Author Topic: THGN801 Radiation Shield Performance (or lack thereof)  (Read 12654 times)

0 Members and 1 Guest are viewing this topic.

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
THGN801 Radiation Shield Performance (or lack thereof)
« on: January 21, 2011, 01:50:17 AM »
First, let me point out that this is only one experiment with data taken over a single day. Although I have done my best to make this a valid comparison, I could have made some mistakes or done some silly things. Please feel free to point these out. That said, and if there is any validity in this experiment, it is a real eyeopener.

My primary outdoor temperature sensor is a THGR810 installed in a home-built aspirated radiation shield. The design uses two coaxial tubes (and as far as I can determine) works pretty well. Various experiments lead me to believe that solar heating is less than 1 degF with this design.

I also happen to have a THGN801 sensor -- the one that comes from OS with a plastic radiation shield. Wondering how well it works, I stuck an aluminum pole in the ground about four feet from the  aspirated sensor and mounted the THGN801 so the sensor was at the same height as the air inlet to the aspirated shield (5 feet above ground level).

The two graphs attached show the results over a single day. Dew points were in the 30's during this experiment and wind at ground level was calm. The first graph shows the temperature reported by each sensor while the second shows the difference between the two. I've marked the approximate times of sunrise, sunset and sun transit on the graphs. Times are UTC and my timezone is UTC-8.

Wow.

Unless I've done something drastically wrong, the radiation shield from OS looks like it may actually be a heat concentrator instead of a shield. I'm tempted to try this again with just the THGN801 (no shield) and see if it is better or worse.

Here's a little commentary on exposure to sunlight throughout the day. I tried to locate the two units such that an equal amount of sunlight was on both sensors most of the time.

* From sunrise to just before 20:00, trees are blocking any direct sun.

* From about 20:00 to 21:00 there is partial sun on the sensors (filtered through tree branches).

* The large peak from just after 21:00 to about 22:30 is direct sun.

* The little "shelf" at 23:00 is back to tree-filtered sun.

* A little after 23:00, the sun starts going down behind more dense foilage and becomes more and more blocked until sunset.

I will leave things in place and continue to record data over the next several weeks at least. I wonder if the OS shield doesn't require at least a light breeze to become even a little bit effective. I'll wait for a day with a little wind and see what happens.

Finally, just some thoughts about what is happening. This is winter (I'm about 38N latitude) and the sun does not get very high in the sky. The shield was oriented so that the long, flat side was facing the sun -- not on purpose however. The sun was roughly perpendicular to the side of the shield which would maximize the amount of radiation applied to the shield. (I should try again with the shield turned 90 degrees.) On the other hand, the instructions from OS do not say anything about how the shield should be oriented.

Does anyone else have data that can confirm or deny this terrible performance I am seeing?
« Last Edit: January 21, 2011, 01:56:49 AM by aweatherguy »

Offline mackbig

  • Forecaster
  • *****
  • Posts: 4128
    • Mackie's Main Street, Unionville, ON Canada Weather
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #1 on: January 21, 2011, 07:54:20 AM »
Interesting test and summary.

I know that all OS products are not equal.  So for your results to be more accurate you would need to remove the 801 from its shield. and put your 810 (without the home built shield) and track each the way you did for a few days.  If they track almost perfectly that will support your findings more.

Then if the above works, I would do your test again, and alternate the 801 in your shield, and the 810 in the OS shield.

If that supports your findings, then its more apparent they are correct.

However one more point would be, having no idea what composes your coaxial shield (perhaps post a pic?), is there any chance this shield is actually trapping the cold air some how.. seems a stretch over a long period of time with temps 60-70, but there's a chance.

A perfect (and perhaps unrealistic) test, would be to take 3 identical sensors.  Run those unshielded, graph them. then rotate them through the various configs of unshielded, your shield, OS shield.  But even this test would not disprove the chance your home shield is thermos'ing the cold air.

I do personally believe that a bit of a breeze helps any shield as no matter what the venting is like with no wind a little bit of heating can be trapped inside, and during low intensity winters months could in fact be higher temps than unshielded.

