OpenRocket is telling me to make tiny fins for a long rocket

finspin

Member
Joined
Mar 31, 2022
Messages
16
Reaction score
1
I'm designing the fins for a 10k feet hybrid rocket for my university's rocket club. Propulsion on our team has extended their fuel grain a lot to achieve the desired apogee, and we also went down on the diameter, giving us a 161 inch long rocket with a diameter of 5.1 inches. OpenRocket is telling us we need fins that are about 3.4 inches in span in order to get a stability of 3.24

When looking at these fins in person (which can be held in the palm of my hand), compared to the size of the rocket (which is taller than two people), they do not look big enough for a rocket of that size.

I've been looking into the following theories:

(1). For a skinny rocket, the CP moves upwards much faster than usual as the angle-of-attack increases, so the rule of 1.5-2.5 stability is not enough. The stability number of 3.25 is the stability at the OTRS at 0 degrees AOA. Due to CP shift, we should not rely on this number. However, OpenRocket can account for the CP shift in simulations if you set the correct wind speed and launch rail parameters; then you need to graph stability vs time and look at it off the rail, which is what we did for the PDR. If we assume wind speeds of 4.45 mph, the stability decreases to ~2 off the rail. If we assume 20 mph, the stability decreases to ~0.9. This theory explains only some of the issue. I'm unsure what stability number we should aim for here, but we know how to calculate it with the component analysis in OpenRocket.

(2). Dynamic instability: Fins that are this small cannot produce enough of a lift force to actually correct the rocket in time. The CP might be in the right spot, but the magnitude of its force is not actually high enough compared to the moment of inertia of the rocket. Most of what I've read suggests that dynamic stability issues are not much of a concern for long rockets.

(3). The CG we're estimating is just incorrect because of unknown mistakes in OpenRocket. Our masses for propulsion components were definitely off, such as missing the mass of the oxidizer and fuel grain. But once these were added, the issue actually got worse. Our CG is approximately ~80 inches from the nosecone. It changed to ~72 inches when I audited the sources for the propulsion masses, but I'm still in the process of auditing these masses. I was told it was typical for a Rocket's CG to be approximately in the center.

(4). For a long rocket, the turbulent boundary layer/wake will grow to a significant size, resulting in little or none of the fin span sticking into the laminar region where they can produce enough lift, and OpenRocket doesn't account for this. I have attempted running an ANSYS Fluent CFD simulation to investigate this theory, however, it is giving me a CP that is even further back then what OpenRocket tells us for an angle-of-attack of 11 degrees. This 11 degree aoa is the effective aoa for a wind speed of 20 mph and and otrs of 100 ft/s; the goal with using this number is to check for wind induced instability. The CFD does have some issues with mesh quality so it is still a work in progress.

(5). The fins aren't actually too small, our intuition is wrong and OpenRocket is right. We are trying to find a COTS rocket that matches our rocket dimensions but have so far been unsuccessful. Another thing we are considering is 3d printing a small scale version of our rocket with a similar reynold's number off the rail and a similar mass distribution.

I'm looking for advice on the best way to approach this problem; are the fins really too small?
 

rocketgeek101

Well-Known Member
Joined
Nov 27, 2011
Messages
2,431
Reaction score
332
Location
NB Canada
It would be helpful to post a few pictures (or even better, the OpenRocket file) for your intended design for folks to take a look at.
 

neil_w

OpenRocketeer
TRF Supporter
Joined
Jul 14, 2015
Messages
14,258
Reaction score
7,665
Location
Northern NJ
I was told it was typical for a Rocket's CG to be approximately in the center.
Hmm, not so sure about that. It certainly *can* be, but very often CG is quite a bit aft of center. I suppose a long skinny rocket may be more likely to have it closer to center. But CG is something that can be modeled and measured very precisely.

Very long skinny rockets do often have small fins, relative to their size. Look at the Estes Mean Machine. For a simple 3FNC or 4FNC (which is what I believe you're talking about, obviously pics and/or ORK file will be helpful), I would trust OpenRocket's CP calculations.

And yeah, tall skinny rockets commonly are designed with higher stability margins. This is why some prefer to measure stability as a percentage of rocket length, rather than in calibers. 10% is a common target (plus or minus).
 

neil_w

OpenRocketeer
TRF Supporter
Joined
Jul 14, 2015
Messages
14,258
Reaction score
7,665
Location
Northern NJ
BTW: OpenRocket will calculate CP, but will not specifically evaluate something such as whether the span is appropriate given the size/length of the rocket or whatever. You need to make such determinations yourself.
 

