|
|
Post by birkie on May 24, 2017 13:13:49 GMT
I was thinking of mixing R123ZD and R-1234yf. Ah, I see. Well, apparently CoolProp blows up with those mixtures. I doesn't look like it has interaction data between those two, and whenever I get near the saturation points, it starts throwing a lot of errors. Oh well, I guess only the real world can tell us what might happen! |
|
|
|
Post by ckfan on May 24, 2017 14:46:32 GMT
Why can't SO2 smell like cinnamon rolls? Or at least not smell like Satan's breath.
|
|
|
|
Post by birkie on May 29, 2017 1:14:02 GMT
Something really interesting happened when I was calculating the characteristics of refrigerants in a CA. When I tried R123 in a CA, I got a halfway decent match on capacity (+21%) and head pressure (+10%) compared to methyl formate. Mass flow was WAY off, it was over 3.3 times more then methyl formate! There's an old thread about trying R123 in a CA. It had a worn-out compressor, but performed well, all things considered. However, a key observation to me was that it wasn't much louder than a normal CA on methyl formate. Now, there are a lot of variables at play. The rotary compressor on a CA is nothing like a CK or DR, volumetric flow is much greater, pressures are much lower, etc. However, the fact that a 3.3x increase in mass flow (or density, depending on how you look at it) didn't appreciable increase the noisiness of that CA raised an eyebrow. Maybe the theory of sound level being proportional to density/mass flow is wrong, or incomplete? To me, there's basis to question it. So.. I looked at the data. Apparently, the molar flow between methyl formate and R123 are much closer. This compares the flow in terms of the number of molecules, rather than mass. R123 has 30% more molar flow than methyl formate, which is a much closer match. Could molar flow have something to do with it? Looking at the SO2 replacement candidates from that lens, I get the following: Relative molar flow (unrestricted): SO2: 1 R134a: 1.94 R152a: 1.74 R227: 1.26 R1234ze: 1.40 R12: 2.11 R1234yf: 2.17 A few things are notable. R12 is still a huge departure from SO2. R152a is still a closer match than R134a, but not by a mile this time. The standout is R227. It’s only 26% off. This contrasts greatly with its mass flow/density, which is 3x SO2 (302%). Hmmm. Doesn’t this seem to sound like an apt comparison to R123 in the CA now? Oddly enough, its capacity is also 20% greater than SO2! So the differences between SO2 and R227ea are freakishly analogous to the differences between methyl formate and R123. Remember, the CA on R123 wasn’t particularly noisy, and it ran well. So. I think our theory on what causes compressor noise in the DRs and CKs needs more data points. I would propose that running R227ea in a CK or DR would be a great experiment. Dropped into an un-modified CK or DR, is it noisy as heck, or tame and civilized? Also as interesting, comparing relative molar flow when restricting for capacity: SO2: 1 R134a: 1.12 R152a: 1.13 R227: 1.07 R1234ze: 1.09 R12: 1.24 R1234yf: 1.22 This would put R134a on par with R152a with a restrictor; both under 15% away from SO2. R227ea is still the closest match with only a 7% difference. HFO1234ze trails slightly at +9%. So.. Is it deciding factor for compressor noise molar flow, mass flow/density, or something else completely? Only experimentation will tell! Experience with R227ea (or maybe even R1234ze) could break the code! |
|
|
|
Post by vintageguy on May 29, 2017 8:07:46 GMT
Why can't SO2 smell like cinnamon rolls? Or at least not smell like Satan's breath. Because cinnamon dioxide already does. |
|
|
|
Post by ckfan on May 29, 2017 13:18:49 GMT
Birkie, you truly are doing the lords work. This is very good stuff. If only Gill could get his hands on the 227...
