Constantin or

Plumdog_2

Not too fast; Not to slow. By the way, I liked the one about the splinter and the tree.

EBEBRATT-Ed

Constantin or

Anyone who wants to put there two cents in. This has been buggin me.

a heating system typical design of 20 deg td we all know 10,000 btu/gpm. Many people will say when troubleshooting that "it's pumping to fast or too much flow you cant transfer enough heat". This could be a boiler, chiller@2gpm/ton 12000 btu/hour whatever.

But mathmaticly it works out that an increase in flow with a given system of heat input or output will cause a smaller temp rise or drop but with more flow the heat transfer rate remains the same. Weather the system is designed at a 1 degree temp rise or a 100 degree temp rise. Maybe I am confusing this but a boiler with 100,000 btu output installed in a system with 100,000 btu of baseboard installed will provide just as much heat if it was designed at 10 gpm or 1gpm or 100 gpm assuming that the pipe sizes, pumps etc are matched for the flow needed. In fact I always thought that heat transfer increased with turbulant flow and was more sluggish at slower laminer?? flow. I know the other flows arn't realistic and are out of the normal range-just trying to prove this out

Thanks for the replies.

Ed

bob_50

Myth

Ed, I know a lot of guys on the wall will disagree BUT the line about too much flow won't transfer enough heat is in my opinion and ASHRAE's an untruth. Sorry about the quality of the graph Y axis is btu transfered and X is velocity. As velocity increases so does heat transfer rate. From ASHRAE fundementals. bob

Mike E_2

I totaly agree with you.

Having "too much" flow won't give you less heat transfer, it in fact will give you more.

If you had 100ft of baseboard with a pump designed for a 20 deg delta with 180 entering/160 leaving, you would be pumping 5.4gpm. The baseboard would put out 54,000 BTU (170 deg average temp, 540 BTU/ft)

If you had the same 100ft of baseboard with a pump designed for a 5 deg delta with 180 entering/175 leaving, the baseboard would put out 59,250 BTU (177.5 deg average temp, 592.5 BTU/ft)

However, there is a point where too high of a flow will cause other problems such as noise, etc.

Constantin

I'm not sure this will help but let me give it a try

There are typically three different ways that heat is transferred in our domains: Radiation, Convection, and Conduction. All of them depend (in part) on the surface temperature of the emitter or HX to do their work. In this case, allow me to re-host an image that can be found at the University of Virginia on a page that describes their heat transfer software.



The graph shows what happens inside a pipe as a fluid flows through it and a "hot" wall wams the cold fluid inside. As with most emitters, a boundary layer forms at the edge of the HX. The faster the laminar flow, the "thinner" these layers will become and the more fluid will pass through the pipe unmolested and cold at the center, and only the edges will warm. Since the boundary layer is warmer than the pipe center, the heat transfer via conduction won't be great as HX depends on ΔT.

In other words, the boundary layer will start to act like an insulator, even though it may be made of an material that conducts heat quite well, like water. As you observed, this would not be an issue in situations where you have turbulent flow, but how much truely turbulent flow do we get through regular pipes at the velocities used in hydronic heating?

The larger the pipe diameter, the greater the problem of getting the center core (which will flow fastest) up to temperature. It's one reason that heat exchangers in some mod-con boilers have very thin cross-sections: A thin cross section offers less material in the middle to stay cold. So, IF I am interpreting the graph correctly, you have to give the fluid inside the pipe enough time to mix internally through convection/conduction - as the assumption that flow is turbulent is not a given.

Those darned boundary layers are also experienced in every day life... A convection oven cooks faster not because it is hotter, but because it breaks up the boundary layer around whatever it's cooking through the use of convection fans. I imagine that a lot of HX's also try their best to mix the internal fluids to get as uniform a temperature within the HX fluid as possible, i.e. eliminate the boundary layer to the extent possible. For example, note the tubing that Wolverine Tubing has made available to the industry, where the tubes are internally grooved, ridged, etc. to increase HX w/o "excessive" pressure drop.

Brad White_9

Damn, I am impressed, Constantin

So what's new? CFD printouts no less...

Ooo-Rah!

Constantin

Now wait a minute...

