The geometry of polymeric collectors differs from standard flat plate collectors due to material properties, the mentioned production methods and the resulting degrees of freedom in design. Various approaches to design a polymeric collector are under discussion. Mechanical stresses

occur during operation and stagnation due to thermal expansion and external loads on the collectors. These stresses have to be determined for investigating the durability of the product.

The efficiency calculations mentioned above lead to operating conditions of the collector which the chosen materials have to resist. Stress distribution and deformation show potential risks for the stability and durability of the collectors. They are analyzed for various material combinations and geometries with the models used in FEM-Simulations.

The temperature distribution depends on the position of the absorber layer whereas the temperature level is mainly influenced by the amount of irradiance, inlet temperature and the thermal losses of the collector. These parameters are varied to simulate different operating conditions. The temperature distributions during normal operation and stagnation are calculated using COMSOL Multiphysics. They are compared and validated with the results from the above mentioned simulations and common flat plate collector performances. They are used as input parameters for the numerical simulation of mechanical stresses due to thermal expansion.

Half a fluid channel of two meter length with glazing and thermal insulation was modelled 3- dimensionally including the possibilities to vary certain geometry properties and materials. The mesh and the model geometry used to calculate the mechanical stresses for an extrudable geometry are shown in Fig. 6. Due to symmetry only half of the channel had to be modelled. At both sides of the model the boundary conditions were set to be symmetrical. The calculation of the temperature distribution was done with the general

heat transfer module in COMSOL. The convective top and bottom losses were described with constant heat transfer coefficients. The ambience does not have to be modelled because of these boundary conditions. The convection in the air gaps was neither calculated instead a heat transfer coefficient was applied, too. These parameters were validated with Fluent calculations and literature references. A 2D model of half a collector crosswise to the channels was used for the simulation of an extrudable geometry. Its temperature distribution was transferred from the 3D model.

The distribution of thermo-mechanical stresses shows high gradients in certain areas of the collector. Therefore the mesh was improved by choosing small finite elements in regions with big stress gradients. Due to higher temperatures of its material, the expansion of the absorber is higher than the expansion of the glazing and back frame area. Since the glazing and the back frame are connected to the absorber, the highest stresses appear at the connection between the absorber plates and the channel dividing bars. Fig. 7 shows a close view on the connecting bars. In this example the whole collector is made of the same polymer e. g. PMMA and therefore the deformation is negligible.

A co-extruded collector of two polymers with different thermal expansion coefficients shows an increasing deformation for increased inlet/fluid temperatures. The variation of the inlet temperature and resulting stress distribution are shown in Fig. 8. The images show the distributions at the outlet of the two meter long fluid channel, where the highest temperature gradients appear. In general, the mechanical stresses due to thermal expansion are acceptable for the considered geometries.

The steady state simulations show the maximum stresses and deformations which occur in the simulated operating conditions at one certain point of time and not the deterioration of the material caused by fatigue stresses. Fig. 8. Examples of simulated mechanical stress

distributions during operation and stagnation for an External loads such as snow load were examined co-extrudable collector geometry of polymers with

with a 2D model, as well. Fig. 9 shows the Van — different thermal expansion coefficients

Misses-Stress distributions and deformations. The variation of the bar thickness indicates that the deformations due to top pressures increase strongly for thin bars. In EN 12975 the maximum permitted deformation at 1000 Pa top load is given as 0.5 %. This threshold can be realised by increasing the thickness of the bars to about 2.7 mm for this particular design. For a load of up to 3000 Pa, the maximum stress is in acceptable range regarding PMMA with a bar thickness of 3 mm. It becomes obvious that external operational demands have to be taken into account when designing polymer collector with maximized efficiency.

4. Conclusion

Computational fluid dynamics (CFD) and FEM-Simulations are a useful tool for the identification of characteristic problems of polymer collectors. The analysis of different collector and absorber layer geometries by simulation can be used for optimisation of the parameters regarding efficiency interrelationships as well as for the discussion of durability aspects in terms of mechanical loads. It shows that the heat transfer into the fluid can possibly be improved at higher fluid velocity and further variations of the absorber geometry. The effect of the collector design on thermo­mechanical stresses could be investigated by FEM-Simulations. Main aspects of a polymer collector are now better understood and the outcomes of this work show that both the efficiency and mechanical stability can be further optimised.

5. Outlook

We want to continue with the evaluation of candidate materials and start accelerated and real time tests on selected materials to identify the most promising polymers. Further, the aim is to build demonstration collectors and to perform further tests on durability and efficiency. We would appreciate to do this research in collaboration with industrial partners.

Using numerical simulations we want to calculate and compare the energy output of different designs and material combinations and optimize these designs to develop price efficient collectors of a good quality. Additionally, we want to identify the load levels at various service conditions and investigate different absorber coatings and their application ranges.

The daily temperature cycles and the stagnation situations in the summer, which the collector materials have to resist, cause material aging. Therefore time stepwise simulations using weather data and system operating parameters can lead to information on material aging, and to an estimate of the energy gain by taking into account the changes in the material parameters due to the thermal-mechanical stresses and other degradation appearing at operating conditions in future simulations.

Acknowledgements

The work was funded by the German Federal Ministry of Education and Research (BMBF FKz:

01RI05201).The authors also want to thank all the partners of IEA SHC Task 39 for their input,

especially the French colleagues from INES and CEA.

References

[1] Clough, Roger; Billingham, Norman; Gillen, Kenneth (editor): Polymer Durability: Degradation, Stabilization, and Lifetime Prediction. American Chemical Society, Washington 1996

[2] Duffie, John; Beckman, William: Solar Engineering of Thermal Processes. 3rd edition, Hoboken, New Jersey : John Wiley & Sons, 2006

[3] Kohl, Michael; Franke, Hannes; Stricker, Eva; Weifi, Karl-Anders: Polymeric materials for solar thermal collectors — a feasibility study. In: ESTIF: ESTEC 2007, 3rd European Thermal Energy Conference (proceedings). Freiburg : ESTIF, 2007, S. 223-229

[4] Jack, Steffen: Simulationsgestutzte Qualifizierung neuer Konzepte zur Gestaltung von thermischen Solarkollektoren auf Polymerbasis. Freiburg : Fraunhofer ISE; FHTW Berlin, 2008