Reinforcing fin fillet with CF

[Dane Ronnow]

CF or FG cloth would be better to reinforce the fillets.

This would be a good idea if I was trying to reinforce the fillets. I'm trying to increase the fin thickness at the area where the cracks are occurring. Thicker fin, less flex.

[Edit] Sorry for the confusion on this. I just looked at the title of this thread—'Reinforcing fin fillet with CF'. I guess it should have been 'Reducing fin flex'. (My bad.) Everyone feel free to verbally castigate me.

 

[waltr]

No, just trying to make helpful suggestions.

You can increase fin thickness by laminating cloth onto fin with epoxy. In fact if laminated the fin thickness will increase.
If the fins were not yet on the rocket then laminating with the CF veneer would work well but maybe not helping the fillets. However, it does not work, as you found, trying to bend it around the fillet. You gave it a try so a learning experience.

The old saying in Engineering is "back to the drawing board". Which means lets try something else.

 

[AHansom]

Slightly larger parachute

 

[Dane Ronnow]

No, just trying to make helpful suggestions.

I know, and I apologize if I seemed abrupt. It's just that I feel like I'm running in circles. Earlier, I mentioned simply repairing the cracks with CA, then launching the rocket and see how it does. (The smaller trailing edge on the fin would help, but by how much, I don't know. Flight testing would tell me a lot.) Then I left that idea in the dust and dived back into what to do, what to do.

I'm very familiar with your engineering adage, not because I am one, or anywhere near one, but because I have been back to the drawing board more times than I can count.

I do appreciate your suggestion. It may well be after flight testing this, that reinforcing the fillet with cloth and resin is what I end up doing.

Thanks.

 

[Dane Ronnow]

Slightly larger parachute

The parachute is already oversized to give it a soft landing on the lake bed. OpenRocket simmed it at 10 fps, and FlightSketch confirmed that on the first flights. The descent from 1600 feet was 2 minutes, 45 seconds. From 2500 feet, more than 4 minutes.

You can see how slow the descent is in this clip:

View attachment G74-02 part 3 shortened - no audio.mov

That was in ~2 mph wind. If I launched it in 10 mph, it would probably land in the next county.

 

[AHansom]

The parachute is already oversized to give it a soft landing on the lake bed. OpenRocket simmed it at 10 fps, and FlightSketch confirmed that on the first flights. The descent from 1600 feet was 2 minutes, 45 seconds. From 2500 feet, more than 4 minutes.

You can see how slow the descent is in this clip:

View attachment 519141

That was in ~2 mph wind. If I launched it in 10 mph, it would probably land in the next county.

Then glue more reinforcement everywhere until you need a bigger parachute!!!!

 

[Dane Ronnow]

After going around in circles—literally chasing my tail—throughout this thread, I emailed John Cipolla, asking for help in determining if in fact fin flutter was causing these cracks, rather than ground impact. He took the time to go through this thread, focusing on the NACA TN 4197 spreadsheet I posted with the specs for my fin.

John pointed out, as did @lakeroadster above, that the spreadsheet does not calculate UF for fin shapes such as mine. But, by measuring the semi-span from the center of the root chord to the center of the tip chord—a distance of 5 inches rather than 2.75—he was able to approximate the extra span length of the highly-swept fin on this rocket.

Using that number, John came up with a UF of 323 mph, not the 700-plus mph I was coming up with.

During the email exchange, I mentioned several aspects I thought would point to ground impact, not fin flutter. Among these were the non-uniformity of the cracks, and the fact that they did not all occur on a single flight, but were spread out over four flights. John graciously pointed out (in so many words) that regardless of anything else that might be happening, fin flutter is occurring at this rocket's VMAX when the calculation employs the effective fin span.



Thanks to John's help, I can move forward, confident that fin flutter is the problem. I want to thank him for his quick reply to my email, for investigating my fin problem, and especially for his patience in educating me in the matter.

Armed now with a method that allows the NACA spreadsheet to accurately predict UF, I've recalculated using the new fin shape (tip chord 1.5, not 2.125; semi-span 4.5, not 2.75; and the correct VMAX altitude of 3400', not 600), and came up with a velocity of 362 mph.

