Elvis Jr.,

You highlight an excellent example of experience versus theory and pure science. This comes up many times in this hobby whether it's fin design, EX motors, electronics or whatever.

I've long heard JW swear off medusa nozzles. In fact he's warned me against them several times. I've still chosen the slower speeds, longer burn times, and lower average thrust advantages that the medusa based motors provide. However, every single one of the minimum diameter rockets I flown with a medusa nozzle has corkscrewed to some degree (some catastrophically) and typically closer to motor burnout.

One theory is that small chunks of propellant start to break away from the liner and move about in the motor potentially clogging a portion of the nozzle and/or affecting the directional flow of gases from the nozzle.

Again, empirical data is worth a heck of a lot more than simulation data. Pretty soon I think I’ll start listening.

I realize there is a mixing chamber at the bottom of a motor that uses medusa nozzles, but if you line the slot up precisely with a nozzle port, it would seem to me that there would be a bit of a tendency for that throat to have more discharge.

My rocket "Joyride" flew on a J135 twice and it was so erratic that I thought it would shred. It was fine on any single throat nozzles. That rocket is big enough and has enough mass that I don't think the off-axis mass would have an effect (personal opinion).

I have also mounted batteries and altimeters on the same side of the board many, many times - and the board is riding an allthread that is centered, meaning there is a lot of off-axis mass. I've never had a problem with that setup using single-throat nozzles.

At some point it becomes a confidence thing. I don't have enough confidence in a medusa nozzle to ever fly one again... at least not on a smallish-finned, minimum diameter rocket.

I wonder if anyone has messed around with a helical offset core? That would give it the same burn geometry (at least in 2D), but eliminate the weight imbalance. As long as it were a perfect helix, the core could be extracted (with effort 😉 ) or milled in after casting(?)

As long as there is a chamber at the bottom and a single nozzle, it should cancel out any inherent spin effects.

Just my uninformed L1 idea!
Ken

To be balanced wouldn't it need to do one complete revolution? That would be very challenging. You would end up making the grain shorter to get the same surface area as a straight core, so your impulse would be less since your propellant mass would go down. This would be with the same diameter core as the straight grain. You could reduce the diameter of the core to take the surface area down, but then you risk making your core less than your nozzle and having a motor with high erosivity. A smaller core would also make up in having a higher propellant mass. If you wanted a compromise to keep erosivity down you wold want a helical tapered core, with the bottom bigger than the nozzle and the top smaller to make up for mass. I'm pretty sure it would be easy to model this in Solidworks, but I can't seem to find the "Print to APCP" option on my version 🙂 Just a few random thoughts.

Edward

I forgot to bring that up Ed. Depending on the motor and more likely with the thick webbed motors, you will have a higher % of Erosion. Every motor has some percentage of erosion, especially with the long small diameter motors. This can create pieces breaking off from the movement of the gases, and clogging the nozzles. Hopefully certified motors do not do this very often. Its hard to say, data on the performance of single throat vs. medusa nozzles is limited, maybe I should machine a nozzle and test it 😛

Take a look at Freeze Frame. It had a 10 pound camera 3 inches lateral distance off of the centerline and it was stable. Another example is cluster where one motor doesn't light. The may have curved trajectories but if your rocket is doing cartwheels, it is likely unstable.
Doug

I've wondered about that, so I ran some quick numbers for my own curiousity. This gets long and geeky, so feel free to ignore:

With the three motors you flew on Freeze Frame, and the total mass of the rocket (about 110 lbs), I get about 14 Gs shortly after ignition. So the offset camera was putting about 30 in-lbs x 14 = 425 in-lbs (or 43 N-m) of torque on the airframe. But the moment of inertia for that beast is probably about (1.4 meters)^2 x 50 kg = 100 kg m^2. So the pitch-over rotational acceleration would be 50 N-m / 100 kgm^2 = 0.5 radians/sec^2. Or, after the first 1/2 second, without any fins or launch rod, the rocket would have pitched over .06 radians, or about 3 degrees, and it would be rotating about 14 degrees/second. Considering that the first .25 seconds or so is on the rail, the actual pitch rate and angle would be less than that. And it's also no big deal, because it won't take long for the rocket to have enough airspeed for those big fins to generate more than 425 in-lbs of torque. Rocksim will provide the torque on your plot if you're curious about that.

