Thermal analysis of a 25" dob (preliminary)

[clive milne]

I have to shoot off to work in the next hour or so, but before I do I thought I would put up some thermal infrared images taken last night showing the heat flow over a few hours of Andrews 25" obsession. I'll have a nibble at giving my interpretation of what is going on, but with consideration of the time I have, this is most likely to be incomplete.

Also... there are a few caveats implicit in FLIR images, so caution in advised in rushing to conclusions.

Anyhoo... here are the pics with time stamps and ambient temperatures added for context:


enjoy,
~c

http://i371.photobucket.com/albums/oo158/clivemilne/IR_3702.jpg

http://i371.photobucket.com/albums/oo158/clivemilne/IR_3704.jpg

http://i371.photobucket.com/albums/oo158/clivemilne/IR_3705.jpg

http://i371.photobucket.com/albums/oo158/clivemilne/IR_3707.jpg

http://i371.photobucket.com/albums/oo158/clivemilne/IR_3708.jpg

http://i371.photobucket.com/albums/oo158/clivemilne/IR_3710.jpg

http://i371.photobucket.com/albums/oo158/clivemilne/IR_3714.jpg

http://i371.photobucket.com/albums/oo158/clivemilne/IR_3718.jpg

[clive milne]

Last but not least is the top cage... perhaps most significant point in this image is that the the kydex baffle is 1.5C below ambient... the spider vanes have also radiated enough heat in to space to drop below ambient as well.

http://i371.photobucket.com/albums/oo158/clivemilne/IR_3716.jpg

[clive milne]

Ahh... looks like my text editing (time and ambient for each frame) didn't upload... I'll have to get on to it when I have time..

btw) how do you embed pictures (rather than just having the link visible)?

regards,
~c

[clive milne]

One quick comment... notice how much heat is trapped in the soil under the telescope... it clearly has more thermal inertia than the primary mirror itself. I suppose the obvious heat flow path is through the ground board (and in to the telescope structure)

[alocky]

Thanks Clive;
I should point out that just about every bit of optics on the top cage has a dew heating strap, which I had switched on last night. You can even see the extra one I keep for eyepieces tucked alongside one of the upper cage tubes.
Given that we were still in shorts and sandals at midnight, I probably could have saved a bit of battery power.
Pity about the transparency, I assume we were seeing all the aerosols from the fires down south.
cheers,
Andrew.

[alocky]

Purely radiative transfer mechanism - I doubt there's any 'flow' per se.
The main culpirit will be soil moisture, there was some rain in the last day up there. There's not a lot else in the regolith that has any significant latent heat capacity.

[clive milne]

Is anyone actually interested in this, or am I just wasting my time elaborating further?

[alocky]

I'm interested of course, but it mightbe worth putting it on the Obsession owners group - with your permission.
Cheers,
Andrew.

[bmitchell82]

its very interesting indeed Clive, as it gives people a good idea on what generally happens with scope cool down times. I guess depending on your mirror substrate the 1-2 deg difference of mirror to ambient wouldn't pose much of a issue when looking at pyrex.

What are your thoughts on that?

[clive milne]

Cheers guys.. I'm in the process of putting all my thoughts down in a separate web age... 1/4 of the way through at the moment. I'll post it when finished.
regards,
~c

[clive milne]

https://sites.google.com/site/binocularnewtonian

[clive milne]

The issue isn't so much a question of heat induced distortion of the substrate as it is the convective plume at the boundary layer distorting the image through variation of the refractive index of the air.

There is a qualifier to that... if the mirror substrate hasn't been correctly annealed then you will see the mirror figure do bat**** crazy things as it equilibrates thermally. (even if it is pyrex) Basically you will see astigmatism at the focal plane come and go as the optic cools.

[bmitchell82]

so if i had the volume of the tube being drawin though and exhusted out the rear of the scope essentially sucking in the cool air this would negate these thermals wouldn't it?

[Paul Haese]

There are several of us on the site that have been onto active cooling with TEC's for the last 6 years. I had an 18" SDM that I peltier cooled via a cover that went over the rocker box and had a bottom sash to keep the cool air in the box. I found that the 18" would take around 2 hours to cool to ambient (0.5 degrees C). Natural cooling with a fan is ok but you would still be exuding a lot of energy with that scope even after many hours of using the fan. The IR images seem to indicate and confirm my thoughts. My suggestion is to look at cooling techniques and if you like I can find the images of the system I made to cool my then 18".

[clive milne]

Yes of course... by all means.

