Air, Surface and Material Temperature: Three Different Limits

Why this matters

A tech reads 52 F on the truck's dash, sees "minimum 40 F" on the tube, and caulks a north-side window that is sitting at 38 F. The joint never bonds, and the callback lands in spring when the bead peels out in one piece. Nobody broke a rule they knew about. They checked the wrong temperature.

Data sheets list air, surface and product temperature as separate lines because they are three different things, they move independently, and each one controls a different part of how the product behaves. The general rule that the sheet sets the limit and the forecast says whether the window fits is owned by 'Weather as a Work Condition: The Rule for Every Trade'. This card is about which temperature the sheet means.

What each temperature controls

Air temperature is the one the forecast gives you, and on a cloudy, still day it is a fair stand-in for the other two. That is exactly why it gets trusted on days when it is not.

Surface temperature is the one most often missed, and it is the one most products care most about, because the bond forms at the surface. A thin film or a sealant bead takes on the temperature of what it touches within minutes. Apply a product rated for 40 F to a 38 F wall on a 52 F day and the product is working at 38 F. The air reading is irrelevant to the bond.

Material temperature matters most for products that must flow or be mixed. A tube of sealant that spent the night in a 20 F truck may be well inside its application range once it is on a 45 F wall, but it will not gun or tool properly until it is warmed, and a tech forcing it will under-fill the joint. Two-part products are worse: the mix ratio is fixed, but the reaction speed and the working time both swing hard with the temperature of the components. Handling materials through heat and freeze is its own card; see related: Materials Riding in the Truck Through Heat and Freeze.

Why the surface departs from the air

Surfaces gain and lose heat in ways the air does not, and the departures are large and predictable.

• Sun. A sunlit wall absorbs radiation and runs well above air. Dark colors and low-mass materials run hottest. The same house can have one wall 20 F or more above air and another a few degrees below it at the same moment.
• Shade and mass. A shaded wall, especially masonry, lags the air. It warms slowly after a cold night, so on a morning with rising air temperature it reads below air for hours.
• Night sky. Any surface facing an open, clear sky radiates heat to it at night and can run several degrees below air. Roofs, car hoods and deck boards are the classic cases. This is why frost forms on a roof when the thermometer reads above freezing; see related: Radiational Frost: Why a Surface Freezes Above 32.
• What is behind it. A wall over a heated room, a pipe carrying cold water, or a slab on cold ground pulls the surface toward that temperature regardless of the air.

     Plan of one house, 10 a.m., clear, air 48 F in shade

                 N  north wall, shaded: 44 F
        +--------------------------------------+
        |                                      |
 west   |    metal roof panels above:          | east wall,
 wall,  |    84 F in direct sun                | sun since
 shade: |                                      | sunrise:
 45 F   |                                      | 61 F
        |                                      |
        +--------------------------------------+
                 S  south wall, full sun: 66 F

   One air temperature. Five surface readings, 40 F apart.

These are one illustrative morning's readings, but the shape is what you will find on any clear day: shaded walls below air, sunlit walls well above, and a metal roof in sun far above everything. At night the roof flips and becomes the coldest surface on the property.

The falling temperature during cure

The temperature that fails a job is very often not the one at application. It is the one six or ten hours later.

Most sheets that set a minimum temperature mean it to hold through some stated part of the cure, not just while the tech has the gun in hand. The exact wording differs: some say the minimum applies during application and for a stated number of hours after, some give a cure-time table at several temperatures, some say "do not apply if temperatures are expected to fall below the minimum within" a stated period. Read the wording, because it decides whether an afternoon start is allowed for that product.

The trap is the sunlit wall. It reads comfortably warm at 2 p.m., the crew applies, and then the sun goes off the wall. A low-mass surface like vinyl, aluminum or thin wood follows the drop within an hour or so; by evening it is near air, and on a clear night below it. The product passed its application check and then spent its first night under its minimum.

Worked example: one window, two walls, one afternoon

A handyman is booked to reseal four window perimeters on a vinyl-sided house in November: two on the south wall, two on the north. The sealant's sheet reads, say, air and surface at least 40 F during application and the product must not freeze or go below 40 F for the first 24 hours. (Illustrative values; use your product's sheet.) The tubes rode in the cab overnight at about 55 F.