Andrew

Andrew - Davis VP2+ 6163, serial weatherlink, wireless anemometer, running Weather Display.  Boltek PCI Stormtracker, Astrogenic Nexstorm, Strikestar - UNI, CWOP CW8618, GrLevel3, (Station 2 OS WMR968, VWS 13.01p09), Windows 7-64

Offline utahweatherbear

  • Senior Contributor
  • ****
  • Posts: 273
    • Weatherbear's World
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #2 on: January 21, 2011, 12:22:55 PM »
the design of the OS shield looks to me like it doesn't allow for very good air flow around the sensor, as opposed to the common slotted or layered shields used by most stations and the NWS. I believe the layered versions are supposed to produce at least a little airflow in and out of the slots even in calm winds, due to the convective effect of uneven heating and cooling. the design of the OS shield suggests the air could become more stagnant or trapped around the sensor in calm winds and direct sunlight.

Offline gadget_guy

  • Forecaster
  • *****
  • Posts: 459
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #3 on: January 21, 2011, 12:45:47 PM »
My experience with the OS radiation shield is that it was not effective at all.  I don't have the data anymore but I did note an abnormal rise in temps during the afternoon hours.  Ended up buying a radiation shield from Ambient and that made a big improvement.
 

KVAMCGAH2 - Massanutten VA Davis
KVAMASSA2 -Massanutten VA Oregon Scientific
WA2ONG

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #4 on: January 21, 2011, 03:09:24 PM »
Here are some photos of the aspirated shield. One view is looking into the inlet side before the whole thing was assembled. The inner pipe is made by Hancor and is a double-wall design that provides some extra insulation. In the drawing you can see there is a gap between the end of the inner pipe and the fan -- this allows the fan to pull air through both the inner pipe and the annular area between inner and outer pipes. The fan (turned on 24/7) draws a pretty good airflow so I don't think it is possible for cold air to hide anywhere in there.

The entire assembly is arranged horizontally -- the inlet sucks air upwards at a 45-deg angle and exhausts downwards at 45-deg on the opposite side. The idea was to keep the warmed exhaust air away from the inlet. The unit is also oriented so that the inlet is not downwind under the most often encountered wind conditions.

Another data point is that my CWOP analysis is usually no more than 1 or 1.5 degF off during the daytime. My nighttime temps are often a bit high but that has been attributed to the micro-climates in my area (lots of elevation changes around here).

I could easily (I think) place the THGN801 sensor (sans shield) inside the coaxial shield at the same time as the THGN810. There's enough airflow from the fan that they should both see the same temperature air. I could even try swapping their relative locations inside the pipe to prove it. I could also wire up the THGN810 inside the OS shield in a separate experiment.

Thanks for the suggestions! Those additional experiments will make it pretty pretty clear what's really going on here...I'll post updates as I go along.
« Last Edit: January 21, 2011, 03:15:57 PM by aweatherguy »

Offline mackbig

  • Forecaster
  • *****
  • Posts: 4128
    • Mackie's Main Street, Unionville, ON Canada Weather
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #5 on: January 21, 2011, 08:30:22 PM »
Now that I see your design....I withdraw this statement. 

However one more point would be, having no idea what composes your coaxial shield (perhaps post a pic?), is there any chance this shield is actually trapping the cold air some how.. seems a stretch over a long period of time with temps 60-70, but there's a chance.

Also, when you initially said aspirated, I did not think it was fan-aspirated for some reason....  pretty high tech design.

Andrew

Andrew - Davis VP2+ 6163, serial weatherlink, wireless anemometer, running Weather Display.  Boltek PCI Stormtracker, Astrogenic Nexstorm, Strikestar - UNI, CWOP CW8618, GrLevel3, (Station 2 OS WMR968, VWS 13.01p09), Windows 7-64

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #6 on: January 21, 2011, 09:05:43 PM »
Too much time on my hands...  :oops:

I tried some other simpler designs at first and finally realized that building a good radiation shield was not trivial. I found a web page from Italy that had a coaxial design which turned into this...

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #7 on: January 23, 2011, 07:18:43 PM »
Okay, so I put the THGN801 sensor (sans shield) right next to the THGR810 sensor inside the fan-aspirated shield. The first graph shows the difference between the two over the better part of a day. They track okay but not as well as I expected.

As part of this test I adjusted the THGN801 readings down about 2 degF to get them to match. So now I know that both sensors are at least matched to each other. The 2 degF bump will make the shield's performance look (very slightly) better.

I noticed that the THGN801 appears to respond more quickly to changes than the THGR810, so I applied a mathematical correction to the THGN801 data that essentially slows down its response a bit. I'm skipping the details as it is kind of complicated (details available on request). The second graph shows the result -- both sensors track extremely well over a range of nearly 15 degF.

Now its back to the shield with both sensors, one at a time to see what happens.