Steve Shannon

Well-Known Member
TRF Supporter
Joined
Jul 23, 2011
Messages
7,933
Reaction score
5,796
Location
Butte, Montana
Small diameter hybrid rockets frequently (always?) have a far forward Cg, but you have to be careful because the Cg shifts rearward significantly during flight as the oxidizer (which accounts for probably 70% of the propellant weight) is used, which is opposite solid motors which have the oxidizer mass and fuel mass mixed.
Long skinny rockets have a very high moment of angular inertia; they resist turning quickly. It’s not uncommon to see them slip backwards at apogee without turning over. Wind shear can cause the angle of attack to change rapidly, but the rocket won’t respond very quickly.
Build a Mean Machine and add forward mass and fly a lot. Then remove the forward mass and fly it and see what happens. Test your design before committing to a high cost rocket.
 

JoePfeiffer

Well-Known Member
Joined
Jun 15, 2013
Messages
372
Reaction score
342
I have frequently been surprised by the small fin spans of successful rockets at the Spaceport America Cup. Spans of less than one caliber are not uncommon; the far forward CG of those rockets fools my intuition on a regular basis.
 

finspin

Member
Joined
Mar 31, 2022
Messages
16
Reaction score
1
Attached is a screenshot of the openrocket and the .ORK file. I can also post some screenshots from the CFD if anyone would like those

And yeah, tall skinny rockets commonly are designed with higher stability margins. This is why some prefer to measure stability as a percentage of rocket length, rather than in calibers. 10% is a common target (plus or minus).
We're pretty close to this target already; 161*0.1 = 16.1 = 3.16 calibers. Our current stability margin at an AOA of 0 degrees is 3.25. Or should we apply this rule to be the stability AFTER the CP shift due to the effective AOA from the wind at launch?

BTW: OpenRocket will calculate CP, but will not specifically evaluate something such as whether the span is appropriate given the size/length of the rocket or whatever. You need to make such determinations yourself.

Right, I just meant we're getting a small fin span when applying the 1.5-2.5 caliber rule of thumb

The Estes Mean Machine(I think that is the name) is 6 feet tall with tiny fins. It flies great.

We looked at the parameters of it and are unsure if it is similar enough to draw conclusions from it since the reynold's numbers for our rocket are 10x-100x those of the mean machine. The mean machine is 79 inches long with a 1.64 in diameter and a max velocity of 171 ft/s (giving a diameter reynold's number of 1.6*10^6 and length reynold's number of 7.9*10^7). Our rocket is 161 in long with a 5.1 in diameter and a max velocity of 1000 ft/s (giving a diameter reynold's number of 3*10^8 and length reynold's number of 9.4*10^8) . I'm using the max velocity here since I'm unsure of the off-the-rail speed for the mean machine.

Long skinny rockets have a very high moment of angular inertia; they resist turning quickly. It’s not uncommon to see them slip backwards at apogee without turning over. Wind shear can cause the angle of attack to change rapidly, but the rocket won’t respond very quickly.

Hm, this is suggesting to me that weathercocking would be less of a concern for long rockets, and that it would therefore be less risky to fly with larger fins. Also, would you say it's worth checking the dynamic stability of the rocket using the equations from Gordon K. Mandell's paper on dynamic stability?

Test your design before committing to a high cost rocket.
I'm increasingly leaning towards this. We'll probably go with 3d printing a scaled down version
 

Attachments

  • openrocket-scale.png
    openrocket-scale.png
    128.7 KB · Views: 68
  • openrocket.ork
    4.1 KB · Views: 8

manixFan

Not a rocket scientist
Joined
Feb 15, 2009
Messages
2,523
Reaction score
1,709
Location
TX
After looking at just the screenshot of your design, two things come to mind: 1 - the fins effectiveness is 'magnified' because they extend past the aft end of the rocket, helping to move the CP even farther back; 2 - those fins will likely not survive landing.

Try moving the fins forward so no part is beyond the aft of the rocket airframe, and sweep the fin trailing edge forward to help with survivability. Once you do that you'll likely see a noticeable increase in the required fin size.

Sounding rockets often used fins like you show, but those rockets did not have to survive landing. For our purposes, the fins might not be as aerodynamically efficient but they have the additional constraint of having to survive ground impact. Just something to consider.


Tony
 

finspin

Member
Joined
Mar 31, 2022
Messages
16
Reaction score
1
Try moving the fins forward so no part is beyond the aft of the rocket airframe, and sweep the fin trailing edge forward to help with survivability. Once you do that you'll likely see a noticeable increase in the required fin size.
Yep we're planning to do that both for that reason and for easier integration. By sweep the trailing edge forward, you mean moving it closer to a clipped delta design or just decreasing the sweep angle?
 

