A cap tube design program for RE members by RE members.

[Chef]

Open letter to all members.

Having looked at many threads and also researched the web a lot it seems there is no real place to find a solid cap tube design program or diagram, there is just so much variance and hence the best advice is to cut, test, cut etc.

So what about the RE members writing their own program for use by all at RE.

I also see that Iceman and Peter_1 are not keen on software as it betrays any fundamental knowledge so an open source format is proposed. This allows any one to see the equations used and change them (in their own copy only) to see how modifications work.

The program would use condenser and evaporator conditions plus refrigerant type and mass flow/capacity.

All properties would be looked up on a table (taken from a PH diagram) and tabulated at the top of the program. This would allow anyone to enter new properties for whatever gas they plan to use. Maybe we could do it in Excel to start with.

The program would use adiabatic expansion, calculate the pressure drop for, say, every 5C drop in temp and the incremental length taking note of quality, plug flow, momentum etc etc.

So if anyone wants to volunteer to help with the thermodynamic equations, programming, testing and validation against actual operating systems please join in.

Maybe if it does get going we should use the PM board to communicate to free up the forums from the detailed or dreary stuff.

Any ideas? Anybody up for this?

Regards Chef.

[Raadster]

Im in full support of this, however I can't offer any help as im not an engineer.I can't help but think that there would be some program that already does this. Then again it would probably cost thousands to purchase.

[Peter_1]

...Having looked at many threads and also researched the web a lot it seems there is no real place to find a solid cap tube design program or diagram,

I also see that Iceman and Peter_1 are not keen on software as it betrays any fundamental knowledge so an open source format is proposed. This allows any one to see the equations used and change them (in their own copy only) to see how modifications work.

Myself and I'm sure USIceman uses software on a daily base, so your statement isn't completely true.
It's all those peoples asking just for software because they're to lazy to understand first the basics.

You're talking already in your post about almost all the variables you need to calculate a cap-tube. This learns me that you know understands already the basics and this is something completely different for me.

I've seen once a formula for calculating captubes in a book of Stoecker. I think it's almost impossible to translate this to Excel because some variables are in my opinion unknown or at least very uncertain.


But, haven't you found the free Dancap program yet?
Or the free software of Tecumseh/Unithé Hermétique?
Or the many tables circling around on our forum?

[Chef]

I understand your policy on the software and I agree with it so that is why I thought we could have data about the gas entered at the top of the program, at least it gets some involvement.

I did find and use the DANCAP software but it was way off. Here are the results for some tables.
All with the same parameters.
Dancap 11"
ACC tables 41.7"
My running system 94"
JB tables 144"

It is these results that make me feel uneasy with much of the published data.

The verification phase of this project predicts a cap tube length for a known set of conditions on a running system, we then measure the actual cap tube length to confirm (or otherwise) its validity and hence give confidence in its predictions. I do not have access to enough systems to be able to do this part of the project.

I dont know the Stoecker book but would love to see the relevant pages.
The solution I am trying does have some difficult variables but these can be obtained quite accurately by iteration, on the first try these did not converge to well and so it became unstable and very unweildy but a new solution method will help. It is tricky in Excel but it can be shared - if we go to a language like basic it might get complex to track changes. Maybe if it does work in Excel we could change to basic to make it more resilient?

Chef

[US Iceman]

I'm not against software. Like Peter said, I use different software programs daily to speed up my ability to do work. My issue with software is that people want it to do the work when they don't know how to do the calculations by hand. If they do not understand the calculations then they will not know what the program answers mean, or even if they are correct!

This sounds like an interesting project you are proposing, and the easiest way to probably do this is in a program called EES. It has the ability solve for unknown vaiables if sufficient data is available. This program also has the refrigerant properties built into to it, so that removes one big issue.

I might be wrong, but I think the problem is easier to solve than what you are describing. A cap tube is a restrictor. So is an orifice or a long piece of pipe that is too small (captube).