[alocky]

How did you avoid the peltier device bringing the mirror below ambient? There's plenty of anecdotes from humid areas about the whole primary dewing suddenly with just normal fan cooling. I've sen plenty of space devoted to just attacking the boundary layer, even putting the fan in front of the primary. Others advocate blowing air across the primary. In the Perth part of WA, it's my opinion that the atmosphere is rarely steady enough for the boundary layer on the mirror to be the main problem with your image - but on those -2 degree winter nights when the jetstream stops I want to be able to exploit it!
cheers,
Andrew

[Paul Haese]

The thermal mass of that mirror will allow small windows of the mirror being at ambient before the core temperature begins to release heat and create the boundary layer again. You would need several periods of cooling for that size mirror in a falling ambient temperature.

I only once saw dewing with my 18" and that is because I cooled it to 6 degrees below ambient and near morning it started to dew up. A portable hair dryer puts an end to this though and the little heat will do nothing to the mirror and quickly acclimatise back to stable mirror views. Without this you cannot really tell what the seeing actually is like. The jet stream view of fast seeing can also be imitated by core heat from a mirror.

For further information and data on mirror cooling see Anthony Wesleys (http://acquerra.com.au/astro/)site. He has comprehensive data on mirror cooling.

[alocky]

Thanks Paul, I guess that's why we're experimenting! On the topic of WA seeing however, it's not an opinion I formed in 5 minutes. The seeing in summer here is usually bad enough for the scintillation to be visible at zenith to the naked eye, and it definitely correlates with poorer views through the eyepiece.
The nights when the big girl really shines are when the temperature is plummeting towards zero, there's ice on the cars, and presumably no chance of the mirror being at equilibrium. The stars are steady, and I can crank the thing up past 800 power, no problem. If there's more to be gained by sorting out boundary layers on these nights, I'm all for it.:)
cheers,
Andrew.

[bmitchell82]

:) Winter with clear skies over here ='s brilliant seeing. go up north it is phenomenal!

[Paul Haese]

Andrew I recommend you talk to Anthony also about cooling. It is quite pain free overall, you will benefit from the removal of the boundary layer issues even on nights when the seeing is less than perfect. If you want I can find my images of the mods I made for the 18" and I can send them via email. Let me know via PM.

[clive milne]

fwiw) Some numbers with respect to the heat capacity of a pyrex mirror blank:

The density of pyrex is given as 2.23g/cubic cm
source:http://physics.nist.gov/cgi-bin/Star/compos.pl?matno=169
The mass of a 25"x2" blank works out to 35kg. (at this density of material)
The specific heat of pyrex is quoted as being between 0.75 & 0.85 kJ/Kg/K (depending on the source) Let us pick the middle ground and work with a value of 0.8 kJ/Kg/K
source:http://www.engineeringtoolbox.com/specific-heat-solids-d_154.html
It follows then that for every degree celsius (or kelvin) you wish to cool a 25" telescope mirror, you need to remove 28 kilo Joules of energy.
To convert Joules to Watts (and thereby add a time component) divide the number of Joules by 3600. (coz there is that many seconds in an hour)
ie) To cool a 25" mirror by one degree celsius in one hour, you need to continuously extract 7.8 Watts of heat over that hour.

Whilst this may seem like an abstract exercise, it may be useful information if you have a clear idea of the ambient temperature gradient over a given night and you also have some idea of the extent to which your telescope mirror will lag the air temperature. Under those conditions you should 'theoretically' be able to tune the quantity of heat removed by an active cooling system to keep the substrate at a temperature close to ambient.

For example, it should be reasonably straight forward to set up a PC based PID controller with a feed forward component that monitors the air temperature and can in essence accurately predict how much extra cooling the mirror needs, converts it to a chopped DC power signal to your peltiers, and Bob's your uncle.

Going by the above figures for the 25", let's say we have a night with a fairly steep cooling profile and the primary mirror temperature is lagging ambient by 4 degrees celsius, you would need to pull out 31 Watts of heat to get that critter under control. Peltiers are typically only 30% efficient so you are talking nearly 100Watts of electrical power, and also the requirement of dumping all that heat somewhere (other than in the telescope structure).

This is perhaps a little complex for your average amateur telescope maker, but for a commercial telescope manufacturer it is entirely possible.

~c

[alocky]

An interesting experiment has occurred to me. If we took two similar scopes side-by-side, and cooled the mirror of one to ambient and let the other equilibrate on it's own, took video through each, we'd be able to demonstrate how much the boundary layer effect contributed to the image. I have a pair of suitable cameras - now where's that chap with the 20" binoculars..?.
Cheers,
Andrew.

[clive milne]

Well, yeah... that has kinda been high on the 'to do' list.

[clive milne]

fwiw) Some more figures related to peltier coolers and air flow.

