I want the smallest possible radiator. Are there any almost silent fan assisted designs available?

[Aidan]

Yes, I should.

If one did know better, how would those who know no better know that you knew better than they?

[Ian White]

You are Donald Rumsfeld and I claim my 5 pounds.

[meow2222]

and would rather justify than go back and think it through properly

NT

[meow2222]

Quite correct. But the conclusions you drew when taking your thinking further were in error.

Any shape of domestic rad at all will dissipate more heat when fanned than when not. Your notion that heat transfer from water to metal was the limiting factor is not so with unfinned rads - nor even with finned. If you pass a huricane over it, it will be, but at the very low air speeds a bank of small silent fans create, not so. One only need triple airflow speed, which isnt much, to see a sizeable heat outut increase. Such arrangements never approach the output per area of a rad designed for fanning, and fanned at speed, but this does provide a way to have a small rad give a much bigger output, which is useful a) to avoid large inconvenient rads being installed, but maintain the look of traditional rads, b) to boost output from already fitted undersized rads. And PC fans are cheap simple and easy to install, running them in series pairs on 12v.

NT

[Aidan]

And now you are going to correct me. Oh, dear.

You have dabbled with a few radiators and you believe that you know something about heat exchangers. You think that sticking fans on radiators is a cunning plan. You are deluded.

The only reason you'd want to boost the output from a rad is that you've made a c*ck-up of the heating installation in the first place, which isn't something that really enhances your engineering credibility. You wound up with a badly designed heating system and lots of little =A33 propeller fans whirring away on top of the undersized radiators. It must have looked really cool. And now you're going to pontificate at me?

I studied heat exchangers a long time ago, stuff about parallel-flow, contra-flow, log mean temperature difference, change of phase, extended surfaces, etc.. I've forgotten most of it. I had a considerable involvement with fans and heat exchangers in my career. The result of this is that I think that sticking fans on radiators is a daft idea.

This is interesting. What instruments did you use to measure the air-flow? Did you fit a hood over the radiator? Where did you obtain the ducting for the fan outlet? How did you ensure uniform air flow over the convector?

There are lots of fanned convector heaters in use. To the best of my knowledge, the last ones that used radiators in the ducting, or as a part of the ducting, date from about the 1900s. If you look at any of the more modern ones you will probably observe that they use finned tube heat exchangers and most do not use axial flow fans. Doesn't this tell you anything?

If you want a convector heater you'd do best to get the radiator out of the way as a first step to installing one.

[meow2222]

no point

ad hominem dont cut it, nor do dumb assumptions.

incorrect assumption

incorect assumption.

But you cant be botherd to acutally think it through properly, hence your assumptions.

why would a domestic rad need uniform airflow??

yeees, we covered that one already I believe.

depends what you want.

NT

[Aidan]

[Aidan]

I have thought it through as far as I want to. I think it is a daft idea. I would have made the heating system work properly. I do not intend to excavate ancient college notes to provide you with some calculations about convector heaters. I really can't be bothered.

Yes, why indeed? So you have a 3" wide fan blowing air up a 2 foot wide double panel radiator; sounds great.

What instruments did you use to measure the air-flow? Did you fit a hood over the radiator? Where did you obtain the ducting for the fan outlet?

The technical knowledge underpinning this scheme does seem to be rather thinly spread.

It does. Do you want a radiator or a convector heater? I have noticed that they are very different.

[The Natural Philosopher]

Er, no, he isn't.

As one who has extensively dabbled with getting rid of heat out of seminbcinductor devices using anything from a piecce of flat blac alumnium to a full fan blown heastink, I can assure you that te rate of heat loss from a plane metal surcfdace is almost TOTALLY a function of not only the area and teh ttemerature differential, but of the airflow over it.

Radiators work by 'natural' convection. So do semiconductor heatskinks. And in both cases forcing the air over allows a huge amount more heat to be transferred. Special desgin - finning and so on - helps, because it increases surface area. This is not done much with convectors as it restricts convection - so beyond a certain point it makes no difference, unless you blow the heatsink or radiator.

That is waht a typical fan blown radiator comes with a lot ofextra finning: It can make use iof it - but it does not mean that fan assisting a simple flat plate radiator will not imnprove its output by several times. It will.

A typical simple flat plate heastink can improve thermal transfer 5-10 times with a modest airflow over it.

You can even feel the effect on your skin - a mild breeze makes you feel colder. Even in 'tropical saturated humidity air' where evaporative cooling is negligible.

No, you are pontificating to him.

