Bad Physics/Thermodynamics?

[sheeny]

I just got the latest edition of Australian Sky and Telescope. On page 82 is a half page article on "Preventing Dew Problems". I am sceptical about the explanation of how a dew shield works. Let me explain...

As a professional mechanical engineer I have studied thermodynamics and physics at undergraduate level, and applied same to numerous problems in twenty something years of engineering practice.

The explanation for how a dew shield works, according to the article, is due to radiation loss to the "chill of space". A dew shield reduces exposure to the sky and so reduces heat loss due to radiation. My initial reaction was "No way!". BTW I have seen this explanation before on the web, but passed it off as just mis-information.

So I did a little bit of searching on the net, only to find this seems to be widely accepted by the astronomical fraternity!

I am prepared to concede I can be wrong. But the more I look at the problem the more I doubt this explanation.

Firstly, the lens is not exposed directly to space. It is exposed to the atmosphere which in turn is exposed to space. Heat loss via radiation from the atmosphere to space is not disputed here!:) ...but to calculate the heat loss from the lens to space it is necessary to analyse a radiation network (in the case of radiation losses) and indeed a full heat transfer network. BUt just to keep it simple, lets just consider radiation for now...

All matter emits radiation. The frequencies of the radiation is determined by the material, and the intensity of the various frequencies is determined by the temperature. Radiation heat transfer equations are based on the difference of temperature between two materials, and other factors for emissivities, absorbtivities, and turbidity of the intervening space, etc.

The radiation network from lens to space is a triangle. Some of the radiation leaving the lens goes to space and some goes to the atmosphere (keeping this simple). The third side of the triangle is radiation from the atmosphere to space. As the temperature difference between the atmosphere and the lens is minimal, there is minimal radiation loss if any.

But the atmosphere contains carbon dioxide and water vapour (also known as green house gases) and these absorb most long wave infrared radiation such as would be emitted from our lens. So while the path from our lens to the sky is optically clear, at infra-red frequencies generated by materials at normal ambient temperatures, it is not. So the large majority of the radiation from the lens goes to the atmosphere, not directly to space, and since the atmosphere is at a very similar temperature, the net radiation loss is insignificant.

Now it's all very well to write this and pick the "radiation loss" theory to bits, but let me follow it up by explaining how I believe a dew shield works (I'm prepared to believe it up until it's proven otherwise!;) ).

The dew shield works by two mechanisms. Firstly it effectively stops the movement of air across the face of the lens. In other words it traps a pocket of air in front of the lens and stops it from moving too much. As the dew point in the air is reached, the water vapour in the air immediately adjacent the lens condenses on the lens but initially this is not visually detectible. The rate of condennsation and evaporation at the lens reaches an equilibrium following Le Chatelier's principle. With the air trapped in fromnt of the lens the rate of transfer of water vapour to or away from the lens is limited to diffusion, slowing the rate of condensation on the lens, becasue as the temperature drops further, water vapour diffuses out into the moving air stream at the end of the dew shield or it supersaturates until a dew droplet is formed in the air (which then falls down).

So the first mechanism the dew shield uses is to reduce the rate of dew formation on the lens by limiting air movement.

The second mechanism is to act as an umbrella. When air cools below the dew point, dew will condense on anything it can. It it can't condense, on a solid like the lens it will eventually condense on dust in the air, etc forming dew droplets. Once formed, these rapidly grow in size and settle due to gravity. So unless the scope is pointing up steeply, the dew shield will catch a lot of the falling dew and so shield the lens.

So that's my explanation of how a dew shield works, but what's more I propose an experiment and some measurements that can be done to see what's what...

If the radiation loss theory is dominant, the lens should be significantly colder than the surroundings. Some references I have seen suggest this is the case, so I suggest this should be measurable. I don't have the means to take these measurements accurately enough, unfortunately, but I propose another experiment that could be easier to manage.

The experiment goes like this: set up a series of glass plates exposed to the night sky.
a) One exposed to the night sky with no dew shield or cover.
b) One horizontal with a dew shield, and
c) One vertical with a dew shield.
d) One horizontal with a cover the same distance away from the glass as the end of the dew shields, but sized to shield the same proportion of sky as a dew shield, and
e) a similar arrangement but horizontal.

If the radiation loss mechanism is dominant, the unshielded glass will dew up sooner and more heavily than the others.

If my theory of how it works is correct, c) and d) will dew more slowly / less than the others.

