Variable RPM compressor/cap tube issue [Archive] - Refrigeration-Engineer.com forums

Chef

07-08-2008, 09:56 AM

Why does a BD35F work at all RPM settings yet has a fixed length cap tube.

After installing a new fridge system the option was available to run the compressor at 2000 RPM up 3500 RPM but what revs should be chosen to match the cap tube. After reading Henry Ehrens excellent cap tube article it would appear that only one setting would work but they all work, this should not be possible.

So noting the discharge and suction pressure, amps draw and cycle time for various RPM settings with no other changes, the system was run at many RPMs and the following table is a synopsis of many tests and it does appear that 2500 revs is the most thermally efficient. Of course pull down from warm is quicker at higher revs but so is the power consumption!

Revs
On time
cycle time
amps
A/hr

3500

10

13

6.2

4.77

3000

6

9

5.9

3.93

2500

5.5

8.5

4.6

2.98

2000

8.5

11.5

3.7

2.73

The discharge pressure changed by only 2 PSI between any test and the suction stayed about the same. It certainly does not fit with any cap tube sizing data one can get so anyone with an idea why this setup is so forgiving?

Bestest Chef

Chef

07-08-2008, 10:09 AM

Sorry - teh table was a mess

Revs On time Cycle time Amps Amp/Hr
3500 10 13 6.2 4.77
3000 6 9 5.9 3.93
2500 5.5 8.5 4.6 2.98
2000 8.5 11.5 3.7 2.73

Hope this is better

Chef

Tesla

07-08-2008, 11:07 AM

A capillary system is considered a critical charge system. It uses the condenser as a buffer for differing loads - if lo load it will be cooler and use more volume of condenser as reciever (filling with liquid). then at hi load uses more of condenser for heat tansfer (via hi temp = hi press = increased condensing temp) hence more space for condensing. By varying the volumetric flow rate of refrigerant within the design operating paremeters would only improve load control and efficiancy. This is why the system is forgiving

Chef

08-08-2008, 06:01 AM

Thanks Tesla and I beleived I had "the backing up into the condenser theory" sorted but maybe not.
This unit is rated at 320 Btu/hr at 2000 revs and a nice 540 Btu/hr at 3500 revs so one would expect to see a rise in the discharge pressure and even a fall in suction pressure as the unit is switched to 3500 revs but the Dx is just 2 PSI higher.

The condenser feels the same temperature and the frost line on the evaporator is in the same place. Non of this seems to line up with published data and so it seems cap tubes are way more forgiving.
:(

icecube51

08-08-2008, 04:42 PM

Yo Chef,
if you want some more data you can go to the Danfoss site and download the free program "Dancap" .
there you can give the lengt and diam of the existing tube and find out on wich presure it works best.

just you get an idea;)

Ice

Chef

09-08-2008, 06:18 AM

Thanks for the link Ice - I tried it.

The Dancap program gives me 11" for 540Btu?hr

I also looked at the excellent ACC tables and that gives me 41.7" for the same conditions.

My system is running fine at 94" - again all conditions are the same.

And JB sizing chart gives 144". All based on TC-26 tube.

So with all this great variation in recommended lengths how does one make an intelligent guess let alone a design without some trial and error experimentation.
For my system I already have the tube and so I just need to select the best RPM to run for max efficiency but for guys actually 'cutting a length' to solder in it must be a gamble.

:( Chef

The MG Pony

09-08-2008, 10:36 AM

cut messure sh cut messure Sh, check load handing, check sh, got it! write down length and cut every tube you need for that bit of kit @ that length, thats why I love TXVs!

icecube51

09-08-2008, 03:14 PM

check temperatures in differend places, just behind the capilair,middle of evap/condenser and at the end of evap/condenser. normal optimum temp between end of capilair and outlet evap/condensor is between 4C° and 8C°.if not, look the amound of refrig,take out or fill in.also check the rpm of motor at the high and low, wat the best result is.

Ice

aaron_k

20-11-2009, 03:54 AM

Chef,

It's been a long time since you posted this - I don't know if you'll see my follow up. I'm curious what ambient air and interior compartment temperatures you were using with your data tables? I may be in dire straits as the cap tube length used in my design comes from Dancap. And my system performance isn't what I need it to be. Throw in the variable speed capability and I am really unsure how to decide what is the most energy efficient way to run the system. Did you pursue this further?

I've been reading all your posts here, looking for more clues.

Thanks for your time,

Aaron_K

Chef

25-11-2009, 11:50 AM

Chef,

It's been a long time since you posted this - I don't know if you'll see my follow up. I'm curious what ambient air and interior compartment temperatures you were using with your data tables? I may be in dire straits as the cap tube length used in my design comes from Dancap. And my system performance isn't what I need it to be. Throw in the variable speed capability and I am really unsure how to decide what is the most energy efficient way to run the system. Did you pursue this further?

Aaron_K

Just spotted your reply. The condenser is water cooled and is at 30C so discharge pressure is about 125PSI on R134a. The evaporator is set to -10C on the plate.

What is it that you are trying to achieve and some details may help to give a clue how to recommend some RPM to run at and the tube size.

Chef

aaron_k

25-11-2009, 02:59 PM

Chef

26-11-2009, 11:05 AM

Chef,

Thanks for your reply. We have a passively cooled condenser system with a fin and tube evap and forced air inside our refrigerator. The sticking point is that the ambient temp is 52C while the inside temp is 5C. As currently designed (1.7 m cap tube, 0.71mm ID) the refrigeration system will only barely maintain the required temp, and does so with a lot of energy usage.

First question is what compressor is it - a BD35 or a BD50 - it will make a differance at high Rpm's? The
52C ambient is confusing - is it inside a 'hot' compartment and can you get cool air to it with a duct system? It may be the best option to improve performance dramatically.

Here's the latest running data:

Mean evaporator temperature (at steady state operation, 52C (126F) ambient, 5C (41F) compartment temperature): 35.1ºF
Compressor speed: 3000 rpm
High side pressure : 206 PSIg (as condensing temp 132.2ºF)
Suction pressure: 29.6 Psig
Superheat and subcooling values : 1ºF superheat & 6.6ºF subcooling.
Estimated refrigeration capacity (W, kcal/h, or btu/h): 133 W (per Danfoss).
180 g of R134a charge.

Very nice to see some good data to work on. How did you get the 133W of refrigeration capacity at your conditions? What is the volume of your box, L*H*W would be best info here?
Are you sure your getting 6.6F subcooling as the other data would suggest that your well into the 2 phase zone at the entrance to your tube.

Below 2800 rpm the internal temperature can't be held. The compressor almost never cycles off. I need it to run for only 30% of any given hour to reach our goal. Charge migration is adding 15C to the air temperature in the box if the compressor does shut off. It's obvious that the evap needs to be colder.

This is a good indicator of how much cold your loosing from the box but without the the unit running at max performance it wont help us yet to decide on how much performance you will get when it is properly tuned, but it will later so it is an important measurement.

Insulation is no problem (we're using a very high insulation value, equivalent to 6 inches of urethane foam).

I need to run the system for 24 hours in a 60 A-hr battery power envelope. We're no where close.

Wow - I am not sure you will get to 30% as an ON time in any hour but we shall try.

Tecumseh and ACC (and you) report much longer cap tubes than the DanCap derived value (what we used in our design). Circumstantial data suggests we're way off. I have no way to know for sure, other than guessing. Worst of all for us, we don't build the system (it's contracted out), and don't have any tools for "tweaking". I do though want to know what we can suggest for changes, then get them put in place and tested. Needless to say the design isn't finished!

It all depends on whether it is a BD35 or a BD50 but as an initial input the tube length is way too short.I cannot estimate a better length without the other information.

What are your initial thoughts on my predicament?

You're great knowledge and expertise is obvious Chef, reading these forums; thank you in advance if you have some ideas for us!

Regards,

Aaron_K

Are you OK with PH diagrams? because it is easier to explain what is going on in your system with them.

If your tube is too short then both gas and liquid enter it at the entrance, the gas makes up for the tube being too short as it creates faster flow rates and more friction. The problem here is you are not sending down the tube as much liquid as you could and therefore your getting less cooling. Basically your just pumping gas around your circuit and that costs amp/hrs.

I am currently working on a unit similar to yours but it has a sight glass and you can see the outlet port of the sight glass is covered to a max of 5% with liquid and remainder of the oulet lets gas get entrained. Very illuminating indeed.

Initial (and I mean initial) it looks like you may need around 2 metres of .63mm tube but that depends on your BD model. Your current 1.7m of 0.71mm is pretty radical!

Any info on size of the condenser and evap in terms of line sizes and lengths would also help.

Chef

aaron_k

27-11-2009, 07:05 PM

Chef,

Many, many thanks! It's a BD35F, and the ambient is equal to desert conditions (where this will be used, outdoors). It's a 5 cu. ft. (rectangular box about 1.5X taller than deep or wide) refrigerated compartment. Capacity was read from Danfoss' RS+3 program. The insulation system performance is such that at the stated conditions only 135 btu/hr of heat lead enters the compartment. I'm not sure (completely) about SC, as the data is second hand from our contractor. This is the data they report.

I agree with you about the system just pumping gas around. There seems to be just enough phase change in the evap to overcome the heat lead (above 2800 rpm) and to cool the gas / liquid with the rest of the energy wasted. I tried Tecumseh's cap tube software and come up with around 4.5m of the 0.71mm ID cap tube... so yes, I agree that the cap tube is a big contributor to our problem.

With my additional data what are your thoughts? I think I've piqued your interest!

I am so grateful for your ideas. Thank you!

Aaron_K

PS - PH diagrams are fine.

aaron_k

27-11-2009, 07:24 PM

Internal condenser tube volume is around 40 cu. in. (3/8" tube ID, one row of tubes). Evap is around 9 cu. in. (same construction).

