Chapter 125: The Shape of the Sky
Location: Shergill Aviation Advanced Design Bureau, Gorakhpur
Date: 14 July 1973 — 08:00 Hours
The monsoon had been threatening since dawn.
By seven in the morning the clouds over Gorakhpur had the grey-green weight of a sky that has decided to rain and is only negotiating the moment. The air was thick and warm and smelled of wet earth from the previous night's shower — the first proper rain of the season, which had left the roads dark and the fields sharp and the whole eastern plain looking like something recently cleaned. The trees at the perimeter of the Shergill Aviation compound stood very still. Everything smelled of decision. Everything smelled of before.
Karan had been awake since five.
He hadn't planned it that way. He had gone to sleep at midnight intending to sleep until six-thirty, which was what a reasonable person would do before the important presentation. But his mind had simply refused the allocation and surfaced at five with the full operational clarity of something that had been thinking while the rest of him slept. He had lain there for twenty minutes trying to negotiate with it — telling it, in the quiet way of a man arguing with himself at five in the morning, that nothing needed deciding before dawn — and then surrendered and gotten up and made tea and stood at his office window watching the eastern sky lighten from black to the bruised deep blue that precedes a monsoon morning.
He had spent those early hours doing what he always did when something mattered enough to wake him: not reviewing arguments or rehearsing answers, but thinking about the people who would be in the room at eight. Their histories. Their decisions. The specific quality of attention that each of them brought to any room they walked into, and what that quality of attention would need in order to be satisfied.
Because today was different from anything Shergill Aviation had faced before.
Today was not about proving the company could build aircraft. The IAF knew that. India knew that. Every foreign intelligence agency with a South Asia desk had adjusted their assessments accordingly after the S-27 Pinaka's performance in 1971. Today was about something harder: showing experienced, intelligent, operationally serious men an aircraft that was not an incremental improvement on anything that existed. An aircraft whose specifications were, by any honest assessment, beyond what any non-superpower aerospace industry had attempted from scratch. An aircraft that would require five years of sustained development effort, enormous capital commitment, and a level of institutional competence that Shergill Aviation had already demonstrated with the S-27 but would now have to demonstrate again at a higher level of difficulty.
The S-35 Tejas-M.
Karan had been living with the design for two years. He had first sketched the configuration on a pad during a flight back from Bangalore in November 1971 — three weeks after the air campaign, in the quiet that followed the war's end, when the immediate urgency was over and the question of what came next had finally found space to exist. He had been thinking about what the next generation of air combat would look like. What threats India would face in the 1980s. What capabilities the IAF would need that it currently lacked. Not what could be built easily. What needed to exist.
The answer he had arrived at was not comfortable.
What needed to exist was an aircraft with genuine multi-role capability — not the compromise multi-role of a platform optimised for one role and tolerantly adapted for others, but a purpose-designed aircraft that was simultaneously competitive as an air superiority fighter, effective as a deep strike platform, and capable of the kind of long-range interdiction that the IAF's current doctrine couldn't attempt. An aircraft with range that reached Pakistan's western cities without aerial refuelling. An aircraft with payload that made it relevant against hardened targets. An aircraft with the aerodynamic and electronic sophistication to survive in contested airspace against whatever the next generation of adversary systems would be.
The specifications he had settled on were ambitious. Karan had spent two years stress-testing them — through the aerodynamics team, through propulsion, through systems integration — looking for the places where the design's demands exceeded what the technology could deliver. He had found several. He had resolved most of them. A few remained open, documented honestly, with development pathways identified. The design was not finished. It would not be finished for years. That was the point: today was not a presentation of an accomplished fact. It was the proposal of a beginning.
He watched the IAF delegation arrive in three vehicles at 08:04.
Air Chief Marshal O.P. Mehra stepped out first.
Fifty-nine years old. Slim, straight, with the quality of unhurried precision that came from four decades of making exact decisions in situations where exactness mattered. He had commanded 1 Squadron during the 1965 war, had logged more than three thousand hours on jets including Vampires and Gnats and Hunters, and had overseen the IAF's transition through three generations of aircraft procurement. He stood in the courtyard for a moment before moving, just long enough to look up at the building, and Karan had the specific feeling of being assessed by someone who was very good at it.
Behind him, Air Vice Marshal Denzil Keelor stepped down from the second vehicle. Keelor was different in the way that certain people are different — not louder, not larger, but somehow more present. He had shot down a Pakistani Sabre over Halwara in 1965 in what became one of the most discussed aerial engagements in IAF history — twelve seconds from visual acquisition to guns, in an engagement where the Gnat had no business winning by performance metrics alone and had won anyway because the pilot was better. He walked with the particular carelessness of physical confidence, the movement of someone whose body has been to the edge of its performance envelope enough times to know what the edge feels like.
Air Vice Marshal Hrushikesh Moolgavkar followed, already talking to the man beside him — Air Commodore Suresh Wagh, who was heading the IAF's Air Staff Requirements team. Moolgavkar had a reputation for strategic precision and the willingness to say in meetings what other people only said afterward in corridors.
Group Captain Brijpal Singh Sikand came next — 47 Squadron, a man who had been watching things carefully and saying little for long enough that when he finally said something, rooms listened. Behind him, Wing Commander Surinder Kumar Malik, currently the IAF's most experienced test pilot, who kept a small book in his breast pocket and wrote things in it when he was thinking through something.
There were others — Squadron Leader Rajiv Mehrotra, logistics and maintenance analysis; two doctrine analysts from Air Staff who would say little but write everything; two staff officers whose function was to record accurately and ask nothing — but these five were the ones who mattered. Karan had read their files. He knew what each of them had flown, what each of them had said in procurement discussions, what each of them was likely to want from today.
He went downstairs.
Vishwakarma met him at the bottom of the stairs with two folders and the expression of a man who has been up since five making sure everything is ready.
