corner radiator infrared view

hot_rod

This is a 22 section L shaped radiator. I supplied it with 120F, at 1 gpm. Supply from bottom right, return bottom left.

Starting at about 70F, I heated it for 40 minute, pic at 5 minute intervals.

I was surprised to see how it warmed from top down even with S&R at the bottom connections.

Steamhead

Nice diffusion. This is why we must avoid over-pumping gravity conversions.

nicholas bonham-carter

Certainly worth a trip to Nebraska to pick up that jewel!!!!--NBC

Hilly

It's too late now but I'd be curious to see infrared shots of a castiron heater cooling. You always hear "they're the best, they hold the heat so well" I don't think I've ever witnessed any of the science showing that 'holding heat' ability. I guess I could say that for all heat transfer formats.

Hilly

BTW I love that you take the time to share your experiments with everyone. I'm grateful for all the resources you provide, thank you.

j a_2

I liked that...A picture is often worth a book of words.,like the ones you took

Mark Eatherton

Cool. Now you can do all kinds of cool things, like triple the flow rate and "see" what effect it has on the overall emissions. Or take it to 1/3 GPM and see what effect that has on emissions. Kind of hard to dispute that kind of evidence, although I am sure someone will.

You also have enough metering on it to be able to verify the EDR and see if it comes close to what the dead men said it should be in the way of BTUH output.

I did the same thing with my Buderus panel radiators. The color change was so consistent, it was rather unremarkable. We did the same thing with our radiant window once, and it was so subtle that you really had to concentrate on the colors to see the changes, and really, the beginning and ending slides were the only ones worth looking at. If you sped the video up, it looked pretty cool. Can't seem to locate those pictures. Is suspect they are on my old lap top, which is on the western slope.

Thanks for sharing HR. From the labs on KIA Lane!

ME

hot_rod

I ran 20 minutes at 2 gpm flow this morning, I will crunch those slides in a bit. I'm also watching a cool down period as J A suggested, it is taking a lot longer than ramp up, and my loaner camera is headed back to Colorado in about an hour.

Harvey Ramer

The way I quickly figure out if a bottom plumbed rad is over pumped, is to feel the return port while the system is loading. The top of the rad should always heat up the whole way across before the return port warms up.

I read about the gravity circulation within a Rad, In one of Dan's articles probably about 8 years ago.

It will be interesting to see what your tests reveal. There should be a fairly wide margin of error between the lowest and highest desirable flow rates.

In the first test it appears the flow is all going up through the first section and then across the top. With an increase in flow I would expect more sections to be used to allow the water to the top. I would say once you reach close to half of the sections with an upward flow of supply water, and the other half with the downward return water, you are at the point where any further increase in flow will have a detrimental effect on performance.

Gordy

I'd be interested in the low end. How low can you go with the flow? Been the latest buzz around here. My prediction would be more time, and not as saturated from top to bottom side to side.

bob_46

reinvent the wheel ?

Steamhead

Here are some we got a few years ago. You have to click on the icons to display the pics:

http://forum.heatinghelp.com/discussion/140472/seeing-inside-an-over-pumped-radiator/p1

Gordy

bob said:

reinvent the wheel ?

Not really been some discussion about how laminar verses turbulent flows can effect heat transfer. In the sub 1gpm range. This stems from various discussions on a new HTP mod/con line with 10:1 TDR. The HTP UFT 80 boiler that has a low end of 8k modulation. Theoretically flows of .8 gpm through that boiler at low end. What does .8 gpm do to, and for convectors heat transfer.

hot_rod

Hilly said:

It's too late now but I'd be curious to see infrared shots of a castiron heater cooling. You always hear "they're the best, they hold the heat so well" I don't think I've ever witnessed any of the science showing that 'holding heat' ability. I guess I could say that for all heat transfer formats.

It took almost 3 hours for the radiator to cool down to 72F at the bottom, about 75F at top. It holds just shy of 15 gallons, weight?? maybe 800 lbs?

hot_rod

Harvey Ramer said:

The way I quickly figure out if a bottom plumbed rad is over pumped, is to feel the return port while the system is loading. The top of the rad should always heat up the whole way across before the return port warms up.

I read about the gravity circulation within a Rad, In one of Dan's articles probably about 8 years ago.

It will be interesting to see what your tests reveal. There should be a fairly wide margin of error between the lowest and highest desirable flow rates.

In the first test it appears the flow is all going up through the first section and then across the top. With an increase in flow I would expect more sections to be used to allow the water to the top. I would say once you reach close to half of the sections with an upward flow of supply water, and the other half with the downward return water, you are at the point where any further increase in flow will have a detrimental effect on performance.

I don't understand the dynamics that caused it to heat across the top, then migrate down?
It clearly has less pressure drop straight across the bottom, I say about an 1-1/4" passageway and the shorter flow distance.

At 2 gpm it showed the same, in 1/2 the time.

I may try 3 gpm to try and determine what the over-pump threshold is that it doesn't allow it to heat evenly like the 1 & 2 gpm runs.

In my case it runs directly from the HTP Phoenix, 80 gallon. With two other zones calling the temperature stays within 3 degrees, burner modulated at lowest rate.

Harvey Ramer

@hot rod

It's gravity flow within the rad that causes the effect. As soon as the supply water enters the rad, the velocity drops so much that gravity circulation takes over between the inlet and outlet. This is essential for proper performance. If the velocity is high enough to be affected by pressure drops within the rad, the supply water will scoot the whole way across to bottom to the return port and stymy the gravity flow within resulting in a lower average surface temp of the rad and a longer duration to warm up.