Density of air = 1.29g/L
Specific heat of air = 1 Joule per gram
1 cubic foot of air = 28.3L (36.6 grams)
A 1cfm fan will shift 2190 grams of air in 1 hour.
2190 grams of air divided by 3600 = 0.61 (temperature rise of the air with 1 Watt applied.
It therefore follows that if you are trying to extract 100 Watts of heat from a bank of peltier
coolers and you wish to keep the air ejected at no more than 0.5C above ambient.
you will need to provide the heat sinks with an air flow of approximately 120 cfm.

best,
~c

<edit> I should point out that the figure of 100 Watts and an air delta T of 0.5C are arbitrary.
The above example is simply to give you a feel for the relationship of air flow and energy dissipation.

[DavidU]

Interesting thread. Lot's of good info.

[bmitchell82]

You can do a very quick look at the tube currents and I have done it before, get onto a bright star eg Rigel, defocus so your star takes up 3/4 of your screen and then with a DSLR put it on live view and then youll see the waffling air coming up the side of the tube. for added affect put your hand in front :). You should also be able to see if there is the boundary layer that you talk of clive coming off your spider vanes as they too will show the refractive index changing you will see it almost like clear smoke coming off!

I sometimes utilize this method to see issues. I think too that drawing the scopes volume of air though on a regular basis sucking in the cool air above the ground and exhusting though the rear would be extremely benificial as it would decrease the depth of the boundary layer as the tube/vanes and the like will not influence the air so far out. If sufficiently strong enough the air would not let a boundary layer really form over the face of the mirror or it may sufficiently stir the air on the mirror surface and hence decreasing that nasty thick layer of warmer air.

It would also aid in thermal equilibrium of the mirror.

[Acrab]

Hi Clive,
Very, very interesting post. I have an Ostahowski's 16" 1,6" thick pyrex primary mirror and the problem is that my Dobson not acclimated enough to perform at their best. I have three 120 mm fans blowing air, two for the boundary layer and the other in the back of the mirror. Here, in my usual place of observation, the temperatures fall in winter about 9-10ºC in three or four hours. I saw my telescope this summer fully acclimated in a privileged observation area at 2000 meters altitude and the images were spectacular.
Is there any magic formula for to bring the mirror at his ideal temperature?

[clive milne]

The subject is more complex than I had originally thought, the ideal temperature is a moving target, and there is more than just the heat stored in the primary mirror to consider.

Go here for some more detailed thoughts on the subject:

https://sites.google.com/site/binocularnewtonian/

I suppose if I were to reduce it to a 2 second sound byte as it were, I would say that on the night where I took those thermal images, even though the primary got to within a fraction of a degree of the temperature of the air surrounding it, this wasn't even close to the ambient air temperature measured a few meters above the ground.

I might just add an addendum to this though and point out how much heat energy is typically stored in the mass of plywood that makes up the telescope structure.

Even though wood has (slightly) less than half the density of pyrex, it has almost double the specific heat, ergo; there is actually more thermal mass in the rocker box than the primary mirror itself. I suppose the huge surface area of the telescope structure allows it to dump heat (in to the air) relatively quickly... this may not actually be a good thing when you think about it. The problem of local seeing is not proportional to the temperature of the telescope itself but the quantity of thermal energy put in to the volume of air inside (and to some extent above) the telescope over time.
ie) A more accurate discussion should include Joules (or Watts) and how this translates to temperature forcing in air.

I think the last chapter on this subject has yet to be written.

best,
~c

fwiw) Some further reading:
http://www.cruxis.com/scope/mirrorcooling.htm

[clive milne]

I have to say... carbon fibre is looking like THE material to make telescopes from.

incidentally,
Check this stuff out:
http://www.dragonplate.com/ecart/categories.asp?cID=109

[Paul Haese]

Interesting read there Clive but I think your assessment of the mirror temperature is erroneous. Anthony Wesley has shown that large mirrors can cool at varying rates on different part of the mirror. His data shows this well. The imager would be picking up surface temperature and not the core temperature of the mirror you imaged. The core temperature will continue to release all night without active cooling. Using a mirror fan will help to some extent, but you cannot get a large mirror to within 0.5 of ambient by simple fan cooling.

My suggestion would be conduct the experiment again and this time use two temperature sensors one on the mirror outside edge and one near the center of the mirror. Each sensor should be isolated by foam with silicon sealant. That way the sensors cannot interact and read the outside temperature. Data log the data from these sensors and compare the information against the IR imager. I think you will get a vastly different result.

[clive milne]

Well... I did put a few qualifiers in the discussion.
It was prefaced with the caution that the temperatures indicated in the images should not be taken as being exact or absolutely accurate, but they are useful for showing temperature differentials and trends over time.


Also, some context is required with respect to the temperature profile of the night the images were taken.