"Projector is calibrated," Vishwakarma said. "The aerodynamics summary is in the left folder. The propulsion and systems integration briefing is in the right. Rathore arrived at six-thirty."
"Rathore being early is unusual."
"I noticed." Vishwakarma fell into step beside him. "Vardhan wants to walk the Air Chief through the Kaveri Mk 2 cell before the main session if there's time."
"After the morning session," Karan said. "Not before. I want the Air Chief in the room when the performance numbers are presented first."
"Understood." A pause. "Ramanathan is nervous."
"Ramanathan is always nervous before presentations."
"More than usual. He found a 12-kilometre discrepancy in one scenario and spent three hours establishing that it was an input error in the scenario rather than a calculation error."
Karan almost smiled. "Tell him the Air Chief likes honest numbers."
"I told him. He said, 'What does the Air Chief like about numbers that are honest and slightly disappointing?'"
"Tell him those are the only numbers worth presenting," Karan said. "And that if our numbers were perfect he'd have no reason to keep improving them."
Vishwakarma nodded and peeled off toward the main hall.
Karan pushed open the door to the primary briefing room.
The room was set for eight people from the IAF side and four from Shergill Aviation. A long table. Water. The projector screen at the far end showing nothing yet — the opening slide had been loaded but not displayed, which was how Karan preferred it. The room should not look like it had been waiting. It should look like it had been ready.
Vikram Rathore was standing at the window, looking out at the compound, arms crossed, with the particular stillness of a man whose mind is somewhere else entirely. He was thirty-four, lean, brown, with the weathered precision of someone who has spent a significant portion of his adult life in aircraft cockpits. He had joined Shergill Aviation from the IAF two years ago, which made him the only member of the design team who had arrived from operational service rather than academic or industrial background.
He turned when Karan came in.
"You've been here since six-thirty," Karan said.
"Six-twenty," Rathore said. "I wanted to sit in the design bay for a while. Before the meeting."
Karan understood this. "How does it look?"
Rathore considered the question seriously. "It looks like an aircraft. Not a mock-up. Not a concept model. The proportions are right in a way that's hard to explain if you haven't sat in a lot of cockpits, but once you've sat in enough of them you develop a sense for it. This one sits like something that wants to fly."
"That's what we need them to feel today."
"They'll feel it," Rathore said. "Feeling it and funding it are different things, but they'll feel it."
The IAF delegation came in at 08:09. The introductions were efficient — Mehra ran meetings the way he'd run squadrons, which was to say without unnecessary ceremony — and within four minutes everyone was seated and the room had the particular quality of attention that distinguished meetings that were going to matter.
Karan remained standing.
"Air Chief Marshal, gentlemen, thank you for making the time. I'll begin with what this meeting is and what it is not." He paused. "This is not a presentation of a completed aircraft. We have a full-scale aerodynamic mock-up in the design bay — I'll take you through it this morning — and we have eleven months of design refinement behind everything I'll present today. But the S-35 is not built. It will not fly for eighteen months at the earliest, and that's the demonstrator, the KW-501, not the prototype. What we're presenting today is a design intent, a development programme, and a case for why the IAF should want this aircraft to exist."
He looked at Mehra. "I'm asking for your directional endorsement of the programme. Not your signature on a procurement contract. Those are different things, and I want to be clear about which one is on the table."
Mehra said nothing. His expression was attentive in the specific way of someone who has decided to let a person finish before responding.
Karan advanced to the first slide.
"The requirement we're designing to is not in your current Air Staff document," he said. "I know that because we've read the current Air Staff document, and what we're building doesn't match what it asks for. We're building to what we believe the Air Staff document will ask for in 1978."
The room shifted slightly. This was not a conventional opening.
"By 1978," Karan continued, "India will face an air environment that is different in three specific ways from the environment we face today." He advanced the slide. "First: adversary ground-based air defences will be more sophisticated and more lethal at longer ranges. The SA-3 and SA-6 systems that Pakistan and China are acquiring now will be updated and expanded. Low-level penetration will be more dangerous. Standoff capability will matter more."
He advanced again. "Second: adversary fighter capability will improve. China is developing its J-7, which is a genuine progression from the J-6. Pakistan will receive upgraded aircraft from whatever source it can access. The engagements of the 1980s will not look like the engagements of the 1960s. BVR — beyond-visual-range — capability will be decisive."
Another advance. "Third: India's strategic interests will have expanded. By 1978, this country will have economic relationships, diplomatic commitments, and territorial considerations that require the IAF to project power over longer distances than the current force structure enables. The IAF's current reach, measured in operational range from Indian airbases, is insufficient for the India that is coming."
He paused.
"The S-35 Tejas-M is designed for that environment. Not the environment of today. The environment of a decade from now."
Wagh, the Air Staff Requirements officer, spoke for the first time. "That's a significant assumption. You're telling us you know what we'll need better than we do."
"No," Karan said. "I'm telling you what I think the threat environment requires. Whether it matches what you'll need is a conversation, not an assertion. I'm asking to have that conversation."
Wagh nodded slowly, the nod of someone putting a point in reserve.
The first hour was the numbers.
Ramanathan presented the aerodynamics — the blended wing-body configuration, the LERX geometry, the specific design choices that had been made to achieve sustained turn rate without compromising high-speed performance. He was nervous, as Vishwakarma had predicted, and it showed in the first four minutes, and then the numbers themselves took over and he became something else: precise, systematic, confident in the way that people are confident when they are describing things they actually understand.
The S-35's design specifications:
Blended wing-body construction. Single engine. Single crew. Length 17.8 metres. Maximum speed Mach 2.1 clean. Supercruise at Mach 1.3 to 1.4 dry — sustained supersonic flight without afterburner, which extended operational range and reduced the infrared signature. Combat radius between 1,350 and 1,600 kilometres depending on mission profile. Ferry range of 3,700 kilometres on internal fuel alone. Nine hardpoints — six wing stations, three fuselage. Maximum payload 7,200 kilograms.
