Help requested about new app to design motor

Discussion in 'Rocketry Electronics and Software' started by jbgust, Apr 8, 2019.

Help Support The Rocketry Forum by donating:

  1. Apr 8, 2019 #1

    jbgust

    jbgust

    jbgust

    Active Member

    Joined:
    Apr 8, 2019
    Messages:
    34
    Likes Received:
    13
    Gender:
    Male
    Hi, everyone,

    With the help of a friend I developed a free web application to design rocket engine. This application uses the same calculation model as Richard Nakka's SRM Excel file. Would any of you be motivated to test it and give me feedback? What is missing, improvements, ...

    meteor.png

    The application is almost finished (preview below). We have planned to publish it at the end of April.

    If you're interested, send me a private message and I will send you the link to the application (not install required).

    Thank you for your help.
     
    Last edited: Apr 8, 2019
    Zertyme likes this.
  2. Apr 10, 2019 #2

    jbgust

    jbgust

    jbgust

    Active Member

    Joined:
    Apr 8, 2019
    Messages:
    34
    Likes Received:
    13
    Gender:
    Male
    Anyone interested? Is it because of the lack of usefulness of this kind of application?
    The application is easy to use, no need to test it for hours. Any help is welcome.

    The app is now fully responsive.

    Just Like the post and I will send you the link :)
     
  3. Apr 10, 2019 #3

    heada

    heada

    heada

    Well-Known Member

    Joined:
    Jan 18, 2009
    Messages:
    2,709
    Likes Received:
    222
    Gender:
    Male
    Location:
    Indianapolis, Indiana
    Have you looked at the product called "burnsim"? It is what most have used to do the same as your application.
     
  4. Apr 10, 2019 #4

    BDB

    BDB

    BDB

    Absent Minded Professor

    Joined:
    Aug 22, 2015
    Messages:
    1,975
    Likes Received:
    281
    Gender:
    Male
    Most people use Burnsim for APCP motors. This is specifically for Sugar motors.

    I think this is cool, but so is Nakka’s spreadsheet. I’d test it for you but I don’t have much time to make and test motors lately.
     
  5. Apr 10, 2019 #5

    jbgust

    jbgust

    jbgust

    Active Member

    Joined:
    Apr 8, 2019
    Messages:
    34
    Likes Received:
    13
    Gender:
    Male
    Yes I know it. Of course my application work solely for cylindrical grain configurations, but :
    • You can use 6 different propellants
    • I can provide the same graph result (KN, Thrust, ...) for the moment on Thrust is display but JSRM can display about 30 results
    • It's free and opensource, so anyone can help to implement new propellants or grain configurations
    • No installation needed
    • Works on phone
    • Import and save your work
    • ... and other planned feature like motor comparison
    I you use burnsim, I will be happy If you want to compare it
     
    Last edited: Apr 10, 2019
    timbucktoo and Zertyme like this.
  6. Apr 10, 2019 #6

    jbgust

    jbgust

    jbgust

    Active Member

    Joined:
    Apr 8, 2019
    Messages:
    34
    Likes Received:
    13
    Gender:
    Male
    No APCP for the moment but I can easly add it if you know where I can find this information:

    • IdealMassDensity
    • Ratio of specific heats, 2-ph.
    • Ratio of specific heats, mixture
    • Effective molecular wt. [kg/kmol]
    • Combustion temperature [°K]
    • Burn rate coefficient by chamber pressure [mm/s]
      Pressure exponent valid at Po
     
  7. Apr 10, 2019 #7

    BDB

    BDB

    BDB

    Absent Minded Professor

    Joined:
    Aug 22, 2015
    Messages:
    1,975
    Likes Received:
    281
    Gender:
    Male
    If this could easily simulate both sugar and APCP, and could be freely available (as opposed to Burnsim) I’d definitely use it. Alas, but I don’t have the info that you need for APCP.

    You may want to also post this on some of the FB and Reddit pages that are dedicated to EX and sugar motors.
     
  8. Apr 10, 2019 #8

    jbgust

    jbgust

    jbgust

    Active Member

    Joined:
    Apr 8, 2019
    Messages:
    34
    Likes Received:
    13
    Gender:
    Male
    Good idea, but first I want to make sure there are no problems of ergonomics or misunderstanding how it works. That's why I'm looking for people who want to try it, to improve the help and ease of use.
     
  9. Oct 31, 2019 #9

    serk

    serk

    serk

    New Member

    Joined:
    Oct 27, 2019
    Messages:
    3
    Likes Received:
    0
    Gender:
    Male
    I strongly recommend to avoid BurnSim and OpenMotor - they are at some point completely innaccurate. I havent investigated the source code but my real life tests and data from these simulations does not match. You can use them to get better understanding on parameters that affect burn speed and pressure in the chamber but thats mostly all.
    If you want to get them a bit more accurate You have to start from scratch - create your propellant and actually measure the real burn speed and configure the software. Additionally you have to measure real pressure in the chamber and again tweak the software. Default setup is highly innaccurate.

