| User | Post |
|
4:58 am December 27, 2009
| rpulkrabek
| | |
| Member | posts 348 | |
|
|
I have been investigating as to how the rocket will be designed and one place I have been trying to find out more information is with the nozzle shape. There are a lot of variables that affect this. For a good start, visit here:
http://exploration.grc.nasa.go…..ozzle.html
and here:
http://www.pwrengineering.com/…..design.htm
I also think we can use the advancements of FEA/CFD to guide us. If we input the right information, it can basically tell us what shape is optimum. The problem is that we need the correct information in the first place. Otherwise it's just garbage in, garbage out. For examples of garbage out, take a look at these pictures I took from a very quick demo I have done with Ansys. The reason I did this was to show a bit of what's possible and for me to learn what things are needed to perform this type of analysis.
 
 
 
 
 
There are quite many things wrong with this. The most obvious is that I have no idea why the fluid is flowing only towards one side and not through the center. Perhaps it's a geometry issue. Another thing is that I don't have any of the right inputs. I also didn't do anything with materials or gases involved. Also, the purpose of the nozzle is at the throat, the fluid velocity is at Mach 1 and then at the exit, it becomes supersonic. I haven't done anything with supersonic flows before. I need to study more and find out how to tell Ansys that supersonic flows are involved. To sum up, I was only playing around with Ansys :)
So what does everyone think? Can we start designing our OHKLA rocket?
|
|
|
8:51 am December 27, 2009
| Rocket-To-The-Moon
| | Altus, Oklahoma, USA | |
| Member | posts 685 | |
|
|
So what does everyone think?
I think that this is nice work. There are obviously some problems, but just showing this tells me that you have the software and ability to do some real design work. I'm excited to see how this develops.
Can we start designing our OHKLA rocket?
Absolutely! There isn't anyone who is going to stop us from doing this. Have you considered forming a local CSTART chapter? Having a physical group of people who can meet will really help you push this forward.
|
Main Workgroups: Propulsion & Spacecraft Engineering
|
|
|
9:59 am December 27, 2009
| Luke Maurits
| | Adelaide, Australia | |
| Admin
| posts 1483 | |
|
|
My response is basically the same as Rockets. There are some very obvious problems with this first play around session (according to the diagram at the Wikipedia article on de Laval nozzles the velocity should be increasing right until the gas exits the nozzle, which doesn't seem to be the case in your simulation) but the simple fact that you've been able to produce these shows that you have access to some really useful software that will be a huge asset to us once you find your feet with it and can handle supersonic flow, etc.
I'm going to pick up a copy of this book from my University library in the next few days, hopefully I will be able to help provide some information to support your simulations and we can start getting some good design work done on OHKLA.
|
Main CLLARE workgroups: Mission Planning, Navigation and Guidance. I do maths, physics, C, Python and Java.
|
|
|
2:15 am January 2, 2010
| rpulkrabek
| | |
| Member | posts 348 | |
|
|
I know there has been talks that it would be easier to hire other rocket boosters, but I would still like to continue to develop our own. I have been researching more and more about rocket nozzles and such and as a result I have been learning more and more. One of the biggest mistakes I made in my prior cfd attempt was the velocity at the inlet, it was extremely too high. I was, of course, just playing around then, but I believe that to be the reason for the result I was given. I have since modified the model and the setup. My results are much more promising, but still need more work. All I am doing is providing a nozzle shape, inlet fluid velocity (60 m/s) and the fluid (N2O, I know this isn't what we are using, I just needed something). To truly understand what will happen in real life, I will need real life inputs, which I hope we can find in the future.
Here are some resulting pictures:
 
 
 
 
 
 
 
 
 
There are still quite many improvements, but one thing that I like is that if you look, the fluid velocity at the throat of the nozzle is near the speed of sound, which is what it should be. Also, the velocity appears to increase as it exits. I'll continue working and keeping everyone up to date. Instead of an inlet velocity, I will try more for inlet pressure and look for a more appropriate fluid. What are chemical compositions we would be using? Any input would be greatly appreciated.
|
|
|
2:57 am January 2, 2010
| rpulkrabek
| | |
| Member | posts 348 | |
|
|
Post edited 11:53 am – January 2, 2010 by rpulkrabek
I updated it again just now to an inlet pressure of 2MPa as opposed to an inlet velocity, which is more realistic. Here are the results. More fine tuning is needed, such as boundary issues. I'll continue from here another time.
 
