CIDCrazyJay
Well-Known Member
Solar Heating of Foam Core Composite Panels.
Composite aircraft articles and publications often mention the susceptibility of epoxy composite structures to serious strength degradation as a result of solar heating.
Invariably, these sources state that white is the optimum color to reduce this effect (matter of fact, it is). And they will often go on to suggest avoiding most other colors, except maybe soft yellow, or possibly something darker on undersides. Although I haven't turned up test data for this, or found any article which produces hard data to draw conclusions from. There are many bright-white composite aircraft having large black lettering and tail numbers, but they don't seem to lose their tails due to this oversight.
The only data given is the well-known chart from Soaring magazine, Sep.1975, which is included in RAF literature and Quickie instructions. But this chart is almost certainly, (quite precisely, actually), calculated from maximum theoretical heating, not from an experiment. As it seems to show inflated estimates which are probably only seen on calculators, after making some simplifying assumptions, such as 100% efficiency, aligned perpendicular to the suns rays, at the equator, on a cloudless day, at solar-noon on the equinox.
In other areas of the above mentioned aircraft construction manuals, it is suggested to shoot some black primer on recently completed components, then place them outside for a few minutes to induce solar heating and facilitate a poor mans post-cure of the composite structure, thereby improve it's properties. This is accompanied by a warning not to exceed a certain threshold. "Too hot to touch for longer than 3 seconds" which probably translates to 140F. (Please just use a temp gun; they are far more accurate and reasonably affordable).
The questions that arise:
1: How hot can a composite panel get by solar heating?
2: How fast will it heat up?
3: What affect does surface orientation to the sun have?
4: What is the weakest link: Epoxy, Micro balloon filler, or XPS foam? (Or 3M 78).
So, I decided a rudimentary test was in order.
But first, a few napkin predictions follow:
"Direct Normal Irradiance" at the surface of Earth is generally published to be 1,000 Watts/m^2. Although several credible sources cite 1,120W/m2, inclusive of atmospheric scattering. (Or 1,360-1,370W/m2 in Low earth orbit, above the atmosphere).
Rough calculation of maximum panel temperature, with several crude assumptions:
1: 1,120 W/m2 Solar irradiance.
2: 95% of energy is absorbed by flat-black paint.
3: Zero conductive losses through foam.
4: Zero convective losses to surrounding air.
5: All energy is balanced by solar radiation in/infrared out.
Using the Stefan-Boltzmann constant 5.67037 x 10^-8 (W/(m^2 x K^4)). ~ Watts/square meter per Kelvin to the fourth.
1120W x 0.95 = 1064W/m2 total energy absorbed.
(1064/Stefan-Boltzmann)^0.25 = 370.1 Kelvin
370.1K - 273.15 = 96.95 Celsius
32+(1.8 x 96.95C) = 206.5 Fahrenheit
The following Test was conducted at 40 degrees north latitude, between 11:15am-Noon. (Solar Noon was at 1:05pm). On a 95 degree July day.
The test panel was made up of 5 plies Hexcel Hexforce 7725 (RAF "BID") and Aeropoxy laminating epoxy in an open wet layup on a glass sheet, trimmed to 8x18"; 1sq.ft. Caliper measured 0.060" thick, +/-0.002". This panel was allowed to cure for several weeks at ambient temperatures between 70-80F and humidity was generally kept between 30-60%.
24 hours before solar heating test, the fiberglass sheet was bonded to 1" thick, 1.6lb/cuft Extruded Polystyrene insulation with Epoxy/3M K20 glass bubble syntactic foam, (AKA "micro slurry"). A temperature probe was embedded in a slit channel extending to the center of the foam sheet and was fully encased in micro. Another 1" XPS sheet was bonded to the back of this foam/composite panel with 3M-78 Polystyrene-specific spray adhesive, ($25/can). In order to test this spray-adhesive, and also to provide some additional insulation to reduce heat transfer. The fiberglass face of the composite panel was spray-painted flat-black shortly before the test.
