Why does my valve seal down to 400psi on hpa, but leak at 750psi on Co2?

[Vdubber]

Hello all.

Been building a bulk fill AR2078 (sub 12) and have a bit of a head scratcher......

I'd done all component testing on air at 55bar (as I hadn't bought a Co2 tank yet) and had proven my valve (45deg peek seat, 45deg delrin poppet lapped, 5mm throat, 3mm stem reducing to 2.25mm through throat) sealed down to below 400psi (I was surprised too!). Running a homemade hybrid oilon hammer (similar to mds) with S/S core/ PEEK face, this valve was sealing perfectly and opening very easily, and producing 13.5fpe (.22 cal) with negative spring preload a 55bar, but......

Tested on Co2 yesterday and the valve leaks from full, slowly reducing overnight until it creates a seal by morning. As the pressure will probably still be close to 750psi, I'm assuming it must be something to do with this valve not liking the Co2 in liquid form, but......I'm all ears if anyone can offer some insight here please?

[Vdubber]

Just to note, once sealed over night I've then tried 'topping up' on Co2 but the valve immediately passes again. Tried this as wondered if the valve was freezing when initially trying to fill from empty (thought maybe that initially if Co2 was making it past the poppet, as it does with air until seated, that there may be ice/ moisture forming between the seat and poppet).

Also tried heating action on radiator to increase pressure, but this also made no difference.....

[JPSAXNC]

CO2 usually uses 90 durometer o-rings , softer o-rings are permeable to CO2. But I wouldn't think they would leak down over night. If the gun holds air it should hold CO2.

[Vdubber]

CO2 usually uses 90 durometer o-rings , softer o-rings are permeable to CO2. But I wouldn't think they would leak down over night. If the gun holds air it should hold CO2.

Thanks JP, Co2 is escaping through barrel so have assumed it's passing through valve of course (although there is a slight chance it could be the oring seals on outside of peek valve seat body, but they're nitrile 70)

[JPSAXNC]

I went back and looked at your pics, the 45 degree wall on your valve is very thin, it could be freezing and causing a leak. Peek isn't the best choice for a valve and poppet. Either the valve seat or the poppet should be a softer material to get the best seal. Also from the stand point of safety no polymer should be used as a valve, it's just not strong enough and could blow out at co2 pressure. Just saying.

[Vdubber]

I went back and looked at your pics, the 45 degree wall on your valve is very thin, it could be freezing and causing a leak. Peek isn't the best choice for a valve and poppet. Either the valve seat or the poppet should be a softer material to get the best seal. Also from the stand point of safety no polymer should be used as a valve, it's just not strong enough and could blow out at co2 pressure. Just saying.

Can understand why you might think this based on the above, because I haven't included all the details of the valve.

The peek body/ seat is pressed into and sealed against a brass outer body so cannot deform outwards or be be blown through-

I had to make this mod as the bulk fill tube came to me with a valve made by someone else that was no good (6.35mm throat and stem/ 8.5mm sealing surface). The same hammer/ spring on my valve barely opened this one so I had to bore it out and mod it).

PEEK is considerably harder than acetal, and whilst I appreciate it isn't conventional to use it as a seat (and may well be part of my issue) it isn't deforming due to a pressure of 55bar....but perhaps may not like the thermal dynamics of Co2 as you say. I only made it from PEEK as I had some in a suitable size, and for now until I've some solid info saying it will move more than a brass/ steel seat I'm reluctant to make a whole new seat and poppet......

Just to add the S/S body came with it and was another valve body (that was also flawed) so I bored it out too and made it into a check valve/ fill port.

[JPSAXNC]

OK, Good looking parts! If you can hear CO2 coming out of the barrel, the valve seat poppet interface must be leaking and if it only leaks on CO2 I would suspect that some part doesn't like the cold. The thing about CO2 is the tiniest leak will cause the cO2 to boil off even a leak you can't see or hear. Your best bet is to just disassemble the gun to just the tube and valves, and submerge them in warm water and watch for bubbles.

[Vdubber]

OK, Good looking parts! If you can hear CO2 coming out of the barrel, the valve seat poppet interface must be leaking and if it only leaks on CO2 I would suspect that some part doesn't like the cold. The thing about CO2 is the tiniest leak will cause the cO2 to boil off even a leak you can't see or hear. Your best bet is to just disassemble the gun to just the tube and valves, and submerge them in warm water and watch for bubbles.

Yeah the tube with valves is removable and it's defo the valve that's passing (ive just been careful with my wording of things....I haven't posted much on here and nobody likes someone coming on, asking for help, then acting like they know it all!)