Maybe he didn't WANT big radiators. I didn't and I get enough from a little 1/4 square meter blower as I would from a meter and a half by half a meter double finned monstrosity...

It shows.

Well its ugly, but in extremis it works.

Its better to buy a unit designed for it, true.

they are not overly complex beasts..

Tht, if you HAD remembered your basic thermal courses, is a fairly basic equation. Rate of heat loss from a plane surface versus flow rate of air over it.

It does indeed show that whislt vertical surfaces area couple oftimes better than flat due to better convection, moving from convection and a farcation of a meter per second to something approaching a meter per second or more makes a huge difference. As anyone who has had a fan belt go in a car can tell you ;)

It tells you that you get more efficiency in the overall package that way.

HOWEVER the best semiconductor cooling packages use axial fans in short internally finned heatsinks. One or two use rotary fans. Its true that once you blow, adding fins makes a lot more difference, but they are not essential to the principle. And they cost money.

Depends. For a quick and dirty bodge, playing a cold air fan on a radiator will increae its output several times.

Obviously a properly designed fanvector is better...

[meow2222]

Yes, and its not enough for this case.

which heating system? And what is improper about a low speed fanned rad that produces the desired output?

One more silly assumption.

no we both have enough of that for this one, what you lack is willingness to pay enough attention to do more than make thoughtless assumptions. Your assumptions are repeatedly right off the target.

NT

[meow2222]

no

yes

no

Never gives up with the assumptions does he.

precisely. And that is what I was discussing with Aiden, an unfinned rad. As he rightly says, finned rads wont give the same amount of gain, and at high fan speed the fin ends will be cold. Unfinned rads do not suffer from this issue nearly as much, and output can be got up to several times with just modest airflow. That can be done at low cost, quietly, and while maintaining a traditional rad look.

Given those points, it has its place.

Pontificating, and making up a lot of nonsense. Straw man and poisoning the well. Classic debate manipulating techniques, with all the logical validity of a lead airplane.

me too

PC fans are small enough to either go out of view, or sit behind a neat and tidy narrow strip.

FWIW its also poss to use a flush in-wall fan behind a rad, but that has a lot less useful applications.

circumstances, as heat output per square meter isnt the only criterion.

indeed. If he'd paid attention he'd have noticed at the start I told him I'd done it already and it did as expected. Its a fool that dismisses all evidence that doesnt agree with his thinking.

NT

[Aidan]

Another one; there must be a rally in town.

approach seems to be to slap on the biggest fan that fits over the heatsink/IC/whatever. So if you were to adopt the standard industry approach and fit a 500mm diameter fan to the radiator, I do agree that we'd probably get some useful heat movement.

Granny, eggs, etc..

Now do try to get a grip on what he's doing. Unlike you, he is using a fan which is about 2% of the area of the heat emitter. He's using 80mm axial flow fans to try to shift an amount of watts approaching or exceeding 4 figures.

OK then, lets have a look at what has been proposed and see if we can anticipate any problems or suggest any improvements.

I think you'd need some ducting or housing for the fan. At present, much of the air moving through the fan is moving from the high pressure outlet back to the low pressure inlet, short circuiting the system. You could fit the fan(s) into a flat plate, maybe with a tapered section to fit onto the top of the radiator. You'd also need some sides on the radiator to reduce air leakage.

This will improve the air flow but it's completely buggered any natural convection that might have been taking place, like those stupid radiator covers. NT's undersized rads would been incapable of delivering the peak load, but would have been adequate for much of the heating season. Not any more. Now you need the stupid little fans on virtually all the time.

The I anticipate a problem with the airflow. If you measured the air flow through the rad, like NT hasn't, I suspect you'd find most of the air movement is directly under the fan (or over it, depending on how you've set it up). You could make up some sort of baffle plate, to rest on top of the fins and restrict the airflow directly under the fan. This would reduce the overall airflow, but would increase the heat transfer by ensuring a more uniform air distribution over the heat emitter.

This, of course, brings us to another problem, doesn't it? A big problem. You're the expert NP, care to fill us in here?

You bought a fan convector, then. Good move.

Yes, it is. Which is what I have been saying.

[Aidan]

Really? Hmmmmmmmmmmmmmmmmmmmmmmmmmm.

I may be confusing you with someone else here, but do you, by any chance, work somewhere near Great Portland Street?

You have done it. You think it works and this confirms to you that you were right. You haven't made any measurements to support this. I think it would increase the heat output, but probably isn't worth the effort.

[Capitol]

Obviously the view of an inexperienced incompetent IMM I presume.

