February 2023 - Nektar-users

Questions about simulations of a flexible cylinder by Nektar++
by Zhipeng Yu 05 Mar '23

05 Mar '23

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Dear Professor Hope this email finds you well. My name is Zhipeng Yu, and I am a PhD student of Tsinghua University, China. Recently, I encountered some problems when I used Nektar++ to carry out simulations of a flexible cylinder. The problems are shown as follows: 1. Can we define the natural frequency of a flexible cylinder in the in.xml file? Besides, could you tell me the relationship between EI and fn? It seems the definition of some parameters is different from other references, such as the mass ratio. 2. During post-processing, the real displacement of a cylinder cannot be seen by using "FieldConvert in.xml in.fld out.dat". Are there any ways to show the transformed coordinates of the cylinder in tecplot? 3. How to turn the DNS to LES by Nektar++? Thank you for your patient. Best regards. Zhipeng Yu

2 1

Compressible Flow Solver - Axisymmetric and specifying output variables
by Kenneth Hinh 28 Feb '23

28 Feb '23

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Hello Nektar++ users and developers, Regarding the compressible flow solver: For solving the compressible NS equations in axisymmetric coordinates, in the Github there is the "NavierStokesCFEAxisym" solver and the Forcing type "AxiSymmetric". I see that the forcing type is sometimes used with other solvers (for example in nectar/solvers/CompressibleFlowSolver/Tests/Nozzle_AxiSym_Swirl.xml where it's used with EulerCFE), but in another case it is used in conjunction with NavierStokesCFEAxisym (in CompressibleFlowSolver/Tests/PipeFlow_NSAxisym.xml). 1. What is the suggested way to solve in axisymmetric coordinates? Is the Forcing supported with the implicit compressible flow solver in 3d? 2. Are the variables u,v,w simply mapped to u_x, u_r, and u_theta when configuring the session file? (and to clarify is the axis of symmetry around x=0?) Secondly, I am working with a simulation with a large DOF and am trying to reduce the size the checkpoint files. By default, 15 variables get saved for each checkpoint file. Is there any way of selecting which variables to print to reduce the checkpoint file size? For example, to not output redundant variables such as rhou, rhov and rhow. Thank you for taking the time to read my questions, and please forgive me if answers to my questions can be found in the Archives or on the user documentation - I have tried my best to look. Kindest regards, Kenneth Hinh Graduate Student University of Calgary Alberta, Canada

1 0

Invert results about a symmetric axis via post processing | 2D flow
by nandan.sarkar＠iitkgp.ac.in 22 Feb '23

22 Feb '23

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Dear all, I want to obtain a two-dimensional channel flow solution (y=-0.5 to 0.5) by simulating half of the channel (y=-0.5 to 0, with symmetry boundary conditions at y=0) in Nektar++. I have the resulting field file (.fld) of lower half channel but need to invert the same about the y=0 centerline to get the field of upper half channel and merge both to obtain full channel flow field. Can this be done in FieldConvert or some other Nektar++ tool? I need to use the full channel field as a restart condition for further purpose. Thank you in advance. Regards, Nandan Sarkar PhD candidate Dept. of Mechanical Engineering Indian Institute of Technology Kharagpur India 721302

3 3

Adjoint stability of Rayleigh Benard Posillue Flow
by Yicheng Cao 21 Feb '23

21 Feb '23

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Hi Nektar-users, I am doing stability analysis on the 2D Rayleigh Benard Posillue (RBP) Flow. I am using the Boussinesq approximation as the buoyancy force. I get the correct eigenvalue for the direct mode. However, when I change the EvolutionOperator to Adjoint, I cannot get the correct eigenvalue. Here are some details, since the Adjoint LNS equation is different from Direct LNS (Adj LNS and Dir LNS equation are attached in the figure for your convenience). The body force for the Direct LNS in v-mom is const*c1 (c1 is the scalar). But in Adjoint LNS, there is a new term (const*v) in the scalar equation. So the body forces need to be modified as, <FUNCTION NAME="BodyForce"> > <E VAR="u" VALUE="0" /> > <E VAR="v" VALUE="0" /> > <E VAR="c1" VALUE="const*v" EVARS="v" /> </FUNCTION> I am wondering if I can put it into the scalar equation (c1) as the above function. Or any idea is welcome. Thank you for your time Yours Yicheng

2 3

Periodic Hill Re=2800 Nektar++ simulation
by Ma, Ting-Hsuan 15 Feb '23

15 Feb '23

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* To those whom this may concern, I have been trying to simulate the constricted channel case posted in the appendix of the Nektar team’s paper (https://www-sciencedirect-com.pitt.idm.oclc.org/science/article/pii/S001046…) specifically section 4.2. Using the configuration files provided as is, I am not able to simulate the case past the 2nd output (don’t remember what physical time this is). The specific error was “Nan was detected”. On further investigation into the configuration file, I also noticed that while the configuration file specifies the fluid is driven by a body force, but it lacks the <FORCING> stuff specified in the documentation. I don’t think this is the root cause of the blowup, but may have contributed? My main concern is that I am not able to get a stable simulation in the Re=2800 case, and I’m trying to move up to Re=5600. Is there some pointers you guys could give me to help me run a stable, turbulent, constricted periodic channel? Thanks, Dustin -- Dustin (Ting-Hsuan) Ma Graduate Researcher Department of Mechanical and Materials Science University of Pittsburgh Tim48(a)pitt.edu<mailto:Tim48@pitt.edu>

1 0

Spurious results using composites with different discretization orders
by Guillermo Suarez 13 Feb '23

