Temperature evolution inside a pot during experimental surface (bonfire) firing

Abstract

Time–temperature evolutions of different parts inside a pot were recorded during three experimental surface (bonfire) firings. The experimental vessels were shaped from a calcareous clay, tempered with 30% vol. of oolithic limestone. The temperature–time recording showed: (1) T_max. was reached after 12–22 min and differed between the firings as well as inside individual vessels; (2) the range of the thermal variation within one single firing was found to be as high as 390 °C, and up to 220 °C on a specific cross-section; (3) the lowest temperature was not systematically recorded in the core of the object, as is generally expected. Under the polarizing microscope, no textural and mineralogical changes were observed in the ceramic bodies. This is sustained by powder X-ray diffraction analyses evidencing no dolomite or calcite breakdown. The presence or absence of specific illite and chlorite peaks can be generally related to T_max. and soaking time, but equivalent firing temperature estimations do not match the measured temperatures.

Introduction

Neolithic pottery was open-fired, without any permanent kiln superstructure, either in a surface firing (bonfire), placing the dried ware directly above ground and surrounding it with fuel, or in a pit firing, placing the dried pottery in a hole in the ground and stacking the fuel around or partially below it. Additionally, the pots can be covered with sherds. In both procedures, the pottery was in direct contact with the fuel and the flames. In the first case, the retention of heat is more difficult than in the second one. Ethnographic firings give some hints on the rate of temperature increase and the length of burning. Shepard (1976) noted a rapid rise in temperature when using dung as fuel, reaching flame temperatures of 900 °C in 20 min. She states (p. 78): “After volatile matter was burned out, as shown by the cessation of flame, the temperature continued to rise to a peak of 940 °C and then fell rapidly because the dung chips do not leave charcoal that maintains heat”. This peak temperature was reached during the following 20 min. Shepard (1976, p. 83) reports a maximum flame temperature of 962 °C with wood “stacked around pottery” for a surface bonfire firing and wrote that 1000 °C can be exceeded in a pit firing with a flue. Ethno-thermometric data of Colton, 1951, Lauer, 1974, Shepard, 1976, Shepard, 1977, Irwin, 1977, Nicklin, 1981, Rye, 1981, Pinçon, 1984, Tobert, 1984a, Tobert, 1984b, Woods, 1984, Miller, 1985, Mpika, 1986, Seehy, 1988, Nicholson & Patterson, 1989, and Kanimba and Bellomo (1990) were summarized by Gosselain (1992). Two thirds of the flame data were in the range 300–900 °C for surface bonfires, 670–870 °C for surface firings with broken ceramics covering the pottery, 620–870 °C for pit firings and 770–870 °C for pit firings with broken ceramics covering the pottery. Gosselain (1992) was the first author to study not only the evolution of the flame temperature during an ethnographic firing, but also to record the maximum gas temperature variations on the same ceramic object by placing 2–4 thermocouples directly on the inner and outer walls of the objects. He showed that within one single firing, differences vary from 94 °C to 295 °C. Such high temperature variations on a single vessel were also recorded by Nicholson & Patterson, 1989, Wotzka, 1991. Livingstone Smith (2001) completed these observations based on 105 ethnographic records from Africa.

The rapid increase in temperature as well as the short firing cycle (= quick rise and cooling down of temperature) in a bonfire firing was replicated in many experiments. With sago fronds as fuel, maximum flame temperatures of 746 °C were reached after 4 min, while the whole firing was finished after 10 min (Lauer 1972). Gibson and Wood (1977) recorded maximum flame temperatures of 700 °C and a firing cycle of 100 min for an experimental open firing done with brushwood or 800 °C/100 min for wood as fuel. Martineau and Pétrequin (2000) placed three thermocouples directly on the walls of a single vessel and found maximum flame temperatures of 620, 720 and 860 °C in the case of a bonfire and 620, 760 and 800 °C or even 900 and 980 °C (experiment with two thermocouples) for a pit firing. Firings were achieved in 80 min. Thermal profiles, heating rates, soaking times, cooling rates and the duration of firings in a pit firing with thermocouples stacked on the inside and outside walls of a pot were also reported by Sestier (2007).

These ethnographic studies and experiments gave many precise results on the thermometric evolution of flames in the pottery stack and on the surface of the walls. However, no attention was given to: (1) the thermometric evolution inside a pottery; (2) the textural changes of the material and (3) the mineralogical reactions in the body. In order to answer such questions, three firing experiments were conducted, placing the thermocouples in the core and the inner and outer rims of the objects and analyzing the fired specimens by optical microscopy and X-ray diffraction.