Design logic of a 3^rd level Passenger Transport/Fire Bomber                         Prof.Proodyut Das

Design team’s note:

The following is an almost verbatim reproduction of a part of our Das Design D 12 Salamander aircraft which was entered into the Royal Aeronautical Society’s 2024 competition. The competition was for a Firebomber with a 2 ton “bomb” load and stringent cruise, minimum speed. range and loiter parameters which could also carry up to 19 Pax. as per part 23. The D 12 Salamander took a top placing in the contest and indeed its design approach of using the past wisely to reduce risks was mentioned in the conference.

 For “students” of the design process we have edited the presentation somewhat to emphasise:

i)                The soft aspects of aircraft lore- beliefs, history, customs, old wives tales, is as important as the ‘hard” technical aspect. Without the compass and navigation chart of the soft knowledge of aircraft lore the design will go irretrievably astray and run onto the reefs.

ii)              There is a trap in the use of technology. “Better” may not be the most appropriate solution and “best” is almost always fatal to realistic outcomes. Usually a patient well thought out frugal effort gives good results.

iii)             The catch in using soft knowledge is that it requires much knowledge of the background to assess what is “most suitable” The perceptive reader will not fail to note that “copying” requires a very deep level of knowledge.

With that as an introduction we repeat below the portion where the approach to ten design and the thought process. Figures quoted are provisional and may change subject to new information. 

Introduction

Firstly, we must thank the Royal Aeronautical Society for another opportunity to participate in the Design Competition, an aircraft design competition of this kind is the food of quiet love. One can think and say what one would not dare to do in other places and one can say so without being told to hurry up because dinner is getting cold. Peaceful holding forth at one’s own pace of one’s own opiate. So once again, a delighted “Thank You” from our team.

If meeting the specifications and conforming to expectations was the way to success then the

He 112A should have been the Luftwaffe’s standard fighter in 1936. However, by comparison the horrid little Me 109 V1 ran away with the trials at Rechlin and Travemunde. Despite Milch’s relentless opposition the Me 109 became the Luftwaffe’s next Jaeger. The Me 109 was horrid, if only because it broke everyone’s comfortable little expectations about what a modern fighter should be. The Me 109 had a tiny wing flogged hard, with tricky slots and slats, a narrow track undercarriage, an uncomfortable sitting posture, a coffin like cockpit canopy, all of which got the Rechlin’s test pilots’ askance hormones into overdrive. What saved the Me 109 was that it was well engineered and reasonably well balanced. Balance not as in the stability about three axes, but design being a process of compromises the Me 109 assured the user that the compromises were beneficial and the best of a difficult job. The He 112A checked the boxes of the gallery’s expectations but was more difficult to modify, the Me 109 checked the often-overlooked soft aspect of aircraft design i.e., ease of production, developability, versatility.

The reason to cite the above is to underline that mathematics or specifications alone does not constitute an aircraft’s design. Indeed, much of design especially conceptual design can be done with no more than with calculators and mental arithmetic. A good design has a personality or a soul and it is this soul which decides the fate of the aircraft, the specifications is merely a contract document for acceptance. The other reason for improvising the specification is that the person(s) preparing the specifications may have had their windows open only in certain directions.

One does not know if that great cynic of the respect for specifications, (“Follow the specifications exactly and dead cert you are a goner”) Sir Sydney Camm, ever got a chance to either give Willy Messerschmitt a pat or tell Dr. Heinkel a “I told you so” but he did believe that irrespective of the pressures of the specifications the Chief Designer should never go against his better judgment. Digressing to speculative mischief, perhaps the British Observer’s Corp should be thankful the Germans did not back the He 112. With both the Spitfire and the Heinkel He 112 having elliptical wings the confusion that would have been caused within the British interception system which relied solely on the Observer’s Corps after the raiders crossed the Chain Home line is an interesting conjecture.