    Nakkas spreadsheet - again, this works for Nakka. Hes way of making propellant is not actually the best one and when following hes guides its pretty hard to get the result hes presenting. I havent been able to match hes results. Chemicals he is using are probably not so clean plus differrnces in melting and cooling process. My burn times are much more aggressive. So it clearly shows that You can use hes formulas that are excellent but you have to start from scratch to make them work - every propellant is different. For example theres huge difference in propellant if using destilled water vs tap water for dissolve process.
     
    Last edited: Oct 31, 2019
  10. Oct 31, 2019 #10

    Rocket501

    Rocket501

    Rocket501

    Well-Known Member

    Joined:
    Feb 9, 2018
    Messages:
    80
    Likes Received:
    25
    I disagree. Yes, you will have to experimentally characterize the propellent, but by starting with propep or some similar program and other people’s recipes you can get a decent initial guess. However, OpenMotor does not and is not meant to do this.

    Once you’ve characterized the propellent (I prefer using 54mm motors and six sizes of nozzles), you plug the data into OpenMotor and as long as you stay within the constraints, future motor designs tends to be within 10% of reality (if you are designing between 29mm to 98mm without too much metal particles).

    As such, I really don’t get what you are saying. While you have some points I agree with, I don’t see how that turns into advice to stay away from OpenMotor. Instead I would say take propep results with an extreme grain of salt and understand that you will have to characterize propellent yourself before using OpenMotor. Once you have good numbers, you will save so much time and money using computer models for the initial planning compared to only using iterative testing.
     
    Last edited: Nov 1, 2019
    dhbarr, Steve Shannon and eggplant like this.
  11. Nov 1, 2019 #11

    dhbarr

    dhbarr

    dhbarr

    Amateur Professional

    Joined:
    Jan 30, 2016
    Messages:
    5,990
    Likes Received:
    741
    Fundamentally, you are not correct.

    Nakka, Wickman, Yawn, Sutton, Chiaverini, Pollino, Kuo, Huzel, Leslie, Huang, Biblarz, McCreary are all fairly well in agreement.

    If your results differ from the accepted results of the last fifty years or so, gently I would suggest error in preparation, measurement, or both.
     
  12. Nov 1, 2019 #12

    serk

    serk

    serk

    New Member

    Joined:
    Oct 27, 2019
    Messages:
    3
    Likes Received:
    0
    Gender:
    Male
    @Rocket501
    " you can get a decent initial guess" -> I do agree 100%, you will get some, some in bold, initial guess.

    Be sure not to blow yourself up when OpenMotors showing you 500psi but in the reality PSI can be times higher (this is the reason why I suggested to stay away from such softwares). Iterative testing, precision pressure measurer and proper high speed data logger -> without these you are blind and you have no idea whats actually happening in your motor.

    My propellant is 490cm3 (25cm width and 5cm diameter)

    Unfortunately admin of this forum deleted my topic where I wanted to shed more light on this. Will start my thread somewhere else and if anyone interested you can stay tuned.
     
    Last edited: Nov 1, 2019
  13. Nov 1, 2019 #13

    serk

    serk

    serk

    New Member

    Joined:
    Oct 27, 2019
    Messages:
    3
    Likes Received:
    0
    Gender:
    Male

    I'll give you simple example.
    I have custom made propellant, volume is approx 490cm3, so it's a pretty big. OpenMotors shows burn time 2.3seconds which is completely wrong. I'm still trying to tweak parameters in OpenMotor to get the formulas working at all in that software. When increasing volume 2x, OpenMotors says that the burn time stays basically same, 2.47sek (LOL, which is completely wrong). When changing configuration so that I am using multiple separate propellants, grains, the same volume, story changes, burn time decreases, which is again wrong, it should be vice versa and also tests prove that :D LOL - you get the idea, hopefully...

    I'm not saying Nakkas and other fellows you mentioned have done wrong physics, not at all, just the implementation in these softwares...

    But as my threads were blocked for some reason, I'm off the forum and going to find some other one thats not so restricted to US. Good luck and all the best.
     
