PDA

View Full Version : How critical are circut lengths when using a distributor?



kengineering
25-02-2006, 02:15 AM
I am building a very large ice pan for a sushi bar. It will require many circuts. Do to its unusual shape, it looks as though it will be difficult to measure the tubing to establish circuts of equal lengths. Any help on layout would be appreciated and exactly how critical is it to have equal circut runs? I have buit single circut ice pans with good results so I have some although limited experience in the concept. I have found that up to 100 feet of 3/8" tubing is fine as a single.

If it were to be unbalanced as follows; Circut 1 =60 feet, 2= 65 feet, 3= 70 feet
would this cause a great effect on the TEV's performance? Ken

US Iceman
25-02-2006, 04:10 AM
If it were to be unbalanced as follows; Circuit 1 =60 feet, 2= 65 feet, 3= 70 feet
would this cause a great effect on the TEV's performance?

Yes it will.

The circuit with 70 feet will have a greater pressure loss. The circuit with 60 feet of length will have the lowest pressure loss.

The unequal circuit length will cause a poor distribution of liquid refrigerant and result in the TXV "hunting". The evaporator superheat will never stabilize and continue to overfeed and underfeed refrigerant into the evaporator.

If you have an unusual pan geometry you will need to find a circuit layout that meets the equal length criteria.

arkay
25-02-2006, 07:59 AM
Hi Iceman,

What would you advise in the given case?
Do we go with 3 TXV's, one for each circuit?

Andy W
25-02-2006, 08:14 AM
Hi Iceman,

What would you advise in the given case?
Do we go with 3 TXV's, one for each circuit? I was just about to say the same, also are you using EPR valves or do you just let it run constantly. I would be interested to see a picture of a finished item as well as an evaporator layout diagram. How do you calculate the duty of it to size the condensing unit out of interest?

Peter_1
25-02-2006, 09:17 AM
We have done this in the past.

The shortest circuit will have least surface to cool, so the heat load will be the smallest on that circuit. So this circuit will determine the superheat and the liquid njection of the other circuits.

What will happen is that the systeem will stabilise but the longer circuits will cool for +/- the same length as the shortest one.

So try to install this superheated section in the part that can be a little bit warmer.
We allways lay the circuits where the superheat section is somewhere around the static evaporator at the back.

The injection starts at the glazings where the heat loads is the biggest.

Using 3 TEV's is not necessary, shame for the money. You also need then 3 suctions coming out the plate to attach the bulbs before they are soldered in a main suctionheader.
Try to do this once so that it looks als nice underneath a counter.

Ken, send me once or post it here a sketch of the counter? If you lay them the same way as they lay radiant heating systems, then you mostly will have +/- equal lenghts.
But... when you use an aluminum plate under the stainless steel presentation surface, you will never measure any difference in temperature.

US Iceman
25-02-2006, 04:45 PM
If you lay them the same way as they lay radiant heating systems, then you mostly will have +/- equal lengths

This is what I would recommend also. These projects can be a bit of a jigsaw puzzle to figure out.

An EPR will not serve any purpose in this application. The EPR only tries to control the pressure in the evaporator outlet connection. With unequal circuit lengths the circuit pressure losses are still different.

Dan
25-02-2006, 09:18 PM
The shortest circuit will have least surface to cool, so the heat load will be the smallest on that circuit. So this circuit will determine the superheat and the liquid njection of the other circuits.

A lot of excellent advice to Ken's question. I enjoy his questions a lot. I think a picture would be worth a thousand words, however. In addition to the length of tubing, the heat load is a discrete consideration.

I would make the perimeter pipes the short ones and the more internal piping the longer runs, for example.

The differences that you are looking at: 60, 65, and 70 will not be significant in a static coil such as you are designing, however, Ken. I would let the circuits gather to a common header and put the TEV bulb there, for starters and then look at individual superheats.

Peter_1
25-02-2006, 11:40 PM
Laying the tubes the way like I sketched will that all the tubes are almost he same length.

You will notice that when a tube has a long turn, that it's next turn will be a short turn and then again a long turn. Sam for the tube which has first a long turn.
That's how we mostly do it.

But again, we now use since +/- 2 years aluminum plates between copper and stainless steel and the result is realy astonishing.

Make no couplings underneath the plate.

In the previous posts, you can see we draw the circuit first on the plate itself.

Like Dan said, this will give not a big difference.

Ken, what cooling capacity do you usually take for the plate cooling?

Peter_1
26-02-2006, 10:57 AM
Sorry, forgot the sketch to add

Andy W
26-02-2006, 12:04 PM
I did not picture it like that, EPR would as you say serve no use, I would personally fit 3 expansion valves though.

Peter_1
26-02-2006, 01:00 PM
How do you calculate the duty of it to size the condensing unit out of interest?

Impossible to calculate.
Can only be done with a practical test. We take +/- 200 W/m² to evaporate at -10°C.

If you fit and EPR, you can adjust the evaporating temperature so that it matches exactly what you need.
If you make it to cold, the fish will stick frozen on the ice.
Keeping the ice at -3°C/-5°C is enough.

3TEV's for an ice pan? Try to keep it as simple as possible.

Dan
26-02-2006, 08:59 PM
But again, we now use since ± 2 years aluminum plates between copper and stainless steel and the result is realy astonishing.

Make no couplings underneath the plate.

In the previous posts, you can see we draw the circuit first on the plate itself.

Of course, now, I have to ask. How is it you use the aluminum and copper and stainless? I imagine you have taken advantage of aluminum's heat transfer capabilities, but how are you bonding the copper and aluminum and stainless?

