Starfire73
OpenRocket Wannabe
If you have any questions or feedback regarding my sims, you can post them here.
Please visit my sim thread if you haven't already.
Please visit my sim thread if you haven't already.
Starfire's OpenRocket Open Files Feedback Thread
- Thread starter Starfire73
- Start date
Help Support The Rocketry Forum:
- Joined
- Mar 27, 2013
- Messages
- 22,536
- Reaction score
- 14,949
Nice looking work!! Congrats on the sims!!!
A few tweaks for current OpenRocket features (fins on transitions). Also, your fins did not have the same dimensions, and your motor tube installation dimensions did not match the instructions.
Nice work.
Nice work.
Starfire73
OpenRocket Wannabe
Thanks for the feedback. I found a few minor tweaks I wanted to make as well. I incorporated your tweaks and mine into a v02 of my file and revised the post.
Beautiful sims. I especially love the Eradicator. Let's talk about that one. Beware, long post follows!
The reason you are getting bad sim results is the use of body tubes-on-pods for the ring fins. Generally I would discourage this practice because it greatly confuses the simulator. Body tubes are presumed to be *external* components, with airflow only on the outside.
Before we try to figure out a way to fix this, let's see what the simulator was computing. I removed the Cd override for the rocket, renamed the body tube components used for the ring fins and did a component analysis (Tools -> Component Analysis). Looking at the "Drag characteristics" tab:
We can see here that the pressure Cd on the forward ring fin is very high. That's because it assumes the front of the body tube is capped with a flat bulkhead (remember, OR assumes all air flows outside the tube, and it has no idea that this body is concentric with the main airframe). Then there's *no* base drag because this BT is not the rear of the rocket, and then a small Friction Cd of .006 due to the surface friction of the tube.
The rear ring, by contrast, has no pressure drag because OR thinks it's sitting behind an airframe of the same diameter (the forward ring). It does have base drag, and the same friction drag as the front ring.
Now, before we go further, let me say that we don't really know exactly what the drag characteristics of those rings would be. All we *do* know is that what's there now is clearly wrong. So we're just going to have to make some guesses, at least until OR is able to handle ring fins natively. Applying some guess directly to those rings, though, should be better than coming up with a wild guess for the entire rocket.
Let's start by keeping the pod/body tube hierarchy as-is. What we want to do is override the Cd of just ring fins to come up with something reasonable. The simplest thing we could do is imagine that the total Cd of the ring is simply twice the exterior friction Cd (since we also have that friction on the interior). That would make the Cd of each ring equal to .012. Annoyingly, the UI only allows two decimal places, so I've used .01 for now. If we apply those overrides to each one, we immediately get much better looking simulations.
In reality, though, those numbers are almost certainly optimistic. The gap between those rings and the body are so small that it's unclear if any air will flow through there at all. So to come up with a more realistic estimate for Cd, you need to start layering guesses on top of guesses. I'll leave that as an exercise to the reader.
Now, if air is flowing under the rings (and honestly we don't know), then they are also contributing to stability, and you're not getting any of that with the use of body tubes. In fact there's no good way to do rings that'll give you a proper CP calculation. So how do we fix *that*?
My standard approach to rings is use an inner tube. Inner components are completely ignored aerodynamically, so my inner rings do not contribute to drag or CP, just mass. That's fine: I handle the other stuff a different way, with a set of invisible fins that have equivalent characteristics to the ring. I use my own method for this, which is guaranteed not to be 100% accurate but so far it has worked well enough for my needs. Because my method is based completely on my own intuition (which is why it is almost certainly wrong) I won't detail the whole thing here. But the most fundamental idea is that the finset is roughly equal to cutting up the ring into a bunch of pieces, flattening out each piece, and mounting the pieces as fins. Then I override the finset mass to 0, because the mass of the ring is already in there.
After doing this, I have a ring that works for visualization and mass/CG calculation, and an invisible set of fins that handle the aerodynamics.
It's still not right, because I strongly believe that the rings on this rocket are going to create more drag and less stability because of their tight fit around the body.
The one feature of this model that argues in favor of the pod/body tube approach is the rail button on the rear ring. OR won't let you attach a rail button (an aerodynamic component) to an inner tube. So, if you want to include the button, it argues in favor of using the body tube approach with Cd overrides. But you might still also need to put in a set of invisible fins to incorporate the stability effects of the rings (if there are any).
Where does that all leave us? Well, it is not possible to get a perfect sim of many complex rockets such as this one, especially if you want to get both appearance and performance right. Oftentimes I will create separate models for appearance and simulation, although with Cd overrides and invisible components in 22.02 you can now generally make one file do it all if you really want to.
Hope this is helpful.