Sustained turn rate of 16 degrees per second at 5,000 metres combat weight. Instantaneous turn rate of 22 degrees per second.
The turn numbers produced the first extended silence of the meeting.
Keelor, who had been listening with the particular stillness of a pilot doing mental flight modelling, looked up from his notepad. "Nineteen sustained. Twenty-four instantaneous."
"Yes, sir," Ramanathan said.
"At what altitude and weight?"
"The sustained figure is at 5,000 metres, combat weight, with two BVR missiles and fifty percent internal fuel. That's the standard test weight we use. Clean aircraft at lower altitude achieves marginally higher, but that's not the relevant comparison for operational assessment."
Keelor looked at Karan. "That's a significant number. Better than anything currently in service."
"It's designed to be," Karan said.
"What's driving it?" Keelor asked. "The LERX?"
"The LERX primarily," Ramanathan said. "Combined with the wing-body blending, which gives us more lift from the fuselage itself — the body generates aerodynamic force, not just the wings. It allows the total lift coefficient to be higher without increasing induced drag proportionally." He paused and advanced to a technical diagram showing airflow patterns. "The specific innovation is the vortex management geometry. The LERX generates a controlled vortex that stays attached over the main wing at high angle-of-attack. Most conventional designs lose that attachment above eighteen to twenty degrees AoA. We maintain it to thirty degrees."
He clicked to the next slide showing computational fluid dynamics imagery. "The vortex creates a low-pressure region over the wing upper surface, effectively increasing lift. At high AoA, this vortex lift becomes the dominant component. The challenge is keeping the vortex attached and stable — if it separates or becomes unstable, you lose the lift and the aircraft departs. Our LERX geometry is shaped to generate the vortex at the right position and feed it smoothly over the wing throughout the turn envelope."
"The blended wing-body," Ramanathan continued, "contributes in two ways. First, the fuselage cross-section itself generates lift — roughly twenty percent of total lift comes from the body rather than the wing. Second, the blending eliminates the interference drag you get at a conventional wing-body junction. The result is that we can sustain a given turn rate at lower total drag than a conventional airframe, which means lower energy loss per turn, which means sustained performance rather than instantaneous-only."
Keelor wrote something in his book. "Thirty degrees AoA with vortex attached. That's exceptional. What happens beyond thirty?"
"Beyond thirty, the vortex becomes unstable and we begin to lose effectiveness," Ramanathan said honestly. "The aircraft can physically achieve thirty-two degrees with the FBW system allowing it, but the useful turn performance begins degrading past thirty. We designed to thirty as the practical operational limit."
"And the instantaneous twenty-two degrees per second — what limits that?"
"G-load on the pilot," Ramanathan said. "At twenty-two degrees per second sustained turn, you're pulling approximately eight and a half G at the speeds we're discussing. That's the human limit. The airframe can handle more, the FBW will allow more, but the pilot can't sustain it. The instantaneous number is what you can pull for a few seconds before you need to unload."
Mehra spoke for the first time since the technical session began. "These are carefully calculated figures. Not aspirational."
"No, sir," Ramanathan said. "These are the figures from the aerodynamic model with safety margins applied. We've wind-tunnel tested the 1:8 scale model and the results validate the computational work within four percent. The full-scale demonstrator will be the final validation, but I'm confident in these numbers."
Moolgavkar leaned forward. "Range."
"Sir?"
"The 1,600-kilometre combat radius. Which mission profile achieves that?"
"High-low-high strike profile," Ramanathan said. "High altitude cruise to the target area, low-level ingress for the final 200 kilometres, weapons release, low-level egress, high-altitude return. That profile achieves 1,600 kilometres with two 1,000-kilogram bombs, two BVR missiles, and full gun ammunition."
"From Amritsar to?"
Ramanathan had prepared for this question. He advanced the slide to the range rings map. "Amritsar to Rawalpindi with 200 kilometres margin. Amritsar to Islamabad clean. From forward bases in Rajasthan, the aircraft reaches Karachi without refuelling."
The room was quiet.
"From bases in the northeast," Karan added, "it reaches Chengdu."
That produced a different quality of silence. Chengdu was where China's aviation industry was concentrated. Reaching it wasn't a threat; it was a statement about what the aircraft was capable of.
Moolgavkar sat back slowly. "The current doctrine does not contemplate operations at those distances."
"The current doctrine," Karan said carefully, "was written for the current aircraft. Doctrine should follow capability."
It was a precise statement and a slightly aggressive one. Moolgavkar looked at him with an expression that was not quite a smile. "I'll remember you said that."
"I hope you will," Karan said.
The second hour was the engine.
Vardhan was not nervous. He was never nervous. He was, in some fundamental sense, the least nervous person Karan employed, which was occasionally counterintuitive given that he was responsible for the component whose failure would be most immediately fatal. But Vardhan had a relationship with his engines that operated below the level of anxiety — he understood them the way a farmer understands land, which is to say completely and without drama, including the parts that were difficult.
The Kaveri Mk 2 was not the Kaveri Mk 1 that powered the S-27. It shared lineage — the same basic architecture, the same team, the knowledge accumulated through two years of operational data from the Mk 1 fleet. But it was a different engine.
"The Kaveri Mk 2," Vardhan began, and his voice had the calm clarity of a man describing something he knew completely, "is a variable-bypass turbofan with full authority digital engine control. Dry thrust is 97 kilonewtons. Wet thrust with afterburner is 115 kilonewtons."
He wrote the figures on the whiteboard in his careful block letters.
"Those figures give the S-35 a thrust-to-weight ratio of approximately 1.15 at combat weight — fifty percent fuel, two BVR missiles, full gun ammunition. At light weight, early in a mission or late after fuel burn, the thrust-to-weight approaches 1.28. That puts it in the class of the best Western fighters currently in development."
Mehra spoke. "The variable bypass. Explain that."