13 Feb '23

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Hello, I want to solve the incompressible flow around the naca0012 with nektar++. I have attached the input file that I am using for such a case. For nektar++ versions v5.1.0, v5.2.0 and 5.3.0 (master) some spurious results, two checkerboard patterns, around half and one chord length right above the trailing edge show up. Such a checkerboard pattern does not appear when using nektar++ v5.0.1 and v5.0.2. I have enclosed a screenshot showing the comparison between the different versions. The checkerboard pattern seems to first appear in the horizontal velocity, u. It develops right at the boundary between two composites with different discretization orders. This is not the only boundary between composites with different nummodes. The pattern does not belong to a transitory state, once it shows up it will not fade or get shed outside of the domain. It will eventually appear in the pressure and vertical velocity fields as well, but their magnitudes are considerably smaller compared to one present in the horizontal velocity. I have tried different approaches to mitigate it: - Increase and decrease nummodes. - SVV - Diferent precodnitiones for velocity and pressure - Different Reynolds number - Differnt operators for SmoothAdvection - Differnt operators for GlobalSysSoln The pattern only does not show up when using the same discretization order for all the composites. Any help would be welcome. Cheers, Guillermo

2 2

Restarting a VIV case in VCSMapping solver
by Chengjiao Ren 13 Feb '23

13 Feb '23

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Dear Nekar++ users, Hope the email finds you well. I would like to know if anyone has tried restarting a VIV problem in VCS Mapping solver? In my case (VIV of a circular cylinder), when restarting a simulation from a flow field (restart2 in the attached figure 1, "*.map" for mapping and "*.fld" for initial conditions), the force in the motion direction has large fluctuations at the initial few steps. These large fluctuations cause subsequent inconsistent evolutions for all signals compared with the original ones. This issue is found for a forced vibration case when the domain has an initial acceleration, as per figure 2 below. Noted that if we compare the Fy evolution (restart 2 in figure 1) with that of the restart 1 (only define the "u,v,w,p" in "*.fld" file for initial conditions), they are quite similar. I therefore suppose if the p field due to coordinate transform is not considered in "*.map", resulting in significant changes on forces when the domain has non-zero acceleration. Does anyone have any suggestions on restarting a VIV case? Since the initial difference doesn't decay during the simulation, it cannot be ignored currently. Thanks in advance. Chengjiao Figure 1, a free vibration case where the cylinder has 1-dof in y-direction Figure 2, a forced vibration case where the cylinder is forced to vibrate in x-direction. Dr Chengjiao Ren Research Associate School of Engineering * Perth WA 6009 Australia +61 413077330 * chengjiao.ren(a)uwa.edu.au<mailto:chengjiao.ren@uwa.edu.au>

1 0

2D geometry but rotates about the z_axis
by Ali Ahmed Abdulrasool - علي احمد عبد الرس ول 11 Feb '23

11 Feb '23

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Hello Stan, I'm grateful for your response. I'm sorry, but I'm new at using Nektar. For me, it doesn't matter which axis the channel spins on as long as it does it from the channel's face. I thought that Nektar only offers the option of rotation about the z axis. Is that correct? I.e., can I rotate the channel around any axis I want? Or can I just insert velocity values for the channel wall instead of using the "forcing" option and all of its requirements? I mean like this <REGION REF="0"> <D VAR="u" VALUE="-2.0*y"/> <D VAR="v" VALUE="2.0*(x-0.4)"/> <D VAR="T" VALUE="1600"/> <N VAR="p" USERDEFINEDTYPE="H" VALUE="0"/> </REGION> In addition, according to your suggestion, can making a homogeneous (Fourier) expansion result in a 3D mesh with a high computational cost? I want to know whether there is a difference in computation time between quasi-3D and normal 3D so that I can just create a 3D model using Gmesh and convert it to Nektar. Thank you in advance, Ali -- *University of Kerbala* website : https://uokerbala.edu.iq/en/ <https://uokerbala.edu.iq/en/> Location : https://goo.gl/maps/pNQ9LkdMbi72 <https://goo.gl/maps/pNQ9LkdMbi72> *This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. *

2 2

Special treatment of WallViscous boundary
by Chengeng Qian 08 Feb '23

08 Feb '23

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Hello everyone, I tried to run CompressibleFlowSolver to resolve some non-dimensional problems. I found that I must set parameter Twall if I wanted to use WallViscous boundary conditions. Twall is referenced in NavierStokesCFE::SpecialBndTreat <https://doc.nektar.info/doxygen/latest/class_nektar_1_1_navier_stokes_c_f_e…>, which is used to calculate diffusion flux on boundary. Are there some references which can show me more details of this special treatment of WallViscous boundary? Any comment is appreciated. Best, Chengeng

1 0

2D geometry but rotates about the z_axis
by Ali Ahmed Abdulrasool - علي احمد عبد الرس ول 07 Feb '23

07 Feb '23

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Hello Nektar community, I want to simulate a two-pass square channel with rotation effect using the Nektar++ solver. First, I tested 2D geometry, and it worked. However, when I tried to add rotation, it didn't work. As far as I'm aware, Nektar's rotation is constrained on the z axis, therefore I built my channel around that. I created "GEOMETRY DIM="2" SPACE="3">" so that a 2D mesh could be obtained in 3D space and rotated about the z-axis. Nevertheless, I keep getting the message "Fatal: Level 0 assertion violation output coord 2 is not defined." Who can advise me on what to do? I truly appreciate it. Thanks, Ali A. Abdulrasool -- *University of Kerbala* website : https://uokerbala.edu.iq/en/ <https://uokerbala.edu.iq/en/> Location : https://goo.gl/maps/pNQ9LkdMbi72 <https://goo.gl/maps/pNQ9LkdMbi72> *This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. *

2 1