The panel was oriented due South. Inclined perpendicular to the suns rays, using available data for this location. This resulted in an angle tilt of 17.5 degrees from horizontal. The panel was later rotated to face South-East as Noon approached, so that the shadow was almost perfectly perpendicular behind it. (Solar noon happened later, at 1:05pm).
This test resulted in a maximum recorded temperature of 206.2F
Steady 205.2F +/-1 degree F. for a duration of 5 Minutes, between approximately 45-50 minutes in the sun. Which agrees quite closely with the back of the envelope estimate, (206.5F). Even though it includes many assumptions. This leads me to believe solar heating of the dark cement patio tile under the panel offset much of the convective losses, by raising the air temperature in that area well above ambient. Or the white pavers might have reflected additional energy onto the panel. Also, flat black paint might be slightly better than 95% efficient as an absorber. Or, more likely than not, the 1,120W/m2 solar energy assumption is a low estimate for Earths surface at this test location, which is 5,000'MSL. Therefore having less radiation blocking atmosphere above the panel than if the same test was conducted near sea level, leading to additional solar energy reaching the panel than initially assumed.
I began to suspect the temperature probe, which might have become inaccurate after layup into the composite. So, the panel was acclimated indoors after the initial test. Where both the panel and ambient temperature probes were within a few tenths of a degree after acclimation. Both temp probes were also reading within 1 degree of the room thermostat after 2 hours. Confirming their relative accuracy. I then placed the panel outside again, at 2pm. Or 1 hour after solar noon. And in this test, it reached 165F from a previously stable 70F. Gaining +95 degrees in under 10 minutes!
What I learned:
Flat-black paint or primer will indeed reach an unsafe temperature for many epoxy resins available on the market, and also most foam core materials including XPS, PVC, and PU. (PIR or Polyiso is the exception, as it is rated around 300F. But it has poor shear properties and exhibits concerning friability, even when compared to regular Urethane foams, which are already rather poor in this regard).
The panel temperature will quickly rise 90+ degrees within 7-8 minutes, it will stabilize over +110F above ambient on a 95 degree day.
It will rapidly exceed 165F* within 10 minutes. *(This is the widely published maximum service temperature of Polystyrene foams). After only ten minutes in the sun!
Taking the panel from perpendicular to the suns rays, and laying it flat. Results in a small ~5 degree temp reduction almost instantly. Placing it in the shade drops 20 degrees within a minute. Cool-down below peak T. happens rapidly, if and when needed.
3M "78" spray adhesive for Polystyrene (not 3M 77, which melts XPS foam). Will soften substantially and peel apart easily with no damage to the foam at a temperature well below 165F. This is nowhere near adequate for foam core composites. BUT it *might* be useful for tacking together XPS foam blocks prior to Hot-wire cutting. Because it melts at a lower temperature than XPS, builders may be able cut through it seamlessly. A hotwire cutting test between 3M 78 and regular hot-glue might be in order. ~ Any takers on that experiment?
Polystyrene foam, and Epoxy/micro mix, both held up to this high temp abuse. I found it was practically impossible to peel the composite sheet from the foam core. The micro did not appreciably soften, even though it was certainly at 206F, as the temp probe was embedded in the micro and was reading this value. Keep in mind, water boils at a lower temp than this, at this elevation. Aeropoxy has a published Tg of 194-203F, depending on specific hardener and post-cure. But of course, glass transition is not an instantaneous melting point. It simply begins to soften above this temperature. The mechanical properties were not evaluated other than attempting to peel the fiberglass from the foam. The structural properties might in fact be seriously degraded, but that is not the point of this test.
The XPS remained the weakest link by a significant margin at all temps, and therefore it is likely the limiting factor for temps in these types of structures.
The darkest color allowable, assuming 165F maximum surface temperature, would absorb no more than 73% of solar energy. Ideally, less. Keep in mind, 140F is a common initial heat-cure target for many epoxies, and is well below XPS service limit. 140F requires only 62% absorber efficiency under similar conditions. This absorption percentage probably corresponds to Medium-Grey, similar to US Navy aircraft.