Vardhan nodded as though he'd been waiting for exactly this question. "A conventional jet engine has a fixed ratio between the air that goes through the core — the combustion section — and the air that bypasses the core and mixes downstream. Higher bypass gives you better fuel efficiency and lower noise. Lower bypass gives you better high-altitude, high-speed performance. A variable bypass engine gives you both by changing that ratio depending on what you're doing."
"How does the ratio change?"
"Variable geometry inlet guide vanes in the bypass duct," Vardhan said. "Controlled by the FADEC — the Full Authority Digital Engine Control — which is the computer that manages all engine parameters. The pilot doesn't manage this directly. The FADEC reads the flight condition and adjusts the bypass ratio to optimise the fuel efficiency and thrust balance for that specific combination of altitude, speed, and power demand."
He paused. "The practical result is that at subsonic cruise — transit to target area, for example — the engine behaves like a high-bypass turbofan. Fuel consumption drops by approximately eighteen percent compared to a fixed-configuration engine of similar thrust. At supersonic sprint or maximum power, the bypass closes down and the engine behaves like a low-bypass turbofan optimised for performance."
"Eighteen percent fuel consumption reduction," Mehra said. "That's where the range number comes from."
"Yes, sir," Vardhan said. "Approximately two-thirds of the range improvement comes from the variable bypass. The rest is airframe efficiency."
"Time between overhaul?"
"Design target is 1,400 hours. The Mk 1's TBO is 900 hours, and we've been exceeding that in operational service with careful maintenance management. The Mk 2 is designed for a longer life from the outset — better thermal barrier coatings in the hot section, improved blade metallurgy from the superalloy work we've been doing with DMRL Hyderabad, more conservative stress margins in the turbine section."
Vardhan advanced to a cross-section diagram of the engine. "The FADEC does more than control the bypass ratio. It monitors 340 parameters continuously — temperatures, pressures, vibration signatures, fuel flow rates, everything that tells you how the engine is behaving. The system learns the engine's normal operating signature and alerts the pilot to deviations before they become failures. We've incorporated prognostic health monitoring — the system doesn't just tell you something is wrong, it tells you something is going wrong and you have time to land before it becomes critical."
Malik, the test pilot, who had been writing steadily in his book, looked up. "Single-engine reliability. For a single-engine aircraft, engine failure is not an abstract concern. What happens when it fails?"
Vardhan nodded as if the question were obvious — because it was. "Three things. First, the TBO target itself. 1,400 hours means fewer overhaul cycles, fewer opportunities for something to go wrong during maintenance. Second, the FADEC's health monitoring system I just described. The pilot gets warning. Third—" he paused "—the aircraft is designed to glide from operational altitudes to a suitable airfield following an engine failure, without thrust. The glide characteristics were a deliberate design requirement. Ramanathan can speak to this, but the wing loading and glide ratio are such that a flame-out at 8,000 metres in clear conditions gives the pilot options."
Malik wrote something. "What options?"
"Range to nearest divert of approximately 45 kilometres per 1,000 metres of altitude in clean configuration," Ramanathan supplied. "At 8,000 metres, that's 360 kilometres of glide range. Not enough to be complacent, but enough to not be immediately fatal. The aircraft glides like a sailplane compared to most fighters."
Malik put a mark next to whatever he'd written. "That's better than I expected."
The third hour was the systems.
Deepak Menon was twenty-seven years old, which was young enough that two of the IAF officers had initially assumed he was a junior staff member rather than a lead engineer. He had corrected this impression within approximately three minutes of beginning the Trinetra radar briefing, after which the IAF officers had adjusted their expectations and begun asking him harder questions.
"Trinetra," Menon began, and there was no hesitation in his voice, "is a mechanically scanned pulse-Doppler radar with a 780-millimetre antenna diameter. The name means 'three-eyed' in Sanskrit — the radar's ability to simultaneously track, identify, and engage. Track capacity is sixteen simultaneous targets. Engagement capacity is six simultaneous targets with BVR weapons."
He advanced the slide showing the radar architecture. "The S-27's radar was first-generation pulse-Doppler with eight-target track and two-target engagement. We've learned a lot since then. Trinetra is second-generation. The improvement comes from three areas: processing power, clutter suppression algorithms, and antenna design."
"Processing power first. We're using the AN/AYK-14 processing architecture that the Americans developed for the F-15 — we licensed the design, modified it for our requirements. That gives us computational capacity to handle sixteen tracks simultaneously with continuous target state updates. The S-27's processor could handle eight tracks but the update rate degraded when it approached that limit. Trinetra maintains full update rate at sixteen tracks."
"Clutter suppression second. Look-down shoot-down is fundamentally a clutter problem. A low-flying aircraft has a radar cross-section that's partially masked by ground return. The question is how well you can distinguish the target from the background. Our algorithm uses Doppler filtering — the target has a specific velocity signature relative to the ground return — combined with pulse compression and adaptive thresholding. The combination gives us reliable track on a 2-square-metre RCS target at 85 kilometres in high clutter conditions."
Wagh spoke up. "Eighty-five kilometres is a significant number. MiG-23 is getting Sapfir-23 with claimed sixty-kilometre look-down capability. What's the specific improvement?"
"Two things," Menon said. "First, our antenna is larger — 780 millimetres versus Sapfir's 600. Larger antenna means more gain, which means more range for a given target RCS. Second, our pulse compression ratio is higher. We're achieving twelve-to-one compression versus Sapfir's eight-to-one. That improves range resolution and signal-to-noise ratio, which directly translates to detection range in clutter."
He clicked to the next diagram. "The antenna itself is the third improvement area. We're using a slotted planar array design that gives us lower sidelobe levels than the S-27's parabolic dish. Lower sidelobes mean the radar is less likely to track the wrong target or be confused by jamming. The scan rate is 60 degrees per second in azimuth, which means the radar can cover the forward hemisphere and update all tracks within two seconds."
