According to Appendix A of the European Insulating Standard EN 13162 ... 13171 the manufacturer has to submit at least 10 test results of thermal conductivity for CE-certification. In the beginning only 4 tests must be carried out externally, later only 1 will be needed annually. The results are used to calculate the declared value of thermal conductivity λD by using the formula:
The factor k decreases with the number of tests done. According to Table A.1 of the European Insulating Materials Standard EN 13162 ... 13171 the value of k with the minimum number of 10 completed tests reads 2.07. With 50 test results k is decreased to 1.56.
Suppose one produces Type EPS S according to EN 13163. Aiming at the product quality of 0.040, one measures the value λmean of 38 mW/(m∙K) with a heat flow meter of the standard ISO 8301. With a maximum measurement error of 2.5 %, the heat flow meter achieves the standard deviation sλ of 0.577 mW/(m∙K). A larger number of measurements (50 instead of 10) alone results in a reduced (improved) λ90,90 value by 0.74 %.
|λmean||k||sλ||λ90,90||Δλ90,90||Improvements in %|
If these measurements for all manufactured products are to be carried out externally, it will cost more than about €35,000 to 40,000 ‒ the price of a top-quality guarded hot plate apparatus of the ISO 8302 standard.
Suppose one chooses not to use a heat flow meter of the ISO 8301 standard for these measurements, but instead a guarded hot plate apparatus corresponding to ISO 8302 as our Thermal Conductivity Test Tool λ-Meter EP500e (guarded hot plate apparatus), which guarantees a maximum error of 1 %, has an average error of 0.7%, and by this achieves a standard deviation of the thermal conductivity sλ of 0.129 mW/(m∙K). With the same λmean of 38 mW/(m∙K), its higher precision results in a reduced (improved) λ90,90 value by 2.53 %.
|λmean||k||sλ||λ90,90||Δλ90,90||Improvements in %|
As the diagram above showing the dependence of the thermal conductivity on the raw density demonstrates, one can reduce the raw density by 0.7 kg/m³ during manufacturing in our example (from 14.6 to 13.9 kg/m³).
With an annual production of 30.000 m³ and a net price of about €1.70 per kg for the granules, the savings will amount to approx. €35,700 per annum! For this sum you can already acquire a guarded hot plate apparatus tool like the high-precision thermal conductivity test tool λ-Meter EP500e, which comforms to established norms!
If however the compressive stress is the key dimension, the raw density will have a lower limit. In this case multiple and accurate tests can bring about substantial benefits for the respective manufacturer. The declared value of thermal conductivity will namely be considerably better than for manufacturers without their own accurate conductivity test tool.
EPS 100 is mostly needed in improved conductivity quality, e.g. with 0.035 W/m·K. A manufacturer with an accurate thermal conductivity test tool can produce the material which is usually foamed up to a raw density of 23 and 26 kg/m³, going to the lower limit and thereby bring about savings in costs.
Successful products such as Neopor and Lambdapor are exclusively characterised by improved thermal conductivity. They are used if thicker layers of insulating material cannot be accommodated. The manufacturing of these products would simply be unprofitable without highly reliable and accurate conductivity testing.
The same applies to mineral wool and other insulating materials.
A heat flow meter generally has to be calibrated by the user. This often has to be done repeatedly! However the test values for the calibration standard are not absolutely accurate and have inbuilt inaccuracies. These errors will therefore be carried over onto the heat flow meter with each calibration.
The European Standard EN 12667 Para 9.q therefore stipulates the following:
A heat flow meter has a sensor, providing an electric voltage as function of the heat flow that exists within the specimen at given surface temperatures and specimen thickness. It calculates the thermal conductivity based on an assumed linear function between the sensor signal and test temperature (mean temperature of the tool) as well as the heat flow density.
The sensor signal however is not entirely a linear function of the test temperature and the heat flow density. It also depends on the thermal conductivity and specimen thickness.
