Anchor rod tolerances are written in eighths of an inch. Structural steel changes length with temperature by amounts in the same range, so on long runs the time of day a measurement is taken can matter as much as the instrument taking it. This article covers how much steel and concrete actually move, where that shows up in anchor bolt checks and steel as-builts, and how we plan surveys and scans so temperature does not end up in the result.
How Much Steel Moves
The standard design value for the coefficient of thermal expansion of structural steel is 6.5 millionths per degree Fahrenheit (about 11.7 millionths per degree Celsius). Change in length is that coefficient times the length times the temperature change. For 100 feet of steel, that works out to about 0.008 inch per degree Fahrenheit.
| Length of steel | 10°F change | 20°F change | 40°F change |
|---|---|---|---|
| 2 ft (one bolt group) | 0.002 in | 0.003 in | 0.006 in |
| 50 ft | 0.04 in | 0.08 in | 0.16 in |
| 100 ft | 0.08 in | 0.16 in | 0.31 in |
| 200 ft | 0.16 in | 0.31 in | 0.62 in |
For comparison, the AISC Code of Standard Practice and ACI 117 allow ±1/8 inch between anchor rods within a group and ±1/4 inch between groups. Inside a single bolt group, temperature is negligible. Across a building, a 20 to 40 degree swing between a cool morning and a sunny afternoon uses up most or all of the between-group tolerance before anyone has made a mistake.
Concrete Moves Too, at a Different Rate
Anchor bolts sit in concrete, and concrete expands with temperature as well. The Federal Highway Administration puts the typical range for portland cement concrete at about 4.1 to 7.3 millionths per degree Fahrenheit, driven mostly by the aggregate, since aggregate makes up about 70 percent of the mix. So the bolt layout in the foundation and the steel that lands on it are not guaranteed to grow and shrink together, and neither one is necessarily at the temperature the shop fabricated to. The practical point is that a bolt survey is a measurement of a moving system, and it should say when and under what conditions it was taken.
Sun and Shade
Uniform temperature change is the easy case. Sunlight heats one face of a member more than the other, and the member bows toward the cooler side. In one 2022 field study of steel beams under solar radiation, the temperature difference between top and bottom flanges reached roughly 9 to 10.5 °C (about 16 to 19 °F), with peak temperatures in the early afternoon and the smallest differences at night. For as-built work this means columns that are plumb at dawn can read out of plumb in the afternoon, and a beam elevation taken in full sun is not the same number as one taken on an overcast morning.
The Instruments Are Affected Too
Electronic distance measurement depends on the speed of light through air, which changes with temperature and pressure. As a rule of thumb, a 1 °C error in the atmospheric correction produces about 1 part per million of distance error, or 0.001 foot per 1,000 feet. At building scale that is small, but only if the correct temperature and pressure are entered or measured.
The bigger instrument issue on steel is timing. A scan or a set of total station shots records the structure as it was at that moment. A scan project with several setups spread across a warming afternoon is recording a structure that is changing while it is being measured, and those setups then have to be registered together. The fix is in planning, not software.
How to Measure Around It
- Measure early or overcast. Temperature is most uniform before sunrise and through the morning, and steady conditions matter more than any single reading.
- Keep control off the steel. Hold the survey to control on the ground or the foundation, not on members that are moving.
- Record time and temperature. Note air temperature and, on steel, surface temperature in sun and shade, so the result can be interpreted later.
- Keep scan sessions short and setups tight. Fewer hours between the first and last setup means less movement inside the dataset.
- Check, and report the check. Independent check shots that were not used to build the result, with their residuals in the report, show the accuracy achieved instead of assuming it.
- Compare like with like. When steel and bolts are surveyed at different times, compare under similar conditions or account for the temperature difference.
How We Handle It
For anchor bolt, embed, and steel verification, we work from site control, scan to 1/8 inch, and add total station check shots on any bolt group near tolerance. Our reports record when the work was measured and the check-shot residuals. When a job needs tighter verification, we plan the timing, control, and checks for it, and the report documents the accuracy achieved.
Sources
- ASCC Position Statement #14, Anchor Bolt Tolerances (AISC 7.5.1 and ACI 117 anchor rod tolerances)
- Federal Highway Administration, Thermal Coefficient of Portland Cement Concrete
- Sensors (2022), Investigation of Temperature Variations and Extreme Temperature Differences for the Corrugated Web Steel Beams under Solar Radiation