The forecast: 41 F at 8 a.m., a high of 50 F at 2 p.m., 43 F at 6 p.m., a low of 31 F at 5 a.m., clear.

At 9:30 a.m. the tech takes readings with an infrared thermometer, aiming at the vinyl and the window frame next to each joint:

• Air in shade: 44 F.
• South wall, in sun since about 8 a.m.: 59 F at the joint.
• North wall, shaded: 41 F at the joint, and 39 F on the lower window where it sits close to a cold concrete stoop.

What the readings say. The south windows clear the 40 F surface minimum by 19 F. The upper north window clears it by 1 F, which is inside the error of a typical infrared thermometer, and the lower north window is 1 F under. The air reading of 44 F would have passed all four.

What the forecast says. The 24-hour line on the sheet is the one that decides the day, and it fails for every window: the night goes to 31 F, 9 F under the 40 F minimum, and a clear sky will put the siding below air. No time of day makes the cure window fit for this product.

The call. The tech has two honest options. One is to reschedule to a forecast that holds 40 F for 24 hours. The other is a sealant whose sheet allows application and cure at lower temperatures; many exterior sealant families include such a product, and its sheet will also state a surface-condition requirement (dry, frost-free, clean) that becomes the real limit. The tech calls the office, the office has a low-temperature sealant on the shelf, and its sheet, say, allows application down to 0 F on a dry, frost-free surface.

With that product, temperature stops being the limit and surface condition becomes it, and the north wall is where that bites: a shaded wall that ran below the dew point overnight can still carry a film of moisture in the corners of the frame. The tech wipes a dry cloth along each joint, looks for darkening, and does the north side after lunch, when it has had the longest to dry. The south side goes first.

What would have happened otherwise. Reading air only, the tech caulks all four windows with the original product by noon, passes every check on the truck, and leaves four joints that spend their first night 9 F under their minimum, and the two north joints on a surface that was 1 F over or 1 F under it from the start.

Measuring surface temperature without fooling yourself

An infrared thermometer is the right tool, but it is easy to read wrong.

• Emissivity. Most IR thermometers assume a surface emissivity around 0.95, which fits paint, wood, masonry, vinyl and most dull surfaces. Shiny bare metal (bright aluminum, new galvanized, stainless) reflects the surroundings instead of emitting, and reads falsely low, sometimes by tens of degrees. On bare metal, put a piece of masking tape or a dab of flat paint on the spot, let it come to temperature for a minute, and read the tape, or use a contact thermometer.
• Spot size. The reading is an average over a circle that grows with distance. Read close enough that the circle is on the joint and not half on the sky or the window glass.
• Read the coldest spot, not the average spot. The bond fails at the coldest joint, so the reading that governs is the lowest one on the elevation, which is usually low on the wall, near grade, near a slab, or in an inside corner.
• Read it twice. Once at application, and once again before the last section on that elevation, because a wall that is losing sun is falling while you work.

What changes the answer

• A coating rather than a sealant. Coatings add the dew-point margin as a fourth limit that also lives at the surface; see related: The Surface Five Degrees Above Dew Point Rule for Coatings.
• A maximum temperature. Many coatings and some sealants also list a maximum surface temperature, and a sunlit dark wall or a metal roof in summer can exceed it. Then the advice inverts: work the shaded side, and follow the sun around the house so each wall is coated after the sun has left it.
• A heated interior. An exterior wall over a heated room may run warmer than the forecast suggests on the inside face and at the frame. That helps an interior application, not an exterior one, because the outer surface still faces the night.

References

• Manufacturer product data sheets for the specific sealant or coating (the governing limits)
• Infrared thermometer manufacturer documentation on emissivity and spot size
• See related: Weather as a Work Condition: The Rule for Every Trade
• See related: Radiational Frost: Why a Surface Freezes Above 32; How to Take a Dew Point and Surface Reading Before Coating
• See related: Materials Riding in the Truck Through Heat and Freeze