"Jamming resistance," Menon continued. "The S-27's radar had limited ECCM — electronic counter-countermeasures. Trinetra has a full ECCM suite. Frequency agility — the radar can hop between frequencies to avoid spot jamming. Sidelobe blanking — we can detect when a jammer is trying to enter through the sidelobes and blank that signal. Pulse-to-pulse frequency change — makes it harder for a jammer to track what frequency we're on."
Mehra looked at Karan. "These are significant advances over the S-27's radar. Where did the ECCM work come from?"
"From the Electronic Warfare Group we established in 1972," Karan said. "After we saw what happened in the 1971 war, we realized EW was going to be decisive in the next war. We hired twelve engineers with signals intelligence and EW backgrounds — some from DRDO, some from other sources. They've been working on ECCM algorithms for eighteen months. The Trinetra implementation is their work."
Wagh leaned forward. "The FBW system. You mentioned second-generation architecture. Explain."
Arvind Pratap Singh, who had been quiet until now, spoke up. He was in his early thirties, thin, with the precise manner of someone who thinks in control loops and state variables. "The S-27's FBW was first-generation — quadruplex architecture, analog control laws, fixed control gains. It worked, but we learned its limitations during flight test and early operational service."
"The S-35's FBW is second-generation. Still quadruplex — four independent channels for redundancy — but digital control laws instead of analog. That gives us adaptive control gains that change based on flight condition. At high speed, the aircraft needs less control surface deflection for a given stick input — the controls get 'heavier' automatically. At low speed and high AoA, the aircraft needs more deflection — the controls get 'lighter.' The pilot doesn't feel this, but the aircraft responds consistently across the flight envelope."
He advanced to a block diagram. "The control laws also incorporate envelope protection. The FBW will not allow the pilot to exceed structural limits or depart controlled flight. If the pilot pulls back on the stick at maximum G, the FBW limits the control surface deflection to keep the G-load below 9.0. If the pilot tries to overspeed, the FBW limits throttle. The pilot can override in an emergency, but the normal operation is protected."
"The carefree handling," Arvind continued, "comes from this. The pilot can focus on tactics and situational awareness, not on keeping the aircraft within safe parameters. The FBW does that. In a turning fight, you can pull full stick and hold it, and the aircraft will turn at maximum performance without departing or overstressing. The S-27 could do this, but the S-35 does it with more precision and across a wider envelope."
Malik looked at Arvind. "The S-27 had FBW oscillations during development. I remember reading about that. What's different here?"
Arvind didn't hesitate. "The S-27's oscillations were in the lateral axis at specific AoA combinations. The root cause was interaction between the control law update rate and the actuator dynamics — they were coupled in a way that produced instability. It took us eleven months to resolve because we had to fundamentally rework the lateral control law and modify the actuator response characteristics."
"We learned from that. The S-35's control laws were designed from the beginning with the actuator dynamics included in the model. We're not designing the control law and then hoping the actuators can follow it — we're designing them together. The oscillation problem shouldn't occur because we've designed it out rather than discovering it in flight test."
"Shouldn't occur?" Malik pressed.
"There are no guarantees in flight test," Arvind said flatly. "Something unexpected will happen. That's why we test. But we've done everything we can in design and simulation to anticipate the known problems. If we find something new, we'll document it and fix it the same way we fixed the S-27 oscillations — thoroughly, no shortcuts."
At 10:30, Karan said: "I'd like to take you to the design bay."
The walk from the briefing room to the restricted design bay was two minutes through a corridor and a security checkpoint. Karan led. Rathore fell in beside Mehra, and they talked in low voices. The security door opened.
The design bay was cold — the sustained chill of a space kept at controlled temperature for measurement instruments and materials that behaved differently in heat. The ceiling was high, four and a half metres, with fluorescent lighting mounted on gantry rails. Along the left wall, benches of instruments, drawing tables, component storage. Along the right wall, fabrication stations and a large-scale surface table.
And in the centre of the bay, occupying a floor space of approximately 22 by 12 metres, the KW-501.
The full-scale aerodynamic mock-up of the S-35 Tejas-M.
It was painted in flat grey-white — a neutral colour that told you nothing about the eventual aircraft's scheme and everything about the object itself. The colour drew no attention, which meant the shape drew all of it. And the shape was immediately arresting in the specific way that aircraft shapes sometimes are: not dramatic, not aggressive in the aesthetic sense, but simply right in a way that was difficult to articulate and immediately apparent. The blended wing-body configuration looked different from every conventional configuration these men had seen — there was no clean separation between the fuselage spine and the wing root, no joint or fillet or transition, just a continuous surface that swept from the nose to the wing tip through a geometry that resolved naturally into both.
No one spoke for about fifteen seconds.
Sikand moved first.
He walked around the left side of the aircraft without asking permission, and Karan let him. Sikand's approach to things was tactile and direct — he looked at the leading edge curvature at the wing root, at the LERX geometry, at the way the fuselage cross-section changed from circular at the nose to the flattened oval of the intake region to the widened wedge of the wing-body junction. He crouched and looked at the underside — the ventral surface, the main gear bay doors, the position of the centreline hardpoint. He stood and looked at the intake geometry from the side.
He hadn't said a word and he'd already told Karan more than most people did in an hour of questions.
Keelor had gone to the cockpit. He was standing on the ladder that the crew had positioned, looking into the open canopy. "Single piece canopy," he said.
"Single-piece frameless canopy," Rathore confirmed, climbing the ladder to stand at Keelor's shoulder. "180-degree forward visibility over the nose, unrestricted to 60 degrees either side, good rear-hemisphere visibility with the head turned. The canopy frame is the one visual element you'll see as a pilot — the ejection initiator rail runs down the centreline — but it's positioned below the line of sight during BVR radar scan."
Keelor leaned into the cockpit. "What's the seat?"
"Martin-Baker Mk 10 derivative, modified for the cockpit geometry. Zero-zero capability. Optimised for the ejection envelope at the higher speeds this aircraft is capable of."