With a 25 percent buffer, the velocity drops to 290 mph, 10 mph less than VMAX on a G74, and 120 mph less than VMAX on a G80.

So, I'm going with my original plan of stiffening the fins to strengthen them using .5mm polystyrene sheet, but over a larger area of the fin than illustrated above. This will not only dampen the effect of flutter, but has the added bonus of increasing fin thickness from .158 to .198.

That takes the UF up to 509 mph—407 mph with a 25 percent buffer. And that, friends, is just right.



Now I can quit hyperventilating. (As exhausting as this has been for me, I can't imagine what it must be like for anybody reading this thread. My sincerest apologies.)

I'll post pics when I get going on the repair.

 

[cls]

Excellent investigation work! Thanks for posting the details. I'm interested to read the follow up on how the new fins work.

Yes, whew!

 

[lakeroadster]

Totally different loading scenario, comparing a cantilevered / overhanging fin design, to a cantilevered only design.

Did John actually revise the calculation to reflect the actual unsupported fin section hanging off the end of the rocket, or did he merely make the fin larger, assuming a fully supported fin at the root, which is not the case?

 

[Dane Ronnow]

Yes, whew!

Whew x 2!

 

[Dane Ronnow]

Did John actually revise the calculation to reflect the actual unsupported fin section hanging off the end of the rocket, or did he merely make the fin larger, assuming a fully supported fin at the root, which is not the case?

Referencing the NACA spreadsheet in post #16, he changed only the fin span, measuring from the center of the root chord to the center of the tip chord, providing what he called an approximation of effective fin span—effective as opposed to actual.

Below is the progression of the fin calculations, beginning with the Original Fin calculation in post #16 (which John referenced for his approximation), the Original Fin diagramming John's approximation (top calculation in post #37), and the New Fin using John's method of approximation (bottom calculation in post #37).

 

[Dane Ronnow]

Totally different loading scenario, comparing a cantilevered / overhanging fin design, to a cantilevered only design.

As much as I like the 'cantilevered only' design, it drops my stability from 1.41, to 1.03. It also places the force of ground impact squarely on the motor nozzle. I don't know whether or not that would weaken the CR's over the long haul.

 

[lakeroadster]

As much as I like the 'cantilevered only' design, it drops my stability from 1.41, to 1.03. It also places the force of ground impact squarely on the motor nozzle. I don't know whether or not that would weaken the CR's over the long haul.

The motor mount is the strongest part of the rocket. Compare the thrust loading (due to motor thrust) to the assumed loading if the rocket lands on the motor extension. No worries.

 

[Dane Ronnow]

The motor mount is the strongest part of the rocket. Compare the thrust loading (due to motor thrust) to the assumed loading if the rocket lands on the motor extension. No worries.

And the reduced stability?

 

[lakeroadster]

And the reduced stability?

Swing test the rocket to check stability.

I'm not a fan of fragile fin cans..

 

[Dane Ronnow]

Swing test the rocket to check stability.

The problem with that is I have to chop the fins, then test. There's no turning back if the rocket flies backward or end over end. I'd have to replace the fins completely if I ever want to launch this rocket again.

I'm leaning toward your idea of the cantilevered fin (no overhang), but testing it with OpenRocket simulations, specifically, the Stability vs Time plot, with a fourth axis for Pitch Rate.

In order to get data I can use, I need to establish a benchmark with an actual flight, looking at the smoke trail. It's subjective, I know, but also easy to see what's going on with white smoke against a dark blue sky.

Below are four stills from the ascent of the rocket on a G74, with the original fins. Conditions on the lake bed were 2 mph wind, 90 degrees temp, 2800 feet ASL.



A little bit of waver, but overall, pretty straight.

Here is the OR simulation plot for the flight with the same conditions plugged in:



The first deviation in pitch is ~9.25 degrees. OR's margin of stability is 1.72 cals. This will be my benchmark for comparing other fin shapes.

Next is the fin in its present state, which I've been referring to as 'new fin'. Same motor and launch conditions. This is the plot:



First deviation in pitch is down to ~8.5, with CG and CP locations still looking good. OR's margin has dropped to 1.6 cals.

