Insulation - real test results...

[bobechs]

This is a big project and a noble attempt to shed some scientific light on a practical problem.

I have one suggestion to make. When selecting colors for paint testing, make at least one a plain bright white on the outside, and compare it with the fancy stuff on the outside to match the current factory practice of painting the top of AS white to beat the heat.

The reason is to act as a control for the white hi-tech Kool-whatever coatings. I mention this because the folks running the monolithic dome institute site ( www.monolithic.com ) once tested a pair of small buildings side by side, one with a fancy-dan microsphere insulating paint the other plain white enamel, with a temperature logger inside each. The result? No measurable difference in performance between them.

One can't say in advance how a test will go with different products, different surfaces and different methods, but it would be a logical flaw to test the hi-tech substance only against black and uncoated aluminum and not account for the effect of mere color in blocking heat.

[markdoane]

Originally Posted by chuck

do you suppose this is "instead of" pink stuff under the floor, or "in addition to"? I wonder what the R-value of a single layer of the bubble-foil is when simply stapled up against the plywood.....

I hope it's 'in addition to'. As I recall, bubble foil has about 0.3 R-value.

[2airishuman]

hi chuck and markdoane.....

as i recall the tour (in my drunken haze) the bubble foil is instead of the pink....

and rated at a higher r value....

while no doubt the bubble foil comes into contact with the wooden flooring....

it was not attached simply my tacking it to the wood, there was an air gap (see it in the pic)....

i don't recall how the bubble foil was fixed in place....

the air gap would add some to the insulation value.....
and provides another cavity for craftsmen to leave debris.......

i too would have liked the pink and bubble foil....

but with such a small space there is only some much potential anyway....

cheers
2air'

[Rivka]

How are things going over at the various research labs?
i had a thought as i was looking through some of my shop. So the foil insulation needs air space right. Do we know the effect of that airspace being larger or smaller? how bout in a side to side measurement? as i was cleaning up i found a sheet of 1/2 in close cell foam that i had cut out a ton of 4in circles from, for another project. so what i am left with is a lattice of the the reverse. so i was thinking that something like this could be spray glued to the foil insulation and maybe even a laminate of a few layers of each? from what we know do you think of a sandwich of say two or three layers of the foil and therefor 3 or 4 layers of the lattice as a single unit then fit tightly in-between the ribs? do you think that the insulating factor would justify the man hours or cost? do you thing that there is a minimum airspace that thinner than that does not work? i know they recommend something like a inch or two, but i wonder what the minimum is and what relation the amount of air has to the final effectiveness?

[mark lillard]

My wife and I are as new to airstream as it gets. We towed our 78 Tradewind home last night, vin# 125D8SO967. Anything ,anyone knows about a 78 restoration would really help. We retire this year and have better than average skill and tooling but no knowlege of airstreams. In the past we have rebuilt : 2 Cessnas ,2 houses,2 houseboats,and a modest bank account. I'm thinking complete rebuild although,the rv is not in "too" bad a shape. We'd sure appreciate any info you might have. Thanks Folks , Mark & Trudy Lillard Williamstown ,Ky.

[Zeppelinium]

Originally Posted by Zeppelinium

Looks like the standard will be 12" square by 1-1/2" thick. I'll test six panels simultaneously and publish the graphs for as many test items as I get in-hand. Target date for testing would be the first week of June, so get your panel proposals in!

Panels I'd like to suggest:

1. Single sheet of aluminum (a possible worst case standard)
2. Two sheets of aluminum separated by 1-1/2" and only air inside
3. #2 filled with 2" spun glass (vintage trailer standard)
4. #3 with 3" length of rib riveted in the center (this would provide about 4% rib area--attempt to duplicate real vintage trailer construction). (two inside temps would be measured, one near edge and one at the rib location)
5. #2 with suspended foil and bubble insulation, sealed at edges
6. #2 with top sheet painted outside with white roof thermal coating
7. #2 with top sheet painted inside with white roof thermal coating
8. #2 with some variety of solid foam fill
9. #5, but with two layers of foil, suspended from each other and from outside sheets

Others?

After a year of doing other things and letting the hot weather get by me (again), I've built the test fixture. It's glued up from 3/4" foam panels to achieve good isolation between the front and back of the test panels. The test panels slip into 1-1/2" deep wells in the face of the fixture and are taped in place to prevent air flow around them.



The test panels, #2, #3, and #5 from the list above, are made from 12" square sheets of .025 and .032 aluminum. The air-filled and one glass-filled panel are the .025, which is moderately polished. The panel with the silver bubble sheet (this particular insulation has two layers of bubbles between the silver mylar outside layers) and the other glass-filled panel are .032 mill finish (a striated finish with the grain clearly visible that creates the appearance of light sanding in one direction). The air-filled test panel should give a pretty powerful contrast in heat transfer compared to the others.



