PDA

View Full Version : Metering device?



Apoc
12-07-2004, 07:10 AM
Basically this is the system I'm working on atm, it's my little project.

http://apocforever.homestead.com/files/liquidpump2.bmp

I'll be using a 3/4hp LPB R404A compressor, with a co-axial water cooled condenser. The refrigerant pump with be a bronze 1/3hp gear pump, and the refrigerant will be R410A.

My question is: What type of metering device would give the greatest performance in a system similar to this? I'm not sure if a tev would work. I was thinking about using cap tube with a solenoid valve on the liquid line and a liquid level sensor in the liquid separator. Simply I have to always maintain the column of liquid on the pump inlet to avoid cavitation.

Any thoughts?

Peter_1
12-07-2004, 11:50 AM
You draw a schematic of a liquid overfeed system, as used in almost all the NH3 systems.

I don't see a liquid receiver after the condensor.

What you need is a level indicator on the liquid separator which 'demands' liquid from the receiver to maintain the same level in the separator.

And then I should use a capillary tube.

What's the purpose of your project?


Peter

Apoc
12-07-2004, 12:09 PM
Thanks for your reply.

It's for delivering liquid to multiple evaporators on my computer :cool:

frank
12-07-2004, 08:23 PM
How can you use a low back pressure compressor with a high pressure refrigerant :confused:

chemi-cool
12-07-2004, 08:38 PM
Very Carefully :D

Chemi

Peter_1
12-07-2004, 08:58 PM
If he can condensate low enough and stay belwo max current , why not?

What will you do with the oil that will stay in the LP separator that the compressor can't suck back to his sump.

Why not use a heat exchanger and cool a secundary refrigerant/fluid (for example glycol) and circulate this with a pump?
We have some application we service (foil laminating with alumini) where we pump pure glycol at -35°C.

Apoc
13-07-2004, 06:22 AM
What will you do with the oil that will stay in the LP separator that the compressor can't suck back to his sump.

There will be a oil separator after the compressor discharge.


Why not use a heat exchanger and cool a secondary refrigerant/fluid (for example glycol) and circulate this with a pump?

I want this system to operate at -45C. I want to push heat transfer coefficients as high as possible through forced-convection boiling. Also, anyone can build a chiller, but not everyone can build a system such as this :cool:

Peter_1
13-07-2004, 09:50 AM
Oil separator will not block all the oil, so that's not sufficient whereby my initial question remains.

You may use whatever technic you wish (forced convection boiling or whatever), going down to -45°C with R404a will let you run below atmospheric pressure.

What do you mean whit forced convection boiling (I can imagine what you mean) and how will you do this?

Have you figured out what the cooling capacity will be taken into account losses of lines ,pumps etc for your desired HP/LP combination? And will this meet your requirements?

What kind of pump will you use? You will need a very little capacity. Viscosity of the refrigerant will be much higher then usual found.

And also, not everyone can build any chiller but you still need to prove that you can build a liquid overfeed.

Let us know what you found/did so that we also can learn from your experiment.

Apoc
13-07-2004, 10:50 AM
Oil separator will not block all the oil, so that's not sufficient whereby my initial question remains.

I'm only a hobbyist. Maybe I should be asking you how to solve this problem.


What do you mean whit forced convection boiling (I can imagine what you mean) and how will you do this?

I'll be using jet impingement on the a micro grooved enhanced surface. I'll talk about this more late, if need be.


Have you figured out what the cooling capacity will be taken into account losses of lines ,pumps etc for your desired HP/LP combination? And will this meet your requirements?

The heat load itself will not be more than 0.25kw. I was told the only heat transferred to the fluid from the pump will be due to mechanical in-efficiency of the pumphead. The capacity of my compressor at -35C is 400w, because I'm using R410A I should get a increase in capacity. Basically I just bought the largest compressor my budget allows.

I do use cycle analysis in coolpack quite a bit, so everything should workout.