Now for the cantilevered shape @lakeroadster suggested. I'm calling it 'minimum fin', mainly because it can't get any smaller. Same motor, etc.



The first deviation in pitch has dropped to 6, which is interesting, because OR's margin is down to 1.19 cals. And this has caused me to rethink my approach to margins of stability.

From the start of this build, I knew the rule of thumb is 'between 1 and 2 calibers'. There are exceptions (long, skinny rockets and short fat ones). I settled on 1.5. No particular reason.

I know also that the higher the margin is, the more likely it is that the rocket will weathercock. And I think that's what I'm seeing in these plots—straighter ascents when the margins are lower.

But how low can I go? The G74 is the smaller of the two motors I've launched this on. With the G80 loaded, OR's margin drops to .93. Six hours ago, I would have run from that screaming. Then I plotted it:



Initial pitch has dropped to ~5.6. CG and CP locations still look good.

So, this rocket, evidently, does all right with a sub-1-caliber margin of stability, at least as far as I can see from these simulations. Does anybody see flaws in my methods or reasoning? Because, in the interest of reducing flutter, I'm ready to cut these fins to the minimum shape, and get this repair out of the way.

And speaking of flutter, I plugged the 'minimum fin' numbers into the NACA spreadsheet using John Cipolla's approach to semi-span, and came up with a UF of 453 mph. It doesn't give me much of a safety margin. But it doesn't require fin lamination to increase thickness either. And that is good news to me.



Anyway, sorry for the long-winded blah, blah, blah, but I do that when I get down into the weeds. I'd appreciate any feedback I can get. Thanks much.

 

[lakeroadster]

The problem with that is I have to chop the fins, then test. There's no turning back if the rocket flies backward or end over end. I'd have to replace the fins completely if I ever want to launch this rocket again.

No turning back? Just add nose weight. I guess I just don't get all the drama.

 

[Dane Ronnow]

I guess I just don't get all the drama.

Yeah, I don't either. It happens when I try and figure things out. Maybe Barrowman or Stine could explain it.

 

[lakeroadster]

Yeah, I don't either. It happens when I try and figure things out. Maybe Barrowman or Stine could explain it.

Don't get me wrong... passion for the sport is good. And figuring things out is good.

Just don't sweat the small stuff.

 

[Dane Ronnow]

Just don't sweat the small stuff.

Trying not to.

Sweep all the words and pictures aside (and the emotion). The point I was trying to make was this:

Despite a) losing almost 40 percent of the original fin area, b) a semi-span that is less than one body tube diameter, and c) a stability margin of .93—each of which would cause concern when it comes to 'rules of thumb'—the last plot above tells me the rocket will fly fine.

Or am I misunderstanding what that plot shows? That's all I care about.

 

[lakeroadster]

Trying not to.

Sweep all the words and pictures aside (and the emotion). The point I was trying to make was this:

Despite a) losing almost 40 percent of the original fin area, b) a semi-span that is less than one body tube diameter, and c) a stability margin of .93—each of which would cause concern when it comes to 'rules of thumb'—the last plot above tells me the rocket will fly fine.

Or am I misunderstanding what that plot shows? That's all I care about.

Can you attached the .ork file so we can take a look at the sim?

 

[Dane Ronnow]

Can you attached the .ork file so we can take a look at the sim?

Here you go.

 

[lakeroadster]

Here you go.

Looks good to me Dane. Nice model and sim.

I did a data export, massaged the format and attached it below also. I enjoy looking at the data.

 

[Dane Ronnow]

Looks good to me Dane. Nice model and sim.

I did a data export, massaged the format and attached it below also. I enjoy looking at the data.

You, sir, are a gentleman and a scholar. How did you get the sim data into that format?

 

[lakeroadster]

You, sir, are a gentleman and a scholar. How did you get the sim data into that format?

Awe Shucks...

Flight Simulations - Plot / Export - Export Data

Then you just pull up the file and manipulate the spreadsheet.

Important Note: Be sure to save the file as an "Excel Workbook"

 

[Dane Ronnow]

Flight Simulations - Plot / Export - Export Data

Then you just pull up the file and manipulate the spreadsheet.