|
|
|
|
Post by ChrisJ on May 29, 2017 13:34:09 GMT
Something really interesting happened when I was calculating the characteristics of refrigerants in a CA. When I tried R123 in a CA, I got a halfway decent match on capacity (+21%) and head pressure (+10%) compared to methyl formate. Mass flow was WAY off, it was over 3.3 times more then methyl formate! There's an old thread about trying R123 in a CA. It had a worn-out compressor, but performed well, all things considered. However, a key observation to me was that it wasn't much louder than a normal CA on methyl formate. Now, there are a lot of variables at play. The rotary compressor on a CA is nothing like a CK or DR, volumetric flow is much greater, pressures are much lower, etc. However, the fact that a 3.3x increase in mass flow (or density, depending on how you look at it) didn't appreciable increase the noisiness of that CA raised an eyebrow. Maybe the theory of sound level being proportional to density/mass flow is wrong, or incomplete? To me, there's basis to question it. So.. I looked at the data. Apparently, the molar flow between methyl formate and R123 are much closer. This compares the flow in terms of the number of molecules, rather than mass. R123 has 30% more molar flow than methyl formate, which is a much closer match. Could molar flow have something to do with it? Looking at the SO2 replacement candidates from that lens, I get the following: Relative molar flow (unrestricted): SO2: 1 R134a: 1.94 R152a: 1.74 R227: 1.26 R1234ze: 1.40 R12: 2.11 R1234yf: 2.17 A few things are notable. R12 is still a huge departure from SO2. R152a is still a closer match than R134a, but not by a mile this time. The standout is R227. It’s only 26% off. This contrasts greatly with its mass flow/density, which is 3x SO2 (302%). Hmmm. Doesn’t this seem to sound like an apt comparison to R123 in the CA now? Oddly enough, its capacity is also 20% greater than SO2! So the differences between SO2 and R227ea are freakishly analogous to the differences between methyl formate and R123. Remember, the CA on R123 wasn’t particularly noisy, and it ran well. So. I think our theory on what causes compressor noise in the DRs and CKs needs more data points. I would propose that running R227ea in a CK or DR would be a great experiment. Dropped into an un-modified CK or DR, is it noisy as heck, or tame and civilized? Also as interesting, comparing relative molar flow when restricting for capacity: SO2: 1 R134a: 1.12 R152a: 1.13 R227: 1.07 R1234ze: 1.09 R12: 1.24 R1234yf: 1.22 This would put R134a on par with R152a with a restrictor; both under 15% away from SO2. R227ea is still the closest match with only a 7% difference. HFO1234ze trails slightly at +9%. So.. Is it deciding factor for compressor noise molar flow, mass flow/density, or something else completely? Only experimentation will tell! Experience with R227ea (or maybe even R1234ze) could break the code! I suspect this isn't an easy question to answer but. Does this suggest methyl formate is actually a fairly good refrigerant vs even modern equivalents? Where's SO2 apparently, doesn't work as good as replacements? |
|
|
|
Post by coldspaces on May 29, 2017 15:09:28 GMT
Birkie, you truly are doing the lords work. This is very good stuff. If only Gill could get his hands on the 227... Refrigerant grade 227ea is over $1,200 wholesale for 30 lbs, and it has a high GWP, so it most likely never will be readily available here. This is truly great work Birkie! |
|
|
|
Post by birkie on May 30, 2017 0:38:42 GMT
I suspect this isn't an easy question to answer but. Does this suggest methyl formate is actually a fairly good refrigerant vs even modern equivalents? Where's SO2 apparently, doesn't work as good as replacements? "Good" is complicated  It looks like most modern flurocarbon replacements I've analyzed at (except for the hydrocarbons) have higher capacity, higher pressures, and higher mass flow than either methyl formate or SO2. So the comparisons are very similar. |
|
|
|
Post by birkie on May 30, 2017 0:57:53 GMT
Refrigerant grade 227ea is over $1,200 wholesale for 30 lbs, and it has a high GWP, so it most likely never will be readily available here. This is truly great work Birkie! I'm going to ask some of my contacts in the chemistry dept to see if I can get my hands on smaller quantities, but yeah it looks hard to get a hold of at a price reasonable for an experiment like this. It's a shame, as it would be a scientifically interesting comparison that's easy to do by ear. R1234ze is also an interesting comparison along those lines, but more subtle. If it's mass flow, noise would fall between R1234a and R152a. If it's molar flow, it'd be fairly quieter than either of them. |
|
|
|
Post by birkie on Jun 1, 2017 21:02:26 GMT
Just an update on the search for R227ea
- I'm waiting for a few calls back from a few sales representatives claim it might be possible to obtain some reclaimed R227ea at $4-$5/lb. The problem is that it'll likely contain on the order of 1% residual nitrogen.
- It looks like old filled tanks are available on ebay, for about the same price. The problem is that they are certainly charged with lots of nitrogen as a propellent.
- There are reputable manufacturers in China willing to sell 30lb tanks of R227EA for arounff $4-5/lb. This is virgin stock, with < 10ppm moisture or acid.
Unless I hear back really good news from the US Halon recyclers/distributors, I'm actually thinking it might be worthwhile to buy a 30lb tank from China. It's only one experiment. What's the worst that could happen.....?
