understanding open rocket and its calculations
- Thread starter UVU_Team_Rocket
- Start date
Help Support The Rocketry Forum:
NTP2
Well-Known Member
+1 it looks like a lesson is in order!
Maybe so.
I am only a non-practicing naval architect so what do I know ...
I know about interference drag and I also know that the traffic at the INTERSECTION of Ranch Road 620 and IH 35 is a DRAG
But I don't know about aerodynamic intersection drag
-- kjh
Rschub
Well-Known Member
Well yeah, if you want to get pedantic about descriptions, then yes interference drag covers the drag at the intersection of the fin and the body, as well as other forms.
When people pull out the "you didn't capitalize that variable correctly" argument that's usually where I stop.
When people pull out the "you didn't capitalize that variable correctly" argument that's usually where I stop.
UVU_Team_Rocket
Well-Known Member
ok, i have a question. I think this is a typo in the tr11 report, but im not sure. It says that SF is the surface area of all the fins. So they calculated the area to be 4.5 on page 51 of the PDF but page 47 of the actual document. They got 4.5 by doing the area of a triangle and multiplying by 3. BUT that would only account for ONE side of the fin and not BOTH sides of the fin. Air flows on BOTH sides, so it should be 6(bh/2) to account for both sides of the fin, and not just one side.
thanks
thanks
Aerodynamic drag, and corresponding coefficient of drag, is only calculated by the frontal area (reference area). TR11 is correct.
UVU_Team_Rocket
Well-Known Member
Looking very briefly, especially earlier on page 21, it appears he is considering the planform area, not the surface area. If the coefficients are twice as large, using a measure that's half the area works fine. I'd have to spend more time than I have available to be sure, but i think the TR is correct.
Cd is referenced to final area, but surface area (or see my response a minute ago, planform area) is absolutely used in calculating it.
UVU_Team_Rocket
Well-Known Member
it probably is correct, im just trying to see why its only the one side. I did see the section on planform area, but it didnt state why its just one side. I may have to start doing some extra digging to get this answered haha.
thanks for your reply
Rschub
Well-Known Member
The fin is a wing. The fin [wing] imparts a force when deflected from the free stream airflow that depends on its coefficient of lift and area of the fin.
The program makes some simplifying assumptions about the lift coefficient and uses the distance from the center of area of the fin to the CG of the rocket to determine the stability contribution of the fin. The lift coefficient of airfoils was usually determined by wind tunnel testing in the era these equations [barrowman] were developed, today we have software that can predict the properties of airfoils pretty well.
Lift is denoted as a force per unit area. You do actually need both sides of the wing to generate the force but the convention is to use the planform area.
How lift is generated is a whole nother religious topic.
The program makes some simplifying assumptions about the lift coefficient and uses the distance from the center of area of the fin to the CG of the rocket to determine the stability contribution of the fin. The lift coefficient of airfoils was usually determined by wind tunnel testing in the era these equations [barrowman] were developed, today we have software that can predict the properties of airfoils pretty well.
Lift is denoted as a force per unit area. You do actually need both sides of the wing to generate the force but the convention is to use the planform area.
How lift is generated is a whole nother religious topic.
UVU_Team_Rocket
Well-Known Member
ok, looks like i need to start looking up the planform area nad drag and see what i find. thanks for that. I have learned a lot of information with this thread.
Rschub
Well-Known Member
Planform area is just the 2D area of the fin no matter what its shape.
Last edited:
Convention.
Cd (coefficient of drag) is usually determined experimentally.
Put an object, like a fin, into an air stream, measure the force on the object and the velocity of the fluid, choose a reference area and calculate Cd for that object.
For flat plate objects, the convention is to use surface area. For wings, planform area. Other objects may have different reference area, or something relevant. For a circular cylinder, it is typically diameter or Re (Reynolds number).
Take figure 40 in TR11. The drag coefficients were determined based on the thickness ratio (by cross section outline)
When you use a Cd to determine Drag Force, you need to use the Cd for the appropriate object and the appropriately associated A (reference area)...