:wink2: Now I really doubt that I'll get around to doing the experiment, but I'd like to. If anyone knows of sound scientific analysis or experimental results that "prove" radiation loss is the way a dew shield works, I'd really like to see it. At the moment, I believe it falls into the category of an "astro" myth...

I hope I have got you thinking about it! Such a humble piece of equipment, the dew shield...

Al.

[avandonk]

It is due to radiation loss (IR). The temperature of a dark sky is ?(3.5K) colder than your optics.

Hence the phenomena of frost when the ambient air temperature is 2 or 3 degrees on a clear night.

No frost on a cloudy night even when the air temperature is 1deg C.

This all depends on RH and Temp.

Microclimate is the major consideration. Even a slight local warming can stop condensation. The mechanism
depends on the design. All your scenarios are correct. Unfortunately there is no simple answer without ruining
seeing.

Even the smallest temperature variation can give you condensation or not. The other critical factor is nucleation
on the surface.

A very carefully worded question.

Bert


Bert

[bird]

G'Day sheeny,

believe it or not, the explanation of heat loss by direct radiation into space is correct. Materials lose heat by radiation at a number of wavelengths , going out into the long infra-red, and many of those wavelengths are transparent to the atmosphere, meaning that the material "sees" the 3 degree kelvin temperature of space more significantly than the warmer air right next to it.

This effect changes with different materials, and it's quite easy to do an experiment for yourself and see this in action, try placing square pieces of steel and aluminium (the same size) next to each other under an open night sky. The steel will start collecting dew very quickly, but the aluminium will stay dry (ie warmer) for a lot longer.

The explanation is that aluminium radiates heat at many wavelengths, some of which are absorbed by air, so it doesn't "see" the open night sky, whereas steel radiates at wavelengths that are mostly transparent to the air.

This is also the reason that dew forms on the ground, and frost can form on the ground well before the air temperature is low enough to have fog.

There's no problem with thermodynamics etc, as during the day we *absorb* heat through the air by direct radiation from the sun in addition to the normal convective warming from the air. The earth is not a closed system, we exchange heat with outer space all the time.

By the way, I have many graphs that show this effect - I have a temperature sensor on my aluminiun tube, and an ambient air sensor nearby. It is quite normal for me to see the tube temperature 1 or 2 degrees C below the air temperature. If you did this experiment with other materials then I believe you would see a wide variation, with some materials going 4 or 5 degrees below ambient air temp.

The graphs also clearly show what happens when cloud comes over - the ambient and tube temperatures close up and become identical as the tube can no longer "see" through the cloud.

regards, Bird

[janoskiss]

Do the experiment Al! There is nothing as convincing as seeing it for yourself. ;)

[sheeny]

Bird! Thanks, mate! Glad to see you have some facts. I'm happy to admit I have learned something. At least you are doing something that demonstrates the effect - it is obviously more significant than I would have thought, and I can't fault your explanation - it fits with the generalised text book data I have available to me.

So radiation is significant... I guess I'm going to have to do some sort of an experiment if I want to determine whether it is dominant - that's a different matter. I would still expect the mechanisms I described to be at least as significant as radiation loss. The experiment I described before may still have some merit, but it will need some refinement. Measurement will be the real issue!

BTW radiation shields are a fairly common device for controlling radiation. If the dew shield is truly primarily intended as a radiation shield, it will perform best if it is polished metal - not a dull matt colour... I guess there's a compromise there to minimise reflections in the OTA!:wink2:

BTW Bird, have you ever tried to measure the difference in your OTA temperature with and without the dew shield? Perhaps remove the dew shield after the scope has stabilised and see if the temp drops???

Al.

[bird]

Al, might sound a bit silly, but I don't have a dew shield, my scope is a 1.8m long newtonian monster with the secondary a fair way down from the open end so it doesn't tend to get dewed up.

I'm quite paranoid about cooling the whole scope down to exact thermal equilibrium with the air for best planetary imaging, and a heated dewshield would upset that, so I decided a while back to just avoid them altogether.

When I do get dew on the secondary I have a hand-held warm air blower (i.e. hair dryer) that removes it in a few seconds.

regards, Bird

[Miaplacidus]

Bird, how long, roughly, after you've used the hair dryer does it take to dew up again?

Brian

[Starkler]

I have also seen mention that the colour of paint used on a telescope tube makes a difference. IIRC, a white OTA will radiate more and get colder than a black one.