A darker, true-grey primer, slightly darker than this, would probably be ideal if planning to place components in the sun to post-cure. Greyscale choice could be used to automatically limit the maximum temperature to a safe level for the foam core (<165F). And by limiting the maximum attainable temperature to the 160-165F range, (72% efficient absorber). This would allow longer-term curing for several hours, or all afternoon, as opposed to only 10 minutes if using black on a hot day.
Also, by using a darker primer for solar heating than the final paint scheme color, this would theoretically ensure the airframe was cured to a higher temperature than will ever be encountered in service.
Hours-long cure times would be beneficial for several reasons. Primarily by allowing this heat to saturate the entire structure. Components painted black, with the resulting 10 minute cure limit, will only heat the skin and not the embedded wing spars. Post-curing these major structural elements for longer durations to obtain the benefit of increased cross-linking, improved strength and resilience, would provide the greatest benefit for the finished aircraft from this extra curing effort, as opposed to simply post-curing the skins.
Maximum benefit would probably be gained by using a medium or darker-grey primer, one which absorbs 60-70%, and performing a 4-6 hour solar post-cure, starting early in the morning to generate a slower ramp time. Ending just after solar noon, or whenever the item reaches the target temp and dwells there for a specified timeframe, this temp and time should be based on the specific epoxy system. More even component heating could be facilitated by rotating the item every half-hour or so. And tighter quality control of the process comes from using a temp gun every 10-15 minutes to ensure the entire structure reaches an adequate post-cure temperature, without exceeding the above mentioned limits.
Note: Black primer may still be useful for post-curing in late Autumn or winter. Where the ambient temperature is below 50F. As black paint in bright sunlight seems to readily reach +110F above ambient temp. Black should probably be exposed to sunlight with abundant caution any time the ambient temp is at or above 50F. Also, post-curing to the 160-165F range should probably be done with wings or flying surfaces jigged to a torsion box/worktable. As any undesirable twist which develops while curing components near their maximum allowable temperature is going to be difficult or impossible to resolve, without re-heating to a higher (possibly dangerous) temperature. So, it makes sense to do it right the first time, and never worry about it again after it's been painted in a cooler, more reflective color.
Food for thought.
Composite aircraft articles and publications often mention the susceptibility of epoxy composite structures to serious strength degradation as a result of solar heating.
Invariably, these sources state that white is the optimum color to reduce this effect (matter of fact, it is). And they will often go on to suggest avoiding most other colors, except maybe soft yellow, or possibly something darker on undersides. Although I haven't turned up test data for this, or found any article which produces hard data to draw conclusions from. There are many bright-white composite aircraft having large black lettering and tail numbers, but they don't seem to lose their tails due to this oversight.
The only data given is the well-known chart from Soaring magazine, Sep.1975, which is included in RAF literature and Quickie instructions. But this chart is almost certainly, (quite precisely, actually), calculated from maximum theoretical heating, not from an experiment. As it seems to show inflated estimates which are probably only seen on calculators, after making some simplifying assumptions, such as 100% efficiency, aligned perpendicular to the suns rays, at the equator, on a cloudless day, at solar-noon on the equinox.
In other areas of the above mentioned aircraft construction manuals, it is suggested to shoot some black primer on recently completed components, then place them outside for a few minutes to induce solar heating and facilitate a poor mans post-cure of the composite structure, thereby improve it's properties. This is accompanied by a warning not to exceed a certain threshold. "Too hot to touch for longer than 3 seconds" which probably translates to 140F. (Please just use a temp gun; they are far more accurate and reasonably affordable).
The questions that arise:
1: How hot can a composite panel get by solar heating?
2: How fast will it heat up?
3: What affect does surface orientation to the sun have?
4: What is the weakest link: Epoxy, Micro balloon filler, or XPS foam? (Or 3M 78).
So, I decided a rudimentary test was in order.
But first, a few napkin predictions follow:
"Direct Normal Irradiance" at the surface of Earth is generally published to be 1,000 Watts/m^2. Although several credible sources cite 1,120W/m2, inclusive of atmospheric scattering. (Or 1,360-1,370W/m2 in Low earth orbit, above the atmosphere).
