Large discrepancies between flow5 and real world results

Questions, suggestions, feature requests, bug reports, feedbackCategory: QuestionsLarge discrepancies between flow5 and real world results
William WoodWilliam Wood asked 2 years ago

Hello all, I am writing to ask the typical discrepancies between experimental and flow5/xflr5 results. I have been modelling the ask 21 glider as part of a university project and I am currently getting very different results from the real world data that is available. I have an exact model of the wing that I have implemented in flow5 as well as a fuselage model that is good but not perfect. I am looking to compare the max glide ratio and glide polars available. I currently get a glide ratio that is approximately equal to the experimental data when only analysing the wing, however adding in the fuselage the max CL/CD decreases massively from around 33 to around 20, (although in line with https://flow5.tech/wp-content/uploads/2019/11/Fuselages.pdf). Does anyone have any idea as to what could be the problem here, it would seem that there is a problem with the original calculation for wing lift, if the contribution of the fuselage is reasonable?  I have tried with all appropriate modelling methods, with ncrits advised for gliders in xfoil documentation and with repanelling the aircraft model and the original airfoils numerous times. If anyone has any advice that would be greatly appreciated. I was also wondering if whether the discrepancy between lift and induced drag could have anything to do with it? I seem to be unable to attach an image however there is no large difference in local Cl along the wing, however induced drag is orders of magnitude higher at the fuselage part of the wing on the graph. I understand that it is including the contribution due to the interaction of the fuselage with the flow, however I would not have expected such a considerable difference. Thanks for reading, I hope I have made myself clear.  

William WoodWilliam Wood replied 2 years ago

Hello again,
I thought I would give a little update as to what I have looked into and how it is progressing. Firstly I have looked at ways of estimating the contribution due to the fuselage via rough estimates. I have looked at a method described in Aircraft Design: a conceptual approach by Daniel Raymer. The estimation methods such as the effective skin friction approach and the component build up approach model the fuselage drag effect as a CD0 term with a recommended interference factor of 1 between the wing and the fuselage for geometries like that of the ASK 21. including this term as an added drag component with a wing only model give a CL/CD ratio that whilst lower than experimental data is a lot closer, around 30. This makes me think that although something may be slightly wrong with the modelling of the wing itself, more of the issue lies with how my fuselage is integrated into the design. I noticed that the interference factor recommended could be significantly higher (up to 1.4) for a fuselage wing fairing that wasn’t aerodynamically optimised. It is possible that my design does not reflect this optimisation.

André Deperroistechwinder Staff replied 2 years ago

Hello William,
It’s difficult to say where the problem may be without having a look at the project file.
I’ll send you a private message.
André