What kind of pump will you use? You will need a very little capacity. Viscosity of the refrigerant will be much higher then usual found

It's a Oberdorfer model N.2000E.S16, helical gear pump of all bronze construction, with neoprene mechanical seal. Will deliver 7.5 L/min at 1.5 Bar total differential head, with an NPSHR of 0.7m. This pump may not be used though.


And also, not everyone can build any chiller but you still need to prove that you can build a liquid overfeed.

I never said I could build one. If I could build one, I wouldn't be here asking you these questions. It's just in my hobby I see a lot of chillers, but so far no liquid overfeed systems.


What you need is a level indicator on the liquid separator which 'demands' liquid from the receiver to maintain the same level in the separator.

Is this achieved through the use of a LSF?

Peter_1
13-07-2004, 01:22 PM
How Apoc, no hard feelings. Everything I said to help you, not to offend you.
It's not always that easy to express correct in another language what's in your mind.

I learned via this forum many times that the hobbyist 'overclockers' have many much, much more refrigeration knowledge then the so called techs.

Some things to be aware: high pressure when you will use R410.

You need a refrigerant of -45°C so you will need to evaporate at appr. -50°C.
I think - haven't looked for it - that you available capacity will drop a lot.

The pump you mention is build for R410A? (sealings, high pressure at stand-still, ?)

In fact, your question is more for the section 'Overclockers'

Have also a look at the posts and the site of Dabbit of The Netherlands.

What is an LSF? I suppose it's some sort of Liquid Sensing Fevice ;)

I think and I feel that you know something or even a lot of electronics. So what you can make is a small heated sensor which current will vary if it's immersed by liquid (cooled) or not. They use this system industrial in liquid level sensors.

Have perhaps a look in the 'Electrical' forum under my last add, microprocessors. This is perhaps something for you.

Even we, who are for years in this business can learn something from you, hobbyist. You try out things that takes to much time for us, you have time to experiment. We can give you the information we posses.

I will think on your oil problem. Perhaps an input from the NH3 guys?

Drain some *****/oil mixture after the pump (higher LP) with a correct cap tube via a small heat exchanger wrapped around the discharge line so that the refrigerant boils out the solution and the oil goes back t o the compressor.

Apoc
14-07-2004, 10:25 AM
Indeed, no hard feelings mate.



The pump you mention is build for R410A? (sealings, high pressure at stand-still, ?)

That's a good question, static pressure will be about 130 PSI. And this (http://www.oberdorfer-pumps.com/Spec_sheets/N2000%20Layout.pdf) is all the info I have on the pump, I might have to email the manufacture to find out. One other problem with this pump is the cost $1000 usd :eek:


In fact, your question is more for the section 'Overclockers'

Yes, but the information that I want is more of a industrial application.


What is an LSF?

Low side float ;)

Thanks for your help.

Peter_1
14-07-2004, 12:50 PM
Check also if Viton and/or Teflon, the materials of the seals can be used with R410a.

1000 US$ is too much for such a pump Apoc. For that price, we install much bigger industrial NH3 pumps.

Have you searched Ebay, for example for a hydraulic pump... can whitstand high pressures, seals easily replacable, relative cheap, ... Or a pump out of the hydraulic steering of a car.

Perhaps stupid ideas but who knows???

Instead of a LSF, you can do it optically or with very small heating probes.

It perhaps even can work the way you draw it, without a liquid receiver. You must be sure that there is allways liquid in the LP separator (to avoid cavitation in the pump)

Apoc
15-07-2004, 06:32 AM
It has a neoprene mechanical seal. Is this okay to use?

$1000 usd is for the complete pumpset. This is the cheapest I could find in new zealand. I have seen the bareshaft online for as low as $250, but I'll have to see how much shipping is and I'll have to buy a motor.


Have you searched Ebay, for example for a hydraulic pump... can whitstand high pressures, seals easily replacable, relative cheap, ... Or a pump out of the hydraulic steering of a car.

I don't think they would handle the low temp, but who knows it might work.

I've seen a lot of liquid level sensors online. I don't know if th would handle the low temp though.


