A companion for my XB70, the XF108 Rapier

[Simon Chaddock]

The remarkable XB70 with it Mach 3 speed and six engines is fairly well modelled. Mine is in Depron.

With a 70mm EDF exhausting through 6 scale jet pipes it flies remarkably well.

Light it is an easy hand launch and the ground effect at a grass belly landing is quite remarkable.

Less well known is the mach 3 interceptor fighter that was also designed by North American.

With obvious similarities I think it should fly well too. It certainly looks the part however those big "under fins" could present a belly landing issue.

It was actually cancelled before even one was built as it became obvious that ICBMs rather than bombers would carry the nuclear threat making an interceptor aircraft redundant.

As there is no finished product, unlike for the XB70, meaning there is no "definitive" layout so the available 3 views show significant differences in detail. The result is no matter what I build it will be wrong!

One useful fact is the intake in the above 3 view is big enough to optimally feed an EDF. It will be long duct but it won't need any cheat holes. 😀 

This will be a UK winter build so it is quite likely there will be no particular rush! 😉  

Edited December 9, 2024 by Simon Chaddock

[Simon Chaddock]

I plan on using a pair of 8 blade 40mm EDFs with the airframe made from 3mm XPS foam (B&Q!).

Following my usual build sequence the first task is to design a "stand alone" duct in LW-PLA.

It obviously translates from a rectangular inlet to the 40mm diameter of the EDF body then to the EDF itself without a bell mouth and finally a tail cone "nozzle" with a final  diameter of 90% the FSA.

The first CAD "picture" of the duct. The gap is where the EDF will go.

The inlet and the nozzle are not quite in line so the duct is made up 3 pieces. The middle piece is where the duct shape changes and also has the "offset" between the inlet and exhaust lines. This means it is "handed" to suit a left or right hand duct. The other bit of the duct are common to both.

Printed in single wall LW-PLA it will rely almost entirely on the fuselage formers to prevent it collapsing inward under the inlet reduced pressure.

It makes a long duct but the inlet area is slightly over 1.2 x FSA so should adequately feed the EDF.  With the EDF so far back it will ensure the battery will be well forward ideally in the cockpit area with a removable "canopy" for access.

How to actually build up the fuselage around both ducts at the sane time is a problem yet to be solved. 😉   

    .  

Edited December 13, 2024 by Simon Chaddock

[Simon Chaddock]

Slow progress but a trial assembly of the complete LH duct showing the offset.

It looks like an lot of duct but all the inlet is sized to 1.2 x the FSA. The nozzle is 90% FSA. 

6 sections, all LW-PLA.  The complete duct including the nozzle weighs 17.4g (0.6oz). The 40mm XFly EDF is 40g (1.4oz)

At this stage the EDF and duct can only be functionally tested at low power as the inlet would collapse inward. Full power testing will have to wait until the majority of the duct is supported by the appropriate fuselage formers.  

[Murat Kece 1]

Great bird.. I look forward to its completion Simon.. 

[Simon Chaddock]

This is where things start to get tricky as I have no fuselage cross sections at all to work with, just 3 views. It means much of the local detail of the fuselage shapes has to be "interpreted" or put another way my best guess!

With all the duct sections printed and both EDFs available the basic formers between the ducts can be added to hold them in the correct relation.

Plan and side view.

The complex top and bottom profiles will be added later.

One advantage is the entire nose and cockpit can be built separately and just glued on between the intake ducts.

As the pair of EDFs weigh about 4 times the duct they will only be permanently added when sufficient skin has sufficiently stiffened the duct assembly.

I will be amazed if it all works out!

 

   

[Simon Chaddock]

The duct and basic fuselage structure is complete/

The underside is fully skinned.

It is remarkably light (40.4g) and rigid considering it is still open topped BUT I have come to a serious conclusion, it is too small. 😵

At that size the XF108 with a narrow delta and a very thin wing will have a span of only 24" (610mm) meaning with two powerful EDFs  it will fly FAST. Not ideal for a foam belly lander.

In comparison my 40" straight wing DH Venom uses only one 40 mm EDF and it no slouch.

My intention is to utilise the same sort of construction technique but make the XF108 50% bigger! 

At 36" span it will have over twice the wing area and using the same powerplants and battery it should be a more docile flyer better suited to a hand launch/belly land.

As the ductwork will be more or less the same it will be easier to incorporate it all into the bigger fuselage.

Well that's the plan.      

Edited December 28, 2024 by Simon Chaddock

[Simon Chaddock]

Just as a trial to see what it looks like a 50% bigger inlet against the original.

It looks rather...er...big!

I have created both plan and elevation CAD images of the full length duct.

To ensure the parts are well within the printer volume and to keep individual part printing times reasonable each duct will be made up of 6 parts rather than 4.