One other glaring problem with the OS shield I just noticed -- it is not even close to being white, but a light shade of gray instead. Maybe it is really just a rain shield instead. Once these tests are done maybe I'll try painting it white.

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #8 on: January 27, 2011, 03:05:40 AM »
This is turning out to be an interesting experiment. Thanks again to mackbig for the suggestions -- keep them coming if anything else occurs to you.

It will take a few more days before I have enough data to post some more comparisons, but here's a preview. The OS unit is still shaping up to be a very poor radiation shield. I have now discovered that the shield has problems on calm, clear nights where radiation loss to the sky is a factor. While not conclusive yet, the effect could be as much as 2-4 degF. Didn't expect that.

It seems fairly obvious that one of the reasons for the shield's performance problems is that it is not white. The attached photograph has been adjusted with Photoshop (as described below) and shows:

1) A sheet of regular white paper -- adjusted for a brightness level of 255 (maximum)
2) Part of an 18% photographic gray card -- adjusted for a brightness of 43 (slightly low, about 17%)
3) A necktie with relatively black fabric -- adjusted to a brightness level of 0.
4) The top cover from the OS radiation shield (and a bunch of lint that I couldn't seem to get rid of).

This is not very scientific, but this suggests that the radiation shield is about 70-75% gray -- zero being black and 100% being white. While I cannot convert these numbers to a percent of energy reflection or absorption, the shield is clearly far from being white. The shield has a glossy finish which may make it more reflective, but I suspect that is not a huge offsetting factor.

Stay tuned...
« Last Edit: January 27, 2011, 03:12:55 AM by aweatherguy »

Offline mackbig

  • Forecaster
  • *****
  • Posts: 4128
    • Mackie's Main Street, Unionville, ON Canada Weather
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #9 on: January 27, 2011, 08:11:24 AM »
More interesting results... and good point.

Maybe it is really just a rain shield instead. Once these tests are done maybe I'll try painting it white.

Actually, no place in the manual does it refer to "radiation" or "shield".  And have you seen the new pic's on the OS site, and ambient, it actually looks black or very dark grey.... the wind cup assembly as well.  Black is good on an anny as it speeds up meltage after f-rain....but not on a temp sensor.

The exact description is "Temperature / Humidity Sensor casing"

The manual also does not appear to mention sun, shade or placement.

Now that I think of it, it might just be a big version of the casing found on my os968 stock sensor... definitely just a rain/snow deflector...



Andrew

Andrew - Davis VP2+ 6163, serial weatherlink, wireless anemometer, running Weather Display.  Boltek PCI Stormtracker, Astrogenic Nexstorm, Strikestar - UNI, CWOP CW8618, GrLevel3, (Station 2 OS WMR968, VWS 13.01p09), Windows 7-64

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #10 on: January 29, 2011, 06:55:37 PM »
I finally found a place where OS calls it a heat shield -- look up the THGN801 as a separate unit on their web site -- under "Features" it says:

"Heat shield prevents direct sunlight from affecting the sensor"

Anyway, I've got enough data now I think to wrap this up. There are three graphs attached. The first one compares the sensor to the fan-aspirated shield after having adjusted both sensors to be in agreement as mentioned above.

Daytime errors are as high as +15F but I was puzzled by the negative nighttime errors (as large as -3F or so). But then I realized it was probably due to radiation loss to the cold sky on the calm, clear nights. All I needed was an overcast night to confirm and that happened last night -- the 2nd graph. Fog started forming around 16:00 - 17:00 just as the sun was setting. However it was pretty thin. Later when it got dark, I could still see stars through the fog but the were "fuzzy". It is amazing how little fog it took to cancel out the cooling from a clear sky. Later around 1AM the fog cleared a bit and then came back in.

The final graph is where I just hung the THGN801 sensor from a white nylon cord -- no shield at all (clear night). The only problem was that some thin cirrus clouds moved in around 14:00 hours and killed the peak error. Even so, you can see that the performance is not much worse with no shield at all.

I'm going to collect data for a couple more days then put the THGN801 back into the fan-aspirated shield to make sure the readings still agree. Other than that I'm thinking this is pretty conclusive...comments?

For a follow on, I was going to paint the shield white until I read that white spray paint actually looks black in the infrared so that would probably make things worse at night. Perhaps I'll try to cover the shield with some aluminum foil instead...?

Offline NeverDie

  • Senior Member
  • **
  • Posts: 87
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #11 on: September 20, 2013, 09:37:42 PM »
Thanks for the meticulous comparison.