In this case, the A (reference area) used to calculate Cd is the planform area.
Hope that helps!
Last edited:
UVU_Team_Rocket
Well-Known Member
ohhhh, ok, i get it. it was setup as part of the metrics and convention then. Since the drag force does indeed include both sides of the fin, convention just made it easier to use planform. Perfect. Thank you very much for this information.
so far all my hand calculations match up to open rocket except 1. My velocity is 551 mph by hand, but OR reports 371.... so i messed up somewhere. Both include the 0.304 Cd etc etc.
Lots of ways to build drag models. In OR, use the "Component analysis" and see how they differ. "Component finish" also influences the drag.
By hand has no dynamic modelling of the forces (CG/CP, velocity changes, atmospheric changes during flight, oscillation, spinning, dynamic interference drag, launch rod length, weather cocking and angle of attack changes, motor thrust curve, finish (a rough surface can increase drag), launch rod drag, etc., etc.
Here is a decent paper that is knocking around in my archives....about using equations in rocket flight simulators that expands the static analysis more toward a dynamic model. I have not reconciled this to the OR code.
https://cambridgerocket.sourceforge.net/AerodynamicCoefficients.pdf
Last edited:
Here is a great example of the dynamic nature of drag. Look at the line for a 1" circular cylinder in the diagram below. The line looks like a parachute with a dot. At Re close to 2x10(5), as velocity increases between M=.4 and M=.6, the Cd trend shifts almost to vertical. If I'm dynamically modelling drag, and am using speed as part of my drag model, in that period of time where the model is going M=.4-.6, my drag will increase for that period of time.
Now look at the 4" circular cylinder line (circle with a dot). At Re = 3x10(5), drag falls off significantly. (Re 3x10(5) ~~ 60mph for a 4" circular cylinder at ISA standard).
Again, lots of ways to model drag. I don't know all the dynamic modelling that is in open rocket. The fact that your speeds are different (and OR is lower) leads me to believe there is much more dynamic modelling that is picking up more dynamic drag.
NACA TN 1941
https://ntrs.nasa.gov/api/citations/19930083676/downloads/19930083676.pdf
Now look at the 4" circular cylinder line (circle with a dot). At Re = 3x10(5), drag falls off significantly. (Re 3x10(5) ~~ 60mph for a 4" circular cylinder at ISA standard).
Again, lots of ways to model drag. I don't know all the dynamic modelling that is in open rocket. The fact that your speeds are different (and OR is lower) leads me to believe there is much more dynamic modelling that is picking up more dynamic drag.
NACA TN 1941
https://ntrs.nasa.gov/api/citations/19930083676/downloads/19930083676.pdf
Last edited:
Oh, and there is this...may answer most of your questions...
https://openrocket.sourceforge.net/techdoc.pdf
https://openrocket.sourceforge.net/techdoc.pdf
UVU_Team_Rocket
Well-Known Member
ill do a longer reply after work. thanks for all of this. Yeah, OR has 0.304 for total drag under component analysis so thats what i used in all my hand calcs, but i think i might have missed something for it to be off by 200mph haha
likely...
All of what you are doing can be a bit, or a bunch of a rabbit hole. Fun, but a rabbit hole. I spent the first week of November going down the drag functions rabbit hole for a particular question/interest...I started with the first large scale tests that Eiffel did in 1907 from the Eiffel Tower (French), then Wieselsberger's work in 1921 (German), then flowed through the NACA work, then Hoerner, then, then, then. Rabbit holes can be fun...I just prefer them in English. Translating/reading in French and German is a bitch.
The other aspect of my research that is interesting is A) references/citations early in the 20th century were shorthand...no APA, Chicago etc. There was sparce work, and everyone knew what everyone was doing, so shorthand worked fine. B) terminology, both across languages and within languages were different. Definitions may or may not be clear. This started to shift in the 1930s/1940s...at least for definitions within the work...less so for citations. Definitionally, what we might now know as planar flow and axial flow as standard terminology was not true then...words were different.
