Cover Image
close this book Boiling Point No. 03 - October 1982
View the document News from Shinfield
View the document New Nepali Chulo Extension in North India
View the document Mechanical Testing of Stove Ceramics
View the document Sexual Division of Labour in the Pottery Industry
View the document Biobriquets and Hybrid Stoves
View the document Sawdust Burning Cooker
View the document Village Studies in Sri Lanka
View the document Peace Corps Work in Upper Volta
View the document The Portable Magan Chula - a closer look

Mechanical Testing of Stove Ceramics

Included in the ITDG Stove Programme is E study of the materials aspects of ceramic stove design and construction. This work has entailed a consideration of alternative types of materials, their most appropriate form and the means by which they may be evaluated. Clearly full comparisons between specific materials cannot be made independently of design considerations: shape, size, type of fuel etc. However, it is reasonable to identify in general terms the critical properties for stove applications, of . given type of material, for use as a basis for comparison.

The low cost and availability of raw materials and skills make fired ceramics an obvious choice for this application. However their susceptability to impact damage, thermal stress and thermal shock are their major limiting factors. This area has been the principal focus of attention in this study.

The main difficulty has been the lack of an appropriate test which would give a reasonable indication of merit. Because of the combination of material properties which are involved, it can be argued that ceramics which show good resistance to thermal shock (rapid cooling or heating) would also be resistant to thermal stress (slow temperature cycling) and mechanical impact too.

In seeking a test method to establish thermal shock resistance we were conscious of some specific requirements:

1) The method should be relevant to the stove application, i.e. shock from temperature levels such as experienced at the surface of stoves, samples of appropriate proportions and direct assessment of shock effects.

2) The method should be fairly simple, being easy to reproduce and not requiring highly trained investigators.

3) It should not require sophisticated or expensive equipment that could not be readily purchased or manufactured anywhere within reason.

4) Results should be quantitative and reproducible so that comparisons may be made between results obtained from different laboratories.

5) Low cost.

At the outset a direct shock method was considered essential. 400C was taken as the highest realistic temperature that might be expected on the surface of a ceramic stove, and was therefore taken as the temperature from which to shock samples. To achieve a severe shock samples were plunged directly into a bucket of cold water (20C-25C). To minimize variation an extrusion method has been chosen for sample manufacture. m e die geometry is shown in Fig 1, the length of samples being about 85mm.

Exploratory tests showed that simply shocking samples even thirty or more times did not by itself cause fracture, but did induce surface cracking. A residual strength test (measuring the loss of bending strength resulting from shock treatment ) would register this type of damage, but would introduce too many problems in terms of the equipment needed and the interpretation of results. So we went for a 'residual impact' test involving a simple measurement which would be a reflection of the toughness of the material as well as the severity of the surface cracks.


The impact test we have adopted uses a light pendulum as the means of subjecting a simply supported test piece to repeated blows of gradually increasing severity. The fracture energy recorded is the energy of the impact which finally causes fracture. The method is open to criticism, especially as to the effect of repeated blows; however, results obtained to date are significant and repeatable.

The way we have used the test to assess the thermal shock resistance of a particular clay mix is as follows. Take a batch of between 12 and 24 test bars, subject half to the test as they are and subject the remainder to the test after 20 shocks from 400C into cold water. With this approach, the before and after measurements are of the same form making comparison straightforward, both between shocked and unshocked samples as well as between samples of different materials.

Fig 2 shows the apparatus we have used. Note the twin suspension strings and cylindrical weight (about 100gms to give increments of impact energy of about 10 Joules for each 10mm of height). To check pendulum height at release we use a rule taped to a weight and a solonoid from an old Post Office relay for smooth release.

Some results with an indication of scatter are shown in Fig 3. For the red clay used, an increase of firing temperature from 800C to -900 C is obviously significant. Wood ash, though a beneficial additive for impact resistance, is no help when it comes to thermal shock, while the samples containing sand (here one third of the total volume) are unaffected by shock.



Another useful test which is also fairly simple -to perform is the measurement of apparent porosity. Apparent porosity is the volume of open voids in a ceramic sample as a proportion (usually a percentage) of the total volume. For a given clay body, as the firing temperature increases, the degree of vitrification increases and porosity tends to fall to a minimum, however it can, of course, be dramatically affected by additives. Cur work to date suggests that while high porosity is generally a good thing as far as impact is concerned, it has a bad effect on thermal shock resistance, most probably due to the effectively increased heat transfer resulting from the penetration of water into the open surface pore structure.

Apparent porosity is found by weighing a dry sample (Ml) and then impregnating the voids with water, either by prolonged boiling or preferably by vacuum impregnator. The sample is then weighed submerged (M2) and finally weighed wet but in air (MB), the apparent porosity (AP) is given by:

AP = (M3 - M1)/(M3 - M2)



It is hoped that simple tests like these can be used by workers in different parts of the world to assess quantitatively the merits of some of the traditional methods, and maybe some novel ones too, for producing low cost refractories. Results may then be given wider circulation for the general enlightenment (or confusion) of all.

Richard Chaplin - University of Reading