My first thought was as yours.....there is a small leak as the valve tries to seat (same as it would on air), but it can't then seal because of it. I might add a spring on the Co2 side to try and prevent this when filling, currently just have a biro spring on hammer side.....

[mikeyb]

NOT an airgun application but I had a recent air sealing problem between different materials due to different coefficients of thermal expansion. I HAD considered thermal expansion but did not run the complete calculations over the FULL range of operating temperatures. Sure enough my material choices had HUGELY different thermal expansion coefficients and there was a big problem at the upper temperature operating point. Was able to modify/machine the existing parts to work over the full temperature range. Learning something NEW every day ;-)

I "think" your problem is related to temperature changes from the CO2 state change. There is a much smaller temperature change due to adiabatic cooling from compressed air.

[Vdubber]

NOT an airgun application but I had a recent air sealing problem between different materials due to different coefficients of thermal expansion. I HAD considered thermal expansion but did not run the complete calculations over the FULL range of operating temperatures. Sure enough my material choices had HUGELY different thermal expansion coefficients and there was a big problem at the upper temperature operating point. Was able to modify/machine the existing parts to work over the full temperature range. Learning something NEW every day ;-)

I "think" your problem is related to temperature changes from the CO2 state change. There is a much smaller temperature change due to adiabatic cooling from compressed air.

I think you must be right as the temp shift is the only difference to go on between Co2 and air. Both the peek and acetal 'should' be stable to -40deg c, but with matching 45deg faces it really wouldn't take much at all to effect the seal. Perhaps adding a spring and changing the poppet angle to a sharp 90deg edge should be the first attempt to resolve this (without having to add a whole load more hammer spring to crack it)......

[Vdubber]

To early to be 100% sure but think I may have got to the bottom of this, and it's so obvious I'm feeling a bit embarrassed!

Just pulled the valve and it was FILTHY. I didn't think to clean the chinese double valve fill station I'd bought prior to use (well, im assuming the dirt came from that and not the extinguisher) but either way.....I've now cleaned/ reassembled, and it's sealing.....

[x3baddad]

Don't forget co2 is a very dirty gas and it is 109 degrees below 0 as it comes off the liquid

[Motorhead]

Having played Paint Ball for many years using Co2, it is by it's composition a VERY DIRTY gas.

[Vdubber]

Thanks gents. Will keep an eye and adapt the valve as necessary, there is no way it'll stand up to continued dirt build up, fingers crossed a 90deg poppet face will handle it better.....

[Vdubber]

Put half a tin through it so far (whilst attempting to set this rifle up, not so easy going from 19deg C inside to 3deg C outside) and the valves holding up OK at the moment.

After a little port work (increased to 3.8mm), further reduction in hammer weight (now 30gr) and with a lighter spring, the RVA now works within a range where I need it at sub 12.

Shot a VERY slow string this morning....15 shots over 2 and a bit hours (in-between household tasks), as it was the only way I could test the output and get a reliable set point at between 17-19deg C. Working on the bases of 1FPE increase per 10deg C this isnt far off right for keeping things legal come summer in the UK....hopefully.

Limited data, but happy enough with the spead so far. I've nowhere inside to check accuracy/ shoot a full string to check shots per charge (and no point trying it outside in winter), so it'll have to wait until spring time now....

 

[Vdubber]

Just realised I haven't included a pic of the actual gun yet    so here it is. Next job is prettying her up.....should keep me busy while waiting for the weather to improve.....

[Rick67]

Just realised I haven't included a pic of the actual gun yet    so here it is. Next job is prettying her up.....should keep me busy while waiting for the weather to improve.....

That’s a handsome rifle 👍

[Vdubber]

That’s handsome rifle 👍
[/quote]

Thank you Rick, it has potential 👍

[Mzq284]

with matching 45deg faces it really wouldn't take much at all to effect the seal.

Vdubber,

When doing a valve job on an automotive engine, the valves are ground at 1° less angle than the seat. This makes it seal around the outside edge instead of trying to get a net fit clear around the valve. If the problem comes up again,  give it a try, it might be self cleaning

[Vdubber]

with matching 45deg faces it really wouldn't take much at all to effect the seal.

Vdubber,

When doing a valve job on an automotive engine, the valves are ground at 1° less angle than the seat. This makes it seal around the outside edge instead of trying to get a net fit clear around the valve. If the problem comes up again,  give it a try, it might be self cleaning

Thank you Peter ill keep that in mind 👍

When I cleaned the valve, amongst the black grease, there were a couple of 'strands' of  something which im pretty certain came from the fill station rather than Co2 itself. When I last pulled and stripped the valve it was still spotless, so hopefully it's already (to some degree) dealing with whatever dirt is traveling with the Co2 🤞