If you know the capacity required and operating conditions it should be fairly easy to find the mass flow required and then use pressure loss calculations to find the tube length for any given diameter to meet that mass flow and pressure differential.

[Peter_1]

I will scan the relevant pages and mail those to you.
PM me your private email address please.

[Peter_1]

I might be wrong, but I think the problem is easier to solve than what you are describing. A cap tube is a restrictor. So is an orifice or a long piece of pipe that is too small (captube).

If you know the capacity required and operating conditions it should be fairly easy to find the mass flow required and then use pressure loss calculations to find the tube length for any given diameter to meet that mass flow and pressure differential.

Problem is that the flow in the capillary becomes a 2-phaseflow.
The place where the 2-phaseflow starts depends on the subcooling of the liquid entering the capillary.
You also have the roughness of the refrigerant and the inside of the copper.
Then, some equations like the Bernoulli's law are only valid within some margins.
In a capillary, refrigerant can flow a sonic pressures and other physic laws then must be used.
Choked, restricted flow then starts.

A capillary is the most simple restricting device but the physic laws happening in it are not that easy to describe.

The theory in the Stoecker book is +/-6 pages long

[US Iceman]

Hi Peter.

I knew it was little more complicated than how I described it and agree with what you posted.

The issue as you have posted it depends on breaking down the physics for each area where something different occurs. This is how we wrote the program for the single screw compressor, and it seemed to be very accurate.

The 2 phase region will make you crazy trying to determine what to do. So... you have to use some assumptions that seem reasonable and then validate the results by testing.

Due to the complexity of the problem it is understandable why Chef is seeing such widely different answers.

[Chef]

Oringinally posted by Peter_1
Problem is that the flow in the capillary becomes a 2-phaseflow.
The place where the 2-phaseflow starts depends on the subcooling of the liquid entering the capillary.
You also have the roughness of the refrigerant and the inside of the copper.
Then, some equations like the Bernoulli's law are only valid within some margins.
In a capillary, refrigerant can flow a sonic pressures and other physic laws then must be used.
Choked, restricted flow then starts.

The start of the 2 phase flow is one of the key areas controlling the flowrate. Its start position is as you say mainly controlled by subcooling but evaporator pressure has a feedback too. The tube will only work as a control device if all of its length are operating at subsonic velocities and changes in evap pressure can indeed influence the tubes flowrate. If we get any part into a sonic region then only the upstream pressure controls the flow and any changes in the evap pressure cannot be transmitted back up the tube to force a change in flow to meet a changing demand. So sonic or choked flow regimes are not wanted and any design needs steer clear of getting even close to this point.

In the 2 phase region there are probably 3 flow regimes of interest, plug flow which is slugs of liquid followed by slugs of gas and then liquid and so on.
Mixed flow where the gas and liquid form a uniform state and its properties can be treated as a single phase. Capillary flow is where the liquid is held to the inner wall of the tube by surface tension and gas flows through the centre. Each regime have different friction coefficients caused by wildly different viscosities.

The fourth regime is boiling flow where heat from the tube causes boiling at the surface and we end up with a liquid core inside a jacket of gas, this is too complex to contemplate at this time. Phew!

It is quite important to get the correct flow regime as about 30% of the total pressure loss is attributed to acceleration pressure loss. If it was just friction loss it would quite a lot simpler.

Also I am very interested in EES and if US Iceman can let me have ........ details, it would be great.

Chef

[US Iceman]

I hope you know what you are getting yourself into.

Here is a link to the EES program. The letters stand for Engineering Equation Solver. I have been using this program for over 10 years and just love it. The Professional version is a little pricey, so you may want to look closely at the capabilities of different versions.

http://www.fchart.com/

As you noted there are many different regimes. I think it is necassary to estimate the type of flow you might expect. My impression is the mixed phase flow would be the closest to what you might expect in a captube. At some length when the subcooling has expired the liquid will start to boil in a homogenous mixture as I see it.

And with this you could possibly get away with using a Lockhart-Martinellni correction factor for the friction losses in the two-phase mixed regime.