I wonder if there have been any comprehensive studies done on materials used commonly in telescopes.

[iceman]

Great thread! Excellent learning stuff in here.

[bird]

I did find some links to information on this stuff in 2004 when I was researching it. I'd just built my temp monitoring stuff and I was wondering why the tube was getting so cold (I had a steel tube painted white, about the worst possible design).

A quick search on the keywords "emissiviry" and "supercooling" turned up this gem of an article from the Gemini telescope project, read section 3.2:

http://64.233.179.104/search?q=cache:fZqecCviZTAJ:www.gem ini.edu/documentation/webdocs/preprints/gpre5.ps+emissivity+astronomy+super cooling&hl=en&ct=clnk&cd=3

regards, Bird

[bird]

The same article is also available as a PDF, with pictures!

http://www.gemini.edu/files/pio/newsletters/nwsltr12_6-1996.pdf

Read section 2 on Image Quality, it's very interesting...

And another article...

http://astro.umsystem.edu/atm/ARCHIVES/JUN99/msg00906.html

Bird

[avandonk]

The main way a dew shield works is by limiting the solid angle of sky 'seen' by any element of an optic surface (lens or corrector plate). Without a dew shield the optic sees a hemisphere (2Pi steradians) of cold sky. With a dew shield this area is reduced to a much smaller area of cold sky. This can then lower supercooling but not eliminate it.
By gently heating the dew shield, radiation from the dew shield to the optic can then compensate for heat loss from the optic to the sky thus eliminating supercooling and as a result no condensation.

Bert

[bird]

Normally I have to do this only once every session, I think the air that started in the tube is the source of the moisture. Because the tube is closed around the primary there is not really any airflow into or out of the tube even at the top end, so once the air in the tube has shed its moisture then things tend to stay fairly stable.

regards, Bird

[davidpretorius]

great question, have no idea, i have sears and salinger out and don't know where to start.

great thread!

[avandonk]

I do exactly the same thing with my 300mm F2.8L , a short hair dryer ($10)blow and all is fine for at least an hour. My GF thought I was about to look after my appearance when I bought the hair dryer. There are too many variables to control it in real time. Just be aware of the problem and the causes and dont over react.

Bert

[sheeny]

Been thinking some more about this (as I do).

Radiation shields are well understood technology. Why is no one (big generalisation here on my part!:wink2: ) using radiation shields for the rest of the OTA? It is very simple to do.

A space blanket is a simple radiation shield that can be easily adapted to reduce radiation loss from the OTA at least in directions you don't have to see through. So if you have a steel OTA and are worrying about radiation loss cooling you OTA, wrap it in a space blanket or two.

A radiation shield can be shown to reduce radiation transfer by 50%. The theory says that if you use n radiation shields between source and receiver, the radiation loss is 1/(n+1) of the radiation loss without a shield.

So, Bird, I reckon this could reduce you radiation loss from your OTA significantly without adding heat to upset your optics...

One possible down side is that the space blanket, if it is sealed to the OTA would also limit convection, which could possibly limit the transfer of heat from the surrounding to the OTA... again I think a trial is what is needed to prove the point.

I did toy with the idea of combining a space blanjket with a close cell foam mat made to suit the OTA (like a huge stubby cooler... but with a zip or something to make it easier to install:) ) but that would reduce conduction as well as convection and radiation, reduceing heat from the air to the OTA. If your losing it by radiation to the sky, you need to replace it in the OTA from the air, so not likely to be a good idea...

So I would suggest just a space blanket or 2 draped over the OTA - it shields the OTA from the sky but not the ground (assuming it is warmer than your scope!)

Al.

[sheeny]

Bert,

You talking about your appearance or the dew?;)

Al.

[bird]

Al, I've thought about trying something like this, not got around to it just yet...

My OTA is 16" in diameter and the primary mirror is 13.1", so there's a lot of clearance between the mirror and the inside of the tube - I reckon this is probably enough to keep the residual tube currents out of the optical path. Also I've spray-painted the interior of the OTA with matt black paint, my dodgy idea is that this paint is a little bit insulating, so the interior surface of this paint is likely to be closer to the temp of the air in the tube than the outside of the tube, but I have no way of measuring this as yet.