... you're the HVAC genius here, I'm just a homeowner. Does my explanation make sense or am I muddling as usual?

steve b_14

I think I need more help understanding this. With laminar flow, as the flow rate increases the insulating boundary layer becomes thinner and at the same time the water in the center of the pipe remains hotter. Wouldn’t a higher water temperature combined with a thinner insulating layer result in better heat transfer?

hr

Engineers

play with the flow rates depending on the situation.

I've read that on long runs between buildings on district systems they will design flow rates to minimize the heatloss in the underground piping.

hot rod

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

Constantin

Keep in mind...

....if you were to look at the cross-section of fluid flow through a pipe, the fluid flow velocity through a pipe is almost never uniform. The layers close to the walls will invariably flow slower than the rest of the fluid due to the resistance they encounter as they bump into the wall. As a result, the fluids at the center can whizz by while the layers at the edges creep along.

Constantin

Thanks Plumdog

...and I'm still waiting for Nurse Ratchet... not a peep out of the FD or the BBRS yet.

Steamhead (in transit)

Time is also a factor

especially in old radiator systems.

If the water moves so fast that it doesn't have time to diffuse thru the radiator, it will not be able to give off its heat as it should. This shows up as a low delta-T between supply and return. This is less of a factor in baseboard and similar systems, but can still cause problems.

Consider a typical radiator with the supply connection at the bottom of one side and the return on the bottom of the other. If the system is over-pumped, the water will follow the shortest distance between those two points- a straight line. When this happens, I've felt a supply riser get warm and the return riser get warm about a minute later.

Read this for a description of an extreme case of over-pumping (and improper piping too, that's a story in itself).

http://www.heatinghelp.com/newsletter.cfm?Id=119

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

ALH_4

Viscosity

And as viscosity (propylene glycol concentration) is increased the boundary layer thickness increases. Another catch is that as the boundary layer becomes thinner, the drag coefficient increases.

-Andrew

ALH_4

And

with condensing hot water boilers, the lowest return water temperature is desireable, so the widest dT allowable should be used to maximize the flue gas condensation. Also what about the boundary layer of air around the heat emitter, particularly in baseboard applications where natural convection is relied upon.

-Andrew

Brian (Tankless)

The magic word,

Turbulence!

I first met Mr. Kim via an industrial radiator that his company had modified, in the North Sea, 1974.

I forget the reason, but the top cap had been removed, and inside all the 5/16" vertical tubes, were 1/4" spirals of twisted "flat-bar", brazed at top & bottom. That was a VERY efficient radiator.

He invented & patented this idea, everything since then, is a variation on a theme, including Wolverine.

Turbulence works (Ergomax), high velocity doesn't. Try removing the thermostat from the cooling system in your V8 powered shrimp boat, my brother-in-law did. I think this where "Dead-in-the-water came from" The lack of a $5 T/stat cost him a new engine

Great stuff Constantin & you guys.

Is heattransferaholics a disease ? If it is, I have it.

I'm off to convince a lady client that the 54*F un-insulated floors are her problem, not just the open fireplaces that I've already blocked off. Her future husband, an extremely wealthy lawyer, is very scrooge-like when it comes to his rental properties (even the one she's living in!!!!)

Turbulence is where it's at.

Is it really 11.0 AM ? Why am I still here.

Later, Brian.

Addendum: Kim's Radiator Service is located in Sulphur, LA.
just 10 miles from me. He's 80 yr's old, and still shows up for "work" every day. He has an 8' x 12' office wall covered in patents & accolades. This guy LIVES for heat transfer, I'd bet that he has "twisties in his arteries .

He belongs on "The Wall".

Google him, "Kim's Radiator Service". He can build any kind of heat transfer device you can imagine.

Steve D._4

Great graph

I have to comment on some interpretations of the graph.

"The faster the laminar flow, the "thinner" these layers will become and the more fluid will pass through the pipe unmolested and cold at the center, and only the edges will warm." - This is the nature of laminar flow. A boundary layer always forms and is microscopic in size. The white "D" shape is the velocity profile and is determined by the fluid properties - Reynolds, Prandtl numbers and the shape of the pipe.

With still water, the velocity profile is flat and you would see heat entering the fluid through convective currents. Now imagine cold fluid just starting to enter the pipe and cool off the water that is heating up. As this speeds up, the cold zone reaches further and further down the pipe. If you only look at the center 1/8" diameter of the pipe along its length you will see the temp of the water rise as you get further through the pipe. This suggests that there is no "insulation" in the boundary layer.

