Winter Build 2018/19: Hangar 9 P-47D Razorback

[willsonman]

After getting lots of input and watching a couple of YT videos, the thought occurred to me that the key component to making these sound systems work is surface area. An I'm not talking about just any amount of area. The area has to be uninhibited by structural support. This is why my test on the fuselage itself was not working well. The stringers and formers make the balsa skin more rigid and I need a loose surface area to sympathetically vibrate. I mounted a transducer to a spare sheet of 1/16" balsa and the sound was amazing. Just amazing. With all that in mind, along with trying to reduce hard corners and a more ideal mounting position of the transducer, I built a 1/16" balsa box.


This box sounded so much better than the foam with a deeper sound that was richer but there was a fair amount of distortion. This indicated to me that there was over flexing of the balsa. With that in mind, I proceeded to cut a hole in the top of the motor box to permanently affix the sound box.


I sized and mounted the box in such a way that all the sides would not come in contact with anything.


The overall fitting is perfect and I tried to mount both transducers but it did not yield any additional volume. This proved that my idea of increasing sympathetic vibration area makes these more efficient. There was still a bit of ringing of the highs but with a tiny bit of fiberfill (pillow fill) that instantly went away without reducing volume.

 

[jaredstrees]

Amazing ideas you come up with!

 

[Mid7night]

Awesome!

Can you mount it with the hole facing down? I think that would give you better sound in the air on flybys.

 

[PsyBorg]

Wouldnt his pillow fluff get pushed out.

Actually it looks like it has to go in pointing up as there us no clearence for the inducer above the motor. I thought the same thing the first time I saw it.

After the head on shot in the test video you can see much more clearly.

 

[willsonman]

I had considered having the hole pointing down but that would make the base as another rigid point. Also, the inability to round out a portion of the box would inhibit the increase of surface area and more optimal geometry. In terms of sound projection, when the cowl is in place, there will be additional resonance within the cowl that will project and distribute sound there. It is more along the lines of sound reflection and reverberation. After I install all of the additional details in the fuselage I hope to see if I can create an additional transducer elsewhere. Really it comes down to the room available. We'll see.

 

[willsonman]

As mentioned previously, I had to re-build the oil cooler doors. I just could not remove the soda can paint using a chemical I had on-hand. A bit of 220-grit paper removed it just fine. With the new doors made, I constructed both parts.


Now, before I could install these parts, I had to coat the parts with foil. The intricate work would be nearly impossible to do with it installed. I applied pieces of the foil using a bamboo BBQ skewer and it turned out quite well. I was also impressed with how well I was able to smooth the curves of the waste gate exhaust exit hole. I did not fully cover the part since I will have to blend the edges with spackle and sand.


I dry fit the part and it looks great. I tacked it in place with thin CA to soak into the wood and the ABS part. Once dry, I applied some spackle and let it cure out. A bit of sanding and we are close to a final finish. A bit more sanding and I'll saturate it with WBPU. Again, the entire fuselage will get a coat of epoxy to harden things up to apply the foil.

 

[TooJung2Die]

Your attention to detail is crazy. I'm impressed.

 

[willsonman]

Thanks! I always love the details and I feel like it is one of my strengths that sets my builds apart from the other fantastic builds here.

 

[willsonman]

I managed to get the servo installed on the first OCWG last night. The first one is always the toughest as you have to sort of "wing it" as you go. I needed a balsa spacer as the linkage needed a specific clearance against the side interior wall to function properly. Getting glue on the balsa spacer was tricky given the internal structure but I managed by putting glue on the tip of a BBQ skewer and place it where needed. Having learned all of this the other side will be able to be installed much easier.

I posted a video of this on FB last night but I'll include it in my next YT update.

 

[willsonman]

I got the other OCWG installed last night and it works too. I told you the second one would go faster. I've included a couple of pictures to describe the internal mounting and mechanics.


I got spackle on the second OCWG and did a little cleanup on the first one.


Very pleased with how this turned out. Moving on to the intercooler doors. Some geometry to figure out but I have a good idea how to attack this.

 

[willsonman]

The intercooler doors on the full-scale airplane do not simply pivot in and out. When closed, the door is a single rectangular panel. When it opens, it first slides forward while recessing into the fuselage and then pivots outward as it moves. To emulate this, I had to derive the geometry of a sliding track. This will be 3D printed to make construction easier. I prototyped a track and used PLA as a quick print mock-up with a piece of balsa to simulate the size and position of the door.

The first attempt shows the door at the closed and then open positions. The open position revealed that the initial recessing of the leading edge of the door is good and the final position is bad. The door moves too far and the angle is wrong for the door swinging out. To correct the swing out a bit, I change the pin at the aft side and got a far better position and noted how far off I was for the final position.


I re-designed the track using these reference points. You can note in the comparison that the aft part of the track is further forward to account for the change in position. Also note the shortened forward curve. I'm extremely happy with this movement.

 

[willsonman]

With the hard part out of the way, I drew up a full setup that I can simply glue the door to. The left side is the aft position and the right is forward.


I designed the parts to key together. The sides have a groove for the central ducting plate. As these key together, they should give strength. I'll also note that in my experience of keying printed parts, you need to offset the slots by 0.2mm to account of variation in printing. It will be snug but material should not have to be removed or minimal sanding involved.


The door mounting bracket has exact holes that metal pins can be glued into, as well as a rounded cross member. The cross member provides and exact distance but will also allow a cleaner look of any gap that may exist at the leading edge of the door. This may be necessary to clear the sheeting of the fuselage.


