So, I exceeded the limits of a 3D printed fin can

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
With the Markforged (if it is the Mark Two) solid strands of carbon fiber (or fiberglass or kevlar) and be inlaid into the print to increase the rigidity and give it a solid core to withstand and fluttering since nylon is more flexible than pet. However the Onyx material on its own may survive landing better than if it was stiffened with CF.
 
With the Markforged (if it is the Mark Two) solid strands of carbon fiber (or fiberglass or kevlar) and be inlaid into the print to increase the rigidity and give it a solid core to withstand and fluttering since nylon is more flexible than pet. However the Onyx material on its own may survive landing better than if it was stiffened with CF.

Print a rectangular void inside each fin for a small G10 plate, should be enough to stiffen any fin. Just cut the plate to size, and glue it in.
I don't see the need to be using CF in the filament, except that it's cool to say you did. The lo-tech G10 plate is probably stiffer.

The failure of the fin can above was caused by melting of the thermoplastic more than it not being stiff enough. Even with a stiffened core, it's still going to melt
 
It isn't just chopped carbon fiber in the filament, the machine can lay continuous strands of CF into the part as it is printing it:
1628774611237.png

The first post in the thread showed that it is possible (though maybe not recommended) to build a fairly high performance fin can without this technique, but I'd be curious to hear what the combination of a higher temp plastic like Nylon and the significant boost in stiffness that the fiber brings would allow you to do.
 
With the Markforged (if it is the Mark Two) solid strands of carbon fiber (or fiberglass or kevlar) and be inlaid into the print to increase the rigidity and give it a solid core to withstand and fluttering since nylon is more flexible than pet. However the Onyx material on its own may survive landing better than if it was stiffened with CF.

Printing / part orientation also comes into play. The Markforge will lay in the fibers / strands, but only parallel to the layers printed. So, you're outta luck if you want to print the fin can vertically, but have the strands parallel to / imbedded in the fin's leading edges.
 
Heating primarily and there was likely some flutter.
Exactly, the heating softened the leading edges, which allowed them to flutter. You can literally see the flutter in the edge as it cooled.

It would be interesting to test different treatments to protect the leading edges. It seems it would have to insulate them from heat transfer to avoid damage. Perhaps a carefully applied layer of JB Weld or something similar would work.

One of these days someone needs to instrument a couple of leading edges with fast response thermocouples or something similar to see how hot they are getting. Or even temperature sensitive paint.


Tony
 
That said, I do use printed fins as the core of a composite construction. I've used a layer of fiberglass on the outside surface of small fins printed with a symmetric Von Karman profile on a minimum diameter rocket, and you probably could have used something similar on that fin can.

I've been playing around with much larger fins, where I print half a fin in Von Karman profile split vertically, laying flat on the print bed, and then laminating (with epoxy) two sides of the fins together around a layer of carbon fiber as core, and then a layer of fiberglass on the outside surface.

To take a look at the performance advantage at supersonic Mach numbers of the Von Karman fin airfoil, the Von Karman fin airfoil (I assume symmetric for both the leading and trailing edges) can be nicely approximated by a selecting the Biconvex fin airfoil input in RASAero II.

1628798161018.png

The rocket supersonic CD can then be compared to the supersonic CD with the traditional Hexagonal airfoil.


Charles E. (Chuck) Rogers
Rogers Aeroscience
 
To take a look at the performance advantage at supersonic Mach numbers of the Von Karman fin airfoil, the Von Karman fin airfoil (I assume symmetric for both the leading and trailing edges) can be nicely approximated by a selecting the Biconvex fin airfoil input in RASAero II.

View attachment 477116

The rocket supersonic CD can then be compared to the supersonic CD with the traditional Hexagonal airfoil.


Charles E. (Chuck) Rogers
Rogers Aeroscience

I have a fin can I designed like this. Yet to be flown.
 
To take a look at the performance advantage at supersonic Mach numbers of the Von Karman fin airfoil, the Von Karman fin airfoil (I assume symmetric for both the leading and trailing edges) can be nicely approximated by a selecting the Biconvex fin airfoil input in RASAero II.

View attachment 477116

The rocket supersonic CD can then be compared to the supersonic CD with the traditional Hexagonal airfoil.


Charles E. (Chuck) Rogers
Rogers Aeroscience

I flew the fin can and it flew great with a G motor. Now to try an H.
 
Here is some data.

The RASAero II simulation predicted going above Mach 1 from about 1.3 to 9.2 seconds with a peak around Mach 2. Max altitude was predicted to be 23k feet.

1631988252097.png

Here is the barometric altitude measured onboard. The melted fins really killed the altitude to just 17k feet.

1631988431546.png

Zooming into the first few seconds, there are two kinks in the baro data. The kink at 1.4 s agrees with the prediction of hitting Mach 1 and a pressure change.

1631988527099.png

Not sure what is happening with the kink at 7.2 s. Is this the rocket again passing through mach, but at an earlier time than the simulation because the fins are melting and losing speed/altitude?
 
Last edited:
Here is some data.

The RASAero II simulation predicted going above Mach 1 from about 1.2 to 9.1 seconds with a peak around Mach 2. Max altitude was predicted to be 23k feet.

View attachment 482427

Here is the barometric altitude measured onboard. The melted fins really killed the altitude to just 17k feet.

View attachment 482428

Zooming into the first few seconds, there are two kinks in the baro data. The kink at 1.2 s agrees with the prediction of hitting Mach 1 and a pressure change.

View attachment 482429

Not sure what is happening with the kink at 7.1 s. Is this the rocket again passing through mach, but at an earlier time than the simulation because the fins are melting and losing speed/altitude?
Pretty solid guess.
 
