- 132
- 301 912
BYU FLOW Lab
United States
เข้าร่วมเมื่อ 18 ธ.ค. 2018
Andrew Ning is a professor in the mechanical engineering department at Brigham Young University. His research and teaching interests are in optimization, aerodynamics, aircraft design, and wind energy.
Getting Started with PyTorch
Walkthrough of using PyTorch for a basic regression problem using a vanilla neural net.
มุมมอง: 293
วีดีโอ
Getting Started with Python
มุมมอง 1935 หลายเดือนก่อน
Install Python, PyCharm, numpy, scipy, matplotlib. Demonstrate a basic workflow of running in terminal vs console (interactive REPL). Create a first plot.
DuctAPE Title Slide Animation Tutorial
มุมมอง 1325 หลายเดือนก่อน
Here's how I put together the title slide animation for my AVIATION 2024 presentation on DuctAPE. I received several questions about how I put it together, so hopefully this answers those!
DuctAPE: A steady-state, axisymmetric ducted fan analysis code for gradient-based optimization
มุมมอง 3115 หลายเดือนก่อน
Recorded version of the presentations of "DuctAPE: A steady-state, axisymmetric ducted fan analysis code designed for gradient-based optimization" a paper presented at the AIAA AVIATION 2024 Forum. The DuctAPE source code can be found at github.com/byuflowlab/DuctAPE.jl
5 Tips for Obtaining Research Funding
มุมมอง 16610 หลายเดือนก่อน
5 tips for grant writing / proposal writing / obtaining funding. Mostly focused on government grants, but also some tips related to industry (private sector) funding.
Drag buildup from airfoils (strip theory)
มุมมอง 542ปีที่แล้ว
As we move from high-level aircraft conceptual design (macro parameters like span and area) towards selecting airfoils and chord/twist distributions, we would like to improve our drag estimation. An airfoil drag buildup (strip theory) approach allows us to integrate viscous drag. This approach then reflects changes in airfoils and chord/twist (via the lift distribution).
Automating Adjoints with Algorithmic Differentiation
มุมมอง 361ปีที่แล้ว
A derivation of how to use implicit differentiation in reverse-mode AD (forward-mode is similar) to compute adjoints for a generic solver. Shows how to use vector Jacobian products for efficient calculation. Ends with a simple example of how you could implement in Julia (or in another language).
Speed Performance w/ Julia Programming Language
มุมมอง 584ปีที่แล้ว
Some examples of some common speed performance issues with Julia, but more importantly an introduction to some tools like profileview, benchmarktools, code_warntype, and others that can help you diagnose your own code.
Using and Developing Julia Packages
มุมมอง 1.4Kปีที่แล้ว
package manager, PkgTemplates, organization, unit tests, project and manifest files, registering packages, continuous integration, documentation
Reverse mode AD / backprop: explanation, Julia example, and custom rules
มุมมอง 332ปีที่แล้ว
Describe basics of how reverse-mode AD (or backprop) works. Derived a couple of examples from scalars to matrices. Then jumped into code using ReverseDiff in Julia. Showed a common type issue and how to resolve. Then showed how to create your own custom rule.
Getting Started with Julia: Workflow
มุมมอง 1.1Kปีที่แล้ว
The workflow for using Julia can feel a bit foreign, especially when coming from Python. This video attempts to provide some tips on using an integrated REPL, a brief intro to multiple dispatch, structs, when to use types, and broadcasting.
Intro to using TikZ and PGFPlots to create high-quality figures
มุมมอง 2.4Kปีที่แล้ว
TikZ and PGFPlots are TeX packages that enable high-quality figure generation using the same environment as your LaTeX document. They are typically more effort than other plotting packages, so not an approach we use every time, but when you want a very seamless look this can be the way to go (note that we generated all the figures in our optimization book with TikZ/PGFPlots, one of which is the...
Deep Neural Networks
มุมมอง 328ปีที่แล้ว
Introduction to vanilla deep neural networks, particularly for regression, activation functions, backpropagation, minibatching, initialization, and various algorithms working up to Adam w/ connection to other algorithms we've learned in this class.
Getting Started With a Text Editor (for programming)
มุมมอง 849ปีที่แล้ว
I use VSCode in this example, although the techniques should apply to any modern text editor. A sampling of the types of keyboard shortcuts available, extensions, snippets, diff, etc.
Principles for Creating High Quality Figures in Scientific Publications
มุมมอง 971ปีที่แล้ว
Some principles and examples to consider as you create figures for scientific publication in conferences and journals. The book I recommended with Tress, maps, and theorems by Jean-luc Doumont.