Now think of how heating units are defined. Amount of energy to raise 1 lb (or kg, depending on upbringing)of water by one degree of your choice. When the fluid increases its velocity in the same pipe, more mass of fluid is moving through the pipe. The same heat transfer is going into the fluid but there is more of it(the fluid). Like if you hit the lottery with 100 other people. Since there is so much fluid going through at the same time, it has to travel further before it reaches a given temperature.

"this would not be an issue in situations where you have turbulent flow, but how much truely turbulent flow do we get through regular pipes at the velocities used in hydronic heating?" You will get turbulence at elbows and restrictions. In a home system this is typically distribution piping so it doesn't help in the heat input you are talking about. Triangle tube's Prestige creates turbulence in their pinched tubes, which are small and share the flow through many tubes. This lowers the velocity and pressure drop compared to a long tube looped back and forth.

"So, IF I am interpreting the graph correctly, you have to give the fluid inside the pipe enough time to mix internally through convection/conduction" YES

Brian (Tankless)

Bump

Bump.

Any sources for 2" solid foam insulation for this Lady's crawlspace joist-bay's ?.

Be happy, keep Trans-Ferrin heat .

Steve D._4

Great graph

I have to comment on some interpretations of the graph.

"The faster the laminar flow, the "thinner" these layers will become and the more fluid will pass through the pipe unmolested and cold at the center, and only the edges will warm." - This is the nature of laminar flow. A boundary layer always forms and is microscopic in size. The white "D" shape is the velocity profile and is determined by the fluid properties - Reynolds, Prandtl numbers and the shape of the pipe.

With still water, the velocity profile is flat and you would see heat entering the fluid through convective currents. Now imagine cold fluid just starting to enter the pipe and cool off the water that is heating up. As this speeds up, the cold zone reaches further and further down the pipe. If you only look at the center 1/8" diameter of the pipe along its length you will see the temp of the water rise as you get further through the pipe. This suggests that there is no "insulation" in the boundary layer.

Now think of how heating units are defined. Amount of energy to raise 1 lb (or kg, depending on upbringing)of water by one degree of your choice. When the fluid increases its velocity in the same pipe, more mass of fluid is moving through the pipe. The same heat transfer is going into the fluid but there is more of it(the fluid). Like if you hit the lottery with 100 other people. Since there is so much fluid going through at the same time, it has to travel further before it reaches a given temperature.

"this would not be an issue in situations where you have turbulent flow, but how much truely turbulent flow do we get through regular pipes at the velocities used in hydronic heating?" You will get turbulence at elbows and restrictions. In a home system this is typically distribution piping so it doesn't help in the heat input you are talking about. Triangle tube's Prestige creates turbulence in their pinched tubes, which are small and share the flow through many tubes. This lowers the velocity and pressure drop compared to a long tube looped back and forth.

"So, IF I am interpreting the graph correctly, you have to give the fluid inside the pipe enough time to mix internally through convection/conduction" YES, because of the additional mass of fluid at higher temps.

I hope I am making sense.

Steve D

EBEBRATT-Ed

Thanks to all for the information. This is a great site and a great source of knowledge.

Constantin comes up in color--what else could we expect!


Ed

Mike T., Swampeast MO

Don't the very well-tested output ratings of fin-tube baseboard show that output does not increase as fast as velocity?

Constantin

.... particularly when ...

... the baseboard was put in when the place still had a hardwood floor and then someone decides to install wall-to-wall carpeting without adjusting the height of the radiators and hence blocks the bottom intake of the fin-tubes. Bye-bye convection!

Constantin

Thanks for the corrections, Steve

I had some difficulty expressing what I meant to say up there, that post got revised at least once. Heh! :-P

But seriously, you make a very good point re: the passage design in the Prestige HX. Somewhat similar results can be achieved with tubes that feature rifled edges sticking into the flow of the water at the core of the tube.

Brad White_9

That was my point,

for a homeowner, you really get into this stuff! The concept of computational fluid dynamics is something we farm out for certain aspects of our work (trouble-shooting or pre-design testing of certain hard fittings).

Great age we are in.