Can't you look at the velocity curve to see if it was passing mach at that time?

Well, no. Not with this baro data. Velocity derived from baro altitude is usually suspect, and it is totally meaningless in this case with the discontinuities in the altitude curve.
 
Last edited:
Well, no. Not with this baro data. Velocity derived from baro altitude is usually suspect, and it is totally meaningless in this case with the discontinuities in the altitude curve.
Oh, OK. I've no experience with Mach data. I just assumed you could smooth the data enough around the glitches to get a good guess.
 
Here is some data.

The RASAero II simulation predicted going above Mach 1 from about 1.3 to 9.2 seconds with a peak around Mach 2. Max altitude was predicted to be 23k feet.

View attachment 482427

Here is the barometric altitude measured onboard. The melted fins really killed the altitude to just 17k feet.

.........

Zooming into the first few seconds, there are two kinks in the baro data. The kink at 1.4 s agrees with the prediction of hitting Mach 1 and a pressure change.

View attachment 482429

Not sure what is happening with the kink at 7.2 s. Is this the rocket again passing through mach, but at an earlier time than the simulation because the fins are melting and losing speed/altitude?

I'd add rail guides to the rocket, or make the existing rail guides larger, and increase the rail guide size until the RASAero II predicted altitude matches the 17,000 ft actual altitude. You'll be increasing the drag of the rocket to match the increased drag with the damaged fins. Then take a look at the RASAero II velocity (and Mach number) versus time, and compare them to the baro plot. Could be interesting.


Charles E. (Chuck) Rogers
Rogers Aeroscience
 
I launched a 54mm min dia rocket with a 3d printed fin can by Landru several years ago, using an I65 motor. Right from the start, I was in trouble. Perhaps the rocket sat too close to the blast plate, because the rocket immediately began to corkscrew and the result was not good. Upon retrieving the rocket, I noticed the fins were melted and resulted in what looked like a boat motor prop. Two fins bent in one direction and the third bent in the opposite direction.
 
I launched a 54mm min dia rocket with a 3d printed fin can by Landru several years ago, using an I65 motor. Right from the start, I was in trouble. Perhaps the rocket sat too close to the blast plate, because the rocket immediately began to corkscrew and the result was not good. Upon retrieving the rocket, I noticed the fins were melted and resulted in what looked like a boat motor prop. Two fins bent in one direction and the third bent in the opposite direction.

That i65 does not have much oomph. Yeah the fins were probably getting cooked trying to get off the pad.
 
It will likely be a little stiffer than with other fillers, and I think the flutter might be less than expected at first thought. The softened matrix would have higher hysteresis and probably not resonate so badly.
 
I'm thinking about it. I don't sell very many so it may be an option. Half the sauce is getting the print settings dialed for a good surface finish and lightweight finished product with good bonding...

There is also carbon fiber PEKK or fiberglass PEEK which holds up to 260C. You can even anneal FG PEEK at 140C to go up to 315C (you might have to not sure). It also has way higher tensile strength than CF PEKK. But you need a serious 3D printer that can extrude at 375-410C with a bed temp of 110-150C
 
Is plain-old-boring ABS an option? There is a feeling out there that PETG for temperature resistance is not that great in real-world situations, tho it certainly is superior to PLA according to my basic testing.
 
Now that I have the fin can in hand:

Post mortem:

The 3d printing process creates a small air "gap" right behind the leading edge of the fins. This is because the 3d printer uses discreet lines and can't fill in the gap smaller than the line width (at least for the slicer I use)

This means the leading edge is thin AND has very low thermal mass behind it.

In this case, the extended time above Mach softened the material and collapsed the pocket of air.

The tips of the fins sloughed off because there was nothing to support them.

I bet carefully drilling the fins and injecting epoxy into the cavity would have largely prevented the distortion.


1639359153498.png


Melted fincan.jpgPXL_20211202_031108773.jpg
 
Is plain-old-boring ABS an option? There is a feeling out there that PETG for temperature resistance is not that great in real-world situations, tho it certainly is superior to PLA according to my basic testing.

ABS and PC Blend are better options.
 
For strength, print fins separately, in spiral outline (Vase) mode. Leave root open, print with top 3 layers solid with 100% infill. Fill the cavity in each fin with thin epoxy. (System Three, West Systems, Tap Plastics, etc.) You can also add a piece of 0.030" fiberglass sheet inside for more strength. Utilize the 3d printing as a nice surface shape. Not strong enough by itself for high velocity applications. Use PLA+ (eSun makes the best) It has better heat resistance than regular PLA and is much tougher than PETG. Also has far superior layer adhesion. run it a bit hot, 215-220C


Nike Fin example: TOP layer shown
1639416302233.png

From build plate up the to last three layers...

1639416320274.png
 
If you use PC Blend or Pc Blend with CF, you can print them with a 20% infill, save weight, and it it very strong.
 
Seeing as how you got the rocket back in one piece instead of confetti, I'd say it was a successful flight.

Now I wouldn't want to fly that rocket again, that would be pushing your luck.
 
Seeing as how you got the rocket back in one piece instead of confetti, I'd say it was a successful flight.

Now I wouldn't want to fly that rocket again, that would be pushing your luck.

I agree. I am claiming some success! Sounds like just a little more reinforcement of some kind will make that fin can work on similar extreme flights.

I didn't fly it again, and the fin can has been donated to the Additive Aerospace Museum of 3D Printed Coolness.

I did rebuild the Wildman Mach2 with the stock fins.
 
Back
Top