 
Edit: 2mPa -> 2MPa. That could have been catastrophic.
|
|
|
5:39 am January 2, 2010
| Luke Maurits
| | Adelaide, Australia | |
| Admin
| posts 1483 | |
|
|
It's good to see this work progressing. I agree that we should continue to pursue the option of developing our own boosters – even if we get the rest of CLLARE ready first and end up using a commercial launch vehicle for our first few missions, it will be cheaper in the long run if we eventually migrate to our own low cost, resuable alternative.
rpulkrabek said:
What are chemical compositions we would be using? Any input would be greatly appreciated.
While nothing is certain yet (hopefully Copenhagen Suborbitals can help with this a lot), I think the most likely candidates are nitrous oxide (N2O) or liquid oxygen (LOX) for oxidisers and paraffin (C(n)H(2n+2) with 20 <= n <= 40) or some kind of plastic/rubber for fuel. Copenhagen Suborbitals are using an "epoxy" fuel grain – one thing I'd like to learn from them is exactly what it is made out of.
Keep up the good work!
|
Main CLLARE workgroups: Mission Planning, Navigation and Guidance. I do maths, physics, C, Python and Java.
|
|
|
6:34 am January 2, 2010
| Rocket-To-The-Moon
| | Altus, Oklahoma, USA | |
| Member | posts 685 | |
|
|
Very cool, I bet you are having a great time learning about this.
What is the atmospheric pressue in these tests? Is it vaccum?
|
Main Workgroups: Propulsion & Spacecraft Engineering
|
|
|
12:32 pm January 2, 2010
| rpulkrabek
| | |
| Member | posts 348 | |
|
|
Rocket-To-The-Moon said:
Very cool, I bet you are having a great time learning about this.
What is the atmospheric pressue in these tests? Is it vaccum?
Yes, it is really nice to learn about what is going wrong and improving on it. There is something great about not having to spend money on materials and manufacturing costs and creating digital prototypes to simulate how it will work.
I was running through the simulations with 1atm pressure. A decision needs to be made about what pressure should be used to do these simulations. Is it at the halfway point in altitude that the combustion will be happening? What pressure is that? I guess we won't know until we understand more about the fuel used.
|
|
|
10:36 pm January 2, 2010
| Luke Maurits
| | Adelaide, Australia | |
| Admin
| posts 1483 | |
|
|
rpulkrabek said:
I was running through the simulations with 1atm pressure. A decision needs to be made about what pressure should be used to do these simulations. Is it at the halfway point in altitude that the combustion will be happening? What pressure is that? I guess we won't know until we understand more about the fuel used.
Everything I have read on the subject suggests that for unmanned suborbital flights the best approach is to do a very high energy very short life burn – combustion should last for 5-10 seconds. During this early stage of the flight the altitude will be quite low (a few kilometers at most), so 1 atm is probably the most suitable value to use for ambient pressure.
|
Main CLLARE workgroups: Mission Planning, Navigation and Guidance. I do maths, physics, C, Python and Java.
|
|
|
2:12 am January 18, 2010
| rpulkrabek
| | |
| Member | posts 348 | |
|
|
Post edited 9:06 am – February 18, 2010 by rpulkrabek
I have done some more work towards OHKLA. My knowledge is limited towards hybrid rocket design, and would like some guidance. I have uploaded pictures from the Pro/E model I am working with to the wiki as a host.

