SA-6 Compressor Balanced Stage Pressures and Temperatures With Other Considerations

[iain/hsm]

@RayfromTX mine are actually having their fans removed. Too ineffective at that slow speed IMO. I am likely going to put a variable speed electric fan on the shroud to keep air flow maximized during slow operation. I do have thermal cutoff switches ready to put on each of the heads and set to 400f. Teflon starts degrading at 500f, so want to keep it significantly cooler than that.
Thankfully with big banks and CCR, I can let it run at 1cfm and just let it go.

At least removing the cooling fan will make it run quieter I guess LOL
Now from memory the SA-6 has used at least six (6) different cooling fan designs over the years that I can remember
And only four main types are available as stock fan designs that I can photograph as examples.

The current SA-6 for sports diving uses a moulded generic white six (6) bladed fan
The Mil-spec SA-6 uses a hand built black eight (8) bladed zero degree pitch
The SA-6 scientific Pure Air SA-6 uses as above but with a 30 degree pitch angle
For low RPM using a paddle shaped blade would be a consideration designed for generating fanned pressure
while the wing tip trailing edge reduces noise and generates a lower threshold of wake turbulance

In a nutshel for Mil-spec use its peddle to the metal and compete with an F-15 in rattling the roof while for recrational use its as quite as it needs to be to pass noise limits and not interfere with your Iphone ringtones.
I should look up the stall speed graphs for the mil-spec units as we built them if l get time

1. The fan inside the specially designed SA-6 shroud work together to incease air pressure from the fan blade tips over to the back of the shrould housing while the increased air pressure blowing around the cooling coils allow for a pretty decent amount of heat exchange between the two.
On cooling coil one off the 1st stage hot side this is to reduce the "Approach Temperature" into stage 2 inlet side.
But critically not so much to the point at which it would induce water vapour to condence out over the 2nd stage head. Then out of the 2nd stage hot side the cooling coil 2nd stage is when most of the water is "knocked out" of the gas stream and condenses out into the collection separator tower due to the sudden loss in temperature along the cooling coil and again another advantage of going oil free over oil lubricated is manifest as water cooler faster and more efficiantly than an oil film.

On the final stage this is to reduce the approach temperature into the filter (if fitted) and bring the temperatures down to within hand hold limits allow.

Now the big question here is:
1. With your required temperature limit (400F)
2 Where and how are you measuring this temperature.

Reason I ask is for consideration that what you are measuring may not be what you need to know
For example measuring the head temperature is not the same as measuring the gas temperature no way close.
And the position or point of measurement is critical for a accurate reading but more important to achieve a relivant reading.
Bauer for example to asses filer life measure the skin temperture of the cool filter shell externally with the B-Timer
Making you totally unaware of the real temperature that makes the real marked effect as to filter life expectancy. Measure the gas temperature not the filter shell.
Others measure the skin temperature of the compressor stage head again with no idea of the real gas temperature.
Others know the temp of a bolt screwed into the metal head. Why?
All are not accurate readings of the real need to measure actual gas temperature inside the gas stream at the point it comes off the cylinder head discharge valve side with a suitable pressure proof RTD
That is the point you dont want to see 350F IMHO

Frankly I think you were ill advised regarding this temperature business.
1. Measure the gas temperature not the head temps
2. OK to set at 350F to start off with but it is critical IMHO to monitor the compressor in real time and see what number you are getting. Then reduce the 350F setting down to say a few degrees above the maximum you are normally seeing. That way and I assume you have incorporated a high temperature cut out switch when the compresor cuts out high temp you can asses the reasons Hot day, extra high pressure, Dog sleeping in front of the fan shrold etc and make the judgement call to wake the dog.

Also get a good quality digital room temperature and humidity reading monitor as well I will try and look up some suitable models. Iain

 

[iain/hsm]

What I do to keep mine happy when filling my 4 cylinder bank to 4,500 is threefold

1 I run it in the air-conditioned indoors and set the thermostat to 70F.
2 I blow a small strong fan into the intake side of the compressor fan blade which helps keep the pesky 3rd stage substantially cooler. It is unfortunately hidden behind the block in a way that it gets the least airflow so increasing it helps.
3 I religiously lube the bearing surfaces as specified but more often than required.

Hi Ray

I will go off the record here and suggest with this 4000 to 4500psi range you guys consider at the next 3rd stage piston change out we discuss some options as we are so off topic for this OP post now.

Also with this "pesky" 3rd stage I'm guessing you dont want to remove it or lower you filling pressures but I would suggest considering a 2nd stage pressure gauge to monitor the interstage pressures in real time.
If the inlet pressure to the 3rd stage is reduced the compresion ratio increases and more heat is generated to compress to your required pressure.