What's your best guess as to how much of a difference it would make if you spray painted the THGN801 "shield" a bright white?  Would it still be lousy?

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #12 on: September 21, 2013, 01:33:18 AM »
I suspect it would not hurt, at least in the daytime. White spray paint is known to be very absorptive in the infra-red, so perhaps there might be more negative errors on a clear night...but that's only conjecture on my part. I would not be willing to bet money on that without an experiment or two...

BTW, I'm currently working on a comparison of the same fan aspirated shield with an Ambient SRS-100LX and so far it appears to be doing a much better job. I'm hoping to be able to post some results here within the next couple of weeks or so...


Offline DanS

  • Chiang Mai weather
  • Forecaster
  • *****
  • Posts: 5434
    • ThaiWx
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #13 on: September 21, 2013, 02:00:57 AM »

"Okay, so I put the THGN801 sensor (sans shield) right next to the THGR810 sensor inside the fan-aspirated shield. The first graph shows the difference between the two over the better part of a day. They track okay but not as well as I expected."

Another possibility to throw into the mix. If either of these 2 sensors uses a thermistor as opposed to the digital therm/hygo sensor chip it may be more susceptible to sunlight ground reflections. My older station uses a thermistor and if I mount it in a loosely spaced plate shield so the sun on the ground beneath is "seen" I will get higher readings during the times the sun is shining. With this same sensor remounted into a shield with closer spacing so you cannot see the ground below (ex. deeper bowls) it tracks a lot better. I don't see this effect with my newer station with the SHT11 digital sensor chip.

Offline NeverDie

  • Senior Member
  • **
  • Posts: 87
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #14 on: September 21, 2013, 11:30:12 AM »
White spray paint is known to be very absorptive in the infra-red...

That I did not know.  Is there a paint that's good at reflecting infra-red?  Apparently some coatings are good at radiating heat that has already been absorbed...  They're often applied to metal roofs to make them cooler.


BTW, I'm currently working on a comparison of the same fan aspirated shield with an Ambient SRS-100LX and so far it appears to be doing a much better job.

By that, do you mean that the Ambient SRS-100LX is doing a much better job than THGN801 housing?  Or did you mean that the fan aspirated shield is doing a much better job than the Ambient SRS-100LX?

Did you consider the Davis 7714 at all?  It looks as though a THGR810 would fit inside the 7714.  I was considering an Ambient SRS-100LX until I read the customer reviews.  Other than the weak bracket failing, do you think the rest of it would survive intact after being struck by a soccer ball?  If it's supposed to be 5 feet off the ground, I have to consider it might get accidently bashed by my kid playing ball with his friends, or by heavy hail.

I guessing that for my situation, maybe an aspirated Stevenson (if there is such a thing) could better survive a pounding...
« Last Edit: September 21, 2013, 12:06:36 PM by NeverDie »

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #15 on: September 21, 2013, 09:15:51 PM »
Dan,

As part of my current experiment in progress, I've discovered that placement inside the shield might matter too -- e.g. center of the inner pipe versus near the edge. I'm working on collecting some data there too and should have some results soon. Both sensors in the previous work were thermistor-based and both protected the thermistors inside a plastic case with small holes to allow air circulation (both were OS sensors). I'll keep an eye out for possible effects of ground reflections as I go forward.

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #16 on: September 21, 2013, 09:27:04 PM »
I do not know of a paint that is reflective in the infra-red -- I'm sure there are but I just don't know about them. I kind of suspect that getting a surface to be highly reflective in both visible and infra-red is not easy and finding a paint to do this may be even harder. I don't really know however so perhaps someone else can comment.

If you want to re-radiate infra-red I think you would want something like the white spray paint. Heat transfer will occur by radiation from a warm object to a cooler one. I think for that purpose you would want the warmer object to be as "black" as possible at the wavelengths over which most of the radiation would occur (which depends on temeprature). I don't pretend to be an expert physicist though, so don't take this to be 100% accurate.

I'm finding that the SRS-100LX works much better than the OS shield. I don't want to quote numbers right now since I have not finished collecting data yet, but so far it is looking like the Ambient unit is a reasonable choice. I agree that the mounting bracket is weak and you would want to make your own bracket. I have completely no idea whether one of these mounted on a good bracket would survive your kids!!!