http://cstart.org/wiki/images/…..cept_1.jpg

http://cstart.org/wiki/images/…..cept_2.jpg

http://cstart.org/wiki/images/…..cept_3.jpg

http://cstart.org/wiki/images/…..cept_4.jpg

http://cstart.org/wiki/images/…..cept_5.jpg

http://cstart.org/wiki/images/…..cept_6.jpg

http://cstart.org/wiki/images/…..cept_7.jpg

http://cstart.org/wiki/images/…..cept_8.jpg

http://cstart.org/wiki/images/…..cept_9.png
I will edit one of these to include arrows pointing to what each component is. Basically, I used the same/similar geometry as I had with the CFD for the nozzle, and created all new components to form an assembly. There is a lot I want to investigate and optimize. For one, I am a bit unsure of the nozzle shape. Sure it looks good in the cfd, but we may lose some of the performance as a result of the drag. I will investigate further and see if Ansys can perform an optimized shape through design exploration.
The fins I am completely unsure of what is best. Is their only purpose to prevent pitch/yaw? Is there a way to optimize to provide maximum thrust or even lift? I modeled these with a draft hinge that is parallel to the leading edge. I noticed Copenhagen suborbitals had a draft hinge as well. What is its purpose? How many should we use? I remember seeing the Ares I have 12 fins: http://www.nasa.gov/mission_pa…..old_1.html
I am still unsure about what geometry the fuel grain should have. Obviously we may need to do some tests to determine what suites us the most. It may be the geometry that provides the most thrust at the beginning of its flight. Also, the material needs to be determined. The dimensions should be investigated also, such as its height and diameter. This goes for the oxidizer too.
The nose will contain the avionics and other instruments. The dimensions of this will be changed to accommodate everything.
Feel free to poke holes in this design and I will do my best to modify it.
|
|
|
2:29 am January 18, 2010
| rpulkrabek
| | |
| Member | posts 348 | |
|
|
I would also like to upload the files to the repository. Luke, is it best for me to get access and for me to figure out how to upload, or for me to send you the files and have you do it?
Once the files are there, I would like to recommend that anyone reviewing the CAD files download a free viewer. My suggestion is to get the one from SolidWorks called eDrawings: http://www.solidworks.com/sw/s…..dcheck.htm. Make sure you check the box for Pro/E before you download. This will allow you to manipulate the model in 3D as well as hide and make components transparent. It is also possible to create a cross section and I think even measure.
The other viewer is from the same company as Pro/ENGINEER, PTC. It's called ProductView. Get it from here, after filling out a form: http://www.ptc.com/appserver/w….._dbkey=482. You can still view 3D, however, you won't be able to measure or create cross sections or transparencies.
FYI: All files with the extension ".prt" are individual parts or components. The extension with ".asm" are an assembly. The full assembly is currently called "ohkla.asm.1"
|
|
|
3:04 am January 18, 2010
| rpulkrabek
| | |
| Member | posts 348 | |
|
|
As I promised earlier, here is a picture with arrows and short descriptions.¨
 