I'm still curious how information traveled back then. I grew up in the age of mail and written material. No computers, cell phones, internet. Between 1900 and about late 1920s, it seemed like communication and evolution took about 10 years. After about 1930, at least for aerodynamics research, what others had been doing elsewhere shifted to center around NACA. WWII certainly impacted the volume and source of research (shifted to NACA) for what is available readily today. Interesting stuff.
Eiffel in French.
https://histoire.ec-lyon.fr/docannexe/file/1870/01056v01.pdf
Wieselsberger in German.
https://babel.hathitrust.org/cgi/pt?id=uiug.30112008570167&seq=1
Eiffel's apparatus:
Last edited:
UVU_Team_Rocket
Well-Known Member
Yup, the rabbit hole has been fun, but very deep and many caverns to explore.
I am doing research and fatigue analysis now with NACA airfoils for large 5MW wind turbine blades, about 65 meters in length. I have been using XFLR5 for airfoil CFD and OpenFAST for simulations. I just submitted a paper to be published on it. Hopefully it goes well haha
I like learning about the stuff because computers have to estimate the Cd and all the other stuff from something, it just cant create the data. Thats why i like doing the hand work because then i see how the software is doing it, but at a much faster rate.
So far i have found some very interesting equations for things that i have not seen. Such as this simplified approach. It looks like the "big 3" from physics with a little bit of manipulation.
https://www.nakka-rocketry.net/articles/altcalc.pdf
i also have some papers i found from NASA in the 50s from barrowman and he references some other works.
Thanks for the link @UVU_Team_Rocket !
I've never seen that one ... saved for further study !
I wonder if that's what @JohnCoker is doing in his ThrustCurve.org - Match a Rocket - Motor Guide WebApp ?
-- kjh
EDIT: corrected alt-text for the link
EDIT: I REALLY like this John's App because I can use it instead of a paper hard-copy sim summary when I am at a launch site ...
Last edited:
- Joined
- Sep 5, 2009
- Messages
- 3,539
- Reaction score
- 1,676
No, Thrustcurve is performing numerical integration.
https://www.thrustcurve.org/info/simulation.html
The "simplified approach" from Nakka might have been useful in 1960. It is strange that he would propose it in 2007, though. Numerical methods and software have been highly efficient and available to any hobbyist with a PC since the 1980's. I never understood why aerodynamicists like to play around with "zero drag" concepts.
Agreed. I see no merit to that simplified approach, other than just noodling around at random. There is much utility in the extended Feskins-Malewicki equations. They are analytic and exact within reasonable assumptions. Tom Keuchler even had some analytic solutions for variable thrust and exponential atmosphere, although he had to use special functions (Bessel functions?). You can't do as much with digital numerical solutions except endlessly run more of them.
I could explain the utility and interest in "zero drag" CFD like aerodynamics, but you might have to buy a beer or two.
Rschub
Well-Known Member
I may have to get this put on a tee shirt.
- Joined
- Sep 5, 2009
- Messages
- 3,539
- Reaction score
- 1,676
Yes, F-M equations are exact and still offer some areas of improvement to play around. In addition to overcoming the constant thrust and density as you mentioned, I think Cd is also assumed constant.
I will argue that numerical solutions are essentially exact (something like errors of 1.0E-14 or machine precision) even with time steps as large as 0.01 or 0.001 seconds.
Nobody is arguing against the precision of properly done numerical integration. My point is that the F-M like equations are analytic, i.e., closed form algebraic equations, from which you can easily derive partial derivatives for creative use in other algorithms. For example, you could efficiently solve for the Cd that gets your rocket to a target altitude.
- Joined
- Sep 5, 2009
- Messages
- 3,539
- Reaction score
- 1,676
Gotcha. I wonder if any student teams try to optimize/solve in this manner?
When RockSim offered a console executable, I called it from my root-finding and optimization program. Each call only took couple seconds and the routines typically converged in less than 10 iterations. That was plenty fast for me and the sims were full-featured.
shockie
High Plains Rocketeer
Estes TR-11 is at Ye Old Rocket Plans / Rocketry Publications
It's been there for about 15 years
It's been there for about 15 years
Similar threads