|
|
|
|
Post by ChrisJ on Jun 2, 2017 1:30:28 GMT
I suspect this isn't an easy question to answer but. Does this suggest methyl formate is actually a fairly good refrigerant vs even modern equivalents? Where's SO2 apparently, doesn't work as good as replacements? "Good" is complicated It looks like most modern flurocarbon replacements I've analyzed at (except for the hydrocarbons) have higher capacity, higher pressures, and higher mass flow than either methyl formate or SO2. So the comparisons are very similar. Stupid question but, what's mass flow? It seemed like the latent heat of vaporization was extremely high with methyl formate. I thought this suggested it was able to transfer a huge amount of heat for the amount used? I'm asking, because honestly I have no clue. You're talking so far over my head I'm about to go get an extension ladder and try to catch up. |
|
|
|
Post by birkie on Jun 2, 2017 4:15:18 GMT
Stupid question but, what's mass flow? It seemed like the latent heat of vaporization was extremely high with methyl formate. I thought this suggested it was able to transfer a huge amount of heat for the amount used? Ah. Mass flow is the amount of refrigerant (as measured by weight/mass) that flows through the refrigerant circuit in a given amount of time. The head of vaporization is usually measured as a energy per unit mass. Let's say a particular refrigerant has a heat of vaporization of 100BTU/lb. If the mass flow rate of this particular refrigerant in a system is 10 lbs/hr, then the capacity of the system is 1000 BTU/hr. Compressors pump a fixed *volume* of refrigerant in a given amount of time. (e.g 10 cubic feet per hour) This primarily determined by their physical characteristics (e.g. bore and stroke for reciprocating compressors), and rotation speed. At a given temperature, each refrigerant gas has a particular density. Let's say that at 15F, a hypothetical refrigerant weighs 1lb per cubic foot. If we know our compressor pumps 10 cubic feet per hour, and our refrigerant is 1 lb per cubic foot, then the mass flow is 10lbs/hr . Knowing its heat of vaporization (100 BTU/lb) allows us to calculate the refrigerating capacity under these conditions; 1000 BTU/hr. Let's say we drop a different refrigerant into our system; one that has a lower heat of vaporization (30 BTU/lb), but a higher density (5 lb/ft3). The amount of volume pumped through the compressor is the same (10 ft3/hr), but the higher density means the mass flow is higher (50 lbs/hr), resulting in a capacity of 1500 BTU/hr. Clearly, the compressor is doing a lot more work with this refrigerant. So methyl formate has a very high heat of vaporization (as you noted), but it also has a very low density, at least compared to "modern" flurocarbons. So for a given volumetric "gulp" when the compressor rotates one revolution, very little methyl formate (by mass) is drawn in, compared to other refrigerants. Even though methyl formate is more "potent" in terms of its high heat of vaporization per unit mass, its density is so low that dropping almost any other comparable "modern" (flurocarbon) refrigerant will result in higher capacity; more mass flows, more power drawn by the compressor, and more refrigerating effect. This can be bad if your're trying to search for a "drop in" replacement. The same is true for SO2. It's potent in terms of heat of vaporization, but very light and low density compared to its alternatives |
|
|
|
Post by ckfan on Jun 2, 2017 14:09:54 GMT
Very interesting Birkie! I wonder if this is why the early refrigerants like SO2 and Methyl Formate were low mass flow (low pressure?) but had a high heat of vaporization to keep the efficiency up? This stuff is over my head too but is very fascinating.
Birkie is dropping some serious science on this forum!
|
|
|
|
Post by birkie on Jun 2, 2017 16:06:43 GMT
Very interesting Birkie! I wonder if this is why the early refrigerants like SO2 and Methyl Formate were low mass flow (low pressure?) but had a high heat of vaporization to keep the efficiency up? See, I think they didn't really have as much choice with the early refrigerants. There was really only a handful that would work at all for the desired temperature ranges. Then they had to balance toxicity with flammability with with operating pressures, etc. There was never really a good all-around choice, everything was a compromise in one way or another. Then CFCs came around, which seemed like a miracle cure until the effects on the ozone layer were discovered (and it was discovered how obnoxiously loud they were when used in monitor tops!). Then global warming.. It's a curse that plagues refrigerants to this day. Anyway, constrained to the natural refrigerants they knew about at the time (before CFCs were a real option), I *think* mass flow or density was one of the last things on their mind, and it just worked out that way. That could be wrong, though. The key scientific question on my mind at the moment is "Does mass flow in different refrigerants directly correlate with compressor noise". If the answer to this question is "yes", then a low mass flow has a distinct aesthetic advantage for monitor tops. That's what makes R227ea so interesting. It's very similar to SO2 on everything *except* density and mass flow, which are much higher. Is it as obnoxiously loud as R12, or more silent than any SO2 alternative tried to date? By the way, efficiency is even more complicated. A high heat of vaporization doesn't inherently mean a refrigerant will be more or less efficient, there are a lot more factors at play! |
|
|
|
Post by ChrisJ on Jun 3, 2017 2:29:17 GMT
Is this why the CA's compressor's displacement is 3 cubic inches?
That, and the fact it has to pull down to 26"+HG in a vacuum?
Wonder if it also has to do with why the passages in the CA's evaporator are huge?
|
|
It looks like most modern flurocarbon replacements I've analyzed at (except for the hydrocarbons) have higher capacity, higher pressures, and higher mass flow than either methyl formate or SO2. So the comparisons are very similar.