Both ducts represents a total of at least 12 hours single wall LW-PLA printing. 😉 

 

Edited January 4 by Simon Chaddock

[Simon Chaddock]

I spent some time over the new year period printing the ducts for the 50% bigger XF108. Afterall the fuselage "box" is built around them.

Apart from the bigger inlet the other major difference is the exhaust nozzle . It looks massive although its internal dimension are the same.

To achieve the required nozzle area for the 40mm EDF the 18 "petals" have been contracted down to slightly more than would have ben used at mach 3 speed on the full size.

Next to start building the fuselage box.

 

[Simon Chaddock]

The fuselage box built in the same way as the first smaller version with the longer 6 section ducts installed.

Note the magnet wire leads from the completely buried EDF run forward towards the nose.

So far so good but the fairing of the box to the twin exhausts is not quite right. Some rather fiddly bits need to be cut away and rebuilt.

Hopefully the wings and nose will be of a simpler and rather more conventional construction.

A view through the exhaust of the 12 blade 40mm fan.

   

[Erfolg]

Simon, it seems that you have moved to commercial fan units?

 

With such long ducts, a couple of cheat holes could/would improve thrust even more? From the I have material cheat heat holes would not be out of place, as close up, it was/is (the remaining example), has many panels etc

 

I have also considered the NA XB70, in Depron. I now have a couple of books on the XB70, I was so enthused.

 

You  certainly do better than myself, as I have abandoned a He 162, as my foam board and 3d printing seems to heavy. Particularly when compared to the commercial foam BAE hawk and hawkish look alikes. 

 

I do fly a Depron Delta, glass covered that flies exceptional well, (although the twin fin and motor flew better still). The original Depron was much better, cutting much crisper and stiffer.

[Simon Chaddock]

Erfolg

For the XF108 commercial units were the only possibility due to the fact it used true turbo jets and thus they are a relatively small diameter. I only use drone motors and multiblade props as fans on airliner types where the turbo fans give a reasonable scale diameter nacelle.

 

My own view on cheat holes is they only work if the inlet duct is aerodynamically inefficient meaning the inlet pressure on the face of the fan is significantly below atmospheric.

Duct theory suggests that provided the inlet duct has a constant sufficient area, has a smooth inner surface and changes direction as little as possible then the duct pressure losses per unit lenght are surprisingly low. Like all aerodynamics the principle the object is to keep the airflow as undisturbed as possible. This also applies to the air stream as it approaches the leading edge of the fan blades and is the primary reason nearly all commercial thrust figures quote the fan operating in 'free air' and with a bell mouth to ensure the fan is receiving a smooth airflow normal to the blades. It can make a significant difference.

Surely a cheat hole in the wall of a duct will disturb the airflow proceeding down the duct and the closer it is the the face of the fan the more the disturbance.

You may have noticed that in many cases I place the fan right at the back with a short thrust tube rather than the conventional midway along the duct. My thinking here is the airflow down the inlet is rather different to the exhaust from the fan.

The inlet airflow will be more or less constant across the duct diameter although a bit slower next to the duct wall due to skin friction. For the exhaust the airflow is not constant across the duct diameter. It will be fastest closest from the fan blade tip and much slower from the fan root. This suggest to me that exhaust duct losses per unit length will be higher than the inlet hence my preference where possible for a rear placed EDF.

 

Of course over riding all duct theory is the simple fact the efficiency (g thrust/Watt) for a given EDF is improved dramatically as the fan rpm is reduced. Light weight airframes with reduced power requirements can result is smaller batteries and a longer flight duration but then I am not a member of the "it must go over 100mph club"! 😉

 

All a bit contentious. I think I had better keep my head down.   

   

  

[Simon Chaddock]

A bit more progress but it is rather slow. Apart from the cold much of the construction method has to be decided as it goes along!

With both EDFs installed and tested the basic fuselage box is largely complete.

 

There is still a bit more to do fairing in the nozzles followed by the fin and part of the fuselage dorsal spine.

By covering the box top surface the issue is ensuring there is sufficient access left open to install the elevon servo wires.

Then the wings can be built and added although their actual method of construction is still to be decided.

When the delta wings are attached it will become quite a substantial plane.

 

 

 

 

 

[Erfolg]

I think I must misunderstood what you have written.

 

Any way, in contrast to the approach (Physics)we were required to adopt, at work, entered into the design file. I now generally adopt  intuition, relying on academic training ,only on calcs if really desperate or pretty bored.

 

considering duct shape, we all recognise (from fluidics) that a circular duct is better than say an extreme letter box shape in most circumstances, for the same unit area. The losses tending to be lower. That is other than say an outlet, where Bernoulli (conservation of energy) suggests that the mass flow, can be traded at an outlet for increased velocity and pressure. So the round  duct is good, other shapes not as good.