Keelor looked at the HOTAS layout — the Hands On Throttle And Stick configuration, the position and function of every control element accessible without removing the pilot's hands from the primary flight controls. His fingers moved over the grip without touching it, tracing the button positions in the air. "How many primary weapon functions on the throttle?"
"Four," Rathore said. "Radar mode select, weapon select, cage-uncage, and master arm. The stick has BVR release, gun trigger, target designator, and countermeasures quick-release. Everything a pilot needs for a BVR engagement without leaving HOTAS."
Keelor was quiet for a moment. Then he stepped back down the ladder and looked at Karan. "I want to fly it."
He said it with the simple directness of a man who has been in enough cockpits to know what he's asking for and to mean it precisely.
"When it flies," Karan said, "you'll be in the queue."
"Not a promise?"
"A commitment," Karan said. "Soon as Rathore finishes the envelope expansion and we're cleared for evaluator flights, you're on the list."
Keelor nodded once. That was sufficient.
Mehra, who had been walking slowly along the aircraft's centreline from nose to tail, turned at the tail section. He had said almost nothing for the last twenty minutes.
"Walk me through the reasoning on the single engine," Mehra said.
Karan moved to the tail section, standing beside the engine nozzle — a variable geometry aperture with the actuator housings visible at the nozzle perimeter. "The decision to go single-engine was contested internally. The argument for twin-engine is obvious: redundancy, particularly for over-water operations, and the ability to continue flying with one engine out." He paused. "The argument for single-engine is performance and simplicity. Two engines means more weight, more complexity, more maintenance, more failure points even if any individual failure is survivable. The thrust-to-weight calculus shifts. The aircraft gets heavier for the same speed, or slower for the same weight."
"And you chose single engine."
"We chose single engine and then designed around its limitations," Karan said. "The TBO target, the FADEC health monitoring, the glide characteristics Ramanathan mentioned — all of those are the design's answer to the single-engine risk. Plus the operational reality: most fighter losses to engine failure happen because of maintenance lapses or undetected developing failures, not because the engine catastrophically fails without warning. Better TBO and better health monitoring prevents more losses than redundant engines in most realistic scenarios."
Mehra looked at the nozzle. "Most realistic scenarios."
"Yes," Karan said. "Not all. Over-water operations beyond glide range to land, extended combat air patrol over hostile territory without divert options — for those missions, the single-engine design has limitations we don't pretend don't exist. We designed for the primary IAF mission set, which is land-based operations over the subcontinent."
Mehra turned and looked at the aircraft again. From the tail, looking forward, the Tejas-M's shape resolved into something that was at once modern and instinctively right — the narrow spine widening to the wing junction, the intake lips on either side of the fuselage, the canopy sitting low in the spine with the minimum of profile that meant minimum of drag.
"It looks like an aircraft," Mehra said, which was the same observation Rathore had made at six-twenty and which meant the same thing.
"That's what we were going for," Karan said.
They ate lunch in the adjacent conference room. It was a functional lunch — food sufficient to sustain people through an afternoon of technical discussion, not a hospitality gesture. Rathore and Vardhan ate with the delegation, which Karan had decided was important: the IAF men should eat with the people who had built this thing and speak to them informally, because informal conversation revealed things that formal presentations didn't.
The conversation over lunch was not about the S-35 directly. It was about aircraft generally — stories, comparisons, the particular culture of aviation knowledge that operated through accumulated specific experience. Keelor told a story about a hydraulic failure over Rajasthan in 1966. Sikand, who had said almost nothing all morning, turned out to have strong opinions about cockpit layout that he expressed with the directness of someone who had been waiting for a relevant venue. Vardhan and Malik got into a conversation about engine instruments that became surprisingly technical and that both of them appeared to enjoy.
Karan ate without speaking much and listened to everything.
The afternoon session covered three remaining areas: the development programme timeline, the funding model, and the specific commitment Shergill Aviation was making.
On the timeline, Vishwakarma presented. He had the quality of a man who had spent his career managing the gap between what engineers said they needed and what programme managers could actually deliver, and he presented the S-35 schedule with the specific honesty of someone who had done this long enough to know where programmes lied to themselves.
"The KW-501 demonstrator first flight is targeted for January 1974," Vishwakarma said. "That is eighteen months from now. It's aggressive but achievable if the engine integration timeline holds and the FBW system validation completes on schedule. If either of those slips, the demonstrator first flight slips."
"The prototype?" Malik asked.
"KW-502, first flight April 1975. Prototype has full mission systems — the complete Trinetra radar suite, the weapons integration, the FADEC in its final configuration. The gap between the demonstrator and the prototype is technology maturation and the production tooling required for the prototype's construction standard."
"IOC?"
"January 1976. That assumes first production aircraft delivery in June 1976 and a six-month operational evaluation period. It also assumes the production line is established and tooled by mid-1976, which requires a production commitment no later than the first quarter of 1975."
"First quarter 1975 production commitment," Wagh said. "Before the prototype has completed its flight envelope expansion."
"Correct," Vishwakarma said. "We will not have complete flight test data before the production commitment is needed. We will have demonstrator data and partial prototype data. This is normal for programmes with IOC targets in the five-year range — you commit to production before the testing is complete, based on demonstrated risk reduction and design confidence."
Mehrotra, the maintenance analyst, spoke up from the far end. "The nine-to-eleven man-hours per flight hour maintenance target. Our current aircraft run twelve to eighteen operationally. The MiG-21 in particular runs high. What's the design approach?"
Vishwakarma deferred to Vardhan. "Engine accessibility," Vardhan said immediately. "Engine removal and replacement time is designed to be 45 minutes for a two-man team. The engine mounts, the connector interfaces, the bleed air connections — all designed for tool access without requiring deep disassembly. In the S-27 programme, we learned that most maintenance man-hours go into access rather than the actual maintenance task. If you can get to the component in four minutes rather than forty, you've changed the whole economics."
"The avionics?"