Paul48

I doubt that it can be over-pumped. Half the radiator at twice the temperature is a wash, heat-wise. Do a time/temp study at 12" above the supply, and see if my theory holds water(hot or cold)

Steamhead

Check the link I posted. Yes, it is definitely possible to over-pump a radiator. Remember, the shortest distance between two points is a straight line, and if the flow rate is too high the water will simply follow that straight line without diffusing thru the rad.

hot_rod

Steamhead said:

Check the link I posted. Yes, it is definitely possible to over-pump a radiator. Remember, the shortest distance between two points is a straight line, and if the flow rate is too high the water will simply follow that straight line without diffusing thru the rad.

I'm trying to find that number for over pumping. From my basic testing this week I don't see how a radiator piped with 1/2" pex, at least this 20 section radiator, could be over-pump, with a 80W typical sixed circ?

I had to series a couple pumps to get up to a 3.5 gpm flow rate, that includes the radiant manifold and around 20' of 1/2" a couple flow meters. I'll run a simulation on that circuit to come up with a pressure drop.

Possibly at some flow rate the fluid would blast through the lower connections, and that may be possible with radiators piped with 1-1/4 or 1-1/2" pipe and a large enough high flow circ.

I'm not following Harvey's explanation, I thought gravity powered system relied on the buoyancy of the warmer, lighter water, and the vertical rise. So upper levels needed to be choked down to get lower level rads to heat. How does gravity work in this case that it allows hotter fluid to flow downward?

As far as velocity drop, we have done extensive testing and engineering on that concept for the hydroseparator design. In smaller size seps the body needs to be 3 times larger that the connection piping, and about the best we get is .46 fps in that wide chamber. There can never be 0 pressure drop, even in a 1" section of pipe, as learned from P/S closely spaced tee "ghost flow"

Gordy

I would be more interested in how low you can go. What happens at .25,.5,.75 gpm. If you can nail down the top end, and the bottom end. It would be safe to predict what flow rates a properly piped, and functioning gravity flow systems flow rates actually were.....with out a circ.

Harvey is correct the gravity circulation is occurring with in the rad. If over pumped each column of the rad starts to act like a pair of closely spaced t's.

Once hot water buoyancy in the rad takes effect it is essentially pushing the cooler water out of the return port. Think hot water, and cold water flowing past each other inside the same pipe like an expansion tank, and the reason for the Bell and gossett airtrol tank fittings invention.

Steamhead

hot rod said:

Steamhead said:

Check the link I posted. Yes, it is definitely possible to over-pump a radiator. Remember, the shortest distance between two points is a straight line, and if the flow rate is too high the water will simply follow that straight line without diffusing thru the rad.

I'm trying to find that number for over pumping. From my basic testing this week I don't see how a radiator piped with 1/2" pex, at least this 20 section radiator, could be over-pump, with a 80W typical sixed circ?

(snip)

Possibly at some flow rate the fluid would blast through the lower connections, and that may be possible with radiators piped with 1-1/4 or 1-1/2" pipe and a large enough high flow circ.

Right. We generally see this effect on converted gravity systems, where the large pipes offer so little resistance to flow that we really have to be careful what circ we use.

Paul48

I think what's interesting to note, although it is harder to see in hot rod's pictures vs. Steamheads, the first section or 2 got considerably hotter, faster. Does an increase in flow result in sequential heating section by section vs. gradual top to bottom. And what is the net gain of either?

Paul48

With the Second Law taken into consideration, what flow rate would hamper convective flow?

vicbrick

Nice photos, thanks for sharing!

One other mechanism that might help explain uniformity of temperature across the radiators and the migration of heat down the columns:
The thermal conductivity (k) of cast iron is about 100X that of water. The hot water rises to the top of the rad, the heat/energy is transferred to cast iron, the energy flows down through the thick walled cast iron columns.

something to think about anyhow

hot_rod

Same radiator at 2 gpm, 120F supply, 17 minute run.

Gordy

I don't think the cast iron is that conductive in a 17 min. Time frame. Only getting hot at the top.

Paul48

Some of my radiators are hot within a minute of the circulator running. My system is a gravity conversion, but I am not sure of the flow.

Gordy

Don't know but it looks like the biggest jump is 72 - 96 degrees initially in what ever time the second image was taken. From there the climb is pretty small 4-6 degrees in each image after.

It would seem the 2gpm test has better results verses the 1 gpm.

Paul48

I'd say so, also. Doubling the flow, cut the time it took to completely heat the radiator, by well over 2X's. Notice the first couple sections getting hot first is even more apparent at the 2 gpm. I think this would always be present because of the change in velocity. It would be interesting to know the btu output at the various stages, regardless of how even the appearance of heating.

hot_rod

Paul48 said:

I'd say so, also. Doubling the flow, cut the time it took to completely heat the radiator, by well over 2X's. Notice the first couple sections getting hot first is even more apparent at the 2 gpm. I think this would always be present because of the change in velocity. It would be interesting to know the btu output at the various stages, regardless of how even the appearance of heating.

No question the increased flow quickens the output. I'd like to try 6 gpm which would be 4 fps in a 3/4" supply tube
and 12 gpm which is 4 fps in a 1" supply tube, and to pump it with a 100W or less circ pump.

Paul48

hot rod......If the radiator starts to shudder, don't go any higher with the flow.