Half-model-balance for the Cologne-Cryogenic-Wind-Tunnel development and test results

After the instrumentation of the Cologne Cryogenic Wind Tunnel (KKK), with internal balances, it has been a demand of the tunnel customers to make half model tests at high Reynolds numbers. The Mach number range from zero to 0.3 and the relative low costs with regard to the high Reynolds number made the KKK very attractive for half model testing. To realize this capability DLR and the Department of Aerodynamics and Measurement Techniques of the Darmstadt University of Technology (TUD) decided to develop a half-model-balance for the KKK in 1994. A very challenging task, because no previous attempt to solve this problem for a continuous running cryogenic wind tunnel was successful. Fortunately the tests in the tunnel with the first balance (W519) we built were very successful. During the tests a lot of experience was made and some new problems appeared. The main problem was the unexpected great temperature gradient inside the balance, although there was a very sophisticated heating and insulating device around the balance. TUD a finite element study was made to design a new balance in such a way that the balance is more insensitive to temperature gradients. Unfortunately the balance W519 was destroyed by an accident in the tunnel and we had to built a second one (W523) very quickly. All the things we learned had to be turned into practice very quickly without any tests in the laboratory. Finite element calculations turned out to be a very useful tool to solve this task. In November 1997 we started to design and built the second balance and we finished it in August 1998. The very short time for the manufacturing of 9 month was achieved by the use of a three dimensional EDM-machine and the change of the calibration procedure. Instead of using weights as true calibration loads, we used the external balance of our low speed wind tunnel as a reference balance to give us the "true loads". By this method no realignment of the calibration model was necessary and the calibration process could be done in three weeks. The accuracy of the new procedure, based on recalculation of the calibration loads, turned out to be better than the former one, because the accuracy of the external balance is one magnitude better than the required accuracy of the half model balance and there are no errors by misalignment. The balance W523 could be delivered to the tunnel just in time and it was first used successfully in September 1998. The comparison between some measurements, made with the old balance and the new one, showed a very good correspondence of the results. The repeatability of polars with the new balance in the tunnel was about one drag and lift count.