The initial thought on going through the specifications was on how much emphasis should be given to the firefighting role and how much on the secondary transportation role. Following that was to question what should be the possible maximum capacity for passengers, nineteen is the requirement but countries like China and India are significant and densely populated markets. Like the ‘Durzee’ tailored clothes of our childhood it was common sense to ask the tailor to leave a little let in the fitting for the inevitable growth. Therefore, a seating capacity of up to forty in an easily modified version is provided for in the design. By the same compulsions given the pressure on land for airfields, a degree of rough field short field capability should be provided.

In addition to the above, the chart on which the aircraft designer is to plot his course is liberally marked with “here be dragons”, a few of the more important ones are given below.

1.     The engineering of an aircraft is a commercial affair, the design must make money. It is often overlooked that the Spirit of St. Louis was designed to win a prize, the Orteig Prize of $25,000 for the first NYP crossing had to be done within fairly tight time and money limits. Lindberg must have reckoned that after spending about $10,000 on the Ryan NYP, he would end up with about $15,000. It was not the only reason why he did the flight, but it certainly affected his choices.

2.     There is a technological boundary layer in every role; one cannot pack too much technology into a blunt plough and retain common sense. 

In the case of the Spirit of St. Louis, for example, suppose it were possible to tele-transport all 2024 technology to Lindberg’s time. It would still come out that the original NYP was, engineering wise, doing for six pence which any fool can do for six shillings, i.e., the best solution. Some of the wonder technologies (GPS, radio, radar) would simply not work because of the operating environment or would be unnecessary. For instance, what would the radar show in mid-Atlantic of 1929. All the aerodynamic refinements such as engine, variable pitch propeller, retractable under carriage, slotted flaps, composites would perhaps increase the cruising speed by no more than 15%, i.e. from 100 to 115 m.p.h. If we remember that the engine power goes up by the cube of cruise speed, we would end up with an arrival time over Le Bourget may be by mid-afternoon instead of late evening.

 Of course, everything would be better, but ‘better’ is the song of the Sirens; it can wreck the total on its shores of hidden cost and complexity. Ask the Americans; the Pentagon is the master of ‘better’. The F 35 is an engineering marvel, but if it weren’t for NATO how many would be sold? If Lindbergh had tarried for the better, someone else amongst the several competitors would have beaten any 2024 technology design with their old technology. The Lindberg approach must be admired because it was a well-crafted minimum effort that got the job done, truly a case of ‘Citiius et Longius commoda periculus sapientas’. Let us translate our shaky Latin to avoid any misunderstanding: faster and further by taking wise risks. 

3.     We must also respect other common-sense factors in the shaping of a design. There was no doubt that Lindbergh’s own air mail pilot experience shaped his single crew design. The multi crew, multi engine option based on the much-favoured tri motor Bellanca was not only too far down on the delivery list and unaffordable, but also in case of failure of even one engine at take-off, would result not in greater safety but a more expensive crash. By the same arguments of in situ desperandum, Willi Messerschmitt perhaps realizing that he was anyway going to be blackballed by Milch for further orders beyond the prototypes, went out to design the Me 109 on what his instincts told him rather than to pander to the approving nods of the Technisches Amt. 

4.     A logical corollary from the examples discussed at two above is therefore that the past may have much to teach. Even today, we would profit from having humility and perhaps empathy to pay great, reverential, and detailed attention to the work of those who have worked before. We should see what can be used from those well thought out proven designs and what must need be modified. A feature is not stupid because we are not patient enough to understand the context. 

5.     Prior to the Boeing 737-Max incidents, certifiers used to be conservative, indeed to the point of being superstitious. Since the design is to be certificated by an outsider with no particular interest in the outcome, it helps (especially in countries where aircraft certification is not an established procedure) to have a successful model. This reassure the certifier that they are not signing for trouble when giving the clearance. If nothing it would be a hygiene as management people say and won’t hurt. 

6.     Aerial Firefighting is like agricultural aviation a rugged world. The aircraft must be rugged beyond what is imagined by the various certification requirements.