Sterk03

Well-Known Member
Joined
Aug 22, 2020
Messages
72
Reaction score
10
Sorry if in the wrong thread but all this talk about CP and CG and small fins. I have posted in another thread in the great deal of trouble flying the Saturn V Estes 1/100. I have two models an older version which will only take a D motor and the newer 50th Anny version with the 90mm motor. All the discussions on CP/CG
the older version would always lean over wind or no wind. So could not find it in open rocket or rocksim which I have neither but used the cut out cardboard for Cp and the balance for Cg and told to get the size of the body tube seperation3 1/2 inches. But in doing this so much weight had to be added and then you have these mini little fins on the back of this large rocket. Endless disaster flights and then buying the 50 Anny and more of the same. Added larger fins on the old Sat V and still no good. So the experienced rocketeers said add more power, yet I was using the motors Estes said would work??? So it was sort of the stability issue that adding all this weight to get CG higher then Cp just weighted down the rocket too much for the said engines. With the input of the forum I took out all the weight and only added one pad of clay and basically the CP and CG were even and I used a AT E30-4T and presto the most perfect flights ever.

So I'm a little divided on when and where cp/cg come into play and small fins where you add too much weight just to get your computer calculations. Just saying I'm dealing with the same issue on the Boyce Gemini
Titan. I got so much weight in the nose it's a dangerous missile not a model rocket and even have used larger engines then what is published because of all of the added weight it will not fly yet!!!! Does anyone have a Gemini Titan on rocksim or openrocket with any numbers?

Just saying,

Sterk03
 

manixFan

Not a rocket scientist
Joined
Feb 15, 2009
Messages
2,523
Reaction score
1,709
Location
TX
Yep we're planning to do that both for that reason and for easier integration. By sweep the trailing edge forward, you mean moving it closer to a clipped delta design or just decreasing the sweep angle?
Yes, more of a clipped delta shape. A fin shape that has worked very well is the one used on the Mongoose and Go Devil kits by Madcow Rocketry, which are high performance minimum diameter rockets:

Mongoose:
fin-shape.png
Go Devil:
fin-shape-2.png

That shape will survive Mach 2+ with the appropriate material with just fillets and due to the long root has good landing survivability (at least in my experience.)

Just more food for thought.


Tony
 

JoePfeiffer

Well-Known Member
Joined
Jun 15, 2013
Messages
372
Reaction score
342
<snip>
So could not find it in open rocket or rocksim which I have neither but used the cut out cardboard for Cp and the balance for Cg and told to get the size of the body tube seperation3 1/2 inches. But in doing this so much weight had to be added and then you have these mini little fins on the back of this large rocket. Endless disaster flights and then buying the 50 Anny and more of the same. Added larger fins on the old Sat V and still no good. So the experienced rocketeers said add more power, yet I was using the motors Estes said would work??? So it was sort of the stability issue that adding all this weight to get CG higher then Cp just weighted down the rocket too much for the said engines. With the input of the forum I took out all the weight and only added one pad of clay and basically the CP and CG were even and I used a AT E30-4T and presto the most perfect flights ever.
<snip>
This is the best example I've come across for why the cardboard cutout method should *NEVER* be used to estimate CP. It very reliably estimates a CP forward of the actual CP; in a rocket with pretty marginal power anyway putting enough weight in the nose to put CG ahead of that estimate means you're too slow off the rod for the fins to matter.

The cardboard cutout method should be decapitated, its mouth stuffed with garlic, a stake driven through its heart, burned to ash, and then buried at midnight at a crossroads.
 

Steve Shannon

Well-Known Member
TRF Supporter
Joined
Jul 23, 2011
Messages
7,933
Reaction score
5,796
Location
Butte, Montana
This is the best example I've come across for why the cardboard cutout method should *NEVER* be used to estimate CP. It very reliably estimates a CP forward of the actual CP; in a rocket with pretty marginal power anyway putting enough weight in the nose to put CG ahead of that estimate means you're too slow off the rod for the fins to matter.

The cardboard cutout method should be decapitated, its mouth stuffed with garlic, a stake driven through its heart, burned to ash, and then buried at midnight at a crossroads.
Don’t hold back, Joe. Tell us how you really feel.😁
 

Sterk03

Well-Known Member
Joined
Aug 22, 2020
Messages
72
Reaction score
10
OUCH! Yes I got that but I'm not doing the string swing test either and I have downloaded Rocksim so as soon as I
get past the learning curve I will use that to see what numbers it will show. If your getting rid of the
cutout method you might as well erase the minimum distance between CP/CG has to be the diameter of the body tube as well from the Rocket Building Bible! I think the cardboard cutout can give you a ballpark number though and I have read quite a few posts on how inaccurate the Rocksim and Openrocket numbers have been off also, just saying...... I had to use what was available to me, I don't think I committed a crime.
I got your point.