My temp probe on the tube is attached to the *outside* of the tube, which may not be the most relevant place. I'm considering attaching it to the inside of the tube so that I can see the real difference between the tube and the internal air.

regards, Bird

[sheeny]

Bird,

Not sure how your matt black paint would preform in terms of "insulating" properties... it depends if it is "black" in the infra-red part of the spectrum or not.

When "black body radiation" is discussed, it means that all incident radiation falling on the body is absorbed. The absorbtivity is defined as 1. But as Kirchhoff shows the emissivity and the absorbtivity are equal, the emissivity of a "black body" is also 1. So a "black body" emits the maximum radiation possible at a given temperature.

If your matt black paint is a heat dissipating paint, such as those used to paint engine parts and exhaust systems on motorcycles, the paint has maximum absorbtivity and emissivity in the infra-red region - which is not what you say you want. As for any other sort of matt black paint, however, your guess is as good as mine! Just because it's black in the visible spectrum, doesn't mean it's black in infra-red...

I'm not sure that you really want insulation on the OTA. If you are losing heat through radiation it is the heat transferred into the OTA by Conduction and Convection that is maintaining the balance... Insulate the tube, and the temperature in the tube will go down till a new balance is found. Shield the radiation loss, and minimise other insulation effects and the balance point comes closer to ambient temperature... which is what you want.

So what you really want on the inside of your OTA is:

a matt black paint (to minimise optical reflections)
that is conductive (to minimise thermal insulating effects - I'm not sure how important this is!)
that is "white" in the infra-red part of the spectrum (to minimise radiation losses)I have no idea if such a beast exists!:shrug: ...but I've enjoyed the mental exercise of working through the problem!;)

Al.

[bird]

Al, on the inside of the tube there are no radiative losses, as there is nowhere for radiation to go except back into a different part of the tube, so no nett loss.

So perhaps the third item in your list is not so important? Not sure.

Bird

[Starkler]

Perhaps velvet type flocking would provide a desireable insulating effect for the inside of the tube ?

[bird]

I used black velvet on the inside of my previous scope - the one I had with me at last years Snake Valley camp Geoff - and it seemed to work ok. The only hassle was a slight amount of shedding of fibres that would collect on the primary mirror.

Bird

[Starkler]

So, ideally we want a dew shield that is opaque to all wavelengths of IR that covers the entire ota , and allows air to circulate between it and the ota. Does this sound right?

[bird]

It could be done as a kind of "outer OTA" that is larger than the main tube, say leaving a 20mm gap between them. It would be longer than the main ota as well to act as a dew shield.

If it were made from lightweight plastic then it would be easy to fit and remove, I might look at experimenting with one, or maybe a partial outer shield.

regards, Bird

[sheeny]

There has to be, Bird. If you can see the inside surfaces of your OTA when you look in the objective end, then so can the sky. Now, of course most of the radiation from the inside of the tube, goes to the other side of the tube, etc but a proportion is lost out the end. That proportion is defined by what's called the shape factor of the tube, and will depend on the open area at the end of the tube, the total surface area inside the tube and the orientation of surfaces inside the tube (which is pretty simple in a newt!)

Al.

[sheeny]

Yeah guys, you're on the money now, but I don't know how "plastic" would perform... I only have experience with metallic radiation shields. I would suspect that is plastic was an option then first aid space blankets would be plastic rather than metal foil (or at least metal foil coated).

A partial radiation shield would work well for a dob or alt-az mount, but for an EQ mount rotating the partial shield around the OTA could be a pain. A full shield may be more appropriate for EQs.

The gap between the OTA and the shield can in theory be as smal as you like. The smaller it is the more significant conduction and convection transfer between the shield and OTA becomes though.

I still think a simple space blanket is a great place to start...

Al.

[janoskiss]

Visible light and the cosmic microwave background are very different wavelengths, so you cannot rely on visual appearance.

[bird]

Yeah, I know Al, but I was just assuming that radiative losses out the open end would be very small, hopefully small enough to ignore (i.e. much less significant than convective transfer in and out the same opening).

regards, Bird

[sheeny]

Yeah Bird,

Its a case of proverbial low hanging fruit. Radiation losses out the front of you OTA will be much less than the radiation losses from the outside of your OTA... you can't argue with the shape factors there!;) I just did some rough calcs based in areas alone, not properly modelled or anything serious like that, but to give a rough idea the radiation loss out the front of the OTA is just 3% of the radiation loss from the outside of the rest of the OTA (assuming the OTA is 5 x D long, and the sky extends 180 deg from horizon to horizon). Very rough calc, but that should put the radiation loss out the end of your OTA into perspective. In shorter OTAs the % loss out the front is higher, of course.