Menon took this. "Line-replaceable unit architecture. Every avionics box is designed to be removed and replaced without calibration by a technician with the appropriate skill classification. The box comes out, a new box goes in, it self-checks against the aircraft system. No bench calibration required. We're applying the black box concept systematically."
Mehrotra wrote something. "That's the theory. What's the verification?"
"On the KW-501 demonstrator, we'll instrument the maintenance process as deliberately as we instrument the flight. We will know the actual man-hours per sortie and we'll report them to you. If we're wrong, you'll know we're wrong."
Mehrotra looked at him for a moment. "That's brave."
"We'd rather be held accountable to accurate data than praised for inaccurate data," Menon said. "You'll operate this aircraft. If the maintenance numbers are wrong and you find out after IOC, that's worse for everyone."
At 15:30, Mehra called a brief recess.
The IAF delegation stepped outside into the corridor. The Shergill Aviation team stayed in the room. Karan poured water and didn't speak. Ramanathan organised his papers. Vardhan drew something on a notepad that may or may not have been an engine diagram.
Rathore sat down beside Karan.
"How do you think it's going?" he said, quietly enough that only Karan could hear.
"Mehra hasn't said much," Karan said.
"He doesn't say much when he's decided. He says a lot when he's uncertain. Quiet means he's already made up his mind."
Karan looked at him. "How do you know that?"
"I flew under him briefly, before I left the IAF," Rathore said. "He's the same in the air. Silent means he's working through something. It's the briefing-room chatter that means he's still open."
Karan processed this.
"If you're right," he said, "the question is where he's landed."
"The mock-up," Rathore said. "He walked the centreline twice. Most people walk it once. He walked it twice and then stood at the tail and looked forward for thirty seconds. That's someone seeing the aircraft, not assessing it anymore. The assessment was done."
When Mehra came back in, he sat down and looked at Karan with the particular directness of someone who has finished a decision and is about to communicate it.
"I have several observations," he said. The room was very still. "First: the design is serious. By which I mean it has been thought through with the care that serious engineering requires. The willingness to separate modelled from tested, the acknowledgement of single-engine limitations, the maintenance accountability commitment — these tell me you know what you're doing."
He paused.
"Second: the programme is ambitious. Correctly ambitious. The IAF's next aircraft should be an aircraft that expands our capability, not an aircraft that replaces our current capability with a marginally better version of the same thing. The specifications you've presented, if achieved, represent a genuine generational advancement."
Another pause. Longer.
"Third: I am going to take this to Air Staff with a recommendation for programme endorsement. Not procurement — programme endorsement. Formal directional support for the S-35 development programme, with committed review at first demonstrator flight. At that review, Air Staff will evaluate the actual demonstrator performance against the design intent, and we'll decide on the production programme commitment."
He looked around the room.
"What I am not doing is giving you a blank cheque. What I am doing is saying that the IAF has a formal interest in the S-35 programme succeeding, and that interest will be reflected in how we engage going forward. We'll provide operational input during development. We'll review the test data as it comes. We'll work with you, not just evaluate you."
Silence.
Wagh looked like a man who had expected to continue arguing and had instead arrived at a destination he wasn't quite ready for.
Keelor looked like a man who was already thinking about flight profiles.
Sikand, who had barely spoken all day, said: "When Rathore flies the demonstrator, I want to be at the field."
It was so simply stated and so completely characteristic that Karan almost laughed.
"You'll be there," Rathore said.
The delegation departed at 17:15.
The vehicles pulled out of the compound into the Gorakhpur evening, which was wet now — the monsoon had made its decision somewhere around mid-afternoon and had begun, not with drama but with the steady determination that the proper monsoon has. The roads were dark and the compound trees were shedding water.
Karan watched them go from the same observation gallery where he'd watched them arrive.
Vishwakarma came to stand beside him. "The Air Chief's endorsement recommendation. That's more than I expected this morning."
"It's what we needed," Karan said.
"The programme still needs to deliver," Vishwakarma said. "Endorsement isn't production commitment. We have eighteen months to the demonstrator first flight, and that's assuming nothing slips."
"Something will slip," Karan said. "It always does."
"And?"
"And we'll deal with it the same way we dealt with the FBW oscillations in the S-27 programme. By actually solving the problem rather than working around it or pretending it's smaller than it is."
Vishwakarma nodded slowly. "Ramanathan wants to run the scenario models again with the actual test weight we'll use for the KW-501. He thinks the sustained turn number might be achievable at 500 metres lower altitude than the design brief assumes, which would be better for the demonstration profile."
"Tell him to run it," Karan said. "And tell him he did well today. After the first four minutes, he was excellent."
Vishwakarma almost smiled and peeled off toward the main hall.
Karan sat at his desk for an hour after everyone else had left.
The design bay was dark — the crew had closed the access panels on the KW-501 and turned off the working lights, leaving the mock-up in the dim security lighting, grey and quiet and present. He could see it from the glass panel in his office wall.
The S-35 Tejas-M. The aircraft that didn't exist yet and was already real.
He thought about what the next five years would require. Not what he hoped they would require — what they actually would. The demonstrator's first flight was eighteen months away, which meant eighteen months of sustained development pressure on a team that was also managing existing commitments. Then prototype construction, which would be more complex and would require tooling and processes that didn't exist yet. Then the flight test programme, which would be systematic and demanding and would produce both the validation data and the inevitable surprises.
He expected something of similar severity to the S-27's FBW oscillations, probably in a different subsystem, probably at a worse moment in the schedule. There was no way to predict what it would be. There was only the established practice of finding it honestly, documenting it completely, and solving it rather than accepting it.
He pulled a sheet of paper from the desk drawer and began to write.
Not specifications. Not schedule notes. The kind of writing he did when he needed to understand where he was.
Mehra endorsed the programme. Not the aircraft — the programme. The distinction is correct. The aircraft doesn't exist. The programme does, and the programme's credibility earned the endorsement.
He paused.
Keelor wants to fly it. That's the best possible response from the best possible person.