Sterk03
 

boatgeek

Well-Known Member
Joined
Dec 27, 2014
Messages
4,975
Reaction score
4,124
Attached is a screenshot of the openrocket and the .ORK file. I can also post some screenshots from the CFD if anyone would like those


We're pretty close to this target already; 161*0.1 = 16.1 = 3.16 calibers. Our current stability margin at an AOA of 0 degrees is 3.25. Or should we apply this rule to be the stability AFTER the CP shift due to the effective AOA from the wind at launch?



Right, I just meant we're getting a small fin span when applying the 1.5-2.5 caliber rule of thumb



We looked at the parameters of it and are unsure if it is similar enough to draw conclusions from it since the reynold's numbers for our rocket are 10x-100x those of the mean machine. The mean machine is 79 inches long with a 1.64 in diameter and a max velocity of 171 ft/s (giving a diameter reynold's number of 1.6*10^6 and length reynold's number of 7.9*10^7). Our rocket is 161 in long with a 5.1 in diameter and a max velocity of 1000 ft/s (giving a diameter reynold's number of 3*10^8 and length reynold's number of 9.4*10^8) . I'm using the max velocity here since I'm unsure of the off-the-rail speed for the mean machine.



Hm, this is suggesting to me that weathercocking would be less of a concern for long rockets, and that it would therefore be less risky to fly with larger fins. Also, would you say it's worth checking the dynamic stability of the rocket using the equations from Gordon K. Mandell's paper on dynamic stability?


I'm increasingly leaning towards this. We'll probably go with 3d printing a scaled down version
A ~40% size downscale should be easy to build out of 54mm tubing and plywood fins. No printing necessary (unless you want to).

One thing I haven't really seen mentioned is that there's a rule of thumb that the fin span should be at least one caliber. That gets the fins out of the boundary layer and into clear air. It's probably not gospel, but rules of thumb exist because they solve common problems. :D
 

finspin

Member
Joined
Mar 31, 2022
Messages
16
Reaction score
1
Update. So I looked into the theory that the masses in OpenRocket were wrong. The masses were not being updated, and the CG of the motor was off (since the oxidizer and fuel are separate, it's not correct to assume its at the center of the overall motor; our solution was to switch to a .rse file since you can specify the CG in those). Once we updated the masses, the CG moved aft to 90 inches. Also, our off-the-rail speed went down to between 65-75 ft/s (depending on whether we use the lower or upper bound for engine thrust). Which is not great, since we were hoping for 100 ft/s minimum.

We also moved the fins up so they don't stick past the end of the boattail.

Now our fins need to be root=4.6 in, span=4.6 in, tip=4.6 in, sweep=4.6in to achieve a target stability of 1.0 calibers off the rail with 20 mph winds (this is accounting for the body-lift/long-rocket effect). But the rocket reaches a max stability of slightly over 6.0 as it flies, which is concerning. Is there a way to design the fins to ameliorate this huge shift in the stability as the rocket gains speed? Like a way to get large performance at a high angle of attack to shift the CP back after going off the rail, but not a large performance at a small angle of attack so we don't get overstable at high altitudes.
 

Attachments

  • Prometheus 2022 OpenRocket.ork
    4.3 KB · Views: 4
  • Prometheus_SRAD-lower-bound-2022.rse
    2.7 KB · Views: 1
  • Prometheus_SRAD-upper-bound-2022.rse
    1 KB · Views: 1

Sterk03

Well-Known Member
Joined
Aug 22, 2020
Messages
72
Reaction score
10
Please explain the difference. Your doing the cutout cardboard to find the Cp and your using the rocket to find the CG and according to the book the Cg and Cp should be a difference minimum of the size of the body tube the way I read it? So I guess I need corrected. When I used this method on the Sat V it added quite a bit of weight to the nose cone to get the difference. After having stable fights I said this is way to heavy so I reduced it back to the original 2 pads of clay in the kit and it flew very stable . So much for the 3.5 inches of difference. I'm open to yur thoughts as I'm just trying to understand it all but this is the info I found?