Another assumption might be; if the pressure reduction does not exceed the limits for non-compressible flow then you can treat the physics in a simpler fashion (such as prior to choked flow).

I don't suspect flow boiling will enter into the issue as this is typically considering heat transfer, and for all practical purposes I think it's safe to assume the heat transfer from the cap tube offers negligble benefits to make it worthwhile.

You must really want to know how to do this to go through this exercise.

[Peter_1]

Wow, this is going a little bit to far for me. I'm really interested in the ongoing of this project.
As said in my first post, Chef knows the basics even far more then I do and only then you can use - and makes- software

[Chef]

[quote]

As you noted there are many different regimes. I think it is necassary to estimate the type of flow you might expect. My impression is the mixed phase flow would be the closest to what you might expect in a captube. At some length when the subcooling has expired the liquid will start to boil in a homogenous mixture as I see it.

I think that is the best starting assumption and certainly applies to larger bore pipes and channels but when we get to really smore ID bores like TC26 then surface fluid interactions might be a dominant factor and change the fluid composition toward the tube center. But would'nt homogenous be super simple.

And with this you could possibly get away with using a Lockhart-Martinellni correction factor for the friction losses in the two-phase mixed regime.

Sounds to me like you have already got a method lined up for EES?

Yes, I like the Lockhardt-Martinelli, it is a good dimensionless correction factor but then you would have to add an inverse Froude Rate to complete it. Thats more dificult and so maybe a Souza multiplier would be easier as it is just based on the liquid Froude Number which is simpler to calculate. Think upon this I will.

I don't suspect flow boiling will enter into the issue as this is typically considering heat transfer, and for all practical purposes I think it's safe to assume the heat transfer from the cap tube offers negligble benefits to make it worthwhile.

Absolutely and glad you agree as that would be way to tricky.

If we assume all the heat is supplied by the fluid itself the fluid quality should be easy to read from refrigerent properties which would give us an X factor from Lockhardt-Martinelli and thence a viscosity, then a Reynolds Number that then gives a friction factor and finally a pressure drop (a quick iteration) = yeilding a length for the frictional loss part.
Then we calculate the acceleration pressure loss, change in enthalpy, determine new properties at the end points and iterate the whole procedure.
What do you think?

You must really want to know how to do this to go through this exercise.

Have'nt quite done it yet but would'nt it be nice to know whats finally going on inside that little tube.
Might even lead to improved designs!!!!!!!

Like a multi diameter cap tube - more forgiving and faster equalisation for instance, but more on that later.

Chef

[US Iceman]

I have not really done anything like this in a very long time. I spent some time investigating 2 phase flow and heat transfer about 10 years ago for a project I was working on. While very interesting, the research was taking too long and I needed to finish the project, so I backed into some safe assumptions that seemed to be reasonable and correlated with accepted data. It ended up being a situation or cut bait or go fishing...

From the viewpoint I have now, it seems you are going in the right direction.

I might add this also... The small bore pipe would I think tend to force the mixture to exist as uniform within the captube. And, if I remember right; doesn't the captube have grooves on the inside of the tubing like a rifle barrel? The grooves may be the way the liquid is transported down the captube, while the vapor flows through the bore.

I should go back and read the Stoeckers book...

[Chef]

I might add this also... The small bore pipe would I think tend to force the mixture to exist as uniform within the captube. And, if I remember right; doesn't the captube have grooves on the inside of the tubing like a rifle barrel? The grooves may be the way the liquid is transported down the captube, while the vapor flows through the bore.

Now thats an interestng idea, the small diameter squeezes the action together. I hope you have'nt complicated things just as it seemed so simple and I will have to look into that before I continue. Initial thoughts are that a film on the tube will have a finite thickness with a predefined velocty profile and if that actually reaches the centre of the tube then we would indeed get turbulent mixing and a homogenous fluid in the centre. Hmmmm

I will get back on that one later.