Interpreting the Requirements

An aircraft designer is not an inventor. Indeed, if one has looked at Teddy Petter designs, i.e., the Whirlwind, the Gnat, the Lightning, excessive ingenuity can be troublesome. The designer’s job is to provide his backer with a reliable product at the lowest possible expense and risk. Thus, as we continue in the same magma of important but undefined thinking, it becomes clear that rather than striving for brilliance, it is better to be humble and copy. Sneering at copying is highest with designers who have the least experience, but for the humble, it has it has great benefits. For example, the requirement for para jumping nudges us towards something from the 1930s, the Junkers 52 or the Dakota both of which were used extensively in para dropping operations. If we keep broadly to their features, there should be no intractable developmental problem about para dropping. Our aim, outside of the specifications, was not only to propose an aircraft but do so at the lowest possible cost and risk. For this, it was imperative to look at history.

The team analysed commercially successful aircraft to detail the requirement specification.

1.     Developments in laminar flow fuselage, blended wing body, tailless design, self-balancing wings should be considered for application on a proven base design.

2.     The engines and propellor to be mounted far forward of the wing leading edge to obtain aerodynamic advantages from the propellor slipstream blowing over the wing.

3.     The flying characteristics should be very good for operations in uncertain terrain and difficult airfields.

4.     It should be able to operate from unprepared airfields with their inherent deficiencies in facilitating servicing, refuelling, navigation aid and communication.

5.     The engine and structure should be well proven and simple to repair on field.

6.     The cabin should be a constant section so that passenger capacities can be increased by adding additional frames.

7.     It should be possible to pressurize the fuselage pressurization in a stretched version.  

8.     The cabin width should allow the cargo version to accommodate LD3 containers.

9.     Regulations require special provisions for operating at height above two meters, which meant the loading floor must be below two meters.

 

Figure 1: D-10, 9 Seater Electric Aircraft

Using the past to secure the future

The initial idea was to create a twenty-seater turboprop version of the Das D-10, a nine seater tailless electric aircraft with a single engine pusher configuration. The team soon realised that the wing sweep required for a tailless design would hamper easy access to the fuselage, and the stub wings would be inconvenient for para jumping. Reluctant to abandon the benefits of self-stabilizing wings and the reduced in empennage drag, multiple simulations were run with varying sweep angles. The conclusion was that about 10 degrees of sweep and aspect ratio of 10 was a good compromise between stability, reducing empennage drag, and ease of access. Interestingly, this was similar to the Jack Northrop wing chosen for the DC-1 by the design team at Douglas. Thus, we arrived by circuitous route to the Dakota as the muse. The road to the Dakota went past the following and we looked calmly, carefully and long at the following:

Miles M.20	Twin engine blend wing/body design with buried engines.
De Havilland Comet	Fuselage shape & nose
Curtiss Commando	Cabin sizing and shaping
Celera 500L	Laminar flow fuselage shape
Scottish Twin Pioneer	short take off
Junkers JU 52	Structure and fixed undercarriage
Cessna Sky Courier	What the modern American industry thinks
Canadair CL 215	Water bomber
Bennet Airtruck	Layout ideas of the “outre” kind which one would not dare to think. Was useful for the DO 335 layout.
Embraer Bandeirante	Benchmark aircraft for checking estimates
Douglas DC-2/DC-1	Benchmark for development process
Saab Scandia	For comparing tailwheel (DC2/DC-3) and Tricycle undercarriage and systems studies
Antonov An. 2	Features of a successful but offbeat Dakota replacement
Ilyushin IL12/14	A very successful Dakota replacement in a particular economic situation

 

The gentle art of copying

Not all of the aircraft were mathematically analysed; many of our questions fell into a level of rustic wonder and surmise.  In many cases, it was simply looking and carefully at the model or picture and trying to think why the designer of almost a century ago made certain choices. The only dictum being that we would never, even under duress, say ‘That’s stupid’ about a feature we did not understand. Some of our typical surmises were,

I)                 why was the fuselage of the Curtis Commando almost elliptical in elevation when Douglas used the more practical straight tube with end plugs?

II)              How much weight and drag would we save in making the Ju 52 without a corrugated skin.