Lastly, the aft brace and duct sheet. This will be revealed as the door opens and appropriately direct air out of the fuselage. Notches were designed to allow the carrier to slip into them and to make sure they do not bind with the carrier, the holes were offset 0.2mm on all faces, giving 0.4mm clearance on the sides. This part will be tricky to print but I think with proper supports, it will be fine. Smoothing will be pretty easy in post-print operations as these doors are 7 x 5.6cm.

 

[willsonman]

I've been sick the past few days but I slowly made some progress. I got some of the parts printed and made additional modifications for fitting and functionality. First assembly revealed a couple of flaws that have been addressed in Fusion. To start a mock up, I had to think ahead about covering parts in foil. Of course the printed parts need smoothing with sanding/acetone. You can see that the results speak for themselves. While not perfect just yet, I'm still learning my technique.


With the parts covered the way that was needed, I assembled the entire mechanism with a door affixed. I cut the door out using my razor saw. Some re-routing of guide tubes for the tail surfaces is required but I'll get to that later. I need to recess the print more to get the door to sit flush but we are on the right track here.


And checking that the door in the open position protrudes, but not too far...


The open look is perfect and being able to clearly see down the duct is a fun touch. The reflections from the foil obscure the view slightly but I think this feature adds the depth and texture I'm after.

 

[willsonman]

Been many days since my last post. While not making much tangible progress I have had to make a lot of fine adjustments and iterations of all the parts for the intercooler door mechanism. The door itself was not entirely flush so I have modified things to make the door more flush. It fits perfectly now. I also opened up the slot for the sliding mechanism so that it does not bind as much. I removed some portions of the prints to reduce weight as well and kept some parts to make printing easier. After getting functionality and fitting right, I tried different linkage positions and I modeled a horn that will print with the door mounting bracket. I hope to get final prints done in the coming days and have a much better progress report.

This has been a real example of how prototyping goes. While the CAD aspect provides a level of precision, There are iterations that have to be taken to ensure the success of others. I want this print to be right so I've had to make, adjust, re-make, adjust, etc. to make the print files available. I think this will be one mod you could easily do to a stock model and even keep the vinyl covering. While taking quite a bit more time than anticipated, I have to say that I think this is an important detail that I think should not be neglected the way H9 has.

 

[willsonman]

 

[willsonman]

Now let's go over how to get the thing inside the airplane.

The first order of business is to install some servos. Two rails installed just aft of the wing spar tube.


The door panel itself is 7.0x5.5mm. I measured location based on my 3-view printout. The installation for my location is pretty general but outlines some challenges you have to accept BEFORE committing to do this. First... make sure these two square holes are on exactly the same location on each side. It would look pretty bad if they weren't.


OK, The next two shots are key to go back and forth making comparisons. You will first notice that one door has already been installed, pictured on the right of each photo. Please also note the internal structure differences on the first and second images. Portions of the structure have to be removed, reinforced, or re-located, as part of this process. Note that the guide tube needs to retain its ply locator but the door mechanism conflicts with its path slightly. Also note the piece of ply added to the left side This is done with thin CA before the razor saw goes in. Some cuts are made inside the fuselage, very carefully, and others are made through the hole in the fuselage. No matter how you do it, be deliberate but know it will be OK. Also note that some structure is removed by simply breaking it free from the balsa sheeting. A necessity to allow the mechanism to fit.



Installation of the part is simple. I highly suggest beveling the balsa on the leading edge of the square. Bevel with the bias on the INSIDE of the fuselage, so that the door will have additional area to clear and keep the gap to a minimum.

Hold the door in the closed position and position the door assembly in the opening. This requires one arm inside the fuselage through the hatch area. You can adjust the opening motion of the door slightly to clear the opening with minimal trim of the hole. I ended up trimming an additional 2 mm of wood and re-beveled the edge. The panel sits flush when fully closed, and clears the fuselage when beginning to open. Hold the mechanism in place and with the door 1/2 open, apply thin CA at the crack between the fuselage skin and the mechanism. Take care not to get CA on any moving parts. Hold in place for at least 30 seconds before removing your hand. Once additional glue is added as you move the door... flip the fuselage over and glue the other side in like manner. The minimal weight and vibration should allow this to be the minimal amount of mounting required for this part.

Last to pics are for comparison as well. They depict the fully open and fully closed positions. Not the the push rods are bent slightly to allow the two rods to cross over without touching at all. Also note the final location of the push rod guide tube that was cut free previously. There will be a little additional rubbing here but it should not be too prohibitive.

 

[JTarmstr]

This is fantastic. i was always curious how the giant scale warbirds you see at festivals get built. Been so much fun watching this and cant wait to see what other stuff you add!

 

[struggleforlife]

Looking very very nice willson!

As someone who's trying to get into the 3d printing stuff (... which really isn't that easy...), I am wondering if you would be so good as to share your 3d design of the intercooler doors. It's especially interesting, since it's not just design, but also mechanics.

PS. for the doors itself I saw you made a slider-pin all the way through (which shows when you open the door). I was thinking: pushpin... (easy to glew in, doesn't have to go all the way through).

 

[willsonman]

JT: many many more of the things will be added. 😜

Struggle: as stated previously in the build, all of my print files will be posted on Thingiverse for everyone to enjoy... when they are all done. With so many designed parts for this, I’d rather release it as a package rather than single parts.

The point of the pin is a good one. I tried this in a previous prototype but with how flimsy the prints were, there were times when the door carrier would bind with the frame when closing. This was due to inconsistent spacing of the sides where the pin was not all the way through. The option was to print thicker walls or to make the pin go through. The latter was the lighter option.

 

[struggleforlife]

Thanks Willson!

PS. isn't it possible to add a vertical cut-out in the sides of the intercooler-slider, to accomodate for the vertical rib inside the plane?

I am not sure about the dimensions and the depth of the slider vs the rib, but if this is possible, it would be better structure-wise.