Magoo

27-11-2009, 07:55 PM

Hi Chef,
by varying speed of compressor, this will lower compression ratio, and vary flow rate through fixed length capilary tube.

Chef

28-11-2009, 02:09 AM

Hi Magoo

Thanks for that but Aaron has posted a new series of questions about his system and so the thread has changed. Now its how to get a BD35 running in a desert at 52C ambient.

Any ideas

Chef

Chef

28-11-2009, 02:51 AM

Aaron_K

First try seaching google for
DEHC.PK.100.C8.02/520N0436

It is the BD manual and has all the details you will need to see the performance and operating conditions.

The max temp of the discharge is quoted as 60C continuous and 70C intermittent - as you are running around 68 to 70C you're right on the edge!! They also say a fan is necessary (compulsory) blowing over the compressor above 55C so you may like to consider that but at 52C ambient its a bit of a question if it will be of much use? Also they strap on an oil cooler to the side of the pot - a sort of aluminium finned heat exchanger. You may need that as well.

The 133W is based on ambients of 32C and 55C condensing - that is not what you have so the cooling will be less. It shows the test conditions on page 12 of the manual

I have done some quick runs on your conditions and these results are for various lengths of 0.71 tube.
4.5m x=0
3.1m x=0.1
2.5m x=0.2
2.0m x=0.3
1.7m x=0.4

This shows what the entry conditions are of the refrigerant as it enters the tube. With a 1.7m tube the quality is 0.4 and h4=h3=338, so you are operating in the middle of the bell. If the tube is too short it can only meter the refrigerant by having excess uncondensed gas flowing through the tube to give the needed pressure drop.

A tube of about 4.5m should have a quality of about 0.0 at the tube entrance and so h4=h3=280Kj/Kg
You gain 58Kj/Kg of cooling and the system should perform way better.

There seems to be no jargon for x=0.4 at the entrance like 'negative subcooling' - probably because it cant be measured! Do you have a good jargon word to describe it?

I'll have a look at the 0.63 diameter tube later to see if it more suitable with more reasonable lengths.

Chef

Gary

28-11-2009, 03:06 AM

Here's the latest running data:

Mean evaporator temperature (at steady state operation, 52C (126F) ambient, 5C (41F) compartment temperature): 35.1ºF
Compressor speed: 3000 rpm
High side pressure : 206 PSIg (as condensing temp 132.2ºF)
Suction pressure: 29.6 Psig
Superheat and subcooling values : 1ºF superheat & 6.6ºF subcooling.
Estimated refrigeration capacity (W, kcal/h, or btu/h): 133 W (per Danfoss).
180 g of R134a charge.

Below 2800 rpm the internal temperature can't be held. The compressor almost never cycles off. I need it to run for only 30% of any given hour to reach our goal. Charge migration is adding 15C to the air temperature in the box if the compressor does shut off. It's obvious that the evap needs to be colder.

Sounds like an inefficient compressor to me... which is to say that the compressor does not have enough cacacity for the evaporator... nor does it have enough capacity for the condenser.

Gary

28-11-2009, 03:20 AM

But then this assumes standard design and I take it this is a system you are designing?

mad fridgie

28-11-2009, 07:03 AM

aaron, I will leave the cap sizeing to chef (more is skill), I do not believe you are getting any sub-cooling.
So what can we do?
Do you have a drain off your evaporator, if so trap the water and run the liquid line (between the condersor and the capillary) in this trapped water (preferably a long high tube). The over flow dripple over the condensor (this may give the effect of dropping the air temp) Do not change your speed, over shoot your set point (large differential) Install a direct acting solenoid valve (no pressure drop) in the liquid line and Non return valve or another direct acting valve in the suction line ( as low a pressure drop as possible) These measures should stop refrigerant migration.

Chef

28-11-2009, 07:39 AM

Internal condenser tube volume is around 40 cu. in. (3/8" tube ID, one row of tubes). Evap is around 9 cu. in. (same construction).

The smallest size normally used with a BD35 evaporator is 10 cu in and they can drive up to 25 cu in quite well especially at 3500 revs and get it down to -15C or so. Plate sizes and volumes can be found on various web sites including Weaco and Frigoboat. So your evap may be a little on the small size!

You can also increase the charge to 300g max so a larger evap may make sense.

The condenser seems to be a good size for your conditions and should help to keep the head pressures down a little.

I am not sure why Gary thinks they are way too large for your compressor, I am sure he will explain later.

Mad - good idea with the condensate, sort of a loose it on the roundabouts but get it back on the swings approach.

Chef

Gary

28-11-2009, 08:13 AM

If the system has a hand valve in the liquid line, it should be fairly easy to pinpoint the problem.

Close off the valve (liquid flow) until the evap outlet superheat is 10-15F/5.5-8.5K, then take a full set of measurements.

If everything comes together, then we know that the cap tube is not restrictive enough.

mad fridgie

28-11-2009, 08:24 AM

Gary

28-11-2009, 08:26 AM

I am not sure why Gary thinks they are way too large for your compressor, I am sure he will explain later.

The system has low TD's coupled with low superheat. If this were a system that had worked well in the past, I would be looking long and hard at that compressor.

An inefficient compressor has the same symptoms as a compressor that is undersized for the evap and cond... which is the same thing as the evap and cond being oversized for the compressor.

But being that this is a work in progress, maybe its the cap tube.

mad fridgie

28-11-2009, 08:41 AM

Studied the above a bit more, superheat is very low, what is your discharge temp, You may find that a lot of the latent exchanger (cooling) is actually occuring in the compressor.
Stating obvious!
So you are overfeeding.
Run your capillay through (around the suction) reduce the vapour contet entering the evap (coil becomes more flooded) reduces the evap pressure drop, increasing LMTD.
With suction pressure and superheat as is the evap does not look undersize. (not knowing your paticular evap) unless the evap internal pressure drop is high, masking the true performance (check the pressure at evap inlet)
Mad are you mad,the capillary is overfeeding, thats why the evap is to small, but Mad if there is no liquid sub-cooling the liquid could whipping the evap and only flashing of with the pressure drop.
Either way measure the evap inlet pressure
cheers Mad and Mad

Chef

28-11-2009, 10:09 AM

The system has low TD's coupled with low superheat. If this were a system that had worked well in the past, I would be looking long and hard at that compressor.

An inefficient compressor has the same symptoms as a compressor that is undersized for the evap and cond... which is the same thing as the evap and cond being oversized for the compressor.

But being that this is a work in progress, maybe its the cap tube.

These may be fine conclusions for a TXV system but you need to consider that in the current operating condition the cap tube is fed with an x=0.4 so the condenser and evaporator are not operating as they would in a normal TXV system. It is classic to have very low TD's in this circumstance.

The normal rules you usually apply do not count in this case and so you need to think 60% liquid and 40% gas flowing through the condenser and into the tube and then 30% liquid with 70% gas entering the evapoator - this is well off the usual situation. Plot it on a PH diagram and you will see the problem.

Aaron also mentions the data about SC and maybe SH came from the contractor so it is suspect.

The SC is obviously wrong as it never passes the saturation line in the PH diagram (h2 - h3) in the condenser.

Your thoughts?

Chef

Chef

28-11-2009, 10:16 AM

Mad are you mad,the capillary is overfeeding, thats why the evap is to small, but Mad if there is no liquid sub-cooling the liquid could whipping the evap and only flashing of with the pressure drop.
Either way measure the evap inlet pressure
cheers Mad and Mad

Your Madness - No there is no subcooling and most of it flashes down the tube so its just passing gas (no pun intended) thats why the evap looks small? Seems your alter ego sees it correct.

Chef

Gary

28-11-2009, 03:36 PM

These may be fine conclusions for a TXV system but you need to consider that in the current operating condition the cap tube is fed with an x=0.4 so the condenser and evaporator are not operating as they would in a normal TXV system. It is classic to have very low TD's in this circumstance.

The normal rules you usually apply do not count in this case and so you need to think 60% liquid and 40% gas flowing through the condenser and into the tube and then 30% liquid with 70% gas entering the evapoator - this is well off the usual situation. Plot it on a PH diagram and you will see the problem.

Aaron also mentions the data about SC and maybe SH came from the contractor so it is suspect.

The SC is obviously wrong as it never passes the saturation line in the PH diagram (h2 - h3) in the condenser.

Your thoughts?

Chef

The rules apply for both TXV and cap tube systems.

Although the mixture is 60/40 at the cap tube inlet, there is no way it is 60/40 throughout the condenser.

The low TD's are not unlike those of an inefficient/undersized compressor.

All of the data comes through both the contractor and Aaron which makes it doubly suspect... but we work with the numbers we are given.

Try plotting an inefficient/undersized compressor on your PH diagram.

Gary

28-11-2009, 03:42 PM

Being a hands-on kinda guy, I would restrict the liquid line flow as outlined earlier and see what the numbers tell me then. A measurement beats a calculation every time.

I am not saying the compressor is inefficient/undersized. There is little doubt the cap tube is not restrictive enough for the conditions... the jury is still out on the compressor.

desA

28-11-2009, 04:10 PM

Gary

28-11-2009, 04:22 PM

Has a set of charge-determination trials been done Aaron?

The charge will affect your system dramatically.

I'd suggest taking your current setup, beginning with a minimal charge, run the system. Record the results, add in additional charge (say 2.5-5%) test again. Continue the tests, recording as many parameters as you can. Plot the results & see if the system sweetens at some point.

If we are to believe the SC and SH measurements, then we are already flooding the compressor. Adding charge would make this worse. Removing charge would give the compressor a break, but would eliminate what little SC we have.

desA

28-11-2009, 04:49 PM

aaron_k

28-11-2009, 05:02 PM

Gary

28-11-2009, 05:13 PM

I've not had much experience with tuning cap tubes, but, with the low SH, I'd expect to want to increase the cap pressure drop (restriction) & permit less fluid to flow.

This would probably back the condenser up a little further.