US Iceman
26-02-2006, 09:29 PM
...but how are you bonding the copper and aluminum and stainless?

From the way Peter described his project I would assume the aluminum plate is next to the stainless steel, and the copper tubes are connected to the bottom of the aluminum plate.

I would also assume the use of heat transfer compound between the aluminum and stainless steel.

This is a good idea if I have interpreted Peters' description correctly. The copper tubes cool the aluminum (which provides very high conductivity) and helps to evenly distribute the cold to the stainless steel.


If you fit an EPR, you can adjust the evaporating temperature so that it matches exactly what you need.

If the EPR is used for this purpose I agree it would allow the evaporating pressure to be fine-tuned to meet the requirements. In a previous thread I mentioned an EPR would not provide any beneficial effect. My comment was related to the unequal pressure losses and circuit loading.

frank
26-02-2006, 09:36 PM
Hi Iceman


I would also assume the use of heat transfer compound between the aluminum and stainless steel.

Here's a quote from Peter's post .
Make no couplings underneath the plate.

It looks like Peter just allows the aluminium to act as a heat transfer medium between the copper (pipes) and the S/S top, although I do think that the heat transfer compound is a good idea.

US Iceman
26-02-2006, 10:08 PM
I would like to comment on the unequal circuit lengths again. I do not take issue with the comments of Peter_ 1 and Dan.

I think if the differences in the three circuits are relatively minor, the results of TXV control may be acceptable. As Dan said, each circuit superheat should be checked (over some operating period I might add to verify stability of superheat control).

Here is another way to look at this. If a refrigerant distributor is used to feed each circuit and the circuit lengths are pretty close to each other, I suspect the impact will be minimal.

If a refrigerant distributor is not used (say the circuit tubes are installed in a larger pipe and brazed together) I would suspect the difference could be somewhat larger due to potentially worse liquid distribution.

The problem we are faced with is the impact of pressure loss in the circuit. If the mass flow is high through each circuit the pressure loss would of course be greater. Conversely, if the mass flow through each circuit is low, the pressure losses would then be lower.

The way I would approach this is to estimate the pressure loss through each circuit. The number of circuits would be based on maintaining a difference of 1 degree F between the saturation temperatures of evaporating temperature and the equivalent saturation temperature at the circuit exit. In other words, the pressure loss of each circuit would be less than the equivalent of 1 degree F.

If you have two circuits the mass flow per circuit will be higher than the same mass flow in three circuits. We need to evenly separate the circuit loads to minimize the impact of circuit pressure loss.

I Ken's example, the pressure loss would be equal to less than one equivalent degree F for the longest circuit length (70 ft).

If the pressure losses of each circuit were within 10% of one another I think any problem would be minimal. At medium to high evaporating temperatures, the pressure loss equal to one degree F is very minimal. At low temperatures the impact is greater.

My original statement essentially said don't mix circuit lengths. The reason I stated this earlier is to prevent someone from getting the idea that it is OK to match unequal circuit lengths in all conditions.

If you calculate the differences you will have a better idea of the impact. The greater the differences... the more potential problems you can create.

US Iceman
26-02-2006, 10:12 PM
Hi Frank,


Make no couplings underneath the plate.

I understood this to mean no brazed joints under the plate because they might be too difficult to repair. A seamless copper tube laid out in continuous serpentine fashion.

Perhaps I read too much into Peters' comments.:confused:

kengineering
27-02-2006, 03:32 AM
WOW! Lots of info to process. Keep it coming.

The size of this "ice pan" is 13ft.L X 13ft.W X 1.5ft. H, with a cut-out making it look like the letter U. The chef will stand in the cut-out. The load will be ice that will be shoveled in to fill the pan.

I'll try to post some pictures as soon as this project leaves the blueprints.

I have tried a tube spacing of 3" and it seems OK what do you all think?

It should be interesting. Ken

Peter_1
27-02-2006, 08:35 AM
Of course, now, I have to ask. How is it you use the aluminum and copper and stainless? I imagine you have taken advantage of aluminum's heat transfer capabilities, but how are you bonding the copper and aluminum and stainless?

Exactly Dan,it's for the heat transfer capabilities.
It's amazing if you don't use it how bad the heat transfer is for SS.

We use it also for heated counters to 75°C. If you use electrical heaters instead of hot water, you can measure temperature differences of 30 to 40°C°C over 10 cm.

I will take some pictures of the counter which is test running for the moment where we have a lump( bump?) due to the TIG soldering of a seem made by the TIG welders.
You can realy see over a distance of 15 cm a temperature difference of 15°C. The lump is wet at +/- 2°C and the water of it is sliding 10 to 15 cm further to freeze there to -12°C. All this because the aluminum isn't contacting the SS and the bad heat transfer capabilities of the SS.

The forum doesn't allow pictures reposting but you can see them at
http://www.refrigeration-engineer.com/forums/showthread.php?t=3877

Indeed US Iceman, Frank, you can see also the heat transfer compound. You interpreted the way I tried to explain it.
Regarding the EPR, I know what you meant but it's only to clarify how you can use it in such an application, only for one unit, especially because the selection of the condensing unit can't be calculated. Size it a litlle bit to big and then install an EPR.

And no couplings is indeed for safety, you can never repair them afterwards and if it's heavy loaded you will see the thickness of the solder printing/pushing through the stainless steel.

For the spacing, we use the radius of the bender which is +/- 7 cm I think. I think going wider will not give much problems.

We learned all this by making many mistakes.