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

[struggleforlife]

@Willson:
I have been contemplating a 'nice' construction for functional cowl flaps. As a (easier) start, I did some drawing for the corsair. See this fusion file. https://a360.co/2EWcyp9

The idea is to have a 'ring' with cowl flaps attached, that can be glued into the cowl (after removing the flaps from the cowl). The only additional thing needed is :
- a steel wire to attach the flaps to the ring and a bunch of springs to have the flaps extend as default
-a steel wire through the flaps (attached to a servo) to retract the flaps.

I am hoping that this explains the design I came up with.

Would this work? Would this be do-able for the p47 also? (meaning: the p47 has a different cowl shape, so the design should be adjusted).

 

[willsonman]

Actually no. If you look at how the slide mechanism is situated, the door slides right into the former about 2mm from the skin on the inside. It just had to be removed.

 

[willsonman]

@Willson:
I have been contemplating a 'nice' construction for functional cowl flaps. As a (easier) start, I did some drawing for the corsair. See this fusion file. https://a360.co/2EWcyp9

The idea is to have a 'ring' with cowl flaps attached, that can be glued into the cowl (after removing the flaps from the cowl). The only additional thing needed is :
- a steel wire to attach the flaps to the ring and a bunch of springs to have the flaps extend as default
-a steel wire (attached to a servo) to retract the flaps.

I am hoping that this explains the design I came up with.

Would this work? Would this be do-able for the p47 also? (meaning: the p47 has a different cowl shape, so the design should be adjusted).

This is exactly how I did the cowl flaps on my Corsair. Do a search for my top Flite build and you cannot miss it. No spring is needed and I would suggest using two servos the way I did. Using one will create a ripple effect as the circumference expands.

 

[struggleforlife]

I found it. Nice job! Just as i intended, sort of.
Did you print that? Or just cut it from nylon material?

 

[willsonman]

As posted in that build, nothing was printed for that.

 

[willsonman]

I was able to draw up and print a part that was very much needed for this airplane. The wires from the wing go into the fuselage well below the plate where the Rx mounts. This is problematic as you have to fish out the servo wires every time you assemble the airplane.


A single prototype was needed for me to correct issues but after another quick couple of prints I was able to install them. Rough up the flanged plate with sandpaper and some medium CA to adhere it in place and you have a nice tube you can route the servo wire up through to a location that is much more suitable. No fishing involved.


I moved onto the tail gear area. I got the strut modified for my retract. You can see where I had to rim additional wood to allow the strut to more easily clear. This looks to be a really bad binding spot that could cause issues. The dremel made quick work of it.

 

[willsonman]

The tail gear installation was finalized. I disassembled the retract unit to apply some additional high-quality grease. These buggers have a bad habit of getting dirt and dust in them and binding up. From the picture, you can also see that the steering control mechanism got some of this grease as well. Rest-assured, there will be an additional precaution to keep dirt and debris out of this area. You can also see that I sanded away additional material inside the fuselage that could also have been problematic for binding.


Moving along, I tried to mark up gear doors as carefully as possible on the cover for the fixed gear. As this was smooth I decided to use it as a mould. I measured and applied the same markings on the retract cover that is ply and ABS plastic. I'll cut this part away and instert fiberglass doors.


The glass layup is simple. Base layer of plastic kitchen wrap to protect the part. Tape it so there are no wrinkles. Two layers of 6oz cloth (hardware store type) and then a layer of 3/4oz to finish it off. You know it has enough epoxy when you can clearly see you marks through everything. I had some left over epoxy so I applied it to some control surfaces.


The part pops off easy from the kitchen wrap. The part was trimmed, but not to fit, after tracing the cut lines onto the dry part. Some additional measuring showed errors but no big deal. This is why I did not trim to fit. Some additional work getting the part right and we have final. Some pin hinges and cut the part in half are all that is needed.

 

[struggleforlife]

You are a world class builder, for sure. Very, very nice work.

As for the bay doors: i've been looking for some nice spring loaded hinges (as to open the doors by default), but cannot source any. How do you go about that?

 

[willsonman]

There are many ways to do so. Some use rods affixed to a separate control horn, adjusting spring tension by a collet at the anchor point inside the fuselage.