Just how much SC would you be looking for?

Ideally, I would like to see 10-15F/5.5-8.5K for both SC and SH, although the question then arises: What happens when the ambient is far less than 126F/52C? I might expect both SC and SH to be higher at lower ambient and lower at higher ambient. A happy medium is needed, thus I would be aiming for about 10F/5.5K for both SC and SH at the higher ambient and then see how that works out for lower ambient.

Gary

28-11-2009, 05:46 PM

Gentlemen,

Thank you all for your discussion about the system. As it's a holiday weekend here in the US I'm not online as much, so I apologize if I don't respond quickly.

I don't have a hand valve in the system, and I think everyone agrees that the cap tube is not restrictive enough. I'm going to try to start with that, re-test and will report back. I'm also going to add a blower to the condenser to help the SC a bit. I like the ideas about using solenoid valves to eliminate charge migration; I looked into that earlier and didn't find a latching solenoid that seemed to be small enough, or that had solder connections. The fridge will see rough service and any threaded connection is more prone to leakage.

Many many thanks to you all for your thoughts and suggestions. I'll update the group after we've actually made a change and can re-measure the results.

Aaron_K

During the off cycle refrigerant migrates to the coldest part of the system, which generally means the evaporator. If the suction line exits the evap at the bottom and goes downhill to the compressor, then gravity carries the liquid to the compressor and you need to worry about migration. This can usually be resolved by running the suction line up to the top of the evap and then down to the compressor, trapping the liquid in the evap.

If the compressor is above the evap or the evap is fed from the bottom up, then migration is not something you would normally need to worry about.

At this point I don't really see a need for a condenser fan given the very low cond TD, although the need may reveal itself as the other problems are resolved. On the other hand, a fan wouldn't hurt.

On an experimental/prototype system it is a good idea to have valves, access ports and sensors everywhere.

mad fridgie

29-11-2009, 04:46 AM

I think we are going to get very lttle natural sub-cooling (additional can be added as stated above), so it is important that we have a liquid seal to the cap.
I would suggest, that you install a "P" trap on the outlet of the condensor, then take the pipe up, to the height of the condensor, have "u' bend, then directly in the drier (vertical position) then into the cap. by nature we would have a little liquid head sub-cooling.
I agree with gary when doing R&D more data reading pionts, the easier it is to make descisions.
(Its been many moons since I have delt with caps)

Gary

29-11-2009, 06:26 AM

I think we are going to get very lttle natural sub-cooling (additional can be added as stated above), so it is important that we have a liquid seal to the cap.)

I agree. The SC can be no more than the cond TD, and in fact must be less... and we have very low cond TD.

Similarly, the SH must be less than the evap TD... and we have very low evap TD.

We may need something like the Magoo rule for both SC and SH.

Chef

29-11-2009, 11:18 AM

The rules apply for both TXV and cap tube systems.

Although the mixture is 60/40 at the cap tube inlet, there is no way it is 60/40 throughout the condenser.

The low TD's are not unlike those of an inefficient/undersized compressor.

Try plotting an inefficient/undersized compressor on your PH diagram.

Here are 3 sets of readings from a system.

Dx=10.6 Sx=2.52 SC=6.1 SH=1.8
Dx=9.7 Sx=2.32 SC=1.0 SH=0.6
Dx=9.22 Sx=2.05 SC=0.0 SH=0.0

Dx and Sx are bara and SC and SH are in K

They are taken for preset temperatures of 0C, -3C and -7C, its just a simple fridge system.

So what your saying is these obey the same rules as a TXV and you can determine if all is well or not from them. That I doubt very much.

Its a precharged system from a most reputable manufacturer.

Cap tube systems do not obey the same SH and SC rules as TXV systems and trying to use those rules will be a big problem. Henry Ehrens puts it very nicely in his paper.

Oh and last-

Please enlighten us on how to plot an inefficient compressor on a PH diagram. An example would be great.

Chef

desA

29-11-2009, 11:32 AM

Here are 3 sets of readings from a system.

0'C : Dx=10.6 Sx=2.52 SC=6.1 SH=1.8
-3'C : Dx=9.7 Sx=2.32 SC=1.0 SH=0.6
-7'C : Dx=9.22 Sx=2.05 SC=0.0 SH=0.0

Dx and Sx are bara and SC and SH are in K

Is this correct ie. taken at those temps (was a little confused by your reference).

Thoughts:
If that is the case, then I feel that the condenser is running out of capacity (large SC), with the evap slightly over-capacity (large SH), at the 0'C operating condition. It seems to have a system imbalance.

I'd not be at all happy with SH=0K heading towards a compressor, unless the line on the way gets hot & evaporates the droplets.

Gary

29-11-2009, 05:16 PM

Here are 3 sets of readings from a system.

Dx=10.6 Sx=2.52 SC=6.1 SH=1.8
Dx=9.7 Sx=2.32 SC=1.0 SH=0.6
Dx=9.22 Sx=2.05 SC=0.0 SH=0.0

Dx and Sx are bara and SC and SH are in K

They are taken for preset temperatures of 0C, -3C and -7C, its just a simple fridge system.

So what your saying is these obey the same rules as a TXV and you can determine if all is well or not from them. That I doubt very much.

Its a precharged system from a most reputable manufacturer.

Cap tube systems do not obey the same SH and SC rules as TXV systems and trying to use those rules will be a big problem. Henry Ehrens puts it very nicely in his paper.

Oh and last-

Please enlighten us on how to plot an inefficient compressor on a PH diagram. An example would be great.

Chef

I have no idea what your point is. Where did I say I could tell if all is well or not from those few numbers? I didn't. Nor did I say that all rules apply the same for TXV and cap tubes. The specific rules we were discussing were low TD's coupled with low superheat being an indicator of inefficient/undersized compressor.

Oh and last:

I was challenging you to plot an inefficient compressor on a PH diagram thereby enlightening us. If you don't want to do this just say so.

Gary

29-11-2009, 05:22 PM

Is this correct ie. taken at those temps (was a little confused by your reference).

Thoughts:
If that is the case, then I feel that the condenser is running out of capacity (large SC), with the evap slightly over-capacity (large SH), at the 0'C operating condition. It seems to have a system imbalance.

I'd not be at all happy with SH=0K heading towards a compressor, unless the line on the way gets hot & evaporates the droplets.

Very likely Chef's system has a suction/liquid HX... which changes everything by adding both SC and SH.

As far as we know Aaron's system does not have a suction/liquid HX.

Why he (or his contractor) would charge the system to the point where it is flooding the compressor (1'F SH) is anybody's guess.

Chef

30-11-2009, 03:16 AM

The rules apply for both TXV and cap tube systems.

.

I thought you did mention that the rules apply for both TXV and cap tube systems?

The point of posting the 3 conditions of a system (not Aaron's) was to demonstrate that a cap tube system operates with varying SH and SC and that it cannot be used to diagnose the systems condition directly.

I dont think its useful for the forum to use TXV based analysis on a cap tube system - much better to use cap tube based analysis. Some of it has been documented but much of it has not. On the basis that they are very very different this may be an opportunity to move forward and start to develop some rules and guidelines.

DesA kindly demonstrated this with a possible analysis of a condenser running out of capacity, a slightly oversize evap and system imbalance. Well this is a production unit with thousands of units worldwide and the system functions perfectly, the results shown are exactly correct for a cap tube system.

Also as it is critically charged running with very low SH is tyipical when the tube inlet is right on the saturation line, no SC and X=0.

Control and balance are more determined from the excess pressure in the condenser which could be caused, for one, by a high heat load. This may show a SC value of between 5.5K to 8.5K which Gary suggests would be his aim point, when in fact it represents a cap tube that is too long - it would force excessive head pressure, refrigerant backed into the condenser, starve the evap and have a lower than optimum COP.

Another example may be low TD's, here balance is achieved by suction throttling the compressor, lowering the flow rate, allowing gas and liquid to pass through the tube (with an x=0.4 perhaps), less gas is condensed and so the TD shows up as low. This indicates a short cap tube. By TXV rules it seems to indicate an inefficient compressor.

The thread should really concentrate on the cap tube issues.

Chef

desA

30-11-2009, 03:27 AM

Think of a TXV as a variable capacity cap tube - the variation driven from a temp signal. If you took the bulb off the evap line & stuck it in a fluid at preset reference temp, you could tune the 'variable cap tube'.

The rules of RHVAC then should apply. If not, then why not?

Actually, you could perform the same function with an AEV, orifice or even fine-tuned hand valve. They are all restriction devices. There is no black art in this.

Why did the 'large manufacturer' get it wrong? Simple. Their systems were designed in a country far from the current environment in which they currently operate, or, ... they messed up... :)

Chef

30-11-2009, 03:35 AM

I have no idea what your point is. Where did I say I could tell if all is well or not from those few numbers? I didn't. Nor did I say that all rules apply the same for TXV and cap tubes. The specific rules we were discussing were low TD's coupled with low superheat being an indicator of inefficient/undersized compressor.

Oh and last:

I was challenging you to plot an inefficient compressor on a PH diagram thereby enlightening us. If you don't want to do this just say so.

Nicely sidestepped. Not sure if it is rhetoric or semantics though.

The specific rule we are discussing then is low TD's and low SH. Well as mentioned in the above post it is TXV rules that suggest an ineffcient compressor and cap tube rules that suggest it is a very short tube. My point again is we should be using and building on the right set of rules.

Oh and last.
I did indeed suggest you plot an x=0.4 on a PH diagram and was also just challenging you, so here is the plot I was alluding to.
3197

Hopefully you will see the h2-h3 is greatly reduced and hence the TD's will fall.

Presumably you'll do the gentlemanly thing and post your inefficient compressor PH diagram.

Chef

Chef

30-11-2009, 03:48 AM

Why did the 'large manufacturer' get it wrong? Simple. Their systems were designed in a country far from the current environment in which they currently operate, or, ... they messed up... :)

Why don't you test it for yourself, instead of slagging off a perfectly sound company?