Now contrast that with my little F10 that cartwheeled last year despite having a CG that was 1.1 calibers (12% of the length of the rocket) in front of the Barrowman-calculated CP. According to Rocksim, the inertia was 570 gram-in^2, or 0.000374 kg m^2. Freeze frame has the moment of inertia that's 266,000 times the inertia of this little rocket. The initial acceleration was about 23 Gs, so if 1 gram were off-center by 1 mm, that would cause 50x the pitch rate that Freeze frame saw. Or in other words, after 0.5 seconds with no launch rod or fins, (the same time that Freeze frame went 3 degrees), this rocket would have just about done a 180, and it would be going 2 revs/second. And that's just with 1 gram (out of about 120) off-center by 1 mm. My altimeter/tracker combination weighed about 20 grams, so if the CG of that combination were off by 1/15 of a millimeter, that's all the torque that was required to send it cartwheeling that quickly, assuming the thrust was perfectly straight and ignoring the fin contribution.

But what about the launch tower and the fins? Rocksim says that when the rocket cleared the bicycle tower, it was going 45 mph. So far, so good. But let's look at what was going on 0.1 seconds after it left the tower. Now it's going 80 mph, but the angle of attack is 7 degrees if the fins haven't helped yet and it was going straight when it left the tower. If the fins have been doing some good and there's 5 degrees angle of attack, the fin restoring torque (from Rocksim) is .017 N-m. The acceleration is still about 20 Gs, so the 20 gram electronics package is imparting about 4 N to the airframe. If it's only off-center by 4 mm, the CG-thrust misalignment is still overwhelming the fin restoring torque, and the rocket is still pitching over. As the angle of attack increases, at some point the airfoil fins will stall, the restoring torque goes down with increasing angle of attack, and we get a skywriter with a 7-second burn time.

What tipped me off on this was that the previous iteration really was aerodynamically unstable (1 caliber Rocksim, < 0.1 calibers Barrowman) until some fuel burned off, and then it straightened out after doing a few flips. The next attempt had larger fins, 1.1 calibers of Barrowman margin, and it cartwheeled anyway. But this time (the one I'm using in the example) it never straightened out; it was still tumbling and burning when it hit the dirt. Then I looked closely at the recorded lateral accel data and I could see it was trying to pitch at ignition, in the same direction it pitched over after it cleared the tower.

But what convinced me 100% were my 2 recent G37 flights. At Tripoli in August I flew a rocket with a G37 and shot video of it when it took off. It had a nasty corkscrew and immediately turned about 40 degrees off vertical. It landed south of the Platte River, but I recovered it with a tracker. When I looked at the video in slow motion, I could see it came straight out of the tower, but the bottom started going sideways right after it cleared. At Oktoberfest, I flew the exact same rocket with an identical motor, but with the CG offset corrected with some extra lateral weight. Same rocket, same motor, same launch tower, and actually more crosswind than I had at Hartsel, but it flew beautifully straight. Sadly, I had a transmitter problem so I never recovered it.

So I think the bottom line is that the smaller rockets with record-attempt sized fins are more susceptible because their time constants are so much shorter and the restoring torque is just too low. Too much angle builds up too quickly before the restoring force can compensate. Clusters rockets are usually built with a lot larger fin area and stability margin than altitude attempt birds, so that's probably why even the small ones do o.k.

Well there's no point in discussing this anymore with someone that is 100% convinced. But for the rest of you, remember those assumptions that are used to calculate the Center of Pressure? I looked them over last night and I won't list them here for brevity, but I recommend you review those from time to time because if you violate the assumptions that the equations were based upon, you cannot rely on the answer for the location of the Center of Pressure.

And if you are designing a high performance rocket and have ~1 caliper stability margin (based on the Barrowman method), you shouldn't be surprised if it does go unstable. It might fly straight but if it doesn't, it won't be because you had 1 gram off center by 1 mm but rather it is much more likely that you did violate one (or more) of the assumptions that those calculations were based upon. I can't imagine trying to mount the altimeter and battery or the recovery method to within 1 mil position in order to precisely get it's center of mass in the center of the rocket. All rockets likely have their center of mass off from the centerline of the rocket even if they are not carrying camera payloads.

Doug

Regarding medusas and off-axis mass... to me, there is one pretty compelling piece of evidence that suggests the multi-throat nozzle is to blame. Loki K350s have been flown *very* straight by several of us. That motor has an off-centered core. While the K350 core doesn't extend to the exterior of the grain like a traditional c-slot, the core is considerably larger than a c-slot, and it is quite a long ways from being placed in the center. That motor uses a single-throat nozzle.

My Loki K350 shot in July was, if I say so myself, incredibly straight. That was in "benchmarK", which has done some skywriting on L330s and K250s. The only common denominator to the less-than-straight flights was the medusa.

Given the cost of motors like K250s and L330s, I've seen enough to not want another one of either 😯 I will probably fly another K350 next year in a minimum diameter rocket. Great motor!