gas at 1atm (14.6959 PSI) and 20°C have a density of 1.860g/L, at 56atm (823 PSI) the density would be 104.198g/L. Liquid CO2, at the same 56atm and 20°C have a density of 773g/L, so the expansion ratio is 7.41:1, at 0°C the ratio goes up to 8.29:1 and at -30°C, to 10.82:1! Bullseye Woody!!! ;)

I noticed that the paintball valve have a approximately 1.2mm (0.047) orifice and two 90° internal bends, therefore it's CV is near to 0.056 gal/min @ 60F, to put in SI units, 0.00355 L/s @ 15.5°C, at lower temperatures, since turbulence tends to lower, the flow might increase slightly.

According to Catalina's Paintball tank info, an 16Oz tank have an internal volume of 0.669L, considering that there will be a pressure line and a manifold between the tank and the buffer, it is safe to assume that the total volume is going to be near 0.75L.

Hence, putting all together, at 0°C the liquid from the tank would take little more than 17s to fill a completely empty system to 56atm (823PSI). The volume of the ram is 0.223L, consequently the time to refill the CO2 used on the previous actuation would be 5.15s, a bit too slow for the purpose.

This number should be <3s, which tells me that using liquid, the system I have in mind needs a main valve with at least 0.00688 L/s, equal to a CV of 0.109 gal/min @ 60F, indicating a minimum orifice diameter of 1.8mm (near 0.07)... With this in mind the Paintball valve is out by now, at least until I can manage to disassemble one and check de possibility to enlarge the bore to 2mm, size which would give a CV of 0.161 and a refill time of 2.02s. The Aquamedic and similar valves became even more interesting! Tomorrow I will make some calls.

As the system is going to be liquid fed, I need to buy a propane torch and some cooper tubing to convert the anti-siphon tube, since nowhere sells a siphon or dip tube small enough to fit a 16oz bottle.

For now is back to the drawing board... err CAD :mrgreen:

Cheers,
Eduardo

Quote:

---

I noticed that the paintball valve have a approximately 1.2mm (0.047) orifice and two 90° internal bends, therefore it's CV is near to 0.056 gal/min @ 60F, to put in SI units, 0.00355 L/s @ 15.5°C, at lower temperatures, since turbulence tends to lower, the flow might increase slightly

---

Um ...yes and no.

The internet gives the definition of Cv as ...

Quote:

---

Cv - expresses the flow capacity in gallons per minute (GPM) of 60oF water with a pressure drop of 1 psi (lb/in2).

---

This is right ...but misleading.

Cv - follows the square root of the pressure drop in psi .... and the above quote is correct because the square root of one is ..yup one!

Q ( flow ) = Cv sqrt(dp/G)

Lower temp will reduce the bottle pressure and the pressure drop ...so the flow will decrease.



http://www.nadi.it/english/utilities_en.html

A quick calc using a CV value of 0.056 gives me an initial flow rate of 8.64 litres per sec @ 1 bar ...reducing steadily as the pressure differential decreases.. giving a refill time of 2.7 secs with a gas feed.


http://www.proautocon.com/tools.asp

download the paculator ...ignore the cubic meters value it's...as I recall .. wrongly converted by a factor of 10.

co2 gas density = 1.52

I've reread you post...

Quote:

---

Since CO2 as a gas at 1atm (14.6959 PSI) and 20°C have a density of 1.860g/L, at 56atm (823 PSI) the density would be 104.198g/L. Liquid CO2, at the same 56atm and 20°C have a density of 773g/L, so the expansion ratio is 7.41:1, at 0°C the ratio goes up to 8.29:1 and at -30°C, to 10.82:1! Bullseye Woody!!!

I noticed that the paintball valve have a approximately 1.2mm (0.047) orifice and two 90° internal bends, therefore it's CV is near to 0.056 gal/min @ 60F, to put in SI units, 0.00355 L/s @ 15.5°C, at lower temperatures, since turbulence tends to lower, the flow might increase slightly.

According to Catalina's Paintball tank info, an 16Oz tank have an internal volume of 0.669L, considering that there will be a pressure line and a manifold between the tank and the buffer, it is safe to assume that the total volume is going to be near 0.75L.

Hence, putting all together, at 0°C the liquid from the tank would take little more than 17s to fill a completely empty system to 56atm (823PSI). The volume of the ram is 0.223L, consequently the time to refill the CO2 used on the previous actuation would be 5.15s, a bit too slow for the purpose.

---

You appear to have forgotten in the above calcs ( I'm not atm checking for flow formulae errors ) to expand your liquid into gas....divide time by 10.

http://www.valvias.com/flow-equations-f ... -cv-kv.php

Try inputting the following values into the calculator Q = 0.06 l/s delta p= 300 psi S.G. =1.1
Cv output set in us gals and psi

I knocked up a spreadsheet using your buffer / ram volumes and a gas feed with a cv of 0.056 then calc the pressure recovery time for two scenario's.

The refilling to full pressure of the buffer and ram = 2.8 secs
The refilling to full pressure of just the buffer tank = 2.1 secs

here's an easier to use calc .. http://www.swagelok.com/tools/CV_Calculator.aspx

co2 gas = 1.53
co2 liquid = 1.1
P1= 870 PSIA
P2= 670 PSIA
Cheers Alan

Here's the excel spreadsheet with slightly modified parameters...hopefully it's correct ...it may however be worth what you paid for it! :lame:

[attachment=0:15kq0jsg]fill time.xls[/attachment:15kq0jsg]