No, I have not considered the Davis shield...If someone wants to loan me one I'd be glad to run a comparison...  :grin:


Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #17 on: September 21, 2013, 09:56:36 PM »
Forgot to mention...here's paper that reports on some tests on a Davis 7714 shield...results look pretty good. The paper is pretty technical and it takes a while to read through it...but the thing you're probably interested in is their figure 6. Looks like errors are mostly below 1C with a very small number of readings reaching 2C error.

Air Temperature Measurement Errors in Naturally Ventilated Radiation Shields, Nakamura and Mahrt,
published in the Journal of Atmospheric and Oceanic Technology, volume 22, page 1046
(American Meteorological Society)

I think you can get a copy of the paper here:

https://ams.confex.com/ams/pdfpapers/78111.pdf


Offline NeverDie

  • Senior Member
  • **
  • Posts: 87
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #18 on: September 22, 2013, 10:57:27 AM »
Great info.

Rather than exposing the 7714 to direct sunlight and expecting it to do all of the work, I should think there would be a way to shade it without impairing its airflow.  That way, the 7714 would still function as a radiation shield, but it would just be shielding against ambient reflected light and IR.

For instance, why not a simple overhang (low thermal mass) made mostly from rigid foam?  So, the tradeoff would be: the lower it is, the smaller it could be, but the more it might interfere with air flow.  How low could it be?  Where's the sweet spot?  My WAG is that it could be pretty low, maybe a foot or two above it.  As a starting approximation, maybe take whatever the height of the Stevenson roof is above the interior instruments.
« Last Edit: September 22, 2013, 11:09:13 AM by NeverDie »

Offline NeverDie

  • Senior Member
  • **
  • Posts: 87
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #19 on: September 22, 2013, 11:21:42 AM »
Forgot to mention...here's paper that reports on some tests on a Davis 7714 shield...results look pretty good. The paper is pretty technical and it takes a while to read through it...but the thing you're probably interested in is their figure 6. Looks like errors are mostly below 1C with a very small number of readings reaching 2C error.

Air Temperature Measurement Errors in Naturally Ventilated Radiation Shields, Nakamura and Mahrt,
published in the Journal of Atmospheric and Oceanic Technology, volume 22, page 1046
(American Meteorological Society)

I think you can get a copy of the paper here:

https://ams.confex.com/ams/pdfpapers/78111.pdf



An interesting thing about that paper is that they use additional instruments to offer up corrective factors based on the sun intensity and/or using IR to measure how hot the 7714 plates are getting.  Since sun intensity is easy to measure (OS and others sell a part for doing just that), I suppose someone could figure a conservative correction factor that may not be perfect but would at least get you closer to an accurate temperature measurement for the case where the 7714 is not shaded at all (i.e. under constant solar bombardment during the day).  In the case where the 7714 is fully shaded (as I'm suggesting above), then a much smaller corrective factor might still offer up some improvement in accuracy.

Has anyone here tried doing that?

As a further refinement, I suppose that in cases of sufficient wind, you might de-rate the correction factor.  When there's little or no wind, you might use the correction factor at full strength.

i.e. you might want to dynamically calculate temperature with wind speed and solar intensity as two of the factors.  Do any of the PC-based weather station software's support that?  I bet you could get pretty good accuracy doing that, and you may already own the instruments for doing it.  i.e. better accuracy for free!  Heck, why isn't everyone doing this already?  It should come already baked into your kit as an option in software that you can enable by simply mouse-clicking  a checkbox.
« Last Edit: September 22, 2013, 11:57:08 AM by NeverDie »

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #20 on: September 23, 2013, 01:57:10 AM »
Some observations from that paper:

Notice their equipment list includes a sonic anemometer (Vaisala WS425). This was used to collect wind speed data simultaneously with other measurements.

Equation (11) defines the independent variable in their "curve fit" to the errors that could correct for solar errors. So, to make these corrections, you must compute "X" from their equation (11).

In the numerator of (11), we have "Rad" which is shortwave radiation in Watts/sq-meter. You may have a sensor that can do that...although I don't think OS offers such a sensor (I know they have a UV sensor but that's not what you'd want here). And it probably does not need to be all that accurate (a few numeric examples would prove or disprove this).

The denominator of (11) includes two constants (rho and C-sub-p), the thermistor temperature and "U" -- the wind speed.  This might be a problem. Since "U" is in the denominator, if it ever goes to zero, then "X" goes to infinity and their curve fit no longer works. At a minimum you would need to place an OS anemometer at about the same location as the shield solely for the purpose of measuring "U". I'm fairly certain that if you used an anemometer at a different height (like 10 or 30ft AGL), these corrections would not work.