But if that is true then the flow along the wall would be laminar and not turbulent and the fluid mixture would then be flowing through a pipe of liquid. That is going to upset the Reynolds Number and friction factor. Do we know what the friction factor of a gas/liquid mixture is over a thin film of fluid stuck to a wall with surface tension.

A film velocity profile needs to be consulted.

Chef

[US Iceman]

I guess what I'm thinking is; the volume of liquid would be very small compared to the volume of vapor. The grooves provide a volume for the liquid flow, so any liquid film thickness would be very, very thin. Therefore the gas moves through the bore of the tube in much the same way annular flow occurs.

Then, to take this a step further; if you assume the liquid film velocity is equal to the vapor velocity you can essentially treat this as single phase gas (where the expansion starts. If the velocities are equal, there is no slip between the liquid/vapor interface, which negates the mess of dealing with 2 phase flow mechanics.

For a flow problem, I think this would be appropriate. If it were for heat transfer, then a more in-depth approach might be warranted.

[Peter_1]

Wow, interesting lecture for me.

[Chef]

When you metioned the volume of liquid might be small compared to the gas I had to check, after all it is the liquid refrigerant we want in the evaporator.

So I ran a quick model and x (quality) is never much above 0.2 so not too much gas really and the volumetric ratio ranges from around 20 at mid tube up to 55 at the end of the tube. It may sound a lot but at NTP it is more like 850. This gives a velocity of 30m/s and 60m/s respectively and so there is some high velocities running down the tube.

Next in looking at ripple effect from a gas passing over a liquid we would at those small tube diameters have a waveheight of 0.5D (tube diameter) in a length of L=5D - ie a really small distance. This occured at a velocity of only 2.3m/s and looking at the expected velocities of 30-60m/s we can expect some serious turbulence.

So now the fluid is thoroughly mixed with the gas part and they must combine in a sort of mist or more likely a liquid with embedded bubbles or a foam.

I also tried this as a laminar flow regime but found we had a Re (Reynolds Nmber) of over 60,000 and laminar dies at about 4,000 so that did not hold up either, and fully developed turbulent flow at Re=60,000 is just 8 diameters down the tube.

Basically the annular flow of liquid with an internal core of gas is not a flow regime to consider, which is great as its one less complication.

The conclusion from that is:- homogenous flow (or slug and plug flow) is the regime to work with.

Interestingly the viscosity is dominated by the liquid. For instance at the centre of the tube the fluid viscosity is 0.25 and the gas is only 0.013 which at the low quality gives a combined viscosity of just over 0.18cP.

The velocity is however controlled by the gas part and, as above, the viscosity is controlled by the fluid part (which stays relatively constant) so it seems the critical factor is the velocity.

Therefore the components that control gas volume and hence the velocity must be the key to a solution as pressure drop is velocity squared function.

Anyway it is nice to remove one mode of flow from the program so thanks USIceman for bringing that one up.

And so we get to entropy - I was hoping to ignore it but??????

Chef

[US Iceman]

Boy of boy does this bring back memories.

If you treat it as gas, but correct the properties for the quality at any one step in the iteration it should work.

For example, when x=0.05 the density and viscosity are at a specific value because the percentage of liquid mass is greater than the percentage of vapor by mass. This changes until x=1, so as the quality changes the properties also change if you assume a homogenous mixture.

I'm trying to remember what this is called and for some reason the term leverage law comes to mind. I think it was called this because the the leverage is determined by the contributing mass percentage of either the liquid or vapor. It does change the calculated results, much as the Lockhart-Martinelli correction factor does for a two-phase friction factor.

[Chef]

It seems that a homogenous fluid approach is the way to go and as you say treat it as a unfied flid and modify the properties by a correction factor. It eases the iterations needed to get at the end sates for each section of the tube. Maybe the best way to do this is just try it see how close the results are to a real live system.

Maybe I can get a first cut working by the end of the weekend. The current version is just way to manual input intensive and has to go, it would probably be quicker to do it on paper!

Tried to use Refprop but it seems to be very expensive and without a lookup table for the properties it will be much more dificult. You sold me on the idea with EES remember.