The reason you are getting bad sim results is the use of body tubes-on-pods for the ring fins. Generally I would discourage this practice because it greatly confuses the simulator. Body tubes are presumed to be *external* components, with airflow only on the outside.
Before we try to figure out a way to fix this, let's see what the simulator was computing. I removed the Cd override for the rocket, renamed the body tube components used for the ring fins and did a component analysis (Tools -> Component Analysis). Looking at the "Drag characteristics" tab:
We can see here that the pressure Cd on the forward ring fin is very high. That's because it assumes the front of the body tube is capped with a flat bulkhead (remember, OR assumes all air flows outside the tube, and it has no idea that this body is concentric with the main airframe). Then there's *no* base drag because this BT is not the rear of the rocket, and then a small Friction Cd of .006 due to the surface friction of the tube.
The rear ring, by contrast, has no pressure drag because OR thinks it's sitting behind an airframe of the same diameter (the forward ring). It does have base drag, and the same friction drag as the front ring.
Now, before we go further, let me say that we don't really know exactly what the drag characteristics of those rings would be. All we *do* know is that what's there now is clearly wrong. So we're just going to have to make some guesses, at least until OR is able to handle ring fins natively. Applying some guess directly to those rings, though, should be better than coming up with a wild guess for the entire rocket.
Let's start by keeping the pod/body tube hierarchy as-is. What we want to do is override the Cd of just ring fins to come up with something reasonable. The simplest thing we could do is imagine that the total Cd of the ring is simply twice the exterior friction Cd (since we also have that friction on the interior). That would make the Cd of each ring equal to .012. Annoyingly, the UI only allows two decimal places, so I've used .01 for now. If we apply those overrides to each one, we immediately get much better looking simulations.
In reality, though, those numbers are almost certainly optimistic. The gap between those rings and the body are so small that it's unclear if any air will flow through there at all. So to come up with a more realistic estimate for Cd, you need to start layering guesses on top of guesses. I'll leave that as an exercise to the reader.
Now, if air is flowing under the rings (and honestly we don't know), then they are also contributing to stability, and you're not getting any of that with the use of body tubes. In fact there's no good way to do rings that'll give you a proper CP calculation. So how do we fix *that*?
My standard approach to rings is use an inner tube. Inner components are completely ignored aerodynamically, so my inner rings do not contribute to drag or CP, just mass. That's fine: I handle the other stuff a different way, with a set of invisible fins that have equivalent characteristics to the ring. I use my own method for this, which is guaranteed not to be 100% accurate but so far it has worked well enough for my needs. Because my method is based completely on my own intuition (which is why it is almost certainly wrong) I won't detail the whole thing here. But the most fundamental idea is that the finset is roughly equal to cutting up the ring into a bunch of pieces, flattening out each piece, and mounting the pieces as fins. Then I override the finset mass to 0, because the mass of the ring is already in there.
After doing this, I have a ring that works for visualization and mass/CG calculation, and an invisible set of fins that handle the aerodynamics.
It's still not right, because I strongly believe that the rings on this rocket are going to create more drag and less stability because of their tight fit around the body.
The one feature of this model that argues in favor of the pod/body tube approach is the rail button on the rear ring. OR won't let you attach a rail button (an aerodynamic component) to an inner tube. So, if you want to include the button, it argues in favor of using the body tube approach with Cd overrides. But you might still also need to put in a set of invisible fins to incorporate the stability effects of the rings (if there are any).
Where does that all leave us? Well, it is not possible to get a perfect sim of many complex rockets such as this one, especially if you want to get both appearance and performance right. Oftentimes I will create separate models for appearance and simulation, although with Cd overrides and invisible components in 22.02 you can now generally make one file do it all if you really want to.
Hope this is helpful.
Starfire73
OpenRocket Wannabe
Thanks Neil. That was an excellent treatise on the ring fin issue. That's worth saving for future reference.
Please allow me to give you a brief history of how I got here.
I originally had the sim designed with inner tubes, but as you pointed out, I couldn't put the rail button on it directly. I made an extended rail button post to make it look mounted on the ring from the outside. I didn't particularly care for the look from other angles, but it worked and was better than nothing. I applied Levison's fin hack from the Apogee newsletters to this version, but I was getting the "too many fins error" and the sims still didn't look even close to right.
K'Tesh gave me the idea of switching to pod sets with body tubes, that worked great for the rail button appearance; however, it still presented sim issues.
Everything I've tried has thrown the sims way off. Whether the body tubes on pods, or the inner rings on the outside, even my phantom fin hacks - the sims were still way off. In my sims I was getting Cds calculated at lift off of over 2.5, which reduced expected altitudes by about 60%.