Blade Element Momentum (BEM) for propellers and turbines: part 3, airfoil corrections
มุมมอง 1.9K2 ปีที่แล้ว
Blade Element Momentum (BEM) for propellers and turbines: part 3, airfoil corrections
Blade Element Momentum (BEM) for propellers and turbines: part 2 blade element (plus momentum)
มุมมอง 6K2 ปีที่แล้ว
Blade Element Momentum (BEM) for propellers and turbines: part 2 blade element (plus momentum)
FLOWUnsteady: An Interactional Aerodynamics Solver for Multirotor Aircraft and Wind Energy
มุมมอง 1.7K2 ปีที่แล้ว
FLOWUnsteady: An Interactional Aerodynamics Solver for Multirotor Aircraft and Wind Energy
Blade Element Momentum (BEM) for propellers and turbines: part 1 linear and angular momentum
มุมมอง 11K2 ปีที่แล้ว
Blade Element Momentum (BEM) for propellers and turbines: part 1 linear and angular momentum
Trefftz Plane, computing induced drag in the far field, with application to the VLM
มุมมอง 2K2 ปีที่แล้ว
Trefftz Plane, computing induced drag in the far field, with application to the VLM
Lifting Line part 2: general lift distributions, winglets, bound vortices
มุมมอง 7762 ปีที่แล้ว
Lifting Line part 2: general lift distributions, winglets, bound vortices
Lifting Line Theory part 1: elliptic lift distribution, induced drag, reduced lift curve slope
มุมมอง 1.5K2 ปีที่แล้ว
Lifting Line Theory part 1: elliptic lift distribution, induced drag, reduced lift curve slope
Gradient-based wind farm layout optimization
มุมมอง 1.1K2 ปีที่แล้ว
Gradient-based wind farm layout optimization
Finite Wing and Induced Drag Fundamentals
มุมมอง 1.4K2 ปีที่แล้ว
Finite Wing and Induced Drag Fundamentals
Turbulent Simulation: DNS, RANS, Reynolds Averaging, Turbulence Models, LES
มุมมอง 1.3K2 ปีที่แล้ว
Turbulent Simulation: DNS, RANS, Reynolds Averaging, Turbulence Models, LES
Head's method: numerical solution of incompressible turbulent boundary layers
มุมมอง 6342 ปีที่แล้ว
Head's method: numerical solution of incompressible turbulent boundary layers
Demo: Using PyOptSparse, primarily with IPOPT, for Nonlinear Optimization in Python
มุมมอง 1.9K2 ปีที่แล้ว
Demo: Using PyOptSparse, primarily with IPOPT, for Nonlinear Optimization in Python
Thwaite's Method: numerical solution of an arbitrary incompressible laminar boundary layer
มุมมอง 7942 ปีที่แล้ว
Thwaite's Method: numerical solution of an arbitrary incompressible laminar boundary layer
Hello Professor Andrew Ning, thank you so much for sharing this great lecture series. I have a Telecom/EE/CS background in terms of my profession, but somehow developed a passion for aeronautics/fluid mechanics at a late stage of life. This lecture series is truly being helpful. Quick question. It seems that some of the lectures (e.g., those covering compressible aerodynamics, and probably some more on CFD) may be missing at this site? If so, is there any place where I can find them? Thank you again. 🙏
Excellent. Concise, complete and brief. One could construct a functional Hess/Smith code from this video alone (except the matrix solver, of course). Thank you.
Hello! I am trying to self-study aerodynamics, and i found your excellent videos! I really learn a lot from you, so thank you so much! 🙂 In your videos, you mention something like "in the texts"... i was wondering which book(s) you are referring to? Which books would you recommend? Thank you in advance!!
Hello! I don't quite understand. Is the torque Q here the torque applied by the propeller to the air or the torque applied by the air to the propeller? And is Ω the angular velocity of the free air or the angular velocity of the propeller's rotation? And why is it Ωr-v here, while the axial part is V∞+u? One is a plus sign and the other is a minus sign?
You might find the text version more helpful. flowlab.groups.et.byu.net/me515/aero.pdf Generally when we write Q and T we care about the forces fluid acts on propeller, but when using the fluid mechanics equations we need the opposite (force of propeller on fluid). Omega - depends on your perspective. In the frame of reference of still air it is rotation of propeller. frame of reference in the propeller it is movement of air.
Very nice.Thank you.
Great derivation of the stress-strain rate tensor with full details. Thanks a lot!!!
Is there a way to install pyoptsparse using pip instead of conda?
yes you can install with pip: mdolab-pyoptsparse.readthedocs-hosted.com/en/latest/install.html
Very good illustration and easy to understand. thanks
How can i know the twist angle?
That's part of the geometric description of the blade.
How do we get from T=2a(1+a) ho\V_inf^2A_d to T ' = 4a(1+a) ho\V_inf^2 pi r
The area of the disk segment is 2 pi r dr and (T/dr is the thrust per unit length)
Thank you so much for this video!