Build a small prototype & test if your theory is correct, or not.

Then tell us the results of your experiment.

So which one of these 2 quotes do you stand by?

Chef

desA

30-11-2009, 03:59 AM

Both, actually.

The first quote is a typical occurrence for foreign companies manufacturing for a market with a local climate different to their own. It happens in the automotive heat-transfer industry all the time. It is the principal reason for CKD pack substitutions with local parts.

The second is that you punted the 'name-dropping' scenario of 'big company is perfect'. I don't know their name, nor do I wish to. No big company is perfect - they are as good as their weakest designer. Your results prove exactly this.

Better to keep the discussion technical - not personal, eh... :)

Gary

30-11-2009, 04:56 AM

The specific rule we are discussing then is low TD's and low SH. Well as mentioned in the above post it is TXV rules that suggest an ineffcient compressor and cap tube rules that suggest it is a very short tube.

TXV rules would suggest an inefficient compressor or TXV stuck wide open (overfeed).

Cap tube rules would suggest the same, the equivalent for TXV stuck wide open being the short cap tube (overfeed).

IOW, inefficient compressor or overfeed, be it TXV or cap tube.

Oh and last.
I did indeed suggest you plot an x=0.4 on a PH diagram and was also just challenging you, so here is the plot I was alluding to.
3197

Hopefully you will see the h2-h3 is greatly reduced and hence the TD's will fall.

Presumably you'll do the gentlemanly thing and post your inefficient compressor PH diagram.

Chef

PH diagrams are your thing, not mine. I don't use them.

I declined your challenge and apparently you have declined mine. Fair enough.

desA

30-11-2009, 05:11 AM

Chef

30-11-2009, 06:06 AM

http://www.refrigeration-engineer.com/forums/attachment.php?attachmentid=3197&d=1259551642

I don't want to be rude, but this diagram is plainly incorrect.

x=0 must, by definition, sit on the liquid saturation line - if you are referring to the LP line. If you are referring x=0 to the HP line, then your condenser is critically sized & any x>0 implies an undersized condenser... :)

Plainly incorrect or misread and misunderstood?

Previously we have been talking about the conditions at the entrance to the cap tube ie at point h(3) on the PH diagram. The arrows point to 3 cycles on the PH diagram for different X values at the h(3) point.

As you can see the red line touches the saturation curve in the top left hand side. Maybe I should have been totally correct and said "This curve shown in red is the PH cycle for a cap tube system with an entry to the tube of x=0 ie at the h(3) point" but it did not fit on the diagram. I just figured you had been following the debate more closely, sorry for that.

If you are referring x=0 to the HP line, then your condenser is critically sized & any x>0 implies an undersized condenser...

Again your quoting some dogma from TXV conditions and this just does not apply to cap tube systems - and the condenser is not critically sized nor is it undersized its just perfect. Whats imperfect is the cap tube length for the system and the PH diagram shows how it chooses to balance itself out even when the tube is wrong. By having entrained gas in the flow to make up more pressure drop. It is not an ideal situation nor one that should be designed for but it happens and it is used as a tool to evalute the correctness of the tube, not the condenser.

Chef

desA

30-11-2009, 06:14 AM

I think that you are confused on the technical definitions of x (vapour fraction).

If you do want to refer to 'x' then it may be wiser to keep this to the LP portion of the graph - this is the traditional place in which it is used in relation to the vapour compression cycle. The reason for this is that the cycle is defined on full saturated liquid at point 3.

There should be no vapour fraction going into a cap tube, or TXV... If there is, then the design needs to be altered - the cap tube is not meant as a total system imbalance corrector.

If you do want to refer to the vapour fraction in some other place, then it may be wiser to state clearly 'x=0 @ HP' or something similar, to avoid confusion.

Chef

30-11-2009, 09:26 AM

Maybe I may respond to post by inserting blue text within your quote.

I think that you are confused on the technical definitions of x (vapour fraction).

Thankyou for your kind interpretation of my understanding of X - the vapour fraction.

If you do want to refer to 'x' then it may be wiser to keep this to the LP portion of the graph - this is the traditional place in which it is used in relation to the vapour compression cycle.

Well, X occurs in the condenser and the evaporator and always has I beleive, in fact in seems to occur everywhere within the bell (inside the saturation line). I take note of your desire to more precisely position the quoted X and propose XC and XE for condenser and evap respectively, XCT will be the value somewhere along the cap tube.

The reason for this is that the cycle is defined on full saturated liquid at point 3.

Not sure that is an absolute! There are instances when point 3 (h3) is inside the saturation curve, for instance in a classic freezer. You might like to give some references for the cycle being strictly 'saturation'. Otherwise to me its a diagram that descibes what is happening and can have h3 wherever it may be.

There should be NO vapour fraction going into a cap tube, or TXV - none...

I hope you are absolutely sure about that! There seems to be some (many in fact) references that show vapour entering the cap tube is part of the whole process especially when operating in freezer temperature ranges (all else being normal (average) world conditions). In fact Supco report on a design that actually introduced vapour into the cap tube entrance.

"In the early 1960s, a manufacturer of window air conditioners took advantage of bubbles entering a cap
tube by placing a small heater around the strainer before the cap tube inlet. The thermostat was in
reality a rheostat that controlled the intensity of heat to the strainer—which in turn regulated the
amount of bubbles entering the cap tube. The first stage of heat to the strainer was to reduce the
subcooled liquid temperature (increasing the bubble length). The second stage was to create a boiling
action, in various heat intensities, to decrease the overall efficiency of the evaporator."
From

THEORY OF THE CAP TUBE AS A
REFRIGERANT CONTROL
By Henry Ehrens, Sr Engineer - Sealed Unit Parts Co., Inc

There are many other referances that you may wish to discover.

If there is, then the design needs to be altered - the cap tube is not meant as a total system imbalance corrector.

You might like to clarify that comment and provide some eveidence that it is not the total system imbalance corrector, preferably something published rather than conjecture.

If you do want to refer to the vapour fraction in some other place, then it may be wiser to state clearly 'x=0 @ HP' or something similar, to avoid confusion.

Chef

desA

30-11-2009, 09:39 AM

Many thanks...

Chef

30-11-2009, 10:41 AM

The heater you mentioned will, in all likelihood, not be found on the refrigerator under review. This is merely a flow control mechanism. The key words here are - measure, control. This is a pure red-herring to try & bolster your argument.

Of course it wont be found on the refrigerator I mentioned - its a quote from the 1960's. It is a reference related to your catagoric statement that NO VAPOUR ENTERS THE CAP TUBE, purely that and not, I reckon, a red herring, its a quotable and reputable source.

The cap tube you are using seems to be a cut-to length version. Perhaps I'm mis-reading things.

I am pretty sure it is cut to a length - yes.

I would refer you to any thermodynamics text book, which discusses the terminology used in vapour compression cycles. This may make it a little less difficult to follow your logic, in future. (I am saying this because of the general confusion already existing in this thread.) There is no need for vapour entry into the cap tube, I'd bet.

But what if there is - and I have good evidence to suggest it does. The whole point of a discussion is to actually look at what the other person is saying and see if it is valid - not to just throw it away because it does not meet your ideas. Try to see where this is going and just for a small amount of time look seriously at it.

Uncontrolled vapour entry into a cap tube will present difficulties.

No mention of uncontrolled anywhere. Fine statement but no reasoning behind it.

Based on the data you have submitted thus far, I'd suggest that the evap & condenser are not totally suited to the compressor under the situations presented.

I have already stated the system works like a charm and is performing fantistically, really good cold produced, excellent efficiency, top notch ... What else can I say except it works perfect.

This will explain some of the variation you see in SH & SC across the range of operation.

I have tried to explain that this is normal in a cap tube system - you can only have perfect conditions for a cap tube at a single point of conditions and elsewhere it will exhibit strange but definable characteristics.

There is no magic, nor mystery to all of this - it's straightforward engineering & system balancing.

Correct - but who is going to determine the system balance and rules governing it. By addage or engineering. I choose engineering.

Why dont you have a read of this
http://www.refrigeration-engineer.com/forums/showthread.php?t=16920
and concentrate on the limiting phase of the system.
You may find it gives you the information you need.

The solution is well within grasp and at the moment I am working on 4 system and 3 of the 4 exhibit gas/liquid at the tube entrance.

Chef

mad fridgie

30-11-2009, 10:41 AM

mad fridgie

30-11-2009, 11:02 AM

Chef

30-11-2009, 11:27 AM

Hi,
X=0 is at 100 liquid, at cap inlet, (it shows very little sub-cooling, also no allowance has been made for pressure drop through the heat exchangers.
A cap system is critical charge and self balances and at different conditions the varying amount of refrigerant mass are held within different parts of the system.

Precisely and these constitute part of the control process.

The mass transfer can be charted by chefs figures
At low Te mass is in the evap, hence little SH, and little mass in the condensor so no liquid seal is apparent in the consensor so no sub-cooling (vapour and liquid mix)

This is the point and your conclusion is the same as we have. No subcooling and so it will be a vapour/liquid mix entering the tube. It then enters the next pahse of its control, gas entrainment to increase the pressure drop.

The opposite at higher evaps.
Using a P/E diagram on its own to diagnose in efficient compressor not possible.

I think this is just one of Gary's jokes, but as he says he does not use PH diagrams it may be rather a cruel one.

As far as the current problem it seems to me that we do not have reliable constant (excludes comp displacement).
It would be nice for calculation if the sucton return had a much higher super heat (from this we could calculate actual mass flow) at present it seems that we could have a vapor/liqud mix (by what % you can not know)

We running simulations on it to get the various fractions and so far they agree well with the test data. Hopefully soon we will have a way to determine the various % vapour/liquid at different parts of the system and begin to develop an analytical tool.