 

With respect to the length of any containment, we will have observed that a long hose pipe that at the discharge point, the flow an available pressure is often much lower than a short hose. Often this observation goes along with noting a coiled pipe is poorer than a straight pipe, which suggesters that changes in direction, shape transitions in direction, cross section, are all unwanted.  This is all based on a unit feed. Long transmission pipes are not the best.

 

Most of us as kids have played with tap flow, noting that at low discharge rates, the flow can be laminar. Increase the flow and it becomes turbulent. Although many text books show this laminar flow, with zero (stagnant) flow at the wall of a surface, with increasing flow velocity at distances away from the surface. The real world experience is most often not the case. I reality there is a stagnant boundary layer, which quickly becomes a chaotic turbulent flow. any imperfection, even dust can initiate this condition. All of this is associated with losses, per unit length.

 

Again relying on our friend Bernoulli it matters little where the fan is on a single dimension duct. However introduce an additional inlet, things change. Particularly if the inlet is near the discharge point, even better if a Plenum chamber can be arranged in front of the fan unit (in this case) and a short discharge pipe.

 

I would like to claim I have put these concepts into practice, to date, I have failed. My He 162 so far is a failure, being to heavy, although the short ducting 3d printed works well. I am now looking to compromise more, with a 3d printed plenum chamber at the front and a short 3d exhaust with a  3dprined outer shell. To gain experience I am building a TN A10, although the wing loading disturbs me.

 

It is interesting that there a good few doors in the XB 70, outlet doors. Just like Concord, these were to facilitate, super cruising above the speed of sound. IMO a so called cheat inlet, is pretty near to scale (different purpose), with the DF right at the back. I am sure that you could minimise the losses from the chaos of the additional  fluid.

 

I am not sure that the ducting is required between the rect inlet to the fan (other than, are they structural). The increase in area, would potentially drop the air (fluid) velocity dramatically, reducing (potentially) losses.

 

 

 

 

Edited January 11 by Erfolg

[Simon Chaddock]

Erfolg

I agree a plenum chamber in front of the fan rather than a duct is an interesting concept. It certainly reduces the average inlet velocity, however EDFs operate a very low pressures. A fan is more a process of air velocity and the energy required to accelerate it than pressure.

In flight the air entering the inlet is, relative to the airframe, travelling at the planes speed. If it passes into a plenum chamber it will rapidly slow down in proportion to the inlet and planum chamber cross sections. The fan has now got to expend energy to speed up the plenum chamber air to the planes air speed before it can generate any net thrust. The question is thus would the air be travelling faster down a constant area duct than in the plenum chamber? The closer the air is travelling to flying speed the less additional energy is required from the fan to generate thrust.

Of course at higher speeds dynamic pressure comes into play and as you point out at supersonic shock wave pressure can be very significant but is not applicable to EDFs.

 

So my hypothesis is to try to maintain the inlet air velocity so as much as possible of the EDF's energy is spent accelerating the air flow to above the planes air speed.

 

You mention your HE 162.

With mine I had more aerodynamic problems with the airframe (directional stability) than with thrust. It actually it used a drone motor with a small 3 blade ducted prop but being built entirely thin sheet foam is it pretty light.

 

Above it is in a "bank and yank" configuration, no rudder, but like that it simply did not turn when bank was applied. Even with oversize fins and working rudders it still requires a very careful application of rudder & aileron to execute any sort of turn.

[Simon Chaddock]

Added a bit of the box top skin, the fin and a section of the fuselage spine.

The fin is a just sheet foam skins with some internal webs.

This is about as far as I can go as I have to leave access for the elevon servo wires.

The delta wing are likely to be of a similar form of construction. We shall see. 

[Erfolg]

Simon

 

We will both agree, that when gliding, lets say a steady state condition, what generally enters the tube is pretty much the same velocity as the model is at.

 

To get thrust we rely on F=Ma, in essence air density is constant (at a particular time), on that basis the thrust, force comes from accelerating the air the fan sees. Gong back, to our youth, when dong fluidics of a multi element pipe system, if we can introduce an additional lower friction/loss common, a low loss source. The calc changes, with potentially higher thrust, as a additional volume of  lower loss fluid being available.

 

Without resorting to a few calcs being definitive of the why, cheat holes actually work and can work well.

 

it is a great pity that Dave Burton is not with us, as I strongly believe he would present a better and clearer argument than me.

 

We also know intuitively that DF are less efficiently than a prop, hence the short flight time, particularly if we try to fly fast. In essence we can easily show the relationships between Work Done etc and ducted systems, yet for most people, they just know from observation.

 

 

Edited January 13 by Erfolg

[Simon Chaddock]

Erfolg

If the inlet has the same area as the Fan Swept Area I would agree the maximum velocity the fan can generate is ultimately restricted.