http://cstart.org/wiki/images/…..ptions.png
|
|
|
3:46 am January 18, 2010
| Luke Maurits
| | Adelaide, Australia | |
| Admin
| posts 1483 | |
|
|
Great to see work on OHKLA going ahead! A few quick comments:
- It might be a good idea to upload smaller images to the Wiki when the originals are this size. I sometimes just monitor the forum RSS feed from inside of Firefox, where all the images are displayed full size – this post gave my computer a heart attack!
- With regards to getting code into the repository – I am happy to do this if you don't have the time or inclination to figure out how to use Mercurial for yourself, but if you do want to learn the best thing for you to do is to create a clone of the repository. There are instructions on how to do this at the Wiki here. You will need a Google account to do so. You'll need a Hg client on your machine to be able to push your changes back to the server. I can easily merge the stuff in your clone into the main repository once you've done that. If you need more specific help with anything feel free to send me a PM or an email.
- I realise I haven't yet posted anything useful from my hybrid rocket textbook for you to work on, I'll try to pick up the pace on this soon. Just quickly, regarding fuel grain geometry: the more ports are in the grain, the higher the thrust generated, so in order to get a short burst of high thrust, we'll need several ports. Unfortunately there are disadvantages to this, I'll try to post more details later.
- I am not sure the fuel grain plate should have holes in it to match the ports. As the grain burns, the ports will widen and combustion will be taking place directly over the plate. I think perhaps it should just have a single large circular hole in the centre.
- I am not sure that the fins (which, yes, as far as I know are just to prevent pitch and yaw) should be attached directly to the nozzle, they may end up getting too hot?
- I think we should start actively advertising OHKLA to advanced amateur rocketry clubs around the place. They will be able to make much more useful input much quicker than any of us will be if we have to keep learning as we go along. We should try to find clubs which have already made fairly large hybrid rockets since they will be of the most use.
I will try to make some more substantial comments later. Excellent work, keep it up!
|
Main CLLARE workgroups: Mission Planning, Navigation and Guidance. I do maths, physics, C, Python and Java.
|
|
|
6:00 am January 18, 2010
| rpulkrabek
| | |
| Member | posts 348 | |
|
|
- Sorry about the images. I realized the effect after I saved the post. Next time I will make them smaller.
- I have created a clone and I will play around to see if I can upload my files there. I think it's good for me that I learn this. I'll inform you when I am ready to merge.
- Looking at this image, do you think it is a good idea to look more into the multi-fin geometry while still considering the multiple core?
- I'll try to modify the fuel grain plate. I'll model it to be a large hole in the center, just enough to hold the grain in place.
- About the fins being attached to the nozzle. You could be right, but I almost wonder if having fins there will help dissipate the heat from the nozzle since we are increasing the surface area which will increase the convection. I can try a simple thermal FEA of this to get an idea.
- I agree about getting rocketry clubs involved. They would have much more knowledge and experience than I do.
|
|
|
6:54 am January 18, 2010
| Luke Maurits
| | Adelaide, Australia | |
| Admin
| posts 1483 | |
|
|
Is that image of thrust profiles vs grain shapes for hybrid fuel grains or solid? I suspect that the details may be different between the two due to the differences in oxidiser distribution. There may be a hybrid appropriate diagram in the textbook, I will have to check soon for you.
|
Main CLLARE workgroups: Mission Planning, Navigation and Guidance. I do maths, physics, C, Python and Java.
|
|
|
7:30 am January 18, 2010
| Rocket-To-The-Moon
| | Altus, Oklahoma, USA | |
| Member | posts 685 | |
|
|
Very good work! I am excited about OHKLA because it is a project that can produce some real results quite soon
I do believe that the image you linked to is specifically referring to solid fuel grains, but if thrust is a function of the exposed surface area then they should also be an approximate for a hybrid motor. I am trying to understand if it is better to have low or high thrust initially while the rocket is operating in the most dense part of the atmosphere. My thought is that it is best to use a fuel grain with a single central port because its thrust profile starts out low and then ramps up. This would cause the motor to burn the least amount of propellant when it is operating in the dense atmosphere down low. Drag increases with the square of velocity which means that it is ideal for the rocket to travel slowly in the densest part of the atmosphere and then accelerate more rapidly once it gets above 10,000-15,000m or so.
Does this make sense or should we have high thrust initially (as in the dual composition grain) so that the rocket's exposure to the lower atmosphere is as brief as possible?