Rough calculation of maximum panel temperature, with several crude assumptions:
1: 1,120 W/m2 Solar irradiance.
2: 95% of energy is absorbed by flat-black paint.
3: Zero conductive losses through foam.
4: Zero convective losses to surrounding air.
5: All energy is balanced by solar radiation in/infrared out.
Using the Stefan-Boltzmann constant 5.67037 x 10^-8 (W/(m^2 x K^4)). ~ Watts/square meter per Kelvin to the fourth.
1120W x 0.95 = 1064W/m2 total energy absorbed.
(1064/Stefan-Boltzmann)^0.25 = 370.1 Kelvin
370.1K - 273.15 = 96.95 Celsius
32+(1.8 x 96.95C) = 206.5 Fahrenheit
The following Test was conducted at 40 degrees north latitude, between 11:15am-Noon. (Solar Noon was at 1:05pm). On a 95 degree July day.
The test panel was made up of 5 plies Hexcel Hexforce 7725 (RAF "BID") and Aeropoxy laminating epoxy in an open wet layup on a glass sheet, trimmed to 8x18"; 1sq.ft. Caliper measured 0.060" thick, +/-0.002". This panel was allowed to cure for several weeks at ambient temperatures between 70-80F and humidity was generally kept between 30-60%.
24 hours before solar heating test, the fiberglass sheet was bonded to 1" thick, 1.6lb/cuft Extruded Polystyrene insulation with Epoxy/3M K20 glass bubble syntactic foam, (AKA "micro slurry"). A temperature probe was embedded in a slit channel extending to the center of the foam sheet and was fully encased in micro. Another 1" XPS sheet was bonded to the back of this foam/composite panel with 3M-78 Polystyrene-specific spray adhesive, ($25/can). In order to test this spray-adhesive, and also to provide some additional insulation to reduce heat transfer. The fiberglass face of the composite panel was spray-painted flat-black shortly before the test.
The panel was oriented due South. Inclined perpendicular to the suns rays, using available data for this location. This resulted in an angle tilt of 17.5 degrees from horizontal. The panel was later rotated to face South-East as Noon approached, so that the shadow was almost perfectly perpendicular behind it. (Solar noon happened later, at 1:05pm).
This test resulted in a maximum recorded temperature of 206.2F
Steady 205.2F +/-1 degree F. for a duration of 5 Minutes, between approximately 45-50 minutes in the sun. Which agrees quite closely with the back of the envelope estimate, (206.5F). Even though it includes many assumptions. This leads me to believe solar heating of the dark cement patio tile under the panel offset much of the convective losses, by raising the air temperature in that area well above ambient. Or the white pavers might have reflected additional energy onto the panel. Also, flat black paint might be slightly better than 95% efficient as an absorber. Or, more likely than not, the 1,120W/m2 solar energy assumption is a low estimate for Earths surface at this test location, which is 5,000'MSL. Therefore having less radiation blocking atmosphere above the panel than if the same test was conducted near sea level, leading to additional solar energy reaching the panel than initially assumed.
I began to suspect the temperature probe, which might have become inaccurate after layup into the composite. So, the panel was acclimated indoors after the initial test. Where both the panel and ambient temperature probes were within a few tenths of a degree after acclimation. Both temp probes were also reading within 1 degree of the room thermostat after 2 hours. Confirming their relative accuracy. I then placed the panel outside again, at 2pm. Or 1 hour after solar noon. And in this test, it reached 165F from a previously stable 70F. Gaining +95 degrees in under 10 minutes!
What I learned:
Flat-black paint or primer will indeed reach an unsafe temperature for many epoxy resins available on the market, and also most foam core materials including XPS, PVC, and PU. (PIR or Polyiso is the exception, as it is rated around 300F. But it has poor shear properties and exhibits concerning friability, even when compared to regular Urethane foams, which are already rather poor in this regard).
The panel temperature will quickly rise 90+ degrees within 7-8 minutes, it will stabilize over +110F above ambient on a 95 degree day.
It will rapidly exceed 165F* within 10 minutes. *(This is the widely published maximum service temperature of Polystyrene foams). After only ten minutes in the sun!