Do you know of a properties program that I can access without the help of the World Bank.

Chef

[Peter_1]

Chef, USIceman, where do you both find this knowledge?
Are there any books you could recommend?

Chef, were the scans usefully?

And then thinking that a cap tube is only a small copper tube.

This is the first time I read posts which seems almost Chinese for me but I'm really interested in it.
This has nothing to do anymore with refrigeartion but pure physics.

[US Iceman]

Just to be clear: EES has most of the refrigerant properties in it, while RefProp is simply a refrigerant properties program. EES can be used with RefProp but you need an add-on extension to EES, plus you alsoneed RefProp. This is the most expensive way to go.

RefpRop is about $200 US and can be linked via a dll to Excel. This is just about as cheap as you can get this, unless you curve fit the single phase properties and paste them inside of an Excel function.

It will be interesting to see how your results compare to those of the manufacturers.

[US Iceman]

Peter.

I don't know if I had mentioned this before; I used to work in research & development so I had some time to pursue different topics. Most of my research started just a few years after the Internet was commercially available, so I spent a lot of time in the libraries.

This subject is further complicated by the fact that two-phase flow problems can be different than two-phase heat transfer. In some respects both individual topics are similar, but can be different also.

The simplest way to calculate these equations is if you treat the problem as a homogenous (or uniform) flow, however this is not applicable to all concerns.

When this uniform flow is not suitable then you need to investigate the concept of slip, which relates the vapor velocity to the liquid velocity. From this it helps to define the flow regime that might exist based on the equations defined by any given author. As the slip ratio increases, there is an increase in the vapor velocity moving the liquid through the different flow regimes.

A good introduction book is:

Two-Phase Flow and Heat Transfer by P.B. Whalley
This book is part of the Oxford Chemistry Primers volume 42 published by Oxford Science Publications
ISBN 0-19-856444-9
PUBLISHED DATE 1996

There are many others but this one is relatively cheap and contains a lot of useful introductory material.

[US Iceman]

Another comment on these captubes.

I suspect this calculation process is also similar to that used for determining the capacity of distributor tubes for refrigerant distributors.

That may provide another source of material for comparing any calculated results.

[US Iceman]

Here is a source from Google books.

http://books.google.com/books?id=Jvu...result#PPR7,M1

I have this book also. It has a lot of information in it that may be helpful.

[Chef]

It seems USIceman has nailed this thing down with whats going on inside the tube and its great we are both looking at the same phenomena - gives one more confidence to proceed.

Both of you, USIceman and Peter_1, are professional refrigeration engineers. For me its just a hobby and an intermittent one at that, when the real job kicks in it takes up weeks or months of non stop effort so with some time at hand at the moment its nice to get back to some fundamental research.

I see Refprop is $200 and this may be excessive until we can prove it works. They also want $2,000 up front for distribution! I have Fluidprop but this has a library for refrigerants by Refprop so the license is also needed. No way out there, shame as its nice to use in Excel.

Having to enter the data just detracts from the fundamental part - the actual code - the solution.

Peter_1 - I am still working on the Stoecker stuff and will let you know what it's status is soon.

Chef

[Chef]

Peter
I can retrieve and print all except P269 and 271. Could I be so obtuse and ask for those pages again - Sorry about that.

They have a nice momentum analysis part which could be interesting and want to see how it pans out

Thanks

Chef

[US Iceman]

I forgot about the distribution fees for RefProp. They have apparently chnaged them frmo when I last looked. There is another avenue though. The professional version of EES can create license free slef contained executable programs. It is just like compiling a program from Visual C++. You develop the Input screen you want, write the code/equations, and then compile it into an executable file.

You can place limits on the time use, etc. Something to thing about.

Or, you could just download the trial standard version, which is fully functional except you cannot save a file I think. Maybe it's print a file. I don't remember now.

It will be interesting to see how this develops and the comparison to other sources.