The European Standard EN 12667 Para 5.3.4 therefore stipulates the following:
Because the dependence is non-linear, the accuracy of the heat flow meter increases with the degree of similarity of relevant properties (conductivity and thickness) between the calibration specimen and the specimen to be tested. This implies that the test error increases with the dissimilarity between the calibration and measurement specimens. Calibration standards are only available as heavy mineral wool and white extruded polystyrene with a high raw density.
This unfortunately shows that not all our competitors’ statements are true. A manufacturer of heat flow meters for example claims in his sales brochure (kept by us on hand) the accuracy for his tool to be better than 1 %. Technically, this simply is nonsense as the standard used for calibration generally cannot possibly have a thermal conductivity accuracy of less than 1 %. In addition to this, there is still the error from the heat flow meter itself.
Even the FIW Munich, one of the most renowned test institutes in Europe, affirms for the test values of the heat flow meter only an accuracy of ± 2 %!
A guarded hot plate apparatus, like the Thermal Conductivity Test Tool λ-Meter EP500e (guarded hot plate apparatus), measures the energy that flows through the specimen at given surface temperatures on the sensor-plates with a given specimen thickness. With these values, it directly calculates the thermal conductivity and thermal resistance irrespective of material and thickness, yielding also the values of k and U. It does not require calibration (ref. EN 12667 page 22 paragraph q) and shows long-term reliability. It will have no errors even after years of operation.
A modern-type guarded hot plate apparatus, like the Thermal Conductivity Test Tool λ-Meter EP500e (guarded hot plate apparatus) does not need more time than a heat flow meter for measurement, and it only requires minimum maintenance. It can be operated without special training and offers convenient options for the processing of test results. These characteristics clearly go unmatched by any other tool.
The Thermal Conductivity Test Tool λ-Meter EP500e (guarded hot plate apparatus) measures thermal conductivity at exactly the desired test temperature. Unlike a heat flow meter, its heat control does not switch off when thermal conditions within the specimen become almost stationary allowing the test temperature to shift. This is common for heat flow meters and in most cases leads to a deviation of the real from the desired test temperatures.
Therefore a guarded hot plate apparatus according to EN 1946-2, ISO 8302, ASTM C177 should always be your first choice and not a heat flow meter according to ISO 8301, EN 1946-3 even if the latter is less expensive!
Heat flow meters range exclusively among the cheap tools. In most cases they are designed for smaller thickness specimens (50 mm or even thinner). Inhomogeneous conditions within the specimen (e.g. for mineral wool with a low raw density) can lead to inaccuracies. Additional measurement errors may occur with thin EPS especially of lower raw density as a result of heat radiation. The error increases as the specimen becomes thinner. For a 20 mm specimen with a raw density of approx. 12 kg/m³ (λ10 = 0.046 W/m·K) the heat radiation will cause an error of 10 %. Even for 040 material (λ10 = 0.040 W/(m∙K)) 50 mm will cause an error of 3 %!
The appendix B of EN 13163 contains a table for error correction (Table B3). However those factors are average values collected in tests over the last 30 years and do not correspond to present day conditions. In applying these factors for correction the resulting thermal conductivity tends to be too high. Therefore EPS with a low raw density should be tested with thicker specimens only (which need to have at least the actual thickness of the product).
Low-end tool manufacturers often unscrupulously claim a high accuracy for their tools (see above) which a renowned test institute would not do.
So-called non-stationary test procedures are also cheaper than guarded hot plate apparatus. Currently they are still permitted for internal quality control of EPS (in accordance with Appendix B of the European Standard for Insulating Materials EN 13163)
It is not economical to use any other tool than the guarded hot plate apparatus for internal quality control purposes. By using an accurate tool compliant to standards you can additionally evaluate the test results to obtain the declared value of thermal conductivity λD. In the end you will get a better (lower) value achieved by more numerous and more accurate test results without additional technical expenses or higher material usage.
This also means better product quality (lower thermal conductivity), which may even result in better prices. Alternatively one can maintain the desired product quality (with a given thermal conductivity) while the material usage and production costs are reduced.