Important Note: Be sure to save the file as an "Excel Workbook"

Got it. Thanks again!

 

[ThreeJsDad]

I read many of the posts and if I have missed an idea or gotten an incorrect impression feel free to correct me, it's the way I learn best.

I am a composites fabricator and work in Aerospace type stuff but your mileage may vary.

I believe the flutter you are seeing is being caused by two things.

The first and glaring issue I see is the fin shape. There is a going to be a very strong tip vortex with that fin shape. The dark side of this is that is also where some of the stability is coming from. With our models drag is a large component of stability.

The second issue is something you cannot fix and would never have known to look for. When the find were cut out for the kit very little if any attention was probably paid to the attention of the grain of the G-10. Yup it has a grain and to reduce flutter/increase stiffness the grain should be at 90deg to the longest span. The vast majority of folks run the grain 0-90 to the BT with the intention of making the fins more resistant to bumps and dings from hangar rash and landings.

You may not like the look this will create but if you used a template and rounded the bottom of the fins to something similar to a Spitfire or Hurricane wing tip you should see a significant reduction in flutter. Since this will clean up the aero of the model you will loose a slight bit of stability created by drag but pick up some from the increase in speed.

 

[Dane Ronnow]

The first and glaring issue I see is the fin shape. There is a going to be a very strong tip vortex with that fin shape. The dark side of this is that is also where some of the stability is coming from. With our models drag is a large component of stability.

I assume you're referring to the 'original' fin shape, not the 'new' or 'minimum' shapes discussed above.

The second issue is something you cannot fix and would never have known to look for. When the find were cut out for the kit very little if any attention was probably paid to the attention of the grain of the G-10.

Actually, these fins are .125-inch balsa, with the grain running parallel to the leading edge. (I pay very close attention to things like this.) The fins were papered with 50 lb. Avery label paper.

You may not like the look this will create but if you used a template and rounded the bottom of the fins to something similar to a Spitfire or Hurricane wing tip you should see a significant reduction in flutter. Since this will clean up the aero of the model you will loose a slight bit of stability created by drag but pick up some from the increase in speed.

I'm not sure how much of the material above you've read, so I'll recap the most pertinent points.

Below (l to r) are 1) the 'original' fin shape, 2) the flutter calculation for that fin using John Cipolla's approximation of the effective semi-span, 3) the 'minimum' fin shape (the latest iteration), and 4) the flutter calculation for the 'minimum' fin, also using Cipolla's method.



Cipolla's method of approximating semi-span showed me the UF for the original fin shape is 323 mph, which is 90 mph below the rocket's VMAX on a G80.

Using the same method of approximation, the UF for the minimum fin has increased to 453 mph, which is well above VMAX on the G80. The safety margin is only 10 percent, but I can certainly live with that.

As far as stability is concerned, the stability in OpenRocket dropped from 1.47 calibers on the original fin (with a G80), to .967 on the minimum fin (G80). At the same time, though, the pitch rate was cut nearly in half. Pitch rate is what I care about, not calibers of stability.

At this point, I'm confident that the minimum fin shape will resist flutter far better than the original fin, and perhaps altogether. On the other hand, if you see a problem with these methods and calculations, I'd be interested in hearing about it.

Thanks for the post, and for the food for thought.

 

[lakeroadster]

@ThreeJsDad's idea looks easy to implement... and looks good too!

 

[Dane Ronnow]

@ThreeJsDad's idea looks easy to implement... and looks good too!

When he mentioned shaping the trailing edge into an ellipse, I think he was talking about the original fin shape. I changed the fin in OR from the minimum fin (center, in your illustration), to the ellipse (right), and I see no change between the two in stability or pitch rate. No change in apogee or VMAX either. [Edit] Aerodynamically, they strike me as essentially the same thing. I can't address the issue of drag. But I can say stability, pitch, apogee and VMAX are the same.[/Edit]

Personally, I like the hard angles of the center shape in your illustration, although the actual shape doesn't have as steep an angle on the outer end as yours does (see the minimum fin in post #58).