  14. Nov 1, 2019 #14

    jbgust

    jbgust

    jbgust

    Active Member

    Joined:
    Apr 8, 2019
    Messages:
    34
    Likes Received:
    13
    Gender:
    Male
    Have you try Meteor? https://meteor.open-sky.fr/#/home
     
  15. Nov 1, 2019 #15

    eggplant

    eggplant

    eggplant

    L3 | NAR 93664, TRA 17791

    Joined:
    Jan 5, 2011
    Messages:
    1,653
    Likes Received:
    109
    If you are still tweaking your propellant parameters, maybe hold off on saying the software is at fault? Volume and burn time are only loosely related through grain geometry, so I don't know why you are conflating them. If you think there's a bug in the software, please make an issue on github with the relevant motor files so it can be investigated. However, what you are describing sounds like uncharacterized propellant (What pressures/Kn were your test burns at? How many did you do? What did you use to determine the pressure?), incorrectly entered data (I'll admit the software could use a manual, but I think textbooks are a better place to learn what Kn or ballistic exponent are), or a misunderstanding of the model that all amateur motor simulators and textbooks I'm aware of use. I think you'll find the same results in Meteor as the models used are very similar.

    For something actually on topic for this thread, great work on meteor! I like the direction of a web app and think it could supersede Nakka's spreadsheets as an easier to use tool.
     
    Last edited: Nov 1, 2019
  16. Nov 1, 2019 #16

    prfesser

    prfesser

    prfesser

    Lifetime Supporter TRF Lifetime Supporter

    Joined:
    May 7, 2017
    Messages:
    870
    Likes Received:
    607
    Location:
    Murray, KY
    I haven't used OpenMotors...but I have some questions.
    1. When you say that the predictions of 2.3 s and 2.47 s are completely wrong, on what are you basing that statement? Did you build the two motors and burn them? If so, what were the actual burn times? If not, what evidence do you have that they're wrong?
    2. Did you determine the burn rate coefficient and exponent of your propellant? If so, how? I have made a fair number of APCP motors over the last 20+ years, and I taught a couple of practical application classes on the process. There were eight people in one class, preparing identical four-grain 38mm motors from a single large batch of propellant I made. When tested on a thrust stand, burn times varied, often a few tenths of a second. Shapes of thrust curves also varied---some rather dramatically.
    3. When you say the "volume is approx 490cm3", is that the total volume inside the empty motor, the volume of the propellant, or the void ("empty") volume in the motor?
    4. If 490cm3 is the propellant volume and you doubled it to 980cm3, did the nozzle throat stay the same size, or did you increase the throat size to get the same area ratio (Kn)?
    5. You say in another post "I havent been able to match hes results. Chemicals he is using are probably not so clean plus differrnces in melting and cooling process" [sic]. In the sciences that is called "speculation." In the professional scientific literature, speculation is labeled as such, because it is neither data nor results. How do you know that your results are "correct" and his are incorrect; is it possible that his are correct and yours are wrong? Or maybe it is possible that different people making the same propellant from (theoretically) identical ingredients will get different results?
    6. In your original post showing the output of your app, it says that you're using an "optimally designed nozzle" with an expansion ratio of 9.65. At the chamber pressure indicated in the app (49 bar is slightly over 700 psi), 9.65 is hugely over-expanded. How is this "optimal?"
    This isn't a question, but to elaborate on #5 above, early in my work I attempted to make some of the same propellants as my mentor did. Same formula, same ingredients from the same source, same kind of mixer, etc. Both the physical properties and the ballistic properties of his propellant and mine were significantly different.

    Finally, 490cm3 of KNDX propellant is approximately 980 grams, assuming a density of about 2 g/cm3. At a specific impulse of 130 seconds, that is (very roughly) 1500 N-s total impulse---a 'baby' K motor, not an L1672 as shown in the app output.

    Best regards -- Terry
    (Author of a book that some rocketeers have found useful...)
     
    dhbarr, Steve Shannon, BDB and 2 others like this.
  17. Nov 2, 2019 #17

    beeblebrox

    beeblebrox

    beeblebrox

    8 C6-0, 12 D11-9, 20 D20-0, 20 E5-0, 3 Cinerocs

    Joined:
    Mar 19, 2016
    Messages:
    387
    Likes Received:
    94
    Gender:
    Male
    Location:
    West Chester, PA
    Interesting read, I would like to talk to the people that created Burnsim... I have reliable but pain in the butt to use software that when compared to the output of Burnsim was completely wrong. My actual static test results compare to the software I use and is accurate. Something is not right with Burnsim. I have not tried openmotors, will need to look into it for curiosity. One example, I ran a test motor using AeroTech blue thunder propellant which I am very familiar with. The theoretical with my software was an E40. 500psi max. 39.8ns Actual was an F39 510psi 40.1ns. close enough for gov't work. Burnsim said 1350psi max about 49ns F60 totally wrong, some of the data variables in Burnsim were calculated and could not be edited and did not jive with RCS/Aerotech datasheets.
     