If we look at Tbones fan removal proposal its acceptable only because of the exceptional low RPM and reduction in flow.

For the rest of us we can use calculation to calculate the temperatures achievable P1 V1 over T1
Assuming a 14.7psi Atmosheric pressure and your not half way up the Andes mountains
And we increase the required dischage pressure to 5000psi and to cover if I have understood KWS out of balance explanation correctly.

With the SA-6 running at 1500 RPM from ambiant air pressure to 5000 psi you need 4.87 BHP or 3.63 Kw
and achieve a 6.5 acfm flow rate. And would need the sunshine off Mercury to run solar cheaply.

Assuming a 70F ambient air temperature the (Calculated) gas temperature (before cooling) work out at
1st Stage 493.5 Deg F balanced nicely with:
2nd Stage at 497.03 Deg F but out of the KWS sweet spot balance at"
3rd Stage at 592.05 Deg F when your at 5000 psi

Now as far as I understand it this would be balance as far as temperature is concerned.
Now if you went to say 3000 psi discharge the 3rd stage calculated hot gas temperature before any cooling would
be "in "balance" with the other two stages.

Conversly if you wanted the sweet spot balance as far as compression ratio is concerned its very similar

And yet again if your sweet spot was based on the rod loading forces again the same principle and results
1st Stage load calculates out at 707.26 Lb balanced with the 2nd stage at:
2nd Stage load 770.55 lb with only the 3rd stage at 5000psi being out of balance/sweet spot at:
3rd Stage load at 1241.89 lb

And again if you lower the pressure to around 3000psi the sweet spots all equal out the same (near enough)
Incidently if you design for 5000 psi balance all engineering design need do is fit smaller diameter piston sets. And note the figures above are calculated at top dead centre TDC

For lower dead centre LDC the return stroke bottom of the piston its
1st Stage 0 Lb
2nd Stage 122.79
3rd Stage 156.04 lb

And this is to answer the effective back pressure setting the compressor creates itself when at the 5000 psi discharge pressure.

 

[tbone1004]

@iain/hsm temp switches are replacing the plugs on the 2nd and 3rd stage heads which will be measuring gas temperature inside of the head itself. Where Rix suggested to put them, but also the most convenient place to put them since the plugs are already there. The logic I was using with fixed temp switches was to have the temp cutoff at the point that the teflon is going to start degrading. Will likely leave that for now and when I start designing the full PLC, then go to thermocouples so I can watch them in real time

I have lab grade temp/humidity sensors that I can use.

You mentioned adjusting the actual balance of the crank for high pressure use. Do you have any documentation on that being done? Like I said above, this is likely going to spend 98% of its life pumping from 2500-4500psi, so if there is something I can do to make the compressor happier running at high pressure, I'm very curious

@RayfromTX perfectly fine with me to split out.

 

[RayfromTX]

Will you be cascading your banks? Why wouldn't that create lower average pumping pressures than you are discussing? (98% at above 2,500?)
@iain/hsm shall we split the thread?

 

[tbone1004]

@RayfromTX cascading yes, but since most of my diving these days is CCR diving, the volume of gas is limited.
My current plan is as follows
LP bank that the compressor won't touch

1x T- bottle "real o2"
1x T-bottle "argox"-95% O2 from PSA machine for open circuit use and nitrox blending
1x t-bottle Helium
1x t-bottle of 50%

2x 4500psi of 30/35 or 30/30. Likely 30/35 because it is 50%+helium, but can also go straight thru the rix
2x 4500psi of 10/70
2x 4500psi of EAN32

The HP bottles may change, that's still not quite set right now, it will be some combo of those three gases. Two banks is enough to top off the bottles after CCR diving. After filling, I have to decide whether I want to run the haskel to boost out of the banks, or run the rix to top the banks up. Will all depend on how much time I have to work with since my Haskel is not hooked up to a high volume compressor, but even then, I can only let it get so low before I have to top up to 4000-4500psi.
If I was doing a lot of OC fills where the bottles actually got down below 2000psi it may be a different story, but it will be rather rare that I'll be bulk filling

 

[iain/hsm]

@iain/hsm temp switches are replacing the plugs on the 2nd and 3rd stage heads which will be measuring gas temperature inside of the head itself. Where Rix suggested to put them, but also the most convenient place to put them since the plugs are already there. The logic I was using with fixed temp switches was to have the temp cutoff at the point that the teflon is going to start degrading. Will likely leave that for now and when I start designing the full PLC, then go to thermocouples so I can watch them in real time

I have lab grade temp/humidity sensors that I can use.