Comment to boost on the algorithm plus THANKS!
Hi, in your equation shown at 18:30, what do the variables r_1x and r_2x mean?
It is the x-component of the r1 vector and r2 vector respectively.
very helpful. thanks so much for this. 'GradObj','on', and 'GradConstr','on' should be use to replace "SpecifyObjectiveGradient", true, and 'SpecifyConstraintGradient', true. in addtion, MaxFunctionEvaluations', le5 is not recognized. I simply remove it.
Much better explanation and presentation than my professor.
typo: "deviatoric"
Crisp explanations with great insights. Loved watching this!
Delbert Harbors
Susie Trafficway
Correction alert :- Cn = Cdsin(phi) - Clcos(phi) & Ct = -Clsin(phi) - Cdcos(phi)
Sandrine Point
I've been stuck for almost two hours until I found your video Thank you so much!
Great video! What is the title of the text book?
Engineering Design Optimization. PDF is freely available. mdobook.github.io
Thanks for this great tutorial.
Thank you so much this was very clear and helpful
Great, well done, many thanks!
Very Interesting, can I get the code for traveling salesman problem.
Thanks for sharing this video, I have problem the during the importing the geometry, maybe there is a problem. Can you check the link is still working or not?
Awesome video, thank you so much. Sometimes I see derivations where there is some pressure component, at this time I had to go over the derivation a bit quickly, but if anyone can shed some light on the -P*delta_ij that I have seen in some forms of the stress tensor please let me know.
Take a look at my writeup here: flowlab.groups.et.byu.net/me515/aero.pdf, specifically equation 1.113. The pressure term always exists, it's just a matter of preference whether you define that pressure term as part of the stress tensor or not.
I'm interesting about this. May i have ur email. Ee can discuss about it . Can yiu help me about my research. Thanks before
How might this algorithm change when the evaluation of f(x) is noisy? In that you can estimate the value at any given step but not know it exactly. More practically, if this situation presents itself is there a clear way to solve this sleekly in python with something simple like scipy? Thanks for the informative video!
Great video! Dr. Ning has always been a master teacher. Are there any rules of thumb for initial edge length on the simplex?
usually just 1, assuming you've normalized all your inputs
Hey, you keep mentioning a book. What book is it exactly ? I can’t see any links in the description.
You can get the PDF here: flowlab.groups.et.byu.net/me515/aero.pdf
@@BYUFLOWLab Thank you very much for the link ! Btw, I used BEMT for a shrouded Ducted Fan optimization code. Works great. Perhaps you might have any additional ressources on a model that also accounts for Lip Design (Inlet Design essentially being reduced to standard Turbomachinery Theory) of an EDF and not only a shrouded EDF ?
Good Job
You have the lift and drag axes swapped in your discussions around 5:30 on
Great explanation, I finally understand the backtracking line search method.
Hi, I'd like to ask, at 15:46, why is my auto mesh (2D) parts option empty inside
I am confused because the Rot region has 2 million more cells than the video, despite having the same parameters.
Great video and great help!!! I just did as same as your video, but the thrust and torque is different with yours. If possible, can i get your sim file...? and there isn't information fluid conditions like Temperature, Pressure, viscousity.. etc. I need help please..
Hello, : import Pkg Pkg.add("NLPModels") pkg.add("NLPModelsIpopt") using NLPModels using NLPModelsIpopt myobj(x) = (1 - x[1]^2) + 100 * (x[2]-x[1]^2)^2 x0 =[1.3, 0.5] Model = ADNLPModel(myobj, x0) .... did not work, and i think there changes by git hut, it is not look similiar like you use. but tthanks, it was very good to hear :)
amazing video
what do u think the other more better methods would be ?
Thanks for making this.
'Promo SM'
Thank you! Waiting to another awsome video :)
Excellent tutorial. The progression through the various topics is carefully graduated, the examples are helpful, and the explanations are clear. Thank you for sharing.
what if i want to plot the contour in a range (-5 to 5) aligning with the colorbar. will that be okay to put level -5 to 5?
Very interesting video. It is like the Chain Rule, attempting to find the derivative in a composition of functions. Thank you so much !!!!!!
great explanation
How does this math relate to the physical design of the aircraft? Can you make a video about how to design an aircraft for dynamic stability? Thanks
It's hard to generalize (e.g., adding a winglet may improve dutch roll for one aircraft while making it worse for another). For conventional aircraft you can come up with some rules of thumbs by analyzing the contributions in the stability derivatives. Those are fun (though limited), and maybe I should put together a video on that sometime. But the general approach is numerical experimentation (or optimization). You make changes to the design and observe the eignevalues in the complex plane. There are various tools out there that can facilitate this kind of study.