Taking the panel from perpendicular to the suns rays, and laying it flat. Results in a small ~5 degree temp reduction almost instantly. Placing it in the shade drops 20 degrees within a minute. Cool-down below peak T. happens rapidly, if and when needed.
3M "78" spray adhesive for Polystyrene (not 3M 77, which melts XPS foam). Will soften substantially and peel apart easily with no damage to the foam at a temperature well below 165F. This is nowhere near adequate for foam core composites. BUT it *might* be useful for tacking together XPS foam blocks prior to Hot-wire cutting. Because it melts at a lower temperature than XPS, builders may be able cut through it seamlessly. A hotwire cutting test between 3M 78 and regular hot-glue might be in order. ~ Any takers on that experiment?
Polystyrene foam, and Epoxy/micro mix, both held up to this high temp abuse. I found it was practically impossible to peel the composite sheet from the foam core. The micro did not appreciably soften, even though it was certainly at 206F, as the temp probe was embedded in the micro and was reading this value. Keep in mind, water boils at a lower temp than this, at this elevation. Aeropoxy has a published Tg of 194-203F, depending on specific hardener and post-cure. But of course, glass transition is not an instantaneous melting point. It simply begins to soften above this temperature. The mechanical properties were not evaluated other than attempting to peel the fiberglass from the foam. The structural properties might in fact be seriously degraded, but that is not the point of this test.
The XPS remained the weakest link by a significant margin at all temps, and therefore it is likely the limiting factor for temps in these types of structures.
The darkest color allowable, assuming 165F maximum surface temperature, would absorb no more than 73% of solar energy. Ideally, less. Keep in mind, 140F is a common initial heat-cure target for many epoxies, and is well below XPS service limit. 140F requires only 62% absorber efficiency under similar conditions. This absorption percentage probably corresponds to Medium-Grey, similar to US Navy aircraft.
A darker, true-grey primer, slightly darker than this, would probably be ideal if planning to place components in the sun to post-cure. Greyscale choice could be used to automatically limit the maximum temperature to a safe level for the foam core (<165F). And by limiting the maximum attainable temperature to the 160-165F range, (72% efficient absorber). This would allow longer-term curing for several hours, or all afternoon, as opposed to only 10 minutes if using black on a hot day.
Also, by using a darker primer for solar heating than the final paint scheme color, this would theoretically ensure the airframe was cured to a higher temperature than will ever be encountered in service.
Hours-long cure times would be beneficial for several reasons. Primarily by allowing this heat to saturate the entire structure. Components painted black, with the resulting 10 minute cure limit, will only heat the skin and not the embedded wing spars. Post-curing these major structural elements for longer durations to obtain the benefit of increased cross-linking, improved strength and resilience, would provide the greatest benefit for the finished aircraft from this extra curing effort, as opposed to simply post-curing the skins.
Maximum benefit would probably be gained by using a medium or darker-grey primer, one which absorbs 60-70%, and performing a 4-6 hour solar post-cure, starting early in the morning to generate a slower ramp time. Ending just after solar noon, or whenever the item reaches the target temp and dwells there for a specified timeframe, this temp and time should be based on the specific epoxy system. More even component heating could be facilitated by rotating the item every half-hour or so. And tighter quality control of the process comes from using a temp gun every 10-15 minutes to ensure the entire structure reaches an adequate post-cure temperature, without exceeding the above mentioned limits.
Note: Black primer may still be useful for post-curing in late Autumn or winter. Where the ambient temperature is below 50F. As black paint in bright sunlight seems to readily reach +110F above ambient temp. Black should probably be exposed to sunlight with abundant caution any time the ambient temp is at or above 50F. Also, post-curing to the 160-165F range should probably be done with wings or flying surfaces jigged to a torsion box/worktable. As any undesirable twist which develops while curing components near their maximum allowable temperature is going to be difficult or impossible to resolve, without re-heating to a higher (possibly dangerous) temperature. So, it makes sense to do it right the first time, and never worry about it again after it's been painted in a cooler, more reflective color.
Food for thought.