|
Main Workgroups: Propulsion & Spacecraft Engineering
|
|
|
11:12 pm January 18, 2010
| brmj
| | Rochester, New York, United States | |
| Member | posts 402 | |
|
|
My understanding is that suborbital rockets intended to reach high altitudes typically have a short, high acceleration burn and cost most of the way. This suggests that the thickness of the atmosphere at low altitudes isn't a compelling reason for a longer burn time, though this might be an oversimplification or bad analysis on my part.
|
Main work groups: Propulsion (booster), Spacecraft Engineering, Computer Systems, Navigation and Guidance (software)
|
|
|
12:02 am January 19, 2010
| Luke Maurits
| | Adelaide, Australia | |
| Admin
| posts 1483 | |
|
|
My understanding is in line with brmj's: very short, very high thrust engines are the way to go for high altitude. I've read John Carmack discuss the option of delaying most of the acceleration until higher in the atmosphere to minimise drag, and he acknowledged outright that there is a performance penalty associated with this (but said they were willing to pay it in exchange for not having to worry about aerodynamic fairings). I think Gary Schnyder later confirmed this on the CSTART forums somewhere.
|
Main CLLARE workgroups: Mission Planning, Navigation and Guidance. I do maths, physics, C, Python and Java.
|
|
|
1:12 am January 20, 2010
| rpulkrabek
| | |
| Member | posts 348 | |
|
|
Post edited 7:18 am – January 20, 2010 by rpulkrabek
I have created a clone of the official repository, named rapulkrabek-cstart-clone. I believe only Luke has access to merge with the cstart repository. So far, the changes I have done are that I added the directories Concept_Pics and ProE under /hg/projects/ohkla/, which contain the pictures of the OHKLA concept I have shown in a previous post and the Pro/ENGINEER files I created.
If you would like, you can now make a clone of these files to your PC and view them. Here are some quick steps that would be needed to view the Pro/E files:
- Download mercurial: http://mercurial.selenic.com/.
- In Windows, click the blue download button.
- Ubuntu: $ apt-get install mercurial
- Fedora: $ yum install mercurial
- OpenSUSE: $ zypper install mercurial
- Gentoo: $ emerge mercurial
- Mac OS X: ???
2. Go to the command line or terminal (Windows: click start, then run and enter "cmd". In Linux: you should probably know).
- navigate to the directory you want to put the files, or stay at the current location and move the files later on: $ cd /go/to/where/you/want
- enter this command (with out the $): $ hg clone https://rapulkrabek-cstart-clone.googlecode.com/hg/ rapulkrabek-cstart-clone
hg clone <source location> <target location>
- You will now have the entire clone in a directory called rapulkrabek-cstart-clone. You can then navigate to rapulkrabek-cstart-clone/projects/ohkla/ to see the two directories that I added.
3. Download either eDrawings from SolidWorks (make sure to check the Pro/E check box before downloading) or ProductView from PTC (after filling in a form).
4. To view the entire assembly, open ohkla.asm.1 in either eDrawings or ProductView (if opened in ProductView, click on "Default View" in the left column). In eDrawings, click and hold middle mouse button to rotate. In ProductView, click and hold right mouse button to rotate.
I will continue to update the design and welcome all feedback to improve on it. One thing to improve is the grain stopper (make it with just a large hole in the center) and the grain geometry (currently thinking about multi-fin). I will also place the FEA files when they get updated as well.
|
|
|
2:18 am January 20, 2010
| Luke Maurits
| | Adelaide, Australia | |
| Admin
| posts 1483 | |
|
|
Sorry I've been a bit slack with regards to helping you with Mercurial stuff, e.g. not replying to your PM.
I have gone ahead and added you as a "Project committer" for the cstart repository, which means you should now have the authority to merge your clone with the official repository. Truth be told, I haven't actually figured out how to do this myself yet, although I'm sure it's quite easy. Whichever of us successfully does it first can explain it to the other. :)
Note that if you like I can make you a "Project owner", which lets you not only commit to the official repository but also add other people to the committers/owners list, modify the project's pages, Wiki, etc, etc. I didn't know if you were interested in doing this kind of work so I only made you a committer at first, but it's no problem to upgrade you.
If anybody else in the core group wants committer/owner status, just yell out and I'll be happy to do it.
To be clear, the process in general for people who aren't well known/trusted members of the group is that they create a clone for themselves and work in there, and then ask a committer for a merge when they want one, which the committer can either do or not do at their discretion (ideally after discussing it with others). Once somebody has proven themselves trustworthy by requesting a few good quality merges then we can promote them to a committer.
|
Main CLLARE workgroups: Mission Planning, Navigation and Guidance. I do maths, physics, C, Python and Java.
|
|