I was think about the iteration part. If you iterate the calcs' on pressure loss per inch of captube and know the initial and final pressures, the final result would be the required inches of captube needed for that operating condition. The captube bore would be one of the inputs, besides the refrigerant conditions.

[Chef]

I downloaded the demo of EES and it is brilliant, just hack the equations in - in any order - and solve presto. Well impressed and would be a useful way to go BUT.....

It does not allow copy, print, save, paste or edit so the code has to entered each time. whew and only 50 equations - not nearly enough.

The idea of EES is good though as Peter gave me a method that uses Fanno line (adiabatic) calculations directly and EES would be able to resolve these equations. It seems a better method than I have been working with. I was using an isenthalpic style and then iterating to an adiabatic solution but it takes 2 or 3 iterations and is messy and sometimes it does not converge.

The Stoecker method solves the change in enthalpy and velocity head as the quality changes in one iteration so that has big advantages. First it will be easier to write and second it will be possible to change the parameters calculated.

The first thing is to calculate L the cap length.

But now we have a way to calculate the pressure for a known L or the flowrate for a known L and a set of pressures. Now would'nt that help with diagnosis work and give a way to see what happens with pressure and subcooling as we change the load.

I had thought about using Octave - an open source language but it is not as good as EES. I will do some more work on this.

Chef

[icecube51]

Yo guys,
you speak of 2 phase flow,

In the 2 phase region there are probably 3 flow regimes of interest, plug flow which is slugs of liquid followed by slugs of gas and then liquid and so on.
Mixed flow where the gas and liquid form a uniform state and its properties can be treated as a single phase. Capillary flow is where the liquid is held to the inner wall of the tube by surface tension and gas flows through the centre. Each regime have different friction coefficients caused by wildly different viscosities.

how do you now in which fase it is ? by exterior heat ? by measuring pressure ?

wat is the optimum gas/liquid mixture at the end of the tube ?

just curious,

Ice

[Peter_1]

IceCube, you ask about the/ideal gas mixture: I think the more you have liquid, the better... which can only achieved by subcooling greatly the liquid before entering the cap tube.
It will give you anyway the biggest enthalpy.
Too many gas can provoke a choked flow through the cap tube.
Anyhow, this is something very specific for cap tubes in refrigeration.

Received today from the US Fluid Mechanics from Frank M. White. Seems interesting and written for technicians.

[US Iceman]

Peter, that is a good book. I have it also. The other one I mentioned previously is a good beginner text on 2 phase flow also.

Chef, I received your message. I'm trying to find those calculations. I thought I had them, but must have left them on a laptop with my old employer. Darn it!

[icecube51]

So Peter,
if i understand it correctly , by messuring the subcooling whe can determen the state of mixture inside the captube..?

than if the subcooling is correctly , the mixture is optimum , the performence is at his best.

i think,

Ice

[Chef]

Yo guys,
you speak of 2 phase flow,

In the 2 phase region there are probably 3 flow regimes of interest, plug flow which is slugs of liquid followed by slugs of gas and then liquid and so on.
Mixed flow where the gas and liquid form a uniform state and its properties can be treated as a single phase. Capillary flow is where the liquid is held to the inner wall of the tube by surface tension and gas flows through the centre. Each regime have different friction coefficients caused by wildly different viscosities.

how do you now in which fase it is ? by exterior heat ? by measuring pressure ?

wat is the optimum gas/liquid mixture at the end of the tube ?

just curious,

Ice

Icecube. We looked at this quite carefully to ensure the correct flow regime was being modelled. In the most serious part of the tube where the maximum pressure drop is realised the fluid is about 20% gas and most likely has a velocity of over 50m/s.

Most capillary flow breaks down at a few meters per second and waves generate on the surface of the liquid mixing it thoroughly with the gas, this gives rise to a homogeneous type of fluid.

In plug style flow the viscosity of the liquid drags and slows the flow near the tube wall and the central part of the fluid is pushed forward and therefore onto the wall. This mode soon becomes a churn style of flow and can only be found again at low velocities.