I agree with you in principle, but... If the ambient air temperature is stable and not falling (as it essentially was on the night in question) then thermal equilibrium will eventually be reached. It is probably worth making the distinction between thermal equilibrium and ambient temperature.. they are not the same thing.. What is implicit in these thermal images is that even on a night with an exceptionally slow cool down rate, the mirror didn't reach ambient temperature after 5 hours (with the fan running), though it did approach some sort of equilibrium. Even so, the optic and the rest of the lower telescope structure was still 1.4C warmer than ambient. The mirror temperature fell steadily until the last hour, after which, its temperature basically plateaued. The fact that this equilibrium temperature was significantly above ambient suggests that it was acquiring heat energy from an external source as fast as it was shedding it through convection/radiation. (ie thermal equilibrium) The last image in the sequence shows the source of heat: the ground reflected in the metal parts of the mirror cell is quite a bit warmer than the primary mirror. Heat only flows one way, from hot to cold..... The ground underneath the telescope is the elephant in the room.



I would expect there to be some difference, but I would not expect the temperatures indicated by the FLIR to be vastly in error (as you put it), In my experience with the camera, you can pretty much take what it says to the bank.. thermal analysis is what it is designed for, and it is an exceptionally powerful tool in that regard as long as you understand its limitations.
Anyway, this misses the point I am trying to illustrate. Consider the temperature of the kydex ring in the mirror cell... even if you discount the absolute accuracy of the camera, what you cannot discount is its ability to show temperature differentials. Irrespective of the actual ambient temperature, a comparison between the kydex in the top and bottom of the telescope shows that the same material (so you can rule out errors by virtue of emissivity) displays a 2, maybe 3 degree temperature variance.
Kydex sheet has basically no thermal mass, so you can pretty much guarantee that it is tracking the temperature of the air that surrounds it in these two locations.

http://i371.photobucket.com/albums/oo158/clivemilne/IR_3716.jpg
http://i371.photobucket.com/albums/oo158/clivemilne/IR_3714.jpg

ergo, the entire mirror box is hot, and implicitly therefore, effective thermal management of a telescope might require more than just actively cooling the primary mirror because it is not the only (or even dominant) source of heat in the equation.

[Acrab]

Clive, thanks for the links, I already knew. I think I read almost everything on the web about primary mirror cooling.
Practical example:
The telescope is inside the house with a temperature of 18-20 ° C, one hour by car to the place of observation, is installed in the place of observation at 18'00 hours with an ambient temperature of 12 º C. At 20:00 hours the temperature dropped to 10 ° C.




According to the simulator, with the data I put on the screen in 1 hour the telescope would be ready work quite well. Well, in my experience this is not even remotely well.

[clive milne]

I'll just put this out there... why do you think cameras like this sell for $23,000 and thermocouples go for fifty bucks a dozen...?


Maybe you are right... you can expect vastly different results.

[clive milne]

Hi Victor... my interpretation of this is that the thermal model of your optic may be correct.... the question I would then ask is what does the thermal model of everything else contributing to your local seeing suggest?

I hope you appreciate that I am not trying to be obtuse, I'm just inviting you to look outside of the conventional thought parameters.

The primary mirror is not the only source of heat that can disturb the wavefront.

~c

[Acrab]

Clive, my telescope is based in a Kriege-Berry design, like your 25" obsession. Wood, aluminum, metal primary mirror cell metal and few more. Mechanically there are few parameters to enter and I am convinced that the problem is in the boundary layer, not in local atmospheric turbulence. I would like to see the behavior of your 2" thick mirror under conditions of winter, where the temperature can fall about 6 or 8 ° C in two hours. In the U.S. the amateur astronomers of Florida boast excellent conditions for observing with large dobson precisely because thermal fluctuation are minimal and the mirror has time to acclimate within acceptable parameters

[alocky]

Clive's scope hasn't a great deal in common with an Obsession - it's actually my scope in the pictures. In winter here, the temperature fall is nowhere near as severe as you describe. We may see a gradual drop from 16 degrees in the mid afternoon to around 0 before sunrise on the coldest nights. On these nights, when the rest of the atmosphere behaves, the performance of this mirror is good. On summer nights such as the one Clive used his thermal imaging camera, the mirror has equilibrated, yet the views are quite poor due to turbulence in the atmosphere. Yes I'm sure it's the atmosphere, because it knocks every other scope (from SCT to million-dollar refractor) around as well.
Next time I take the scope out, I will try an experimental cooling system I have devised that should be able to pull an additional 250KJ out of the back of the mirror in 15 minutes or so. Worth a data point or two.
cheers,
Andrew.

[Acrab]

Hi alocky,
I'm sorry, in the thermal pictures appears a Kriege-Berry dobson design, and I thought that was Clive's instrument, is your telescope then?
https://sites.google.com/site/binocularnewtonian/
I would be very interested to see your fast cooling system
Best regards.
Víctor.

[alocky]

No need to apologise Victor. If it works I will post details. Otherwise I will say no more out of embarrassment !
Regards,
Alocky.