You mentioned adjusting the actual balance of the crank for high pressure use. Do you have any documentation on that being done? Like I said above, this is likely going to spend 98% of its life pumping from 2500-4500psi, so if there is something I can do to make the compressor happier running at high pressure, I'm very curious

@RayfromTX perfectly fine with me to split out.

The two plugs are standard 7/16 x 18 UN straight plug thread. Putting a temp switch in that position I cant see how your going to have enough clearance from the drive belt pulley belt guard. Further for the 1st stage you could also consider a tee off from the 1st stage relief valve its 1/4NPT taper and depending on your build number the NPT port for the relief valve just clears the belt guard. Whilst in older models its a side porting pointing back to the crankshaft in the 4 o clock position.

Now you could also alter (Drill a load of holes or remove the belt guard (but I could'nt possibly suggest that) and you would reduce the head temperatures a far amount. Also by removing the solid (albeit balanced) twin belt SPZ 180mm driven pulley and replacing it with an open spoke and hub off the shelf type pulley again you would reduce heat build up behind the flywheel pulley side. But you will also need to arrange a second counterweight if you replace the solid balanced pulley or you will cause considerable vibration.

Alternativley consider going direct drive and you have plenty of room at the driven end. We do this with a "desert chiller" for the IDF. Also with a crankshft extention and you could keep the pulley and use it for the flywheel effect

We use small 3 wire RTD's on a 10 foot stainless braided cable 1/8 NPT Platinum wire RTD's and 5000 psi rated stainless shell capable of inserting into a standard Swagelock or equivalent tube fitting, 1/4 pipe fitting or tube fitting with a female branch or female run tee

Iain/HSM also perfectly fine with me to split out.

 

[tbone1004]

@iain/hsm I may consider changing the 220v to direct drive, but can't direct drive the 110v portable one. It was a serious thought, but it would require gear reduction to get the RPM's right.
Good call on the clearance, may put a probe in the plug side and remote mount the switch.
I'm trying to leave the block itself relatively close to stock, just in case.

That all said, the reason I'm removing the stock fan is so I can replace it with an electric fan that will effectively run at the same 1500rpm speed whether the compressor is going 1rpm, or 400rpm

I tried reaching out to you about this specific issue directly, but haven't heard back via email in several months, but if you have any other suggestions on what to do to keep this thing happy at pressures over 3000, I'm all ears.

Oh, one of the other reasons for the VFD and going slow. Even if I put a full speed motor on there if I had the power available, I would have the VFD on there to slow it all the way down for nitrox blending. I have PSA machines that can supply 5lpm and it is ideal for me when blending to be able to use that direct into the stick vs. having to compress it first, then dump it back down to ambient. Having the VFD and being able to go slow allows me to get the nitrox mixes that I need much more easily.

 

[iain/hsm]

@iain/hsm I may consider changing the 220v to direct drive, but can't direct drive the 110v portable one. It was a serious thought, but it would require gear reduction to get the RPM's right.
Good call on the clearance, may put a probe in the plug side and remote mount the switch.
I'm trying to leave the block itself relatively close to stock, just in case.

That all said, the reason I'm removing the stock fan is so I can replace it with an electric fan that will effectively run at the same 1500rpm speed whether the compressor is going 1rpm, or 400rpm

I tried reaching out to you about this specific issue directly, but haven't heard back via email in several months, but if you have any other suggestions on what to do to keep this thing happy at pressures over 3000, I'm all ears.

Oh, one of the other reasons for the VFD and going slow. Even if I put a full speed motor on there if I had the power available, I would have the VFD on there to slow it all the way down for nitrox blending. I have PSA machines that can supply 5lpm and it is ideal for me when blending to be able to use that direct into the stick vs. having to compress it first, then dump it back down to ambient. Having the VFD and being able to go slow allows me to get the nitrox mixes that I need much more easily.

@iain/hsm I may consider changing the 220v to direct drive, but can't direct drive the 110v portable one. It was a serious thought, but it would require gear reduction to get the RPM's right.
Good call on the clearance, may put a probe in the plug side and remote mount the switch.
I'm trying to leave the block itself relatively close to stock, just in case.
.

Good point I also should have a go at calculating out the RPM values for you against power and torque down to say 1cfm flow I did the power requirements earlier just not the RPM values.
Again I need to take some photos of the RTD and positioning options we have for the SA-6
Its not above possibility to add additional side ports to the heads for RTD mounting
or even a small blind screw thread for a bolt on RTD mount. Easy done by yourself even if not specifically accurate of gas temperature at least we could compare position variance readings

That all said, the reason I'm removing the stock fan is so I can replace it with an electric fan that will effectively run at the same 1500rpm speed whether the compressor is going 1rpm, or 400rpm .