As the velocity rises to around 10m/s this churn develops to a whispy flow mode with lumps of liquid lying within the gas.

With a Re number as high as 64,000 the flow at the velocities of 50m/s is thus fully turbulent and very well mixed, it is this homogeneous flow regime that dominates the two phase flow sections. Its properties may be calculated as a combination of the fluid and gas viscosities modified by a correction factor as noted previously by US Iceman.

So the flow regime is assumed from the fluid velocity.

You also asked what is the optimum gas/liquid mixture and Peter is correct in that the more liquid we have the higher the enthalpy will be and the more efficiency we get, but a little gas is needed to realise the high velocities that provide the pressure drop.

US Iceman - looking forward to the equations.

Peter - I cant get p269 and 271 printed as my file was corrupted - could I ask you to scan them again and resend them. Thanks.

Chef.

[icecube51]

i think i understand all of this,but is this not all theory and some assumption ?? is it not so that there is a difference in horizontal or vertical lines ?? wat about the through in working order lines, who are about 1,5 or 2mtr length and wind up like a haspel?
is there not a loss of velocity , or changing in mixture after each bent? or just the opposite?

just thinking out loud.

Ice

[Chef]

Ice It is a good point that much of it is theory and also some assumptions to simplify things. The theory though is quite well founded and calculations done like this do indeed produce useful results. The assumptions part is more of a problem and we dont really know if it is correct so any assumption needs to be tested to see how important it is.

For example:- Lets assume that subcooling of the liquid is not important.
Well that would mean the fluid becomes a mixture as soon as it enters the cap tube and the typical length would be say 1.5 metres long, but if we include subcooling then the tube length ends up being more like 2.5 metres long. So subcooling is important and our assumption that it was not is invalid. So for each of the other assumptions they need to tested. It takes time.

As far as tubes being horizontal or vertical. This is important in suction lines where the velocity is quite low and in evaporators. These have differant design parameters for H or V tubing but with the rates of flow in a cap tube it is supposed to be less important, and may even be negligable. No models seem to incorporate it.

(Except of course the cap tube entry point should be down hill to ensure a liquid seal other than that it is quoted as not important.)

So, as you say, thinking out aloud lets see if it is important. The static head for a vertical tube would be about 0.7% of the total pressure drop for an X factor of 0.2 but as we discussed above the liquid might be in 1 m of a 2.5 cap tube and then it would raise to 1.9%. That might be significant. I will have to check that out further and I will let you know. Nice one Ice.

Chef thinking.

[icecube51]

thinking about the cap tube entry point. Is there not a form of cavitation wich can effect the viscosity of the refrig at surtend point, so that the inner tube has to have axial or helical inner finnes to solve this or at least try to.??

Ice

[US Iceman]

...is this not all theory and some assumption ??

That is exactly what this is. To solve a problem you have to define the variables you think may affect the answer. So, we try to find the best fit for the theory that may be involved, base some on this on assumptions and then determine the answer.

Sometimes you may have to change the assumptions or theory, but...every time you do you should come closer to the right answer.

It is just like analyzing bad systems.

[Peter_1]

And all this for the most simply form of expansion, a long, narrow tube
Someone opened a can of worms
It's indeed theory and some assumptions Ice but many of it is described very precise in well based theory, the fluid mechanics.
But I have to agree, this is something beyond my present knowledge and I'm eager to learn more about this although I will never use it later.
Sorry Chef for not responding faster, will scan them again.

[icecube51]

maybe thats why they started the TEV's , but already come back to the tube's, because when they are correctly measured and placed they are more reliable?

don't put all your thrust in electronics ,mechanical somethimes proves to be far better.

Ice

[Peter_1]

Couldn' agree more on your statement about electronics Ice.

[Chef]

Peter thanks for the other pages. It came through nicely. I see in the table that he gets the last section length as a negative and assumes it is choked flow which should at those conditions be 157m/s so a bit confused at why the method starts to break down.

Can only think its because the method does not include kinetic energy losses?

Chef