  18. Nov 2, 2019 #18

    jbgust

    jbgust

    jbgust

    Active Member

    Joined:
    Apr 8, 2019
    Messages:
    34
    Likes Received:
    13
    Gender:
    Male
    In my screenshot, I use the same example data as Nakka SRM. There are 2,812kg of propellant not 0,980kg. And It was a L1672 motor made by Nakka. When I post this screenshot you can't tweak the advance config. Here is the missing config :

    upload_2019-11-2_20-48-41.png

    About the optimal nozzle expansion of 9.65, it the same as SRM (https://www.nakka-rocketry.net/softw.html) see C13 cell:

    upload_2019-11-2_20-50-31.png
     
  19. Nov 3, 2019 #19

    prfesser

    prfesser

    prfesser

    Lifetime Supporter TRF Lifetime Supporter

    Joined:
    May 7, 2017
    Messages:
    870
    Likes Received:
    607
    Location:
    Murray, KY
    Ah, I see. I thought you were talking about just two motors in your discussions above: one with 490 cm3 of propellant and one with twice that volume. Now it makes sense, thanks.

    Hmmm... two questions and a comment:
    1. Is the chamber pressure actually 49 bar in this example? Neither maximum nor average chamber pressure is shown in the spreadsheet.
    2. Makers of high-power hobby rocket motors---Aerotech, Kosdon, AMW, Gorilla, Loki---use nozzles with expansion ratios typically around 4-6 at chamber pressures around 1000 psi (69 bar). I think it unlikely that they would ALL use expansion ratios that are much lower than the optimum. After all, their motor sales depend (to some degree) on maximizing performance.
    3. Although ProPEP certainly has its limitations, it was written, not by amateurs, but by scientists at Martin-Marietta. It was a tremendous advance at the time, and has been used successfully by professionals (Thiokol and others) since the 1980s or earlier. It can calculate Isp at various expansion ratios. At 1000 psi chamber pressure, the expansion ratio A(e) / A(t) to obtain maximum Isp is typically around 4-6.
    Perhaps there is something else in Richard Nakka's calculations---something I don't know about and therefore have not considered---that makes 9.65 an optimum ratio for that particular motor (for example, maybe the chamber pressure is much higher than 49 bar). I'd be interested in seeing an explanation.
     
    dhbarr likes this.
  20. Nov 3, 2019 #20

    jbgust

    jbgust

    jbgust

    Active Member

    Joined:
    Apr 8, 2019
    Messages:
    34
    Likes Received:
    13
    Gender:
    Male
    Yes it's the average pressure in this example. This information is not present in Nakka file (I add it to help me on my motor design) but you can find the max chamber pressure see pressure sheet K18 (in MPa) or P18 (in PSI).

    For each iteration (around 900) of the computation, the optimum expansion ratio is given by this formula : https://github.com/jbgust/jsrm/blob...nfra/performance/PerformanceFormulas.java#L34 or https://buildmedia.readthedocs.org/media/pdf/proptools/latest/proptools.pdf page 14.
    So what is called "optimum expansion ratio" is in fact the average optimum expansion ratio of the 900 iterations.

    hope it helps
     
  21. Jan 22, 2020 #21

    jbgust

    jbgust

    jbgust

    Active Member

    Joined:
    Apr 8, 2019
    Messages:
    34
    Likes Received:
    13
    Gender:
    Male
    Hi,

    Now you can use METEOR with APCP or any other solid propellant. Since this weak you can also use 7 grains configurations:
    • end burner
    • moon burner
    • star
    • finocyl
    • c slot
    • hollow cylinder
    • rod and tube
    Enjoy
     
    dhbarr likes this.
  22. Jan 22, 2020 #22

    Chad

    Chad

    Chad

    Well-Known Member

    Joined:
    Jul 23, 2018
    Messages:
    215
    Likes Received:
    72
    Gender:
    Male
    Location:
    Dallas
    I really like it, i don't make my own motors (yet) but the application is nicely done. Can you add some details on your development stack?

    So, JSRM is a Java lib so that implies a least some of your backend is Java based. You know what would be really cool? A modern, web based, front end to OpenRocket. OpenRocket is a Java app so, in theory, you could put a REST API between the UI and the underlying application. You could have motor design and airframe design in the same app!
     
  23. Feb 19, 2020 #23

    jbgust

    jbgust

    jbgust

    Active Member

    Joined:
    Apr 8, 2019
    Messages:
    34
    Likes Received:
    13
    Gender:
    Male
    Hi Chad,
    About the stack, the frontend use Vue.js. The back is in Java (pretty simple, no cloud, docker or anything else). The back is only use to convert the frontend request to the JSRM format.
    It's a good idea to make a web version of OpenRocket but it's a lot of work.
     

Share This Page

Group Builder