If you take a careful look at the ABS Fan Guard and shroud there is a front "grill" made either to look like a plastic grey version of expanded metal on the older models or on the later units its a heavy weight black plastic mesh I will take photos later at works. Both are secured using glue compond from a sealant gun. We used to weld finish using a Fungi tool. Point is it's easy to bash out and retain for later replacement and the 305mm ABS rim at the front end may have just enough room and material strength to mount a big 300mm Papst type ventilator fan blower. Now I need to check the dims and clearences but the fan model I would consider to kick off would be the EBM-Papst W2E300 Series Axial Fan using a 350 watt (in my case 230Volt 50Hz AC pancake motor) blowing 3740 m3/hr I will report back.

I really like the idea of full data logging all stages both temperature and pressure on the SA-6 especially if you guys go down the nitrox route. The more I think of this the more I like it.
Further if we all agree the most practical cost effective method and position for the measurement this could be incorporated into future European new builds while data logging each at the same point gives us all a valid comparable point of reference.

We will agree to build a new test bed unit add the full data logging of both temperature and pressure and produce a technical file publishing the technical results of any proposed change or adaption or even suggestion in order to nail this question of performance criteria for the SA-6 scuba builds out in the wild and in order to give you guys some headers as to the improvement (or not) in some or both our and your suggestions. Just save me from the how much nitrox can I pump before I set the house on fire sort of question.

I tried reaching out to you about this specific issue directly, but haven't heard back via email in several months, but if you have any other suggestions on what to do to keep this thing happy at pressures over 3000, I'm all ears.
.

I’m in the engineering side of designing and building diving stuff we are kept mostly far away from you lot and in a dark corner, mostly away from the public. As for getting a prompt email reply from any engineering lot………No chance LOL we are always way too busy and always so far behind I think I get to about 200 emails before I begin to make a start answering them. Welcome to my world of pragmatic compromise on a shoe string budget.
I have a number of suggestions about how to compress up to the 5000psi mark with ease but its a long list of options. Give me a little time to address this point more fully.

The one position I should make clear is that all my posts here are mine in a personal sense and do not reflect or represent any company held view or position. That way I can say what needs to be said pragmatically to offer practical help while not offending corporate or management lawyers.

Note I refer to these compressor as a generic SA-6 and add no other designate manufacturing identity unless it is a specifically identified or described 3rd party vendor component.

I think if we could also consider removing the manufacture identity name from the heading and revise the title to say SA-6 compressor balance stage pressure etc etc you may find a lot more information available to an open forum discussion platform.

Oh, one of the other reasons for the VFD and going slow. Even if I put a full speed motor on there if I had the power available, I would have the VFD on there to slow it all the way down for nitrox blending. I have PSA machines that can supply 5lpm and it is ideal for me when blending to be able to use that direct into the stick vs. having to compress it first, then dump it back down to ambient. Having the VFD and being able to go slow allows me to get the nitrox mixes that I need much more easily.

Give me some more time to clear the decks so to speak i have a couple of modification you could do yourself to bring down the interstage temperatures dramatically on the SA-6 Iain

 

[tbone1004]

@iain/hsm 230vac is a non-issue. The VFD used is an inverter VFD so 110v in, 230v out. It was a requirement when I built this to have everything running on 230 in the event that I want to upsize the motor or VFD and wouldn't have to touch anything else

 

[iain/hsm]

@iain/hsm 230vac is a non-issue. The VFD used is an inverter VFD so 110v in, 230v out. It was a requirement when I built this to have everything running on 230 in the event that I want to upsize the motor or VFD and wouldn't have to touch anything else

You could say that...... I couldn't possibly comment LOL

But we do need to fully explain where you lot are with this 230 volt and where the rest of the civilized world is trying to work out our differences. As a Brit our lot get 230 volts at 50Hz straight into the house. We call this a singe phase domestic supply. Now by contrast you lot as I understand things get your domestic power as a split phase at 60Hz. That is each house is supplied with 240 volts 60Hz. The 240 voltage is then split in each house at the circuit breaker into the "two legs" each giving 120 volts 60Hz. And you call this your 110 volt supply. Clear as mud so far.
So as far as this independant and separate cooling fan is concerned why not you lot go for a 110 volt 60 Hz version and us lot we will go for the 230 volt 50hz version. I just dont see the need to use a VFD in with the separate cooling fan. Power to the low RPM motor agreed yes, I just don't see with a separate fan the need to hook it up with the VFD unless you want to hook in an overrun timer I guess and for that you could simply use a relay timer control. Iain