Then if "U" gets too small you would need to "cap" it at some safe limit so "X" does not blow up on you. As mentioned above, since your anemometer of record is either 10ft or 30ft AGL, this anemometer is a separate unit -- you don't get to make double-use of your anemometer-of-record. Now you need a weather station that can process readings from two anemometers.

Finally, I've not worked through the numbers, but there will be some minimum wind speed you need to discern for the corrections to be accurate (the "cap") -- I don't know what that is or if the OS anemometer can read a wind that slow (I'd give even odds on that one). The Vaisala sonic anemometer is capable of measuring very low wind speeds more accurately than the OS unit...you'd have to work through some numeric examples to figure out if it really matters or not.

The night-time corrections they show are based on "net radiation" instead of shortwave radiation which would require yet another sensor...however, they are so small that you could just skip that part.

So yes, this might be do-able, but it sure seems like it would be less expensive and less work to purchase a fan-aspirated shield instead.


Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #21 on: September 23, 2013, 02:31:30 AM »
Regarding your other post about a rigid foam shield...I dunno. Sounds tempting. Lots of folks have worked on these naturally aspirated shields and I haven't seen anything like that on the market. Maybe you have something there...let's run some numbers...

What you'd want to work out is how much heat would go into the foam versus how much would be convected away from it. Peak sun is 1000W/sq-meter; then multiply that by what percentage your foam absorbs. Then how much would that heat up in calm air? There are lots of numbers around for free air convection but it is probably in the ballpark of 7 Watts over a square meter per degree of temp difference between surface and free air.

So how much of the 1000W/sq-meter can you avoid absorbing into the foam? Let's say that you can reflect 95% -- I don't know if that's even reasonable, but it's a number. So per square meter of foam, you absorb 50 watts. With free air convection and only considering the top surface of the foam (one sq-meter), the foam surface would be heated about 7 degC to get rid of that 50 watts. If these numbers are reasonable, then having a "shield" above the shield that's 7degC warmer than ambient air is probably not good. Even though the foam might prevent that heat from appearing on the bottom side of the foam...I'm not sure that would be enough. Clearly you would want the piece of foam to be as small as possible.

It might be better to use a thin, conductive plate instead -- that way you are convecting heat away from both sides of the plate and the overall plate delta-temperature is only half as much.

I think the key though is the reflectivity over the range of wavelengths where that 1000W/sq-meter lives -- I don't know what typical numbers are but you would probably want better than 95% total reflected energy. What whould it take to maintain the high reflectivity? Would residue from rain, dirt, etc ruin the high reflectivity? Would there be a lot of maintenance?


Offline NeverDie

  • Senior Member
  • **
  • Posts: 87
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #22 on: September 23, 2013, 10:54:01 AM »
Krylon flat white paint has high reflectance and high emmisivity and is readily available.  If interested, I can post a reference to the measurements.  Apparently Magnesium Oxide White Paint would be better, and is widely referenced, but I did a google search yesterday, and I wasn't able to find a reputable vendor.

In either case, you are right: dust or dirt or the like would de-rate it.

Offline NeverDie

  • Senior Member
  • **
  • Posts: 87
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #23 on: September 23, 2013, 11:04:20 AM »

In either case, you are right: dust or dirt or the like would de-rate it.

...and I expect the same would be true for the 7714 or similar.

I've remember hearing about "self-cleaning" house paints you can buy that are supposed shed off dirt, but I don't remember the particulars or whether they have good reflectance and emissivity.  It's probably a long-shot, but it might be worth checking....

Offline NeverDie

  • Senior Member
  • **
  • Posts: 87
Re: THGN801 Radiation Shield Performance (or lack thereof)
« Reply #24 on: September 23, 2013, 11:14:19 AM »
I used a 2-foot diameter piece of hole- perforated metal (below); top painted white, bottom painted flat-black. Worked OK except at sunrise & sunset. A semi-circular (half dome) shape is probably better.



Did you use it in addition to the 7714, or instead of it?

How would you orientate the dome relative to your temperature gear?  i.e. how high, and (if applicable) how far around?  Fabricating a dome doesn't sound easy, but an arch out of that material might be easy.

Maybe a partial dome would be good enough?  I must admit, though, a full dome does have a certain aesthetic appeal to it.  As a conversation piece, it would certainly beat most other yard ornamentation I've seen.   

Can you add a bit more color to "worked OK"?
« Last Edit: September 23, 2013, 11:44:44 AM by NeverDie »

 

anything