He paused again, looking at the dark design bay.
The work that begins tomorrow is longer than the work that ended today. The endorsement is the beginning of the obligation, not the end of it.
He wrote one more line.
The S-35 will fly. The question is only what we learn between now and then.
He capped the pen, folded the sheet, and set it on the corner of the desk.
Outside, the monsoon was doing what monsoons do — remaking the surface of things, feeding the rivers, turning the fields dark, running along every channel and crack that had been waiting for it through the dry months. It fell on the bureau building, on the roof where water pooled and ran in channels to the downpipes. It fell on the design bay where the KW-501 stood in the dark with its access panels closed, its leading edge catching no light, its shape resolved in shadow into something that was absolutely itself.
Tomorrow the teams would return. Ramanathan would run his revised scenario models. Vardhan would continue the Mk 2 test cell run that he'd paused for today's meeting. Arvind Singh would work through the FBW architecture review that was scheduled for the following week. Menon would brief the radar prototyping team on the range test programme beginning in August.
All of it was work that had been underway before today and would continue after it. Today was not a beginning. Today was a confirmation that what had already begun was worth continuing.
That was sufficient. More than sufficient.
India was building things.
Not for the first time. Not for the last.
End of Chapter 125
Principal Characters in Chapter 125
IAF Delegation:
Air Chief Marshal O.P. Mehra — Chief of Air Staff,
Indian Air Force Air Vice Marshal Denzil Keelor — Decorated pilot, Sabre kill in 1965 Air Vice Marshal Hrushikesh Moolgavkar — Commanding Western Air Command
Air Commodore Suresh Wagh — Air Staff Requirements Branch
Group Captain Brijpal Singh Sikand — Commanding No. 47 Squadron
Wing Commander Surinder Kumar Malik — IAF Senior Test Pilot
Squadron Leader Rajiv Mehrotra — Logistics and maintenance analysis
Shergill Aviation:
Karan Shergill — Founder and Chief Designer Major
Vikram Rathore — Chief Test Pilot
S.P. Vishwakarma — Programme Director Dr. Harsh Vardhan — Chief Propulsion Engineer (Kaveri Mk 2)
S. Ramanathan — Chief Aerodynamicist Arvind Pratap Singh — Flight Controls and FBW Architecture
Deepak Menon — Radar Systems (Trinetra) Technical Reference: S-35 Tejas-M Specifications (Design Intent, July 1973)
Airframe:
Configuration: Blended wing-body, single engine,
single crew Length: 17.8 m
Maximum Speed: Mach 2.1 (clean configuration)
Supercruise: Mach 1.3–1.4 (sustained supersonic without afterburner)
Combat Radius: 1,350–1,600 km (mission-dependent)
Ferry Range: 3,700 km internal fuel Service Ceiling: 16,500 m
Rate of Climb: 280 m/s at sea level
Performance:
Turn Rate (sustained): 16 deg/sec at 5,000m combat weight
Turn Rate (instantaneous): 22 deg/sec
Maximum AoA (vortex attached): 30 degrees
Maximum G-load: +9.0/-3.5 G
Thrust-to-Weight: 1.15 to 1.25 at combat weight, 1.4 at light weight
Weapons & Hardpoints:
Hardpoints: 9 total (6 wing, 3 fuselage) Maximum Payload: 7,200 kg Internal Gun: 23mm twin-barrel cannon, 220 rounds BVR Missiles: 4-6 depending on configuration Air-to-Ground: Up to 4,000 kg bombs/precision munitions
Propulsion:
Engine: Kaveri Mk 2 (variable bypass turbofan with FADEC)
Dry Thrust: 97 kN Wet Thrust (afterburner): 115 kN Engine
TBO: 1,400 hours (design target)
Fuel Efficiency Improvement: 18% at subsonic cruise vs fixed-bypass
Avionics & Systems:
Radar: Trinetra mechanically-scanned pulse-Doppler Antenna: 780mm diameter
Track Capacity: 16 simultaneous targets
Engagement: 6 simultaneous BVR targets
Look-down/Shoot-down: 85km range (2m² RCS in clutter)
FBW: Quadruplex digital, 2nd-generation with adaptive control laws
EW Suite: Full ECCM, RWR, chaff/flare dispensers
Cockpit: Single-piece frameless canopy, HOTAS controls, MFDs
Maintenance:
Design Target: 9–11 man-hours per flight hour
Engine R&R Time: 45 minutes (two-man team)
Avionics: Line-replaceable unit architecture (no bench calibration)
Mean Time Between Failures: Design targets exceed S-27 by 30%
Development Timeline:
Demonstrator (KW-501) First Flight: January 1974 Prototype (KW-502) First Flight: April 1975
Initial Operating Capability: Jan 1976
Design Philosophy: The S-35 Tejas-M represents a generational advancement over the S-27 Pinaka, incorporating lessons learned from two years of S-27 operational service and combat experience. The blended wing-body configuration, LERX vortex management, variable-bypass engine, second-generation FBW system, and improved radar/avionics suite position the aircraft for the air combat environment of the 1980s rather than replicating the capabilities of the 1970s.
Note on Nomenclature:
The aircraft designation "S-35 Tejas-M" reflects its position as Shergill Aviation's third major combat aircraft programme (following the S-27 Pinaka and interim variants). "Tejas" (Sanskrit: तेजस्, meaning "radiance" or "sharp brilliance") was selected to represent the aircraft's role as a multi-role platform combining air superiority and strike capabilities. The "-M" suffix denotes the baseline multi-role variant, distinguishing it from potential future naval or specialized variants.
Historical Context:
Development of the S-35 began in late 1971, immediately following the successful combat deployment of the S-27 Pinaka. The programme reflects India's strategic decision to maintain indigenous control of advanced aerospace capability rather than relying on foreign suppliers for next-generation combat aircraft. As of July 1973, with IAF endorsement secured, the programme enters its critical development phase with demonstrator first flight targeted for early 1974.