III)           What if we put an additional turboprop at the back of an AN2, would a firefighters hit the rear propeller when he para-jumped if we kept the same “q” forces and the same distance between the door edge and the strut that braces the AN2 stabilizer?  The AN2 has been used for para jumping and not too many have been injured in the process. So, what do we need to do in case we have a push pull layout?

IV)           The intrigue about the Dornier DO 335 Pfeil type layout is that it would reduce the drag particularly with a 3 or 4 abreast Britten Norman Islander type seating, with an unquantifiable loss of versatility due to a less adaptable cabin volume e.g. Short Stirling vs Avro Lancaster bomb bays. The real interest in the layout is in the possibility of shutting one engine down during loiter time without any asymmetric flight worries. This will need discussions with customers, pilots and certification authorities. The idea has promise and is discussed in the appendix.

The perceptive reader will note that it requires a great deal of background knowledge to copy.

At one point, our team felt that if we could somehow get the Dakota’s P&Ws to run on Jet A1 whilst giving the same power and TBO as a turbine engine, we could just add a dump system for the retardant and submit our proposal. Unfortunately, the turbo prop is a game changer which affects design just as much as, say, composites or plastics. It requires a different thinking to harmonize. This is why adding turboprops never really worked; because substantial mods were required to balance the Cg and utilize the power economically. The other Dakota replacements failed because they were better Dakotas. 

Any proposal talking about replacing the Dakota is met with knowing sighs but the truth is no one really made any thoughtful attempt to replace the Dakota. Only the Russians got the thinking right with their IL-12 and IL-14 but to us the best Dakota replacement was the Antonov ‘Annushka’ AN2. It was the simpler Dakota and exceeded in produced seat mile capacity the original Dakota. We mention this because in aviation lore the Dakota replacement has attained the same status as the Bermuda Triangle, full of accounting wrecks and mystery.

The Evolution of the Design

At this stage- after a great deal of reflection and searching of History- it was possible to begin defining the aircraft. Two possible approaches emerged from endless but exciting discussions. One was a Dakota based solution and the other was a bit more offbeat solution which is covered briefly in the appendix.

At the core of the design is the water dumping system, and the sizing and design of the cabin for this is the part that earns revenue. A versatile rather than an optimized aircraft will always be able to earn its keep.

A minimum dump of one ton per pass is required and the idea is to have at least a 0.5 mm film of on the ground of water or retardant. The initial idea was to use the wing centre section as an integral tank but given that fire retardants or wetting agents such as the Soviet (NP-1 Sulphnol) would also be added meant that there would be airframe corrosion over time. The difficulty of providing adequate inspection hatches in a highly stressed area worked against the idea of using integral tanks. Finally, it was decided to use the rubberized bag tanks used in the MiG 21, for fuel, but with different leak proofing compounds, mildly pressurized i.e., 0.2 bar (3 psi). These tanks could be installed in the cabin by suspending it from the longerons and frames. Refilling would be via a hatch on the top of the fuselage using a crane-necked device, that used to be popular in the days of the steam locomotives. 

The size of the dump hatches would be decided by the aircraft velocity, and as a start, we would use the CL-215 hatches, measuring 155 cm x 78 cm. Venting the tank would be as important as the dump hatch sizing. An over pressure release and a gas charging cylinder to maintain the bag pressure as the water is dumped is considered. When not in use, the tank and pressurising system could be folded and unloaded. 

The other system studied was to replace the dump hatch with a set of nozzles and pressurising system that can give an ejection speed of 50m/sec. This speed was found to be able to penetrate the hot air rising from a fire and hit the core and not be blown away and may be useful in certain cases. The aimed for density is approximately 1 litre/m². An image recognition triggered bomb sight to automate the release can also be added to improve accuracy.

 

Figure 2: D-12, 20 Seater Twin Turboprop Utility Aircraft

The question of high wing in contemporary aircraft of similar size versus low wing was discussed. A high wing with a boat hull would allow modifications to add water scoops. The D-12, expected to operate from semi-prepared fields, benefits from the ability to use a wide track undercarriage in a low wing aircraft. Additionally, the wing can pass through the underfloor without the draggy bulge in the fuselage, similar to the Hawker Siddeley HS 748.