Thanks

Sterk03
 

neil_w

OpenRocketeer
TRF Supporter
Joined
Jul 14, 2015
Messages
14,258
Reaction score
7,665
Location
Northern NJ
Assuming you were replying to my previous post...
Please explain the difference. Your doing the cutout cardboard to find the Cp and your using the rocket to find the CG and according to the book the Cg and Cp should be a difference minimum of the size of the body tube the way I read it? So I guess I need corrected. When I used this method on the Sat V it added quite a bit of weight to the nose cone to get the difference. After having stable fights I said this is way to heavy so I reduced it back to the original 2 pads of clay in the kit and it flew very stable . So much for the 3.5 inches of difference. I'm open to yur thoughts as I'm just trying to understand it all but this is the info I found?
The difference is: the cardboard cutout is a very crude, and often very inaccurate way to calculate a single very important number, which is CP. There are few instances when it is useful anymore, now that we have access to sim tools.

The 1 caliber rule, by comparison, is a guideline that remains quite useful, once you have a handle on CG and CP. It is not gospel, but it is broadly applicable. Some prefer the percentage of length rule: CG/CP margin should be roughly 10% of the rocket length (the new version of OR can report stability this way if you like). For the typical 10:1 rocket, both are the same, but the percentage rule tracks with long skinny rockets better. Generally, it is very helpful to know the limits of the 1 caliber rule, and apply modifications and caveats as needed. But there is no need to throw it out. Ultimately, rocket stability is quite a bit more complex than the 1 caliber rule, but we as hobbyists don't yet have any readily-available tools to analyze more deeply.

As for your SaturnV: I haven't built or flown one so I can't comment.
 

JoePfeiffer

Well-Known Member
Joined
Jun 15, 2013
Messages
372
Reaction score
342
Please explain the difference. Your doing the cutout cardboard to find the Cp and your using the rocket to find the CG and according to the book the Cg and Cp should be a difference minimum of the size of the body tube the way I read it?
The one caliber rule is based on sound physics. It's a good (but not perfect) rule of thumb based on the principle that stability requires a restoring force to be applied when the rocket has an angle of attack other than 0, and suggesting that putting the CP one caliber behind the CG applies adequate restoring force. There's a good (OK, I find it persuasive, but for rockets that aren't absurdly long or stubby it doesn't matter) argument that a rule based on a percentage, like 10%, of length is better.

The cardboard cutout method isn't. It's an intuitively appealing bit of folklore that, upon analysis, turns out to vastly overestimate the effect of the body tube on the rocket's CP and quite reliably gives terribly misleading results -- as you yourself encountered, If you've got a slide rule and an abacus, or even ten fingers, you're better off doing the math. And there are at least three readily available simulation programs that will do the math for you.
 
Last edited:

Sterk03

Well-Known Member
Joined
Aug 22, 2020
Messages
72
Reaction score
10
Ok I understand. I used this method which is all I read about since I have returned to building rockets. As a kid we put the engine in the Estes model and flew it. I never got to this level and there are quite a few references out there saying this is one option to use. That's why I was questioning the method. I will learn Rocksim.
Thanks,

Sterk03
 
Joined
Jun 30, 2015
Messages
1,668
Reaction score
1,833
Location
Australia
Your Nitrous(assumed oxidiser) tank is almost entirely in front of the CP(regardless of how it's calculated) as it empties your CG will move significantly backward. As you get to and go through Mach/s your CP will move forward.
make sure you know that your CG/CP wiil remain in a stable range.
Divide your oxidiser into 4 masses. for each quarter of tank and place them along the tank. Do the same for fuel section but a weights will be in the middle of that section assuming an even burn. Remove one at a time and plot the range of CG. Check your range of CP( available using a RAS Aero simulation. Obviously these are only for burned masses. I'd expect there to be fuel left but not oxidiser. Unless you're planning to burn through your casing wall and use that as fuel. Did that once. Highly recommend you don't do it. It's expensive....... :(
You can divide into more increments if you want but check your ballpark values first.
Good luck with the flight.
 
Joined
Jun 30, 2015
Messages
1,668
Reaction score
1,833
Location
Australia
Your Nitrous(assumed oxidiser) tank is almost entirely in front of the CP(regardless of how it's calculated) as it empties your CG will move significantly backward. As you get to and go through Mach/s your CP will move forward.
make sure you know that your CG/CP wiil remain in a stable range.
Divide your oxidiser into 4 masses. for each quarter of tank and place them along the tank. Do the same for fuel section but a weights will be in the middle of that section assuming an even burn. Remove one at a time and plot the range of CG. Check your range of CP( available using a RAS Aero simulation. Obviously these are only for burned masses. I'd expect there to be fuel left but not oxidiser. Unless you're planning to burn through your casing wall and use that as fuel. Did that once. Highly recommend you don't do it. It's expensive....... :(
You can divide into more increments if you want but check your ballpark values first.
Good luck with the flight.
Mine's only 3.5m long 65mm dia 26,240ft
 
Top