The 3 systems we are working on all exhibit gas entrainment and we have a way to actually see it for comparison to the predictions.

Chef

desA

30-11-2009, 11:40 AM

A cap system is critical charge and self balances and at different conditions the varying amount of refrigerant mass are held within different parts of the system.

At low Te mass is in the evap, hence little SH, and little mass in the condensor so no liquid seal is apparent in the consensor so no sub-cooling (vapour and liquid mix)

MF, does it sometimes occur, in cap systems, that the charge holdup occurs in the opposite way ie. the condenser bleeds off over the course of part of the cycle?

I've seen this in some TXV systems & put it down to orifice size & system balances. In other words, is it a given to expect that a system will always move to empty the evap & fill up the condenser?

Gary

30-11-2009, 04:55 PM

One point I did not bring up was that in many cases the cap is in contact with suction, this is to ensure ad excess liquid in the suction is boiled of and it reduces the pressure drop across the cap (reducing liquid to vapour change)

I suspect the zero SC and SH system Chef keeps referring to has just such a suction/liquid HX arrangement... and this changes everything.

Gary

30-11-2009, 05:13 PM

Using a P/E diagram on its own to diagnose in efficient compressor not possible.

Using a P/E diagram on its own to diagnose anything is questionable.

There is a time for calculations (before the system is built) and a time for measurements (after the system is built). But those who calculate want to keep on calculating.

If you are a hammer, everything looks like a nail.

Gary

30-11-2009, 06:41 PM

mad fridgie

30-11-2009, 08:37 PM

From my experience (many years ago) when producing cap systems, a single method was not used,
Simply you calculated (theory) cap size and charge (this would normally bring you within a reasonable starting point) and is needed, then was tuned, by the practical test method.
I suspect simulators now a days are a lot better than simple steady state calcs that we used in the past.
Cap systems are really designed for fixed load systems. (think of you old fridge with freezer plate in side and skin condensor. fairly constant.)
As long as it worked OK at max design ambient, we never worried about, performance at lower ambients. At the lower ambients the the need for nett refrigeration effect is reduced, so Te normally dropped, with little concern about efficiency. The only concern was liquid flood back.

desA

30-11-2009, 11:07 PM

I suspect simulators now a days are a lot better than simple steady state calcs that we used in the past.

The trouble is - they're not.

A simulator generally gives a static snapshot of what is a thermodynamically-stable system i.e. one that has been stable & unchanging for a long, long time (infinite time). Coolpack, for instance works this way. In this sense, all they do is to automate the hand calculations that any good fridgie could do, anyway.

Things like dynamic response, mass charge & so forth, would be found in some academic research codes, but, they are not common in the public domain.

I agree with the folks who refer to the use of simulators to get through the initial design stage & then hands-on tuning to settle the real system. This is the strength of it. Sometimes, referring back to the simulator to assist analysis of the real operational data can be useful, but, in the end, where the rubber hits the road is whether the unit actually works as per original design, or not.

In many cases, if the original design is too tight, then real performance can suffer. On the other hand, if the spec is too loose, then system stability may suffer, as well as system costs. It's a balancing act all the way.

Chef

01-12-2009, 11:46 AM

Imagine that your fridge is not functioning properly. You call a service company and the tech walks in the door saying, "Hmmm... I wonder what's wrong with the cap tube sizing?"

Does this seem like an insane approach to anyone but me?

We can't see the system from here, we can only see the OP's description of it. Where is the detailed description of the system? Pics/diagrams would be nice.

Where is the step-by-step process of elimination by which we have narrowed the problem(s) down to the cap tube?

Or maybe its just me who is insane.

Completely correct, no one is going to visit a running system and say oooh looks like the tubes wrong. They will look at leaks, blocked condensers etc.

But the original post was handed over to Aaron and the aim is to try and help him sort out his problem.

He has had built, by a subcontractor, a unit to operate in some very unforgiving conditions and it seems the decision on the cap tube was made by using a freely availble progam. It is a new system and all new components.

Now his problem is poor cooling and well...... you have all seen the data he has posted.

As its a new design and a cap tube system (and using a software program freely available) one might say:-

Is the condenser big enough - well at quoted size maybe but a fan would help as you suggested.
Is the evap Ok well it seems small but it is fan assisted so that makes a big differance to its perfromance. So that seems OK
Is the tube OK - well at its stated length and diameter it is really short.
The compressor is new so that should be OK in performing to specs. Is it sized right - only have to look at the heat load and the quoted cooling effect to see its in the ball park.
Is the charge OK - well cant see any data on that yet to suspect it but its always a suspect
So first conclusion is - wait a minute - I think the cap tubes wrong.

Its a differant scenario to the one you have pointed out but it is Aarons scanario.

Chef

Gary

01-12-2009, 03:27 PM

Assuming the system does not have a suction/liquid heat exchanger to add SH, he is currently flooding the compressor.

He needs to remove refrigerant until the compressor inlet SH is about 20F/11K and then take a full set of pressure and temperature readings, so we can see how it runs without a flooded compressor.

desA

01-12-2009, 03:34 PM

What would the impact be of flooding the compressor, in this case?

Catastrophic, or mainly a power-consumption issue?

Gary

01-12-2009, 03:51 PM

What would the impact be of flooding the compressor, in this case?

Catastrophic, or mainly a power-consumption issue?

It could be catastrophic. I would not rule out damaged compressor at this point.

Gary

01-12-2009, 04:10 PM

Completely correct, no one is going to visit a running system and say oooh looks like the tubes wrong. They will look at leaks, blocked condensers etc.

But the original post was handed over to Aaron and the aim is to try and help him sort out his problem.

He has had built, by a subcontractor, a unit to operate in some very unforgiving conditions and it seems the decision on the cap tube was made by using a freely availble progam. It is a new system and all new components.

Now his problem is poor cooling and well...... you have all seen the data he has posted.

As its a new design and a cap tube system (and using a software program freely available) one might say:-

Is the condenser big enough - well at quoted size maybe but a fan would help as you suggested.
Is the evap Ok well it seems small but it is fan assisted so that makes a big differance to its perfromance. So that seems OK
Is the tube OK - well at its stated length and diameter it is really short.
The compressor is new so that should be OK in performing to specs. Is it sized right - only have to look at the heat load and the quoted cooling effect to see its in the ball park.
Is the charge OK - well cant see any data on that yet to suspect it but its always a suspect
So first conclusion is - wait a minute - I think the cap tubes wrong.

Its a differant scenario to the one you have pointed out but it is Aarons scanario.

Chef

To take this step-by-step:

We don't have enough info for dT's, but since dT must be less than TD and TD's are low, we know that evap airflow is not the problem, nor is condenser airflow... although these conclusions may change when the system is actually handling a decent load.

The subcooling is not excessive (liquid not backing up into the condenser).

Next on the checklist would be low superheat. We must assume that the system has no form of suction/liquid heat exchange, therefore 1'F SH is much too low and is flooding the compressor. Refrigerant must be removed in order to raise superheat at the compressor inlet.

The troubleshooting procedure must halt at this point until the excess refrigerant is removed. Once this is done we need a whole new set of temperature and pressure measurements and (this is the part that everyone hates) we then need to start the troubleshooting procedure all over again, taking it from the top and checking the airflows.

icecube51

01-12-2009, 07:44 PM

why not try to put a sightglass just before the captube and see what happens.
if the glass is just "a point" full, take notes and put them in a PTchart.
thats wat i would do.

Ice

Chef

02-12-2009, 02:22 AM

why not try to put a sightglass just before the captube and see what happens.
if the glass is just "a point" full, take notes and put them in a PTchart.
thats wat i would do.

Ice

Ice, yes its a great idea.

Another system we are working on (not the Aaron system) has a sight glass just before the cap entrance. As the unit gets colder and colder so the outlet side of the sight glass shows less and less of the hole covered by liquid. At the end of the pulldown maybe only the bottom 5% is covered by liquid and rest is gas entrainment. Needless to say the system is not performing well but what a totally excellent tool to see what is happening.

The main drawback is that a service engineer could be fooled into charging until it is solid liquid which would be a disaster, instead of putting in a weighed charge. Usful for R&D work and commissioning of a new build though.

Although you see the glass shows a liquid/gas mix there is no way yet to get an actual value of X from it.

Chef

desA

02-12-2009, 03:35 AM

Another system we are working on (not the Aaron system) has a sight glass just before the cap entrance. As the unit gets colder and colder so the outlet side of the sight glass shows less and less of the hole covered by liquid. At the end of the pulldown maybe only the bottom 5% is covered by liquid and rest is gas entrainment. Needless to say the system is not performing well but what a totally excellent tool to see what is happening.

This does not sound right at all. It would surely be extremely difficult to ask a cap tube to manage this?

Is this flashing situation occurring at the same time as the SH~0K situation?

mad fridgie

02-12-2009, 03:55 AM

desA

02-12-2009, 04:05 AM

At this point refrigerant mass is held up in the evap, and very little in the condensor hence low SH and Liquid sub cooling.

Ok, but, is this a safe operating situation for the compressor? Would it not perhaps be useful to have a receiver on the hp side, to retain some of the liquid during the off-times?

I'd imagine a vertical riser outlet of evap could help a little, to keep some of the liquid out of the compressor inlet.

Seems a little precarious.

Chef

02-12-2009, 04:18 AM

This does not sound right at all. It would surely be extremely difficult to ask a cap tube to manage this?

Is this flashing situation occurring at the same time as the SH~0K situation?

This is another system, not anything to do with previous posts except as an example to Ice's sight glass proposal. But as it is a general discussion and this system does need sorting we can have this as the sight glass system.

The SH is somewhere in the region of 3 to 5K but has not been accurately determined on the sight glass system. No measurable SC and suspected as being SC=0.