As a result, the design decision favoured the low wing configuration. The additional crash protection to the passenger cabin provided by the low wing was another factor. Regulations require special provisions for operating at height above two meters, which meant the loading floor must be below two meters. This was examined against the need for sufficient clearance for the props. The floor, it is emphasized, does not have a sill as in the Dakota and is flush, for ease of loading and unloading, similar to the Pilatus PC Turbo Porter.

Much discussion centred around the cabin cross section and width because it is the cabin that decides the aircraft’s commercial potential. The circle encloses the most area per unit perimeter (weight) and if the wing passes underneath the chord of the circle would be a convenient floor and for passenger carrying the circle is ideal because a seated human is widest at some distance above the floor level. The Dakota’s cabin was 2.3 meters and apparently sufficient in the ‘50s for BEA to provide four abreast seating on its Pionair Leopard Dakotas but then Britain was still under rationing. Considering the epidemic of obesity, it was decided to increase the fuselage diameter to around 2.75 meters.

Such an increase of diameter immediately triggers thoughts of an increase in drag but we have to remember that the drag is a function of both the drag coefficient and the wetted area. A comparison of the rather sharpish Fokker Friendship and Handley Page Herald noses of the 1950s, with the ATR 42s, and the Lockheed Hercules blunt nose and locomotive streamlining indicates that streamlining and drag consists of a bundle of drag components and the canny designer can obtain great freedom when careful with the combinations. A four abreast seating for nineteen passengers can be handled in five rows whereas with three abreast seating will require 7 rows. It was found that the wetted area decreases by 19% for the wider fuselage, so CdxA probably be the same or slightly lower and structural weight lighter due to greater moment of inertia, thus confirming the wide body choice. It allows for a greater versatility and a stiffer structure for the same unit (kg/m²) weight.

 The cabin has a constant section so that passenger capacities up to forty can be accommodated by increasing the tube length while retaining the end caps. Unlike the Dakota’s not quite circular fuselage the D-12 has a circular section to allow for pressurization for future development. To address concerns about excess, drag from the larger cross-section, biomimicry studies will explore natural models like the Japanese Fugu-Puffer fish and the Boxerfish (as studied by Volkswagen).

 

Figure 3: D-12, 20 Seater Twin Turboprop Utility Aircraft

This constant section cabin is 5.5 mts long in the nineteen- passenger aircraft and using a De Havilland Comet nose and a wide Dakota cargo door, keeping the same distance between the rear edge of the cargo door and the LE of the tailplane to cater to the paradrop requirement. We ended up with a fuselage, rather porcine but remember the glorious Grummans? They were porcine too but had adequate performance because at relatively lower Reynolds number blunt shapes turn in surprising performance because of their lower wetted areas and weight.

The lower drag of the turboprop nacelles against the big radials of the Dakota, the better manufacturing practices, puts the total drag of the D-12 closer to the smaller but more angular Bandeirante rather than the larger Dakota which is the muse. This despite the much wider fuselage.

The undercarriage breaks from the Dakota’ tailwheel and is tricycle but retains the same semi external retraction for damage protection in case of a wheels up and the undercarriage retracts forward as a bit of engineering housekeeping. In case of hydraulic failure, the undercarriage notorious for defying g forces, can be persuaded by the drag forces.

The wing largely follows the DC-1 pattern in terms of the general layout and structure. At this stage NACA 2215 has been used at the root tapering to NACA 2406 at the tip as there is a need for a docile stall. These aerofoils are proven in rugged service. A quick calculation shows that use of laminar flow sections can reduce the power consumption at maximum speeds by about 25 kW. The fact that the water bomber is required to loiter for about an hour and that turbine engine fuel consumption is not proportionate to power, so the penalties of the NACA 2215/2406 versus easier manufacturing and field maintenance demands guided the choice to retain the old for the proposal.

The fire bomber should have considerable energy performance to get out of tight corners. Though dumping the load improves matters but thrust to weight and wing loading figures should be in the region of a professional firefighting aircraft and a fair number of back of the envelope calculations were made to utilize the great reduction in the propulsions system weight due to turbine power.

