Initial Points of Interest

  1. The peak temperatures are staggered corresponding to the depths of the respective temperature sensors
  2. The sensors closer to the surface registered higher temperatures than the ones further into the wall
  3. The heat was slowly released from the wall after the fire died out (over ca. 40 hours)

Data & Hypothesis

The above left chart ("Maximum Temperature Values") shows the peak temperatures reached for each temperature sensor with the timestamp label below.

We find that the average time difference between temperature peaks for each temperature sensor is 22.5 minutes.

(20 + 30 + 20 + 20) / 4 = 22.5

The sensors are placed at one inch intervals into the cob wall so our data is telling us that the heat moved one inch for each 22.5 minutes.

Our hypothesis states that heat moves through a cob wall at a rate of 1 inch per hour. Our test results disprove our hypothesis. The heat moved through the wall at 1 inch in 22.5 minutes or 2.7 inches per hour.

(1 in / 22.5 min) * 60 min = 2.7 in

Additional Observations

  1. Slow release of heat
    The data shows that after heating the cob wall, using the wood-burning stove, it took about 40 hours for the wall to return to an unheated or ambient state.

  2. Temperature difference
    The weather data tells us that the average temperature difference between outdoor and indoor temperature was 40-45 degrees Fahrenheit.

    Average heating temperature - (maximum-minimum) outdoor temperature = average temperature difference
    90 - (43+((52-43)/2)) = 42.5


  1. Hypothesis
    The rate of heat flow through a material is dependent on (a) the conductivity of the material and (b) the temperature differential on each side of the material. Our hypothesis does not take both sides of the equation into account. This means that it is difficult to test as it is (currently) phrased.

  2. Methodology
    There is a possibility of (a) air infiltration on the sides of or inside the straw and (b) thermal bridging through the plastic straw and/or the temperature sensor cable (see methods). Temperature changes that we assign to heat transfer through the cob may partially be due to one or both of these factors. If either of these factors were at play, we would expect to see a relatively rapid increase in temperature for the deeper sensors - faster than what would be possible by conductivity through the wall. The current data does not seem to be sufficient to allow us to draw any conclusions on this topic.

home | abstract | intro | hypothesis | methods | data
analysis | conclusions | appendix | acknowledgments

Cob Comfort - a study of heat transfer through cob walls
Kathy Bevers | Per Kielland-Lund | Arch 591 Fall 2003 | GTF: Sara Goenner