It perhaps even can work the way you draw it, without a liquid receiver. You must be sure that there is allways liquid in the LP separator (to avoid cavitation in the pump)

There will be a receiver. Sorry, that was just a quick and nasty drawing.

Andy
16-07-2004, 06:03 PM
Hi Apoc :)
why bother with a pump, you can liquid overfeed with a low pressure receiver. Basically this is a suction accumulator with a refrigerant boil off coil in the bottom. Circuit would be liquid off the condenser, thru two solinoides with capillaries after them. these switch to control the liquid feed to the evaporator on a level switch at the condenser outlet (one capillary is bleed and one main liquid feed. After the sol feed the liquid into the evaporator as in DX then trap the liquid comming down the suction line in the LPR, boiling it off by indirect heat exchange from the warm liquid passing off the condenser.

Kind Regards. Andy.

Peter_1
16-07-2004, 07:51 PM
SELF- HEAT" SENSING APPLICATIONS
To "self-heat" a thermistor, it must be subjected to power levels that raise the thermistor's body temperature above the environmental surroundings. Self-heat applications include the sensing of liquid and air level and flow rates. This application is dependent on the fact that the enviro nment surrounding a thermistor directly affects the amount of power the thermistor can dissipate. For example, submerged in liquid, a thermistor can typically dissipate 500% to 600% more power than it can in air.
Therefore, a thermistor being "self-heated" in air is able to dissipate much more power when transferred to a fluid environment. This increase in power dissipation generates a significant increase in resistance. It is this change in resistance which makes it possible to sense the fluid level.

A simple liquid level control system can be designed by putting a thermistor in series with a coil (Figure 8) which operates a valve that releases the liquid in the tank. The thermistor is placed in the tank and operated in a "selfheat" mode.

In air, the thermistor's resistance is low and allows enough current flow to energize the relay coil and keep the relay contact closed. When the fluid level in the tank surrounds the thermistor, its resistance increases and de-energizes the relay which opens a valve and releases the fluid. As the fluid is released from the tank, the thermistor's resistance decreases and the relay coil energizes and closes the valve.

Fuel injection systems in automobiles utilize the thermistor in the "self-heat" mode in order to properly control the air/fuel mixture. Forced air heaters may use the NTC thermistor in the "self-heat" mode in order to maintain proper air flow characteristics. This technology is utilized to monitor the flow rate and level of air and fluids in a variety of applications.

Andy
16-07-2004, 10:34 PM
Hi Peter :)
we use thermistors with a simple voltage sensing relay to control liquid levels. The thermistor is wired in series with a 63 ohm resistor, (12vac sensing relay 13-15 ohm thermistor).
The relay can be bought from the RS catalog, the thermistor we import from somewhere in america.

Kind Regards. Andy. :)

Peter_1
17-07-2004, 07:48 AM
I found this via my son who's studying electronics.
In the past, I sometimes used a simple Kriwan INT69 to control a liquid level. Connected two stainless steel bolts on the thermistor entrance and it worked fine.

Andy, the relay you talked about is a special relay I suppose because you need somewhere a Wheatstone measuring bridge to measure the liquid level.

And the thermistor is a special one?

What's also interesting literature is that you can measure vacuum perfect with a thermistor. They heat a thermistor and the more air there is around the thermistor, the more it can cool. So in a perfect vacuum, there are no air particles anymore which can cool the thermistor and you will have maximum heat. This is measured by some sort of amplified Wheatstone bridge.

They called it Pirani thermistor gauges.



Peter

Andy
19-07-2004, 08:24 AM
Andy, the relay you talked about is a special relay I suppose because you need somewhere a Wheatstone measuring bridge to measure the liquid level.

And the thermistor is a special one?

Peter

Peter the thermistor is a special item, with a 1/4" bsp thread on it for ease of fitting.
The relay is a standard octal plug in type, simply with a small voltage triming pot on top.

Kind Regrards. Andy. :)

Apoc
20-07-2004, 09:36 AM
Thanks for your replies.

It looks like I've got some more things to think about.