Cap tube sizing rule

Good question. unfortunatly they are not designed for this. The input parameters are condensing pressure, degree of subcooling and the physical/thermodynamic properties are taken at saturation point for inlet temperature. I could run them with the inlet temperature 28K lower than the condensing temperature but the results would be kinda meaningless. If you specify condensing temperature and degree of subcooling I can run the models and show you the results.

Im currently working on another model which should be a bit more robust and will take into account the non-adiabatic effect. It s gonna take awhile, you can see why if you read the paper.
http://www.russellpetrie.plus.com/Do...pTubeModel.pdf

im not sure if thats connected to this but here we go
http://www.captaincascade.com/forums...3&st=0#entry91

Perhaps I approach things differently than most. I would load up evaps to find the 'drop off the cliff' point for various lengths of cap tube. This would give me both the centerpoint and the span curve, with which I could develop my "useful tool". It would then be my inclination to let others figure out why it works. For me, the useful tool is the goal.

Thats good in practice, but I dont have the equipment to do this so building a theoretical model is a lot simpler. All I need is books and a pc, heck even just a pc and the internet :D

Quote:

---

If you specify condensing temperature and degree of subcooling I can run the models and show you the results.

---

Okay, lets try 30C condensing and 28K subcooling. All else the same as previous.

Code:

---

Tube Length (m)        Mass Flow (kg/s)        Cooling capacity (watts)       
1.5                0.002562663                276.8        219.4
1.75                0.002378424                256.9        205.9
2                0.00222957                240.8        194.8
2.25                0.002106024                227.5        185.6
2.5                0.002001321                216.1        177.7
2.75                0.001911097                206.4        170.9
3                0.001832285                197.9        164.9
3.25                0.001762658                190.4        159.5
3.5                0.001700555                183.7        154.7
3.75                0.001644707                177.6        150.4
4                0.001594126                172.2        146.4
4.25                0.00154803                167.2        142.8
4.5                0.001505792                162.6        139.5
4.75                0.001466898                158.4        136.4
5                0.001430929                154.5        133.6

---

First model is Wolf, second is Choi.

Ok, edited the numbers. I will look back over the last sets and see if any have changed

I seem to have made a mistake in the autocalculation script, Choi's numbers are out. Give me 5 minutes. Wolf's should be ok though.

After the vapor condenses, it starts gaining subcooling, limited by the temperature of the medium with which it is exchanging heat.

In order to know the correct amount of subcooling to input for the model, we need to know at what point we are measuring subcooling, as this value can and does increase, especially if there is suction/liquid heat exchange.

I assume they are asking for subcooling at the cap tube entrance. It would seem more realistic to use a subcooling value midway through the cap tube. That's why I suggested 28K.

Subcooling is measured at the entrance. These models were based on adiabatic capillary tubes unfortunatly, so no suction line heat exchange is accounted for.

I have sorted out the numbers in Choi's model estimation. Starting from here:
http://www.xtremesystems.org/forums/...807#post375807

I've changed the values in the tables. These should be correct now. My bad :slapass:

To complicate matters more, we are finding that due to uneven heat distribution (heat source concentrated at center of block face), a somewhat higher superheat is optimal. This would change not only the superheat input, but also the average (midway) subcooling within the cap tube, if/when suction/liquid heat exchange is assumed.

Are they asking for superheat at the evap outlet or compressor inlet?

Evap outlet...

The optimum evap outlet superheat varies with evap block design, as well as orientation (vertical or horizontal), but I would assume at least 10K superheat for our model. This, in addition to imperfect heat transfer, would also change my assumed midtube average subcooling.

Let's try the model with 30C condensing, 25K subcooling, and 10K superheat.

Russel have you tried the calcs holding the mass flow constant and watched the changes in wattages for different lengths of captube? We are assuming a liquid column on the inlet side of the cap tube but systems that are currently in the field assume two phase flow at both the inlet and outlet. This occurs when a system is under a load and properly charged, Like during start up or a low temp reachin freezer. To minimize system charge the manafactures take this into account.

Thanks for the input chilly1. Im working on a model that uses 2-phase flow and single phase flow inside the capillary tube. As the pressure falls inside the capillary tube, the saturation pressure is reached and 2-phase flow begins before this happens we have only a liquid phase.

What you say about 2-phase flow at inlet, this would suggest that there is no subcooling. Is this the case? I dont think you can have sub-cooling at the cap-tube inlet and 2-phase flow all at the same time.

Quote:

---

Russel have you tried the calcs holding the mass flow constant and watched the changes in wattages for different lengths of captube?

---

The evaporator pressure is kept constant as well as superheat, thus at a fixed mass flow the wattages are constant. What would change would be condensor temperature/pressure.

In running systems in the field I see the cap tube being hot til it contacts the suctionline. I have been told by a manafacturers engineer that there is bubbles in the cap tube until it is subcooled by the suctionline.

The guy I talked to at hobart told me that for system effeciency they cool the gas from the time it leaves the compressor til it is subcolled by the suctionline and will not be fully condensed until it reaches the outlet of the cap tube.


So the design of the captube/condenser /compressor as a system. Displacment and condenser sizing as well as captube diameter and length all come into play this is a four sided coin.