I've frequently use the torsion spring method. It can be finicky but is dead simple. Basically, use a thin steel wire, like a guitar string, and then is bent on each end. The bent ends are affixed to the fuselage and the door. How long the wire is determines the spring strength as well as the angle at which the wire is bent. You basically want it bent slightly over the fully open point. As the door closes, the wire is twisted, thus the torsion part. It works quite well, but like I said it can be finicky.

 

[willsonman]

I got the hinges installed first and then positioned the fiberglass part to mark where to cut a recess for them. After the recesses were cut I then parted the two halves and glued them to the hinges. I used medium CA here as it bonds well to FG and the ABS part. The curve in the doors means there is a potential for a bit of gap so thin CA would potentially be problematic.


From here, it all went down hill. I installed a bracket to make a mechanical arm for each door that would be fixed to the gear strut. The problem with this that the strut is not fixed and would move the doors whenever I would need to steer. The second issue was that there was so little room to do this that if I used appropriately-sized parts for strength it would be too bulky for the area and exceptionally overweight. Having spent a couple hours on this in frustration (gear doors typically do this to every modeler), I set off to the LHS to get some fresh perspective and ideas. I thought back to @struggleforlife and his post about sprung doors and considered both of the options I mentioned. A torsion spring would likely be too weak for this airplane. @wilmracer has this airplane is she is no slouch in the speed department. My fear is that the potential speeds these doors would see would be problematic for a torsion spring so I opted for a sprung rod solution.

I picked up some smaller collets that I knew I did not have while at the LHS and popped over to the $store for some sprung click pens. A single spring was first used with a collet but then I quickly realized I needed a longer spring for the amount of compression I needed. The collet was removed and the spring going all the way up to the connector solved the issue. The arm is a spare I robbed from a crashed airplane and trimmed down and rounded the base to fit the door. I drilled a hole through the ply plate for the retract mount and while the bit was in there, I moved the drill side-to-side as this would be needed to prevent binding. I extended the rods far enough to clear this plate by 1cm with the doors in the open position. This solution is very light and holds the doors firmly open. The thing FG doors need to have a bit of carbon tow added for stiffness, as they bulge a bit when closed but that is no big deal. Some tow and thin CA will do the trick. A simple wire will extend from the arms and will be caught by the retract. At this location close to the pivot of the retract, it should have enough strength to pull the doors closed. If not I'll need to source some slightly weaker springs.

 

[struggleforlife]

Despite the struggle, it looks crazy tidy, again.
There is one small thing, and I hope I am not somehow mistaken, but the p47 gear doors I know have a somewhat different shape. The rear end is a half moon (when looking from the top onto the closed gear doors).
See my (amateur) work below. PS. I made the doors from thin aluminium (my glassing experiment didn't work out).

I still see a market for a decent spring loaded hinge (where the spring doesn't show from the outside). I cannot understand why robart or Hobbyking doesn't jump to that chance. The @Willson way works nice (I didn't think of that), but is rather complex.

 

[willsonman]

Yes, this is true. If you look back at the stock mould for the cover, making the correct shape was problematic. I had considered trying to patch in part of the glass and blend it but that would have been a lot of work. While this model will be highly detailed, I am not going for a true-to-scale in every aspect build here. Its a minor look change that honestly I'm glad its there as a conversation starter for "those in the know."

 

[willsonman]

Made tremendous progress wrapping up the tail. Spent time aligning the gear to retract perfectly. Just minor geometry tweaks in the linkage. I raided a pen drawer in the lab at work for some different springs. Found some that were more gentle and ended up with the collets and it resolved the bulging stated previously. No carbon tow needed. I added a 1/64" ply lip to the leading and trailing edges of the doors to aid in seamless closure. 5mm wide and pressed in place. Add a few drops of thin CA and the wood wicks it right in. I realized that the door geometry was a little off with the hinges so I drilled new holes. I used a steel wire and fire to heat it up to make holes for the closure string, made of spiderwire fishing line. It took a couple of attempts to get the length right but as you can see from the picture, closure is perfect.


With the tail mechanics wrapped up its time to move on. There is an elephant on the airplane that I will address next in order to proceed further with the cowl. So, I'll be moving back to the front end of the fuselage. At the same time, the hatch needs to be addressed as part of this. So, without a reference picture or two, I'll just leave it at that. Once these are addressed I'll call the fuselage done and move onto the wings.