However in my XF108 the inlet and all the ducting to the fan is 1.2 times the fan gross area or 1.4 times the true Fan Swept Area. With this configuration the fan can increase the exhaust velocity by at least 1..4 or even more at the 90% FSA nozzle before the inlet is creating any velocity restriction. In this case a cheat hole has very little benefit.

I dd agree that if the inlet & duct is below this sort dimension a cheat hole will reduce the inlet choke effect and improve the EDF performance.

 

It one reason I use relatively small EDFs and light airframes so that true scale inlets and exhausts are big enough.

 

 

[Erfolg]

Simon, I have argued that fans increase the thrust or force, that is by doing work. You get nothing for free, you have to supply energy, the resulting effective work done, depends on various factors.

 

My only issues are that ducts absorb energy, supplying air from just Infront of the fan, should be more efficient, than a long duct.

 

I see that you stick to your principles of minimal build weight, to allow lower flight speed, at the expense of robustness and potentially robustness. This is what floats your boat, so why not. I just believe in this particular instance you can/could  improve the power system, by a few tweaks (I always know all my models could be better in some respect, ahh, hind sight).

 

With gliders in particular that the so called lift, term has a V^2 in the relationship. On that basis once the model is at height, not much more velocity is required to achieve a lift value, that is the same as a lighter model, that makes little or no practical difference to duration or distance, the pilot matters generally more. On this basis many favour a robust model, even if the weight is higher than others favour. Epoxy models will often provide robustness and light weight.

 

Your Philosophy  meets your wants and needs. 

 

The mainstream DFs, use power at a rate, that normally results in short flight times, often heat issues of various types

Edited January 15 by Erfolg

[Simon Chaddock]

Erfolg

I wouldn't disagree for a minute that a short duct has less losses than a long one which incidentally is why my VTO V-2 rocket has in effect a virtual 360 degree cheat hole and as close as possible to the ducted fan to ensure the maximum possible thrust from the power available.

 

To me it comes down to what constitutes an acceptable variation from scale appearance and the best way to achieve it. Cheat holes are a last resort.

 

I would suggest the HK Vampire is an example of a rather visible non scale appearance to incorporate a big EDF but as it is on version 3 it must sell well. 

 

There is some progressing on the XF108

 

The inner lower wing skin. The completed wing will be an all foam structure that hopefully will get way with no reinforcement as was the case on my the rather bigger XB70. That is strong enough to do a "very non scale" loop and roll. 😮                                    

Edited January 16 by Simon Chaddock

[Erfolg]

The book ""Valkyrie the North American XB-70 by Graham M Simons, pub Pen & sword aviation,  isbn 184884546-4, on page 141 has a picture immediately prior to the demise 62-0207, has an under surface picture in flight. It shows aa series of what appears to be holes prior to the engines. 

 

Super Cruise equipment is not directly mentioned, as on Concord, I guess it was perceived as a bit of an area of secrecy. Indirectly discussing the system, mainly from the perspective, is it visible. Air was slowed from Mach 3 to I by a series of adjustable internal ramps, plus exhaust doors in the upper surface of the wing immediately Infront of the Engines (to avoid using after-burners, all of the time). Discussed P86-87 amongst other.

 

There seems a number of excuses to have cheat inlets, and still be semi-scale.

[Simon Chaddock]

Some more progress.

The top skin of the RH inner section of the wing added.

All foam and simply glued to the fuselage box with no spar. Full wing stiffness will only be achieved when the top skin is added to the fuselage.The elevon servos have to first added and their wires run forward through the fuselage before the fuselage box can be completed. 

This is a very tentative build as I still have no real idea if it will actually be strong enough but it is light!

 

 

[Simon Chaddock]

Both L&R wings complete and glued on.

The wing tips, with slight anhedral, will add quite a bit to the total wing area. They will be added next, then the elevon servos so the fuselage box top surface can be completed. 

This is the image I am using as a "guide" for the appearance of the top surface of the XF108.

It is an "artist's impression" and differs in detail from other views including even the North American mock up! 😗  

 

[Simon Chaddock]

Some progress. The wing tips, control surfaces and elevon servos added.

I cannot go any further until the servo extension wires and the Little Bee ESCs arrive. Only when they are installed can the nose section be added and some idea reached as to where the battery needs to go. As the EDFS are "right at the back" the LiPo position will be the only available "balancing" item! 

[Simon Chaddock]

At long last a full power test of the QX 40 mm 12 blade EDFs.

https://www.youtube.com/watch?v=TbHSBI8VHPA

On a 4s it is drawing 36A giving 511W. This is far more than it will ever need as the FX108 only weighs just over a pound ready to go!

It does shut the door smartly though!

The EDFs are beautifully balanced and can run at very low RPM.

 

 

[Simon Chaddock]

With the nose built on it is starting to look like a XF108 Rapier.

Still got to squeeze in all the RC stuff in the nose.