Nicolas Aubin asked 3 weeks ago
Hi there,
when doing a sail analysis operating point export into a file or clipboard, it seems to showcase a strange progression for the Cp. The last column associated to Cp in the example below is repeating itself by block of 3 rows alongside the entire data set.
Panel CtrlPt.x CtrlPt.y CtrlPt.z Nx Ny Nz Area Cp
M1 medium resolution_Cp Coefficients
0 0.631401 -0.0322982 0.0731327 -0.0863423 -0.995727 0.0327577 0.000285918 0.748747
1 0.637785 -0.0326767 0.0823715 -0.10581 -0.994214 0.018491 0.000285852 0.748747
2 0.61164 -0.0304312 0.0737465 -0.0892 -0.995116 0.0422749 0.000278211 0.748747
3 0.617879 -0.0306895 0.0829804 -0.103337 -0.994126 0.0321652 0.000278032 0.299569
4 0.591371 -0.0284925 0.0743452 -0.0912455 -0.99462 0.0490469 0.000292442 0.299569
5 0.597971 -0.0287084 0.0835627 -0.100801 -0.994024 0.0418871 0.000292204 0.299569
6 0.571101 -0.0265485 0.0749228 -0.092441 -0.994306 0.0530048 0.000285058 0.462338
7 0.577548 -0.0266945 0.0841409 -0.0981568 -0.993973 0.048825 0.000284781 0.462338
8 0.550993 -0.0246265 0.0754821 -0.0927089 -0.994177 0.0549231 0.000285093 0.462338
9 0.557457 -0.0247385 0.0846973 -0.095462 -0.994026 0.0529112 0.000284796 0.511041
10 0.530872 -0.0227243 0.0760369 -0.0920507 -0.994214 0.0553701 0.000285216 0.511041
11 0.537352 -0.0228152 0.0852491 -0.0926608 -0.994182 0.0549245 0.000284909 0.511041
12 0.510749 -0.0208564 0.0765948 -0.0904884 -0.994383 0.0548947 0.000285067 0.496036
13 0.517236 -0.0209335 0.0858031 -0.0897446 -0.994421 0.0554368 0.000284763 0.496036
14 0.490628 -0.0190331 0.0771596 -0.0885034 -0.994624 0.0537611 0.000284937 0.496036
15 0.497122 -0.0191021 0.0863641 -0.0868038 -0.994706 0.0549976 0.000284613 0.485266
16 0.470527 -0.0172591 0.0777292 -0.0864451 -0.994896 0.0520433 0.000284471 0.485266
17 0.477019 -0.0173251 0.0869318 -0.0839067 -0.995016 0.0538846 0.000284094 0.485266
Is the export done at the node of the triangular mesh or the center point of each triangular panel? If it is the node of the triangular mesh, then you may expect identical pressure on the triangle element but thise points may be repeated with each Cp values for their neighbor surface. If it exports the centre of each triangular panel, then there might be something to look at as Cp repetition is kind of unexpected.
When plotting the Cp field distribution by previewing the exported file into Paraview, it then showcases a strange discontinuity and periodicity along z. I have put screenshots of the preview in: https://1drv.ms/f/c/f5481a46a90a15ef/EsV5V47al_hHhsamLo9eqsUBbp5DZG53UUisgTEBbWVb_w?e=YNQ35Y
Also, the label value next to the color bar when showing Cp are not changing and remain at 0.00 for sail analysis. These observations are made under flow v7.50
I encountered the Cp observation as I was trying to extract some Cp curves in sliced plane. This kind of analysis is quite handy in sail analysis / engineering and may be a good feature to implement into flow5: plotting the Cp value against the sail chord or sail curvilinear absissa for defined slicing planes by normal and location (quite often fixed in the boat frame, so following the heel and pitch angles, but may be useful to have fixed in the terrestrial frame as well). Having this feature with the ability to export the 2D graphs (Cp vs chord or curvilinear abscissa) would facilitate comparison with experimental cases.
Thank you again for your great tool,
Nicolas
Nicolas Aubin answered 1 week ago
Hi Andre,
just to confirm as well, the Cp mentioned here is already reflecting the delta of Cp on both side of the panel, is that correct?
I cannot remember if the computation panel method behind gives access to the overall Cp (delta Cp) only or has access to each Cp on both side as well (suction and pressure or leeward and windard side). Having the Cp as described now is of great value, but if the method is using it and computing it in the background, having the Cp on each side would also be really useful for comparison with experimental data. Experimental data can either measure the delta Cp or the Cp on each side of the skin of the sail depending on the publication. Having the Cp on each side information would facilitate comparison with experimental data for some available literature.
Thank you very much
Nicolas
Hi Nicolas,
By nature, thin surface methods can only evaluate the Cp difference. To access the value on each side, the sail must be modelled as a thick surface.
André
Hi Nicolas,
Will check and post back.
André
Hi Nicolas,
Bug confirmed. Will fix it in the next version which is planned in June.
Will also add the plot of Cp = f(chord, z) to a future version, hopefully the next depending on the work required.
André