 

[willsonman]

OK, Here we go... down the rabbit hole.

The one thing, from the start, that set me off about this ARF was that something just did not look right in the nose. Back when I installed the cowl I quickly realized what it was and knew it had to be addressed. From the first two images today we can see the real deal compared to the model from Hangar 9. What's there that is wrong!? Well, the slope of the top of the nose. Note that the scale image is flying with cowl flaps closed and the model is full flaps open as part of the cowl mould. From there, move back. The model has this drastic hump that protrudes upward as well as having an abrupt slope off into the cowl. It makes the cowl look puny and there is only a 2mm gap between the top of the hatch and the cowl flap. Contrasting with the sides, there is a 6mm gap. This subtle difference completely changes the look and is just entirely wrong. Even 3-view drawings are not this bad. I have NO idea why H9 did this as the hatch is really what needs to be fixed.


Full disclosure... I failed to account for the cockpit floor when I installed the intercooler door servos. I removed the prior mounts but the CA was pretty stuch on there for one, breaking the former. Some simple ply was cut and the servo mountings were lower to give clearance for the hatch.


Time for carnage! I first cut the balsa skin at the former just in front of the cockpit. I chose this area because I did not want to run into fitting issues with the canopy when it is installed. I can blend and change in slope at this joint as well. From there I removed all the sheeting by breaking it away and then cut away the excess glue. I found it was some sort of aliphatic glue, quite soft, and was easily cut away. I created a template of cardstock from the profile of the firewall on the fuselage. Measured 6mm down and freehand drew the curve to blend it out. I cut the template and applied the curves to the hatch and the firewall. I used my razor saw to cut the two stringers at the sheeting joint and used a #11 blade to poke through glue joints. I carefully removed them as well as the two formers.


I sanded the hump out of the stringers and I was satisfied that I could re-use them. After fitting the stringer with the larger former I realized that I could re-use the larger former but not the smaller. The larger former I was able to cut the slots deeper and sand away the excess material for the new curve in the top. The smaller former had to be created anew. I used the inside curve from the existing former and cut. It was wrong but I learned what size I needed and how far I was off. The second one was great. I just used some scrap ply I had in my leftovers box. With the parts fitting well, I used medium CA to install the formers to the hatch and thin CA for the stringers. The fit is good, the shape is much better and the construction is sound. While there is some shadow in the picture, you can start to see how the hatch is now blending with the cowl, and not the open cowl flaps.


Aside from correcting the shape, the other reason I NEEDED to do this was to cut the cowl flaps to make them functional. I really liked this feature on another model I built (Corsair) and I'm excited to try it again.

 

[struggleforlife]

Wow... drastic measures! You are the energetic type for sure!

You're right about the shape of the model being off.... Although your measures are improving the 'fault', I thing the real pain is in the cowl: it is too small (among other things).
The older H9 p47 model (which I own) is much better imo: you can see a straight, slightly curved line from the cockpit to the front of the plane. No humps.

 

[willsonman]

The thing is... It's not straight. Look back at the first reference picture. There is a curve but not as drastic at the current H9 airplane. Its the angle at which it tapers down from the canopy.

 

[struggleforlife]

A agree, it's not straight, it's a very slight curve. But the model seems to hump that curve at the point where the fuse goes into the cowl. The latter seems to narrow (it looks as if it closes only due to the fact that the cowl flaps are spread).

It's a shame the 'old' (green) cowlings are not sold any more. I think the shape is better. H9 is cutting corners if you ask me (same goes for the missing intercoolers).

 

[F106DeltaDart]

Excellent work on the nose! I knew something seemed off with the stock model, but never put my finger on it. The improvement is definitely dramatic. Struggle, I agree. I keep eying up their Hellcat for a project but it’s the same thing. They seem to have cut detail and features wherever they can.

 

[willsonman]

Yes, the Hellcat does not have flaps... WTH?! Still, I love the Hellcat and would love to build one.

 

[wilmracer]

Yes, the Hellcat does not have flaps... WTH?! Still, I love the Hellcat and would love to build one.

Hellcat would be a fantastic candidate for another B-ARF. I WILL build one some day, either B-ARF or from scratch... but I have no desire to fight with rotating retracts again any time soon. Maybe a 2000mm handlauncher like @nerdnic 's Corsair .