Even Rocksim (with the supposed aerodynamic support for ring tail fins), was way off. RS ring tail fins were no better than what I could do in OR. I was getting altitudes of 400' on motors that should been putting it up 1,500' or better, in both OR and RS.
In the end, I ran multiple sims based on Sirius Rocketry's altitude/motor recommendations in the Eradicator instructions, and adjusted the overall Cd override until three or four benchmark motors were about right. It seemed like anywhere from 0.67 to 0.72 was the sweet spot.
With all that said, I really like what you did in your post. It's going to take me a little time to digest it all, but I like your idea keeping the pod/body tube approach for the ring fin and adjusting the Cd on just the ring fins instead of the entire rocket, it makes more sense. It's elegant (as all design should be). I expect I will update the file at some point after I incorporate your feedback.
Thanks again,
Doug
Please allow me to give you a brief history of how I got here.
I originally had the sim designed with inner tubes, but as you pointed out, I couldn't put the rail button on it directly. I made an extended rail button post to make it look mounted on the ring from the outside. I didn't particularly care for the look from other angles, but it worked and was better than nothing. I applied Levison's fin hack from the Apogee newsletters to this version, but I was getting the "too many fins error" and the sims still didn't look even close to right.
K'Tesh gave me the idea of switching to pod sets with body tubes, that worked great for the rail button appearance; however, it still presented sim issues.
Everything I've tried has thrown the sims way off. Whether the body tubes on pods, or the inner rings on the outside, even my phantom fin hacks - the sims were still way off. In my sims I was getting Cds calculated at lift off of over 2.5, which reduced expected altitudes by about 60%.
Even Rocksim (with the supposed aerodynamic support for ring tail fins), was way off. RS ring tail fins were no better than what I could do in OR. I was getting altitudes of 400' on motors that should been putting it up 1,500' or better, in both OR and RS.
In the end, I ran multiple sims based on Sirius Rocketry's altitude/motor recommendations in the Eradicator instructions, and adjusted the overall Cd override until three or four benchmark motors were about right. It seemed like anywhere from 0.67 to 0.72 was the sweet spot.
With all that said, I really like what you did in your post. It's going to take me a little time to digest it all, but I like your idea keeping the pod/body tube approach for the ring fin and adjusting the Cd on just the ring fins instead of the entire rocket, it makes more sense. It's elegant (as all design should be). I expect I will update the file at some point after I incorporate your feedback.
Thanks again,
Doug
When trying out all your different methods, be sure to keep using Component Analysis to see where the crazy CD values are really coming from. That can be quite illuminating. I would definitely be curious as to why you were getting such crazy CD values.
Confession: it was only about a week or two ago that I became aware of the Component Analysis functionality.
Ultimately, you will sometimes have to decide between ultimate visual fidelity and making a model that is useful for flight simulation. With some designs it is a challenge accomplish both with same file.
Confession: it was only about a week or two ago that I became aware of the Component Analysis functionality.
Ultimately, you will sometimes have to decide between ultimate visual fidelity and making a model that is useful for flight simulation. With some designs it is a challenge accomplish both with same file.
Starfire73
OpenRocket Wannabe
Agreed, of course my preference will always be to have one file that can do both appearance and an accurate sim, but in some complex rockets or rockets with sim issues that’s not always possible.
In some cases I have three (or more) versions:
1. Appearance optimized - public
2. Flight sim optimized - public
3. Appearance or physical mods from original - private
- Joined
- Mar 27, 2013
- Messages
- 22,536
- Reaction score
- 14,949
Your stuff is looking really nice! And you're doing great at picking subjects I don't know about.
Starfire73
OpenRocket Wannabe
Thanks K’Tesh, I appreciate that, and I appreciate the help.
- Joined
- Mar 27, 2013
- Messages
- 22,536
- Reaction score
- 14,949
Bum!!! Lucky Bum!!! I want a BLU-97
I used to load CBU-87s when I was in England. This kid is clearly an homage to that weapon.
I used to load CBU-87s when I was in England. This kid is clearly an homage to that weapon.
Starfire73
OpenRocket Wannabe
That's awesome! Always a fan (and mine) favorite with those plastic cluster bombs falling down. This is one of those rockets that gets launched at least once or twice a year. It doesn't sit on the shelf like many of the others do.
Starcrusher looks nice... I *think* you need some sort of finnage underneath, don't you?
Starfire73
OpenRocket Wannabe
I considered placing a pair of fins on the underside of the rocket at the rear; basically, the same canard fins that are up front. However, I have flown this rocket and it flies just fine without them. I think the rudder fin and the outboard tube fins provide enough aerodynamic stability on that axis.
- Joined
- Mar 27, 2013
- Messages
- 22,536
- Reaction score
- 14,949
That Rocketry Show rocket sim is SWEET!!!
Similar threads
- Sticky