Its a complex situation where we see the exit to the tube is sonic so it has a maximum flow rate and is limiting the through put. As a consequence the suction pressure falls well below its 'normal' and the mass flow goes way down. So as you say it is not normal. Once the tube oulet goes sonic no lowering of the suction pressure will increase the flow rate - its just the laws of sonic flow.

The question is, now the cap tube has entrained gas at the inlet the gas velocities at the outlet are much larger than if it had more liquid in the inlet, so the outlet becomes sonic and limiting. So is it the high XC entering the tube or the sonic limitation that is causing the problem. Its a bit of a chicken and egg situation.

Normally the first step would be increase the tube diameter and select a new appropriate length. This should stop the sonic limitation and allow a new set of measurements to be taken.

Chef

Gary

02-12-2009, 05:20 AM

This is another system, not anything to do with previous posts except as an example to Ice's sight glass proposal. But as it is a general discussion and this system does need sorting we can have this as the sight glass system.

The SH is somewhere in the region of 3 to 5K but has not been accurately determined on the sight glass system. No measurable SC and suspected as being SC=0.

Its a complex situation where we see the exit to the tube is sonic so it has a maximum flow rate and is limiting the through put. As a consequence the suction pressure falls well below its 'normal' and the mass flow goes way down. So as you say it is not normal. Once the tube oulet goes sonic no lowering of the suction pressure will increase the flow rate - its just the laws of sonic flow.

The question is, now the cap tube has entrained gas at the inlet the gas velocities at the outlet are much larger than if it had more liquid in the inlet, so the outlet becomes sonic and limiting. So is it the high XC entering the tube or the sonic limitation that is causing the problem. Its a bit of a chicken and egg situation.

Normally the first step would be increase the tube diameter and select a new appropriate length. This should stop the sonic limitation and allow a new set of measurements to be taken.

Chef

Does the sight glass system have a fixed speed compressor or variable speed?

Gary

02-12-2009, 05:35 AM

desA

02-12-2009, 05:53 AM

Or... the cap tube could be a little more restrictive, assuming you are designing for the current evap load and cond load.

Judging from Chef's comments above, it seems like there is already a sonic choke on the current cap tube. Can't get much more restrictive than that... unless you can limit the vapour entry into the orifice.

Seems like a new orifice selection is due.

Gary

02-12-2009, 06:08 AM

I have no idea what a sonic choke is.

desA

02-12-2009, 06:46 AM

http://en.wikipedia.org/wiki/Choked_flow

This should explain the concept of a sonic choke.

Chef

02-12-2009, 06:49 AM

Adding refrigerant would give you some subcooling, but would probably drop the superheat down... unless the heat load were increased (more airflow) to add SH, in which case it should balance out.

Or... the cap tube could be a little more restrictive, assuming you are designing for the current evap load and cond load.

A sonic limitation is where the local velocity in the pipe, usually the end of the cap tube, is at the speed of sound in the medium. The gas cannot go faster than the speed of sound unless it is in a specially designed nozzle. Also if the suction pressure is reduced it has no influence on the flow, it is effectively at its maximum throughput.

So with sudden stop in flow I think adding gas may not help - it may put a little more in the condensor but it is not going to cure this problem.

The cap tube is already too restrictive and so needs to be made of a larger diameter and a new appropriate length.

We did try changing the gas charge a little and no effect, I suspect the sonic limitation is starving the compressor and so reducing flow into the condenser and now we have a classic scenario for the XC to be around 0.3 or even 0.4 - very strange indeed.

Chef

Gary

02-12-2009, 06:56 AM

A sonic limitation is where the local velocity in the pipe, usually the end of the cap tube, is at the speed of sound in the medium. The gas cannot go faster than the speed of sound unless it is in a specially designed nozzle. Also if the suction pressure is reduced it has no influence on the flow, it is effectively at its maximum throughput.

So with sudden stop in flow I think adding gas may not help - it may put a little more in the condensor but it is not going to cure this problem.

The cap tube is already too restrictive and so needs to be made of a larger diameter and a new appropriate length.

We did try changing the gas charge a little and no effect, I suspect the sonic limitation is starving the compressor and so reducing flow into the condenser and now we have a classic scenario for the XC to be around 0.3 or even 0.4 - very strange indeed.

Chef

Apparently there is enough refrigerant getting through to handle the heat load and end up with a few degrees of superheat at the coil outlet, so how restrictive can it be?

If you are reducing the suction pressure by speeding up the compressor, then you might expect increased flow. If you are reducing the suction pressure by reducing the heat load, then the flow would be reduced.

mad fridgie

02-12-2009, 07:28 AM

A cap is not really a control device, as itself does not change, so you can not compare to any form of modulating valve. It is the inlet and outlet conditions that determine the caps performance not the other way around.
If you use a reciever you are unable to flood the condensor (without loosing lquid seal)
In this case sizing and charge needs to be dertermine by max operating conditions (when less than this performance and efficiency go out of the window.)
The unit is over sized as the OP wants a 30% run time.
The cap needs to be larger diameter and longer, in contact with suction.
An accumulator can fitted if an over charge is required. (always a safe option) as far RD goes I would install one (this could save the compressor)
Particular with danfoss compressors.

Gary

02-12-2009, 08:02 AM

The key to any experiment is to control all of the variables except the one being tested.

But here we are discussing one component in isolation as if the rest of the system had no effect on it.

desA

02-12-2009, 08:51 AM

The cap tube is already too restrictive and so needs to be made of a larger diameter and a new appropriate length.

We did try changing the gas charge a little and no effect, I suspect the sonic limitation is starving the compressor and so reducing flow into the condenser and now we have a classic scenario for the XC to be around 0.3 or even 0.4 - very strange indeed.

I wouldn't be surprised to find the system stabilising at solutions outside of the traditional refrigeration operating window.

Each major item in the system is inherently non-linear in its response - even the piping. The system can definitely stabilise at such local solutions. The problem is how to kick it off there, & back to a vapour compression cycle. (I say 'kick', because sometimes it will not willingly flip over to the next solution - it needs coercion.)

I would suggest that the condenser be boosted in your experiment. Take a few fans & blast the thing with cold air - observe. If this helps, then your condenser is undersized for the climatic conditions in your hot, humid, wet part of the world.

Add an exit lift from the condenser, to force a liquid seal at exit.

Get those in place, then begin backing off on other things. Mass charge either low, or high, can play incredible tricks on you.

Chef

02-12-2009, 09:09 AM

Apparently there is enough refrigerant getting through to handle the heat load and end up with a few degrees of superheat at the coil outlet, so how restrictive can it be?

If you are reducing the suction pressure by speeding up the compressor, then you might expect increased flow. If you are reducing the suction pressure by reducing the heat load, then the flow would be reduced.

No the system is running poorly and that is why we are looking at it. It should be producing at least 30 to 40 percent more cold. The main reason we beleive is that as the XC is around 0.3 or maybe even 0.4 we are only getting about 60 percent of the cooling possible if we had condensed all the gas.

Also the flow rate is down as the tube is sonic so we have less Kj/Kg of cooling and less Kg/s passing around the system - so combining these it gives a really poor performance.

The suction pressure is not being reduced by RPM change or heat load, the suction pressure is being reduced because the tube has gone sonic at the end.

Chef

Chef

02-12-2009, 09:30 AM

A cap is not really a control device, as itself does not change, so you can not compare to any form of modulating valve. It is the inlet and outlet conditions that determine the caps performance not the other way around.

Thankyou Mad, this is exactly the case and it the conditions in the condenser and evap that change to meet a system balance. The main item is obviously pressure and comes SC and then comes x=0.1 or 0.2 etc. The evap has much less control because it is limited to pressure only.

If you use a reciever you are unable to flood the condensor (without loosing lquid seal)
In this case sizing and charge needs to be dertermine by max operating conditions (when less than this performance and efficiency go out of the window.)
The unit is over sized as the OP wants a 30% run time.
The cap needs to be larger diameter and longer, in contact with suction.
An accumulator can fitted if an over charge is required. (always a safe option) as far RD goes I would install one (this could save the compressor)

Most systems have a little accumulator on the outlet of the evap that must help (production units), but for and new designs and cut it try it stuff an accumulator is a good idea, cant hurt but it might save.
Particular with danfoss compressors.

The point about the tube is well meritted and after all it is only a tube. It will pass whatever you throw at it and how fast it passes it depends on the pressures.

Chef

Gary

02-12-2009, 02:55 PM

No the system is running poorly and that is why we are looking at it. It should be producing at least 30 to 40 percent more cold. The main reason we beleive is that as the XC is around 0.3 or maybe even 0.4 we are only getting about 60 percent of the cooling possible if we had condensed all the gas.

Also the flow rate is down as the tube is sonic so we have less Kj/Kg of cooling and less Kg/s passing around the system - so combining these it gives a really poor performance.

The suction pressure is not being reduced by RPM change or heat load, the suction pressure is being reduced because the tube has gone sonic at the end.

Chef

Compressor manufacturers require at least 8.5K SH at the compressor inlet. You have said that your system has 3-5K SH. You have also said that the compressor was starved. Which is it? Is the SH borderline low (3-5K) or is the SH high (starving the compressor)?

Or perhaps you are talking about coil outlet SH and there is heat being added between the coil outlet and the compressor inlet? For diagnostic purposes, we should be talking about the compressor inlet SH, not the coil outlet SH.

You seem obsessed with the "sonic limitation" theory. I have seen no evidence of any such limitation.

Gary

02-12-2009, 03:31 PM

The suction pressure is not being reduced by RPM change or heat load, the suction pressure is being reduced because the tube has gone sonic at the end.

Any reduction in suction pressure gives you a reduction in mass flow, the single exception being increased compressor RPM... yet you seem to be expecting the mass flow to increase as suction pressure is reduced?

Chef

03-12-2009, 01:05 AM

Compressor manufacturers require at least 8.5K SH at the compressor inlet. You have said that your system has 3-5K SH. You have also said that the compressor was starved. Which is it? Is the SH borderline low (3-5K) or is the SH high (starving the compressor)?