Why two glass-filled test panels? As soon as I put the other three out in the sun, a simple hand test of the surface temperature revealed that the mill finish was immediately warmer than the polished finish. The mill finish has about the same coarseness as a 25 year old shell that's lightly corroded. So I decided I needed a direct comparison between the bubble panel with the mill finish.

Heat transfer is going to be inferred from the back side temperatures of the panels. I will measure the front side temperature of the two glass-filled panels to see if my initial finding that the mill finish is hotter is actually born out by data.

I hope to get some data by the end of the week. If temperatures here in Colorado stay low, I'll use a heat lamp to get hot temp data. The cold data should come from the environment pretty soon now.

I have 6 additional 12" square cover sheets, enough for three test panels. If anyone has a construction they'd like to test after these first tests are done, let me know. I intend to paint one of the glass-filled panels white on the outside to see what effect that has. What paint should I use? I may paint the inside of the outside panel white on a future test--I think white is generally a good reflector, so it's excellent for the outside, and also a poor emitter, so it would also improve (reduce) heat transfer from the panel towards the inside.

Zep

[Aerowood]

I am excited about the results to come our way

[doorgunner]

Nice ZEPP!
My solar convection loop heater is just a bigger box!
I'll guess that you will see temps behind the glass at temps over 140 F. That would simulate the windows on the trailer. I use reflectix in my windows in the sun facing windows to keep inside temps down.You are going to test glass aren't you?
I'm very interested in your test results.

[Zeppelinium]

the "glass" I'm referring to here is the glass wool batting similar to the stuff used in Vintage airstreams between the shell skins. I'll do a "behind window glass" test with and without solar film later.

Zep

[davidz71]

Originally Posted by doorgunner

Nice ZEPP!
My solar convection loop heater is just a bigger box!
I'll guess that you will see temps behind the glass at temps over 140 F. That would simulate the windows on the trailer. I use reflectix in my windows in the sun facing windows to keep inside temps down.You are going to test glass aren't you?
I'm very interested in your test results.

doorgunner,
I just buttoned up the rear belly pan area after treating the rear frame area with POR. I tore out the 2" glass insulation and replaced it with reflectix bubble foil insulation. I glued the first layer to the particle board flooring with weatherproof/waterproof construction adhesive and then overlapped the 26" tall foil back and forth about 4 times. That should give me some dead space in between each layer but particularly near the frame area where it is needed most. The Reflectix I have left will be cut into window sized pieces to be placed between the windows and the screen to cool the interior or I will attach velcro to the cut pieces and the other side of velcro to the window frame on the inside.

[NevadaGeo]

Hey Zep. I used my heat radar gun inside the trailer at the Burnstream Court, on one of those hot days at the B'man. Noticed 121F on the inside of the trailer, at the blue stripe area next to the bed. 105 or so inside at the silver parts in the sun. In the shade areas was 89, in side with the air on. I really gotta concentrate on getting shade for next year. E's black truck had 174F, and Shadow's Hummer was 150's. Was amazed at the difference of the temp inside at the blue stripe.
Perry

[Zeppelinium]

Originally Posted by NevadaGeo

...Noticed 121F on the inside of the trailer, at the blue stripe area next to the bed. 105 or so inside at the silver parts in the sun. ...Perry

Perry, good to hear from you. I guess the blue stripe is quite an IR absorber (or is that adsorber?). I agee it doesn't make much sense. I'll add other kinds of paint to my list of what to put on the outside of the test panels. It will be interesting to see what the outside skin temp is doing. My sensors are only good to 100C, 212F. I'm guessing the skin could get close to that.

Zep

[HiHoAgRV]

Any chance of doing a shade verses full sunlight for a typical 'stream finish? I would have lost a bet that the reflective surface would tossed the energy back instead of absorbing it, but it's obvious that our precious obsession's do a pretty good job of sucking in solar power. My surfaces are hotter to the touch than a SOB's white box .

[Zeppelinium]

Originally Posted by HiHoAgRV

Any chance of doing a shade verses full sunlight for a typical 'stream finish? ... .

Not a side-by-side test, but it would be easy to put the test box in the sun for an hour, then shade it and watch the temperatures.

One thing I am surprised by is my "hand" test where the mill finish (that's the typical unpolished finish) seemed hotter than the moderately polished panel. I had always thought that even though metal looked relfective (polished) to the eye at visible wavelengths, that at IR it really didn't make much of a difference. We will find out...

The important thing is to get every good idea, like yours, on the record before I start modifying the panels, like painting them.

