Printable View
Thats good enough for me, I need to goto bed now. Have to work out a better way for separation. What about larger diameter pipes on the condensor exit? Redesign the phase separator? I will let you think about that ones. One more thing, I need the machine to run my pc. If i let some charge out so the compressor is warm, then we will be ok. yeah?
Is this your phase seperator?
http://www.russellpetrie.plus.com/im...ade/autoc4.jpg
I would think the closer the r290 exit tube is to the bottom the better chance the r23 gas in there has a chance to escape.
Regards
John
The R290 exit is about 1 inch from the bottom. So it is quite close. I imagined the separator to be flooded so the R290 level was above the exit pipe, but this does not appear to be the case. There might be too much turbulance caused by the stream entering the phase separator that the liquid cant settle and as a result the R23 and R290 are getting sucked out the bottom.
John, The following links should enable us to build a mathetical model of this system. I get as far as figuering the enthalpy and mass ratios but the capillary tube calculations are a little beyond where I am comfortable yet. Wanna give it a shot...
Use the charts log(p)-h and select the refrigerants from these you can calculate the enthalpy (kJ/kg) at the pressures in this system you will also need the mass for each refrigerants. R23 is 70.01 and r290 is 44.10. Article on capillary flow in a tube...
http://www.fys.uio.no/~eaker/thesis/node11.html
http://www.blairwing.com/images/Refrig_Data_Sum.pdf
http://www.et.dtu.dk/Coolpack/Files/RefUtil.exe
On the seperator we need to calculate the flow rates and turbulance of the refrigerants to get proper seperation, we may need to cool the seperator down to the condensing point of the r290 at that pressure. This alone may solve the liquid r290 issue..
YesQuote:
If i let some charge out so the compressor is warm, then we will be ok. yeah?
The condensing point of R290 is 62ºC at 22bara (this is the pressure at the top of the separator) the separator is around ambient temperature and no where near 62ºC. Unfortunatly this does not fix the separator problem. In the separator the inlet stream sprays downwards, you cant see this in any picture so I though I would point it out.
I cant see from that pic where the r23/r290 is entering the phase seperator. this could be a combination of r23 exiting at the r290 exit point and liquid r290 being sucked through the r23 exit line.
Is there anyway of knowing how far up the phase seperator the liquid r290 is?
Chilly1, 1am here in the UK so I will look at those figures tomorrow, getting very sleepy, but learning a great deal, thanks guys.
Regards
John.
maybe u could fill the phase seperator w/ steel wool to force the liquid to the bottom
The C/S idea is looking better all the time. Massive drop in velocity, with no problem settling and separating.
With the conventional condenser, we could very well have a buildup of R23 in the condenser, having difficulty pushing its way out the bottom along with the liquid R290. And if we don't have a liquid seal in the bottom of the condenser, then we may need to do some more condensing in the separator.
c/s?
Good God I should be in bed now:D
Regards
John.
That could add to the problem. It should spray straight to the opposite wall.Quote:
In the separator the inlet stream sprays downwards, you cant see this in any picture so I though I would point it out.
Parallel condenser/separator. The big thing on the right:Quote:
c/s?
http://www.blairwing.com/images/Autocascade14.jpg
Gary I think the addition of the subcooler and a mid condenser tap for the r23 with both the discharge of the condenser and the mid tap into a tube in a shell with the shell so we would have is a desuperheating phase seperator with two inlets and two outlets.
Never would have belived that refrigeration would lead to soil physics. Check out Darcy's equation ... Trying to find some useful description of the flow of refrigerants in a capillary tube.. For something so simple it gets very complicated,
Could you go clockwise from the compressor and explain the components starting with the orange rectangle. I havnt learned so much in such a short amount of time since I was at Uni:cool:
Regards
John.
High Pressure 19.8 bara 1980000 Pa
Low Pressure 2.8 bara 280000 Pa
Viscosity 0.000125 kg/ms
Radius 0.0155 inch 0.0003937 meters
Density 1.9558 dm3/kg 511.2997239 kg/m3
Length 48 inches 1.2192 meters
Velocity 216.1248958 m/s
Volumetric flow 0.534626006 m3/s
Mass Flow 273.354129 kg/s
I punched these numbers for propane into Darcy's Equation. Maybe I missed something obvious, but look at the mass flow of propane? That cant be right.
CompressorClock wise desuperheater,oilseperator,condenser,capillary tube,interstage heatexchanger,capillary tube,evaporator (the brass colored thing) Pressure expansion Valve, Suction line subcooler,compressor
The velocity is also greater than the sound velocity data for propane. Now if I remember my thermodynamics/fluid flow, then we can't have a fluid travelling faster than sound in an adiabatic process (in this case its the pressure drop across the capillary) pipe. Am I right?
I see your point. If we install a 1/4 line from a middle return bend in the condenser to the middle of the separator, this would provide another path to prevent buildup of R23 in the condenser.Quote:
Gary I think the addition of the subcooler and a mid condenser tap for the r23 with both the discharge of the condenser and the mid tap into a tube in a shell with the shell so we would have is a desuperheating phase seperator with two inlets and two outlets.
The middle tap would tend to favor R23 vapor, while the bottom connection would tend to favor R290 liquid.
I'll annotate it if that would help. I threw in the PEV so we could raise the pressure in the R23 evaporator while not effecting the suction pressure effectively changing the balance, Basically if I were to factor in the enthalpies and densities of both refrigerants and used this for the only midigating factor in the system balance and based it on current pressures I would need a delta across the epr of .8 bar. But I think there is more to this than only the heatcarrying capacity and the densities. I think the viscosity and misicibility of the refrigerants and the oil's density and flow rates also need to be considered, published research in this field is scarse and what there is is modeled on other fields of study. I am still looking .....
Exactly then Russell_hq would need only modify his current condenser by putting in a tap a ways down from the top of the condenser this would still allow a liquid seal in the condenser and would stop the liquid from being blown out. I am looking for that condenser. Theres a coil manafacturer in texas that may have it.Quote:
Originally posted by Gary Lloyd
I see your point. If we install a 1/4 line from a middle return bend in the condenser to the middle of the separator, this would provide another path to prevent buildup of R23 in the condenser.
The middle tap would tend to favor R23 vapor, while the bottom connection would tend to favor R290 liquid.
http://www.blairwing.com/images/Autocascade16.jpg
hey guys, why dont you stop talking about how easy it is to build and build one. You can do the calculations while your sleeping, and then build it in the living room so you can watch dawsons creek, while you build. Or maybe you can close your eyes while you build it, so its a bit more of a challenge.
I just couldnt help it:D
Also, if necessary, we could soft solder the common suction line along the side of the separator.
I have all the pieces to build one but why build when we have a guinie pig to watch...LOL!Quote:
Originally posted by captaincascade
hey guys, why dont you stop talking about how easy it is to build and build one. You can do the calculations while your sleeping, and then build it in the living room so you can watch dawsons creek, while you build. Or maybe you can close your eyes while you build it, so its a bit more of a challenge.
I just couldnt help it:D
I couldn't resist either..http://www.blairwing.com/images/PhaseFive.jpg
Future project.....