I mentioned we think the SH is around 3-5K but not sure as accurate measurements were unavailable. I understand your point that if the compressor is starved then the SH should be higher and worth getting better measurements. Its a new build from a hotch potch of parts and so I dont think they ever thought about what the manufacturers would like in terms of SH.

Or perhaps you are talking about coil outlet SH and there is heat being added between the coil outlet and the compressor inlet? For diagnostic purposes, we should be talking about the compressor inlet SH, not the coil outlet SH.

Before the suction receiver.

You seem obsessed with the "sonic limitation" theory. I have seen no evidence of any such limitation.

Not obsessed with it but as its happening and you dont accept it it is in an attempt to get you too realise it could be a phenomena and then offer some solid advice.

If you google 'choked flow in capillary tubes' you will get pages of references alluding to this, both from fridge companies and research institutes.

Chef

Chef

03-12-2009, 01:15 AM

yet you seem to be expecting the mass flow to increase as suction pressure is reduced?

No I dont expect that and did not say it.

If the tube has gone sonic then it cant pass any more refrigerant, the compressor suction requirement is higher than the tubes flowrate and so the suction pressure falls till equilibrium is reached.

Its part of the characteristics of choked flow (sonic) and the references via google in the previous post explain it fully

Chef

Gary

03-12-2009, 01:43 AM

mad fridgie

03-12-2009, 02:31 AM

If refrigerant was added until the subcooling was about 15F/8.5K and the superheat was then high, you might be able to convince me that the flow is restricted, sonic or otherwise. With zero SC, there is not enough refrigerant there to restrict.
The ambient is 50C, SCT is 55C then max possible SC could only be 5C (unless an external force is used)
Refrigeration in mathmatical terms is a " circular reference" there is not a start or finish point.
So at some point you start with your best guess, to get things started. (we do this without even thinking)
In this case best guess seems to be rather out, thus we are unable to resolve the problem. Regardless of where you start.

Gary

03-12-2009, 02:42 AM

The ambient is 50C, SCT is 55C then max possible SC could only be 5C (unless an external force is used)
Refrigeration in mathmatical terms is a " circular reference" there is not a start or finish point.
So at some point you start with your best guess, to get things started. (we do this without even thinking)
In this case best guess seems to be rather out, thus we are unable to resolve the problem. Regardless of where you start.

Good point... but we are talking about two different systems. You are talking about Aaron's system and I am talking about Chef's system.

Chef has not told us what the temps/pressures are for his system. But he knows that the cap tube is sonically limited.

mad fridgie

03-12-2009, 03:55 AM

I have no knowledge of sonic chocking, but have come across damage done to valve seats, from sonic explosions (may have used the wrong term there), so presume that the 2 are the same/similar thing

desA

03-12-2009, 03:56 AM

If refrigerant was added until the subcooling was about 15F/8.5K and the superheat was then high, you might be able to convince me that the flow is restricted, sonic or otherwise. With zero SC, there is not enough refrigerant there to restrict.

Chef's situation is far worse than this.

1. Under-performing condenser;
2. Condensation process incomplete, still 30-40% vapour remaining;
3. Vapour & condensate exit condenser, move to cap tube;
4. Vapour component moves at high velocity through cap tube;
5. Vapour component velocity reaches sonic speed (local speed of sound), cap tube chokes;
6. Once cap tube chokes, no further increase in mass flow will occur;
7. Compressor keeps pulling - lowering suction pressure until system equilibrium is met.

In the end, the system settles with a condenser not completely condensing; a locked fixed choked mass flow;
lower than expected suction pressure; under-performing system.

This is not the vapour compression cycle as most RHVAC personnel would know. It is an intermediate system.

The question then should be:
How to move Chef's system towards the recognised vapour compression cycle?

Ok, gentlemen? :)

Chef

03-12-2009, 04:07 AM

Good point... but we are talking about two different systems. You are talking about Aaron's system and I am talking about Chef's system.

Chef has not told us what the temps/pressures are for his system. But he knows that the cap tube is sonically limited.

Actually there are 3 systems, my system which I posted early 3 operating conditions.

Aarons system and a third system which is called the sight glass system.

It is the sight Glass system that has a sonic tube. The Dx on this is 150psig and the Sx is 7psig, ambient is 32 to 33 and fan cooled condenser, SC=0 and SH not too sure but some.

Here is nice quote from an excellent tutorial on cap tubes.

Critical charge is a definite amount of refrigerant that is put into the refrigeration system so that in the eventuality of all of it accumulating in the evaporator, it will just fill the evaporator up to its brim and never overflow from the evaporator to compressor. The flooding of the evaporator is also a transient phenomenon, it cannot continue indefinitely. The system has to take some corrective action. Since the capillary tube feeds more refrigerant from the condenser, the liquid seal at the condenser exit breaks and some vapour enters the capillary tube. The vapour has a very small density compared to the liquid; as a result the mass flow rate through the capillary tube decreases drastically.
http://www.nptel.iitm.ac.in/courses/Webcourse-contents/IIT%20Kharagpur/Ref%20and%20Air%20Cond/pdf/R&AC%20Lecture%2024.pdf
This the link to full tutorial and is very good reference.

It also discusses choked flow (sonic flow)

Gary you mention 'With zero SC, there is not enough refrigerant there to restrict.'

Well as the above quote suggest any time gas entrainment happens the mass flow rate decreases drastically. This is a pretty key point and once x=0.1 or x=0.2 at the inlet to cap is used in a diagnosis things start to be more apparent.

Chef

desA

03-12-2009, 04:12 AM

If you really want to beef up your condensing capacity on the research unit, install an inline de-superheater (water, or air-cooled) & see what happens.

mad fridgie

03-12-2009, 04:21 AM

Well, that is what I thought I said, (ealier thread) perhaps thats why I am not in the business of writing tutorials, explains it very well!

Gary

03-12-2009, 04:44 AM

Actually there are 3 systems, my system which I posted early 3 operating conditions.

Aarons system and a third system which is called the sight glass system.

It is the sight Glass system that has a sonic tube. The Dx on this is 150psig and the Sx is 7psig, ambient is 32 to 33 and fan cooled condenser, SC=0 and SH not too sure but some.

Here is nice quote from an excellent tutorial on cap tubes.

Critical charge is a definite amount of refrigerant that is put into the refrigeration system so that in the eventuality of all of it accumulating in the evaporator, it will just fill the evaporator up to its brim and never overflow from the evaporator to compressor. The flooding of the evaporator is also a transient phenomenon, it cannot continue indefinitely. The system has to take some corrective action. Since the capillary tube feeds more refrigerant from the condenser, the liquid seal at the condenser exit breaks and some vapour enters the capillary tube. The vapour has a very small density compared to the liquid; as a result the mass flow rate through the capillary tube decreases drastically.
http://www.nptel.iitm.ac.in/courses/Webcourse-contents/IIT%20Kharagpur/Ref%20and%20Air%20Cond/pdf/R&AC%20Lecture%2024.pdf
This the link to full tutorial and is very good reference.

It also discusses choked flow (sonic flow)

Gary you mention 'With zero SC, there is not enough refrigerant there to restrict.'

Well as the above quote suggest any time gas entrainment happens the mass flow rate decreases drastically. This is a pretty key point and once x=0.1 or x=0.2 at the inlet to cap is used in a diagnosis things start to be more apparent.

Chef

Shall I assume R134a?

The quote says exactly what I have been saying. The system has stopped performing because it has reached its charge limit. If you add refrigerant it will have a different charge limit and will drop to a lower heat load/higher ambient combination before it once again has zero subcooling and stops performing.

Chef

03-12-2009, 04:46 AM

Chef's situation is far worse than this.

1. Under-performing condenser;
2. Condensation process incomplete, still 30-40% vapour remaining;
3. Vapour & condensate exit condenser, move to cap tube;
4. Vapour component moves at high velocity through cap tube;
5. Vapour component velocity reaches sonic speed (local speed of sound), cap tube chokes;
6. Once cap tube chokes, no further increase in mass flow will occur;
7. Compressor keeps pulling - lowering suction pressure until system equilibrium is met.

In the end, the system settles with a condenser not completely condensing; a locked fixed choked mass flow;
lower than expected suction pressure; under-performing system.

This is not the vapour compression cycle as most RHVAC personnel would know. It is an intermediate system.

The question then should be:
How to move Chef's system towards the recognised vapour compression cycle?

Ok, gentlemen? :)

Well that is the very best synopsis I have seen and is the situation. Thankyou DesA.

As you say RHVAC'ers dont normally see these sort of conditions and the complexity of mixed gas and liquid in the entire cycle causes many problems.

It is not actually my system - this is the sight glass system and is for another person so just trying to help him out.

Now you have posted the real scenario we may get to move forward.

Chef

Gary

03-12-2009, 05:02 AM

Gary

03-12-2009, 05:09 AM

Chef's situation is far worse than this.

1. Under-performing condenser;
2. Condensation process incomplete, still 30-40% vapour remaining;
3. Vapour & condensate exit condenser, move to cap tube;
4. Vapour component moves at high velocity through cap tube;
5. Vapour component velocity reaches sonic speed (local speed of sound), cap tube chokes;
6. Once cap tube chokes, no further increase in mass flow will occur;
7. Compressor keeps pulling - lowering suction pressure until system equilibrium is met.

In the end, the system settles with a condenser not completely condensing; a locked fixed choked mass flow;
lower than expected suction pressure; under-performing system.

This is not the vapour compression cycle as most RHVAC personnel would know. It is an intermediate system.

The question then should be:
How to move Chef's system towards the recognised vapour compression cycle?

Ok, gentlemen? :)

What you have described here is what happens when there is no subcooling.

Chef

03-12-2009, 05:16 AM

Shall I assume R134a?