Zep

[Zeppelinium]

It was obvious right off the bat, like in two minutes, that the mill finish is 20 degrees hotter than the polished finish. The sun angle was pretty low and it was a moderate day here today, so I think it will be even hotter on a hot day in July.

The rest of the data is for your interpretation. You can see from the notes on the graph that there were a couple of things that happened--a tree shadow crept across the test fixture starting around 1:30 and completely covered the mill finish panel by 1:50.

Just for fun I did a quick experiment with a wet cloth over the other panel that has an outside panel sensor. I put the cloth, a thin cleaning cloth you get in a bundle at the auto shop, on about 1:15. It wasn't sopping wet, just wet. You can see that it held the panel temp in the low 80s until 2:00. The variation in temperature is due to the changing breeze. It looks to me like you could significantly cool an Airstream with two long pieces of cheap fabric along the roof and a couple of gallons of water per day. There was no shade on the wet rag panel until 2:10.

I thought I would see a greater difference between the inside panel temperatures. However, it's clear that the yellow line is for the bubble-filled panel. Considering that it has a mill finish outer panel, it performed very well--almost 50% better than the panels with glass wool or air. The mystifying thing is why the air panel didn't do worse.

The lowest data line in the chart is the temperature inside the box. You can see that the outside panel with the wet rag got down to a lower temperature than even the inside temp (the inside temp is buffered a little by the dirt). I've noticed that when doing photo development, my chemicals will cool rapidly and get down to something like 5 degrees below the air temperaure, just due to evaporation in this very low humidity climate in Colorado.

The Y axis temperatures are in degrees times 10. So 1000 on the scale is 100 degrees F. Samples were taken every two minutes.



No more data until next week. The Caravel is meeting a SOB on Vail Pass for the weekend.

Zep

[HiHoAgRV]

Great info. it would be interesting to see the 'wet rag' data down here in the good ole humid south. I suspect the results would be moderated a bit due to the lack of evap' cooling effect.
Do you think the air cavity did better due to a horizontal position and stratifying of the cavity?

On another note, perhaps your studies could allow us to apply for a Federal energy tax reduction for polished 'streams

[Zeppelinium]

Originally Posted by HiHoAgRV

Great info. it would be interesting to see the 'wet rag' data down here in the good ole humid south. I suspect the results would be moderated a bit due to the lack of evap' cooling effect.
Do you think the air cavity did better due to a horizontal position and stratifying of the cavity?

...

Good thoughts.

BTW the fixture wasn't horizontal, it was against a burm at about 20 degrees of tilt--into the east so it got the most direct sun around 10 AM. After that the sun swings higher but to the south.

Zep

[Zeppelinium]

Well, the hot weather seems to have passed me by. As an alternative test (for now) I decided to blast the outer panels with hot air from a heat gun, holding them approximately 200 degrees F and see how fast the inner panels heated up. This first chart is a time history of all the tests:



The individual tests are charted here. I overlaid the two panel tests with glass wool insulation, figuring that since this was a convected heat, not radiated heat test, that they woudl perform very similarly, which they did.







You have to draw your own conclusions, but the foil/bubble insulation doesn't seem to perform appreciably better than glass (in this case). My conclusion above was wrong because I had assumed which channels were which, but later discovered that I was wrong about which panel was which in the chart in the earlier post.

As the tests progress, I'd be glad to send the raw data in the form of Excel spreahsheets to anyone who asks.

Zep

[Aerowood]

Zep' would you mind doing a test with blue or pink foam Insulation, with and without airgap

Kip

[Zeppelinium]

Originally Posted by Aerowood

Zep' would you mind doing a test with blue or pink foam Insulation, with and without airgap

Kip

Isn't that stuff kinda stiff? I sure can do it in my flat panels, but what about real applications in the shell? And what is the purpose of the air gap? I know the foil/bubble insulation needs an air gap to achieve its advertised R value, but I am unaware that the foam stuff suffers from conduction (contact) problems.

Since the season of high solar radiation is behind us for this year, my plan for the winter is to do simple insulation performance tests, such as you propose. I'd like to do the white paint test, but that's really a radiation test, so either I get a heat lamp and calibrate the radiation between the panels, or wait till spring. I am considering the heat lamp...

Tell me more about the blue/pink stuff.

By the way, I'm going to do a more extended test per post #38. More time, hold the face panel temperture more steady, etc. The heat conduction, as indicated by the slope of the inner panel temperature line, could be more precisely measured if I extend the test. Post #38 was just a quick and dirty to see if there were any startling results. I'm going to try for a 15 degree temperature rise on the inner panels, rather than the 5 degrees I got in the post #38 charts.

Roger