Regards Capitol

[Capitol]

Not IME. Fan convectors suffer from the lack of thermal capacity in the very small radiator. ie when the water flow stops, there is no adequate heat output.

Regards Capitol

[meow2222]

nope

A bank of them is used. The rad already shifts that much heat with the fans off.

nope

Only right at the ends this will occur, even the shortest available airpath is long enough for the job to be done for all the other fans.

If you want, I just used the fans. A row of fans doesnt allow any air to short circuit them, as theyre square chassis not round, and large enough to fully cover the available space.

Youre so busy making so many assumptions.

no.

no, stationary fans permit that, though it will reduce some. This of course enables greater room by room thermostatic control: fan off, reduced output, fan on, higher output. Better control results in fuel savings.

nope. Yawn. It was precisely to deal with peak requirement that I first did this.

no and no.

Consider the behind-rad airpath as a uniform resistive block. Not hard to see the pattern that emerges.

who?

we agreed those are more effective way back. That doesnt of course make them the prime choice in every case, since output per area is not the only criterion.

NT

[Aidan]

You have a line of fans along the top of the radiators. You cannot be serious.

You are serious?

Yes. I had assumed some intelligent life forms.

Maybe not you. Everyone else does. You must write to the manufacturers of air-cooled chillers, they've been doing it all wrong for a long time.

Uniform resistive block? What block? What resistance?

Ah-ha! I can diagnose your complaint now; Overclockers' type geek-speak bullshit. Caused by lurking at computer forums. Sufferers believe that computer fans are the only true fans, that adherents of other fans are heretics and that they know what they are talking about. You should know better than regurgitate such bollocks. It made little sense in a computer and it is nonsense in this application.

There is no block, there is negligible resistance and there is bog all heat transfer.

NP, Natural Philosopher. He went silent after being asked to share his technical knowledge with the forum.

You need to try to look at the thing you've created with unblinkered eyes. It's your baby and although you may love it, it's an ugly mongrel and an embarrassment. Can we see a picture?

[Capitol]

No, it's dIMM

Regards Capitol

[meow2222]

Nearly forgot about this one.

no

no.

no, theyre doing it the way that gives max result for min cost to them. Nothing wrong with either way

The airpath behind the rad has resistance to air movement. If you know what you claim, I'm sure you could then readily see what path the air will take when there are no sides fitted. But given your comments, apparently not.

Hint: Its the same pattern current takes when passed through a uniform resistance block of similiar shape. The 2 are analogous.

Hint: approx the same path as magnetism in a smiliar shaped iron block... getting there yet?

just one small correction: the above nonsense came out of your mind, not mine.

hmmm

there is enough to determine the airflow path

Strange how I was so warm then.

There is no big problem. It worked fine.

nope and nope.

nope, not ugly in the least. Not even visible.

nope.

If I had a pic, all youd see would be a normal radiator.

NT

[Aidan]

What gems of wisdom can have taken you 2 days to concoct?

That airpath would have some resistance.

Yes, I can see perfectly. The air takes the path of least resistance. Air flows in from the sides, avoiding the constrained narrow channel behind the radiator. Heat transfer is reduced.

There is no "similar shaped iron block". There is no block. There is an open-sided channel formed by a steel panel an inch or two away from the wall.

All yours, not mine. I'd never before heard the term "resistive block". A Google search suggests it does not appear to have any connection with real-world thermodynamics, but it does feature in some geeky amateur computer cooling forums. In that context, it seems to be a homemade copper block heatsink with a central airchannels and smaller radial drillings. The concept bears no relation to your radiator set-up, which has no block, no channels, no resistance and bog all heat transfer, except that from the flat panel forming the back of the radiator. You've spent too long lurking on those forums and have tried to apply their small-scale, half-baked ideas to a much larger scale.

There are many narrow air channels between the fins on finned tube heat emitters, which are what are used by most manufacturers of convector heaters and who, presumably, know what they are doing. Unlike you in all respects.

Yes. A large proportion of the air by-passes the heat emitter. What are you telling us? The air obligingly follows the path you want it to, the path of maximum resistance? You should borrow a fluid dynamics book from the library.

Probably due to the electrical consumption of all those fans and the heat generated by their churning the cold air around. Or the warm glow of your embarrassment, sitting around with your ear-defenders on.

Big problem. You just have no clue as to what it might have been. You think it worked, you made no objective measurements. You could have made it work a lot better if you'd known what you were doing. You have no idea.

Shame we'll never see it then.