The quote says exactly what I have been saying. The system has stopped performing because it has reached its charge limit. If you add refrigerant it will have a different charge limit and will drop to a lower heat load/higher ambient combination before it once again has zero subcooling and stops performing.

So a critically charged system that has the evap full of charge to its limit needs more gas? I think not.

DesA has put it in a nutshell and there is much more to consider than just adding charge. It is not a charge issue at all but cap tube and evaps and condensers not balanced to perform correctly.

I said earlier we added charge and it made difference.

Chef

Gary

03-12-2009, 05:23 AM

So a critically charged system that has the evap full of charge to its limit needs more gas? I think not.

How do you know the evap is full to its limit?

desA

03-12-2009, 05:42 AM

...there is much more to consider than just adding charge. It is not a charge issue at all but cap tube and evaps and condensers not balanced to perform correctly.

It would be helpful if the condenser, evaporator & compressor operating curves could be set up, with the appropriate balance curve for the cap tube superimposed & evap air pull-down curve.

We could start with what is known about the size of each part, operating conditions - design & measured, & slowly progress these towards some useful system insights.

This is a lovely challenge - a hybrid between design & commissioning (set-up).

We could tackle all 3 systems - give each one a name e.g. Chef1, Chef2, Aaron_k - each with its design, physical & measured parameters. I think we'd all learn a great deal from the process.

Gary

03-12-2009, 07:57 AM

It would be helpful if the condenser, evaporator & compressor operating curves could be set up, with the appropriate balance curve for the cap tube superimposed & evap air pull-down curve.

We could start with what is known about the size of each part, operating conditions - design & measured, & slowly progress these towards some useful system insights.

This is a lovely challenge - a hybrid between design & commissioning (set-up).

We could tackle all 3 systems - give each one a name e.g. Chef1, Chef2, Aaron_k - each with its design, physical & measured parameters. I think we'd all learn a great deal from the process.

Okay... let's start with the full evaporator at zero SC.

If we increase the airflow through the evap it can hold more liquid because it is boiling it off faster. Conversely, if the airflow is reduced the evap will hold less liquid.

On the other side of the system, if the airflow through the condenser is reduced the pressure will be elevated, pushing more liquid through the cap tube, thereby hastening the moment where there will be zero SC. IOW, the system will reach critical charge at a lower ambient temp.

It is absolutely essential that the airflow through both coils is checked before making any judgements concerning the charge or the cap tube.

mad fridgie

03-12-2009, 09:50 AM

Okay... let's start with the full evaporator at zero SC.

If we increase the airflow through the evap it can hold more liquid because it is boiling it off faster. Conversely, if the airflow is reduced the evap will hold less liquid.

On the other side of the system, if the airflow through the condenser is reduced the pressure will be elevated, pushing more liquid through the cap tube, thereby hastening the moment where there will be zero SC. IOW, the system will reach critical charge at a lower ambient temp.

It is absolutely essential that the airflow through both coils is checked before making any judgements concerning the charge or the cap tube.
Hi Gary, I think you have this wrong (or maybe just worded wrong) With more airflow you will hold less liquid, this because the liquid is boiling so the Liquid to vapor % changes, how ever the mass flow through the evap is going to increase, the reverse happen with less air flow.
The very first stage is just a steady state start point
"best guess" Which I would use max ambient for condenseing, and just above chilled chamber design temp.
Using these we can then use manufactor data, to balance the refrige system Comp, evap and cond.
We should then have a set of conditions to which we are able to select a cap with. This should bring us within +/- 15% of perfect. I would then start the system, slowly adding refrigerant (ensuring that Ct is close to design) ignore Te during pull down. When close to design take measurements, make adjustment with charge or cap length. When at this point is correct (over a time cycle). We then run the system out design conditions, and see if the changes within the process are within acceptable limits. At this stage we are not optomising the system in any form, we are just protecting the equipment. If optimum is required then you do not use a cap.
Forgot to mention speed, this is done in the last process

Chef

03-12-2009, 10:36 AM

By adjusting the charge, we adjust the heat load/ambient temp combination which causes zero SC at the cap tube inlet.

When the cap tube inlet SC is zero, the SH should be almost low enough to flood the compressor... but not quite.

If the SH is high at the zero SC condition, then the cap tube is too restrictive. If it is low, the cap tube is not restrictive enough.

When the SC is zero and the SH almost zero - perfect - your at the point of where the next phase begins.

All the charge is in the evap, it continues to cool and so the pressure slowly falls.

The SC is still zero but now as the Sx falls the mass flow rate falls. Less gas enters the condensor.

But the pressure it enters at is only just enough to condense some of the gas and some is left to pass into the tube. It is how almost all freezer systems work that are equiped with a cap tube.

So the only way to restict the flowrate is to have some gas and some liquid - not all liquid in the tube.

If we can somehow convince you this is a real world situation and not all systems can have or are designed to have 5.5 to 8K SC then it will be day of days. Here is a nice graph which shows how the tube flow decreases with reducing SC and then after it hits 0 it goes into the quality mode. It starts at x=0 when SC=0 and then steadily increases as the mass flow falls.

Of course the SC will always measure zero even when the quality has positve values. Maybe thats why no-one has seen it from measurements.

3202

Chef

03-12-2009, 12:09 PM

Mad and DesA
If we are to progress forward we need to get Gary to accept these concepts. Not every system on the planet has to have loads of SC!

You are both on the right track and see it as it is and it will be a great discussion.

However.

Tommorrow I leave for uncharted waters where they have no internet or any sort of life at all. I shall be there for a month and may be able to get a signal on the odd occasion.

I am sorry but I will not be able to continue in this lively discussion - bit of a bugger really as was just getting to the good stuff.

Maybe we can start it all rolling again in January?

Duty calls I'm afraid.

Chef

PS Best wishes for Christmas and the New Year

aaron_k

03-12-2009, 02:57 PM

Chef, Mad, DesA, Gary,

I sort of feel bad to have kicked the hornets nest (inadvertently) but I think everyone appreciates the discussion; I certainly do.

Some news and updates on my system. Chef is correct - this is a new machine, so its actual design is what is in question. I have contacted Richard Farr, the originator of the R&R Supply simulation software used by our contractor. He's likely the most experienced refrigeration engineer in the world (he's 91 and still active). You can read about him here:

www dot appliancemagazine dot com/editorial.php?article=967&zone=1&first=

(sorry - I'm still not allowed to post URL's - fill in the dots)

He has re-run the analysis for our system and determined that as expected the cap tube length was way too short. The new value is close to what other published literature would suggest. We're in the process of fitting the new cap tube for testing.

I mentioned to him that the Danfoss cap tube program was used to pick the cap tube length. He replied that the published data in the ASHRAE fundamentals is very accurate - there's another source for everyone to refer to.

By the time Chef returns I will have the results of the changes we've made.

Thanks again for your passion for this topic. I will update as new data presents itself. Happy holidays to everyone.

Aaron_K

desA

03-12-2009, 04:04 PM

Chef, you take care on your travels. We'll have lots to discuss when you return in the new year. Excellent topic.

We can all work on Aaron_k's system in your absence & apply the findings to your system on your return.

I suspect that each type of condenser system will have its maximum sustainable SC. For instance, plate-type condensers seem to prefer SC less than 4K. Some run comfortably at 1-2K. A lot will have to do with the liquid seal, at the condenser discharge, as MF has mentioned previously (not sure if it was this thread).

Gary

03-12-2009, 05:13 PM

If we can somehow convince you this is a real world situation and not all systems can have or are designed to have 5.5 to 8K SC then it will be day of days.

I have never said that all systems under all conditions should have 5.5-8.5K SC. This is obviously not the case.

Our entire disagreement centers around your insistence that your system is sonically challenged. This may or may not be the case and you have presented zero evidence to demonstrate that it is so.

You say that the cap tube cannot increase its flow because it is sonically limited. I say that given increased SC at its inlet the flow would indeed increase.

Gary

03-12-2009, 05:45 PM

Hi Gary, I think you have this wrong (or maybe just worded wrong) With more airflow you will hold less liquid, this because the liquid is boiling so the Liquid to vapor % changes, how ever the mass flow through the evap is going to increase, the reverse happen with less air flow.

Apparently I have explained this badly. My entire point was that we cannot determine if the charge is right or the cap tube is right without first determining that the airflow is right.

And just because the SC has dropped to zero does not prove that the evap is full. For that we need to know the compressor inlet superheat.

mad fridgie

03-12-2009, 07:46 PM

Hi Aaron, there nothing like a hornets nest to keep the mind simulated, if we all agreed all the time it would be a bit boring!
Good luck with your trails.
Chef I hope you have smooth waters.
Des and gary clash horns later

aaron_k

19-01-2010, 12:02 AM

Hello everyone,

I have an update for you on my predicament. We increased the cap tube length and reduced the condenser size, while increasing charge to match the maximum compressor rated charge size. I ended up finding **** Farr, the founder of R&R Supply (no one has heard of them). He's 91, and been a refrigeration engineer for a long long time... anyway....

Our system is now able to operate for over 3X the original duration while running at the lowest compressor speed (2000 rpm) - the original system wouldn't even cool if run below 2800 rpm, and used 3X the power. We're really thrilled with the results. I'm literally within 5% of where I need to be to prove out the design, and we think it can be achieved with a little more adjustment to the software.

So it turns out that the published cap tube lengths are pretty close to what we needed. The charge size is also vital, as we have our vendor telling us that 40% less refrigerant is the way to go, but the resulting lack of subcooling ends up causing the system to consume 50% more power.... we have one of each configuration running side by side, with the data to prove it. And Danfoss is OK with our over all system volume to prevent compressor damage at lower operating ambients.

So now we can keep 60 pounds of material cold (5C) while running in a 52C ambient for 24 hours on one car battery, off the grid... pretty nice.

I want to say thank you very sincerely for everyone's help. We couldn't have really understood the cause and effect without your passionate discussion.

All the best,

aaron_K