[mlpack] Rocket injector demo

[Ryan Curtin]

This is super cool!  I really enjoyed reading the writeup in the
notebook also.

Is there any plan to integrate the visualization with the pareto front
given by ensmallen?  It would be cool if I could, e.g., click a solution
on the 3D pareto front and then update the visualization to use those
resulting parameters.  I am not sure how easy or realistic that is,
though. :)  Maybe something like "click to pick a new random point on
the pareto front" could be easier?

Also a couple little grammar/writing issues I found... I think in "The
Achilles Heel" section you mean "the flames would sway uncontrollably",
not "the flames would sway controllably".  Also, in the Background, at
least in my experience "D-day" typically refers to the Allied landing at
Normandy in 1944; I haven't heard it used to refer to other events.  And
just one more, I think the Rocketdyne design of the F-1 started prior to
June 1962, but the background seems to imply that they started designing
it then (but that is a pedantic point!).

Overall I really enjoyed reading through the example. :)

On Sat, Jul 24, 2021 at 12:19:46AM +0530, sai_ng wrote:
> Some updates on Visualizer. The objective function simulation is also done!
> See this Visualizer by FieryBlade (itch.io)
> <https://fieryblade.itch.io/busteringaaz>. The blog is about 90% done, so
> expect that soon :). I'm currently working on SoS kernel integration to use
> bokeh as our plotting agent.
> 
> S
> 
> On Mon, Jul 19, 2021 at 9:49 PM sai_ng <jonpsy101 at gmail.com> wrote:
> 
> > Oh, snap! I totally forgot. The particle simulation has been a success!
> > I've uploaded a WebGL based animation in itch.io
> > <https://znat.itch.io/busteringaaz>. Next is the combustion simulation.
> >
> >
> > S
> >
> > On Mon, Jul 19, 2021 at 9:45 PM sai_ng <jonpsy101 at gmail.com> wrote:
> >
> >> Hi again
> >>
> >> Great insights, I'll add some more detail. For the curious, the paper
> >> used CFD based simulation with multiple trial-and-error to get those
> >> coefficients relating objective function to design variable. Again, I
> >> encourage you to drop in a review or two in the future PR for the notebook.
> >>
> >> Reg those other articles, I found them interesting but the notebook will
> >> get too verbose. I'll do something about them in my blog though :).  Fun
> >> fact: Sirisha Bandla's hometown,Guntur, isn't very far from mine (like 8
> >> hours or so).
> >>
> >> S
> >>
> >>
> >>
> >> On Mon, Jul 19, 2021 at 8:22 PM Benson Muite <benson_muite at emailplus.org>
> >> wrote:
> >>
> >>> Hi Sai,
> >>>
> >>> It is not immediately clear where the optimization equations come from,
> >>> you give references which are great, but may want to see if some of them
> >>> explain how the resulting equations are derived from more complicated
> >>> models. Probably a few sentences on this are helpful, typically for
> >>> studying instabilities, people solve combustion equations that include
> >>> fluid mechanics and chemical reactions. A full explanation would
> >>> probably take more space than is available, but a few hints would be
> >>> helpful.
> >>>
> >>> Probably also interesting:
> >>>
> >>> https://en.wikipedia.org/wiki/Indian_Space_Research_Organisation#Launch_vehicles
> >>>
> >>> Maybe also relevant in the billionaire space race:
> >>>
> >>> https://www.gulftoday.ae/news/2021/07/19/india-born-woman-helps-design-rocket-that-takes-jeff-bezos-to-space-tomorrow
> >>>
> >>> Regards,
> >>> Benson
> >>>
> >>> On 7/19/21 5:21 PM, sai_ng wrote:
> >>> > Hello Benson
> >>> >
> >>> > About the background, I suppose you're talking of the "Framing the
> >>> > problem" section? The notebook is in alpha stage, I'll be making a
> >>> > series of PR to improve it,  you could directly review it on the pull
> >>> > request :)
> >>> >
> >>> > Thanks for the link for combustion instability, I will add that as a
> >>> > hyperlink reference in the notebook. About the Soyuz model, currently,
> >>> > we're only simulating the injector design of a typical rocket agnostic
> >>> > of F-1 engine or in this case Soyuz model. This will keep things very
> >>> > simple.
> >>> >
> >>> > Hope I answered your queries.
> >>> >
> >>> > S
> >>> >
> >>>
> >>

> _______________________________________________
> mlpack mailing list
> mlpack at lists.mlpack.org
> http://knife.lugatgt.org/cgi-bin/mailman/listinfo/mlpack


-- 
Ryan Curtin    | "Rock-and-rollers don't bathe."
ryan at ratml.org |   - Sneed