The compressor displacment will play a large part in the mass flow so compressor size will greatly effect the cap tube size for a given load. On th eflip side if the cap tube is sized for a given load then using a larger compressor should give you colder temperatures.

Gary, got your numbers for you. Took a little while but I should be able to get you numbers much quicker from now on.

Code:

---

Tube Length (m)        Cooling capacity (watts)       
1.5                256.4        209.9
1.75                238.0        197.0
2                223.1        186.5
2.25                210.7        177.6
2.5                200.3        170.1
2.75                191.2        163.6
3                183.4        157.8
3.25                176.4        152.7
3.5                170.2        148.1
3.75                164.6        143.9
4                159.5        140.2
4.25                154.9        136.7
4.5                150.7        133.5
4.75                146.8        130.6
5                143.2        127.8

---

Finally got all the parts I need to convert my water chiller over to direct die. Made my cpu block and got some propane to replace the R134a. My only problem is that I had to get 0.036 cap tube as that was all they had available. My cpu i putting out 180W and I worked out I need around 11.5 ft of 0.031 cap so what is the 0.036 cap equivalent?

Alot :D

Quote:

---

Originally posted by chilly1
The guy I talked to at hobart told me that for system effeciency they cool the gas from the time it leaves the compressor til it is subcolled by the suctionline and will not be fully condensed until it reaches the outlet of the cap tube.

---

This may be the case, but why then when charging the systems we are working on do we look for a sightglass full of liquid? If there is 100% liquid it must be saturated or subcooled, more likely the latter due to above ambient condensing temperatures. In these systems we run to cool cpu's, we aim to have a liquid seal at the capillary entrance and a full sight glass if one is installed, therefore we are feeding subcooled liquid to the capillary. We only get 2-phase flow developing inside the capillary tube at a certain distance into it.

Quote:

---

Originally posted by chilly1
The compressor displacment will play a large part in the mass flow so compressor size will greatly effect the cap tube size for a given load. On th eflip side if the cap tube is sized for a given load then using a larger compressor should give you colder temperatures.

---

I agree that the compressor must be able to accomdate the volumetric flowrate in the suctionline due to evaporator load. If we install a larger than required compressor we can shift more volume but if mass flows are kept constant due to constant condensing temperature then the evaporator pressure must drop to allow the specific volume of the gas to increase hence volumetric flowrate increases to accomodate increase in compressor size. So as you say, we get lower temperatures with the same load :)

Quote:

---

This may be the case, but why then when charging the systems we are working on do we look for a sightglass full of liquid? If there is 100% liquid it must be saturated or subcooled, more likely the latter due to above ambient condensing temperatures. In these systems we run to cool cpu's, we aim to have a liquid seal at the capillary entrance and a full sight glass if one is installed, therefore we are feeding subcooled liquid to the capillary. We only get 2-phase flow developing inside the capillary tube at a certain distance into it.

---

In almost all cap tube refrigeration systems everywhere, there is 2 phase flow at the entrance to the cap tube, 3-5K subcooling would be typical, and sightglasses are not used. I assume people around here use sightglasses on cap tube systems for the gee whiz effect of watching the refrigerant flow... LOL

The notion that subcooling cannot occur in 2 phase is one of the great twin misconceptions of refrigeration, the other being that superheat cannot occur in 2 phase. These can and in fact always do occur in all systems. These come under the heading of: Nice theory, but it doesn't work that way.

Saturation exists where liquid and vapor contact each other. It is entirely possible to have a vapor bubble that is superheated at its core, saturated on its surface, and surrounded by subcooled liquid. It is also possible to have a liquid droplet that is subcooled at its core, saturated at its surface, and surrounded by superheated vapor. These systems are dynamic, which makes static gas laws rather difficult to apply.

Mr Deed, you could try 5m but that may not be enough, though it is the maximum stable length unfortunatly. Otherwise try and get some smaller tube. 0.028" would be ideal

How do you know there is 2-phase flow at the entrance? My sight glass's main purpose is the moisture indicator. The Gee Whizz effect is an added bonus :D

If we have a liquid seal at the entrance to the capillary tube, then the vapour phase should not enter. Its been shown that the longer the liquid column in the capillary tube the greater the mass flowrate hence greater capacity. If we have 2-phase flow from start to finish then we are not making the best use of the capillary tube.

Quote:

---

Gary, got your numbers for you. Took a little while but I should be able to get you numbers much quicker from now on.

---

Thanks, Russell. These numbers seems more credible, especially in describing the range curve. I'm thinking the column on the left probably needs to be shifted, though. Unfortunately, without real world testing, this is all speculation.

Quote:

---

How do you know there is 2-phase flow at the entrance? My sight glass's main purpose is the moisture indicator. The Gee Whizz effect is an added bonus

---

Your autocascade is a whole different ballgame. Solid liquid is necessary in order to ensure complete separation.

Quote:

---

If we have a liquid seal at the entrance to the capillary tube, then the vapour phase should not enter. Its been shown that the longer the liquid column in the capillary tube the greater the mass flowrate hence greater capacity. If we have 2-phase flow from start to finish then we are not making the best use of the capillary tube.

---

Having 2 phase flow at the entrance allows us to compensate for variations in cap tube length by adjusting the refrigerant charge. It gives us room for tuning, or more specifically, a means of tuning.