One area we worked on was to give the D12 soft field strip capability. The undercarriage has twin wheels on both the nose and main landing gear. The wing has single slotted flaps and it should be possible to have lockable leading endge slats on the outboard. This is for dedicated firebomber versions to be able to operate from water sources close to the fire.

The above thoughts also emphasize how the aircraft can largely be intuitively put together by looking around, thinking and looking around again. Rumination almost of the bovine variety is what we did for long.  Until this stage very little calculations were involved. 

Appendix: The Ant-eater layout

A “Wily Messerschmitt” approach to the design of this competition would be to consider the push pull layout a la Dornier Do 335 Ameisenbar or the Fokker D XXIII. This stems from the fact that almost one hour loiter is required in the sortie. Unlike turbines the pistons do not have fuel consumption directly proportional to power settings. Flying with one engine asymmetric shut down or at flight idle on a conventional layout would be fraught with various risks including certification problems.

Briefly the logic is that two engines in ‘push pull’ arrangement, asymmetric effects are zero, power installation can be distributed for optimized for continuous cruise and MTO conditions. Any nose propeller installation has to take cognizance of the amount of blockage the fuselage will offer to the propwash. Fortunately use can be made of the fact that the central half of the propeller does not offer too much of thrust e.g. the Me 109 F/G spinner and can be blocked without major detriment to the flow. We can have a fuselage cross section about 1.8 mts wide without too much of loss of thrust. Pushing out the boat a bit, the idea was to use the Britten Norman (BN) layout of coupe seating, 4 abreast facing each other as in a Victorian Railway carriage but with entry doors as on the BN to reduce torsional stiffness problems. Each cell (how appropriate!) seating 8 pax in 2 benches there being a total of 2 full cells and the remaining 3 Pax being located in the front or back. Alternately 2 cells of 3 abreast with 2 benches per cell i.e. as in Pilatus Porter or DO 27 and the remaining 7 pax distributed front and back. Of course, toilet facilities would be difficult to provide.

The fuselage can be conceived as a shoebox but streamlined in the fashion of the DH Dragon i.e. tapering in the front and back on a basic box.

We were keen on a biplane layout because a biplane is structurally extremely efficient saving about 50% of the structural weight of a monoplane wing of the same area and yet completely suitable for the kind of airspeeds we are looking at plus giving very good handling and safety. However, the lower wing would get in the way of freight loading and passenger egress, and so the final aircraft emerged as a AN2 “Annushka” or perhaps an AN3TP, but with a DO 335 engine arrangement with all the design aims mentioned in the previous text. The massive Shvetsov radial is replaced with a carefully streamlined PT6A installation front and back, the Fieseler Storch’s Balkon cockpit glazing is retained for special roles, certainly for the specialized firefighting versions, and because of the exposure to the fire and the ultra-low flying speed the wing covering is changed from fabric to light alloy. The undercarriage follows Dakota practice and is partially exposed.

The accompanying sketches are provisional with respect to the position of the pusher engine which is also PT6A 114 identical to that of the Cessna Caravan. The same applies to the ducting for the engine air intake.

The design team spent much time discussing the case of the para jumping. As shown the parachutists would be exposed to the danger of hitting the propeller though given that the AN3 can stagger around at a speed of 40 kts it is possible that jumping can take place safely. The math’s of the situation is as follows, given a prop dia. of about 2.6 mts and a cabin width of 1.8 mts there is approximately 0.4 mts zone the jumper must clear. However, the following solutions are being examined:

1.     Moving the cargo door from the end to a position closer to the lower wing trailing edge keeping sufficient clearance to allow safe cargo handling on the ground.

2.     Use some form of ‘Davit cum fireman’s slide’ to allow the jumper to clear the prop disc in a guaranteed manner

3.     Use a floor hatch 

4.     The other direction of solution would be to raise the center line of the rear engine to safely clear the jumper but at the cost of change in pitch trim with engine.

 

 

Figure 4: An Alternate Utility Aircraft

 

 

 

Figure 5: Side View of an Alternate Utility Aircraft