How this compares with the conduction and convection gain (it must be a gain if the OTA is cooler than the atmosphere!), through the opening... I don't know - it would require some proper modelling and real measurements. I'm not sure that that really matters, however, the point is, since you convinced me that the OTA is cooler than the atmosphere, heat transfers to the OTA from the air via conduction and convection, and is lost to the sky by radiation.

So, the biggest bang for your buck should be in reducing radiation heat loss from the rest of the OTA, and what is lost out the front is a very small amount.... which takes us damn near full circle to my original post:wink2: ! ... except that I have learned that radiation loss is real, because you can measure it.

I will play with a space blanket in future, but I'd love you to do some experiments with your rig, Bird, since you are set up to actually measure the effect!

Al.

[sheeny]

Very good point Janokiss! You forced me to go and do some homework;) ...

Materials at normal ambient temperatures, say -10 deg C to 40 deg C emit radiation in the mid-infrared band, roughly 3-8 microns. Now this is well below the diameter of an OTA, so this frequency can be transmitted through the opening of the OTA.

FTWDK a hole of comparable size or smaller than the wavelength can't be resolved by that frequency of light, so no hole is detected. This is why radio telescopes can get away with perforated or wire mesh reflectors.

BTW, Wikipedia has some great graphs on atmospheric absorbtion and transmission of IR... better than in any of my textbooks!
http://en.wikipedia.org/wiki/Infrared
The internet is a great thing! I wish it was around when I was going to uni...

Al.

[bird]

Al, I thought you'd like to see this - it's the temperature graph from my observing session on the 18th of this month.

Legend: RED is the mirror, GREEN is the tube, PURPLE is the ambient air.

Notes: You can see the tube tracks consistently lower than the ambient, not by a lot but measurable, about 0.5C fairly consistently.

Just before 9pm (2100) a bubble of warmer air came across, I see this happen quite regularly in Canberra summers. These bubbles can be a couple of degrees warmer than the rest of the air, and sometimes they even show up on the weather radar as "rain" even though there's not a cloud in the sky... very interesting.

The great thing to see from this graph is the lag in responce from the tube (green) to the air (purple) temp changes - this shows the thermal intertia in these systems very nicely!

At about 11.30 (2330) the temperatures leveled off, and the mirror cooling system switches off.

The red line jumps a little before it levels off because the temp sensor is on the outside of the mirror, and when it's cooling rapidly the temperature inside the glass is a little higher. When the cooling is switched off then the mirror temp evens out, and so the skin temperature goes up slightly.

regards, Bird

[sheeny]

Thanks Bird. Very nice!:prey2: With the setup you have I reckon you should be able to see the effects of a radiation shield over your OTA very clearly! I would expect one carefully placed space blanket should drop that difference to 1/4 of a degree...

I assume you are using thermocouples and A/D converter(s) into your laptop for capture?

When the mirror temp starts getting below the tube temp by, say a degree, you're not tempted to turn off the mirror cooling? I guess there's often better things to do about these times... like astronomy!...;)

Al.

[bird]

Actually Al, I'm using digital sensors, part number DS18B20, look em up on google, they're very cool.

They are connected to a microprocessor kit from www.microzed.com.au, and from there a serial port goes back to my main PC to record the temperatures.

Mirror cooling is a bit like altitude for aircraft pilots - the more you have the better. I have been bitten many times by turning off the cooling too soon and then watching as the ambient temp drops past my mirror temp and so I have to turn it back on again...

It's much better to overshoot by a fairly long way, more than 1 degree even, and then just switch it all off and wait for equilibrium.

Also, there's always some residual heat inside the mirror, so overshooting helps to compensate for that.

When the cooling is running the scope is pretty much unusable. The combination of cold air currents and mirror flex from the uneven thermal gradients in the glass turn the images in a nice blurry mess. It takes about 30 minutes *after* I turn off the cooling before the scope starts to work properly again, so when the cooling is running I'm usually asleep inside (on the lounge!) an the computer is in control of watching the temps.

regards, Bird

[sheeny]

They are pretty cool aren't they? I haven't done any microprocessor programming since I used to hack about with a MicroBee!:) They both look like impressive bits of gear. +/- 0.5 degree is great!



Ah, well that explains it... you do have more important things to do... sleep.

Al.