Chapter 107: The Sea Finds Its Wings
Location: Shergill Aeronautics & Defense Complex — Gorakhpur
Date: 24 December 1972 — 06:15 Hours
---
The engine had been running for forty minutes when Karan arrived.
Not in the aircraft — in the test cell adjacent to Hangar Three, where the second Kaveri Mk 1.5 unit was bolted to its static thrust stand, instrumented with 340 sensors, and being pushed through a sustained high-power run by Dr. Harsh Vardhan and his team at six in the morning on December 24th because the data acquisition window for the turbine inlet temperature profile was only valid at specific ambient conditions and December mornings in Gorakhpur provided exactly those conditions and the calendar was irrelevant.
Karan heard it before he saw it.
The sound a jet engine makes at high power on a static test stand is not like anything else — not the distant thunder of an aircraft overhead, but something more immediate and more physical, a sustained pressure against the chest that you feel before you register as sound. Through the test cell's thick walls it came as a deep, organised roar with harmonics in it, the specific voice of a machine working at the upper edge of its design.
He stopped outside the observation window.
Inside, the Kaveri Mk 1.5 was running at 94% power. Dry — no afterburner, just the core engine pushing through its high-pressure turbine sequence. The exhaust plume was visible through the rear port, shimmering in the morning cold. The instrument panels along the wall showed numbers that scrolled too fast to read individually but whose patterns Vardhan was reading with the fluency of a man who had been watching these specific patterns for eight months.
Vardhan hadn't seen Karan yet.
Karan watched him for a moment — the way he stood, the specific quality of attention he gave the turbine inlet temperature readout versus the compressor delivery pressure readout, the small adjustments he made to the fuel flow settings with the economy of motion of someone who has done this enough times that the interface between man and machine has become nearly automatic.
Eight months Vardhan had been chasing this engine.
Karan pushed the door open.
Vardhan turned. "You came early."
"I heard it from the car park," Karan said Jokingly.
"At 94% it carries." Vardhan looked back at the panel. "We're two minutes from the sustained run completion. Then I can give you numbers."
Karan stood beside him and watched the readouts.
The turbine inlet temperature — the measurement of how hot the gas was immediately before it hit the turbine blades, the number that defined the engine's thermal limit — was holding at 1,410 degrees Celsius. In 1972, no production engine in the world ran its turbine this hot during sustained operation. The single-crystal blade alloy, the film cooling channels bleeding protective air through microscopic holes in the blade surface, and the thermal barrier ceramic coating on the blade exterior combined to make 1,410 degrees survivable where 1,150 would have been the previous limit.
The number held.
The run completed.
Vardhan reduced throttle smoothly and the sound changed quality — dropping in pitch, losing its chest pressure — and the test cell went quiet in the relative sense that rooms go quiet after significant noise.
Vardhan printed the run data sheet and set it on the console between them without a word.
Karan read it.
Dry thrust: 97 kilonewtons.
He read it again. Not because he doubted it — because it was the kind of number that deserves to be read twice.
"Ninety-seven," he said.
"Ninety-seven," Vardhan confirmed. He had the expression of a man who has been running toward something for eight months and has now arrived and is standing inside the arrival, not entirely sure yet what to do with his hands.
"The F-4 Phantom's engine," Karan said, "the J79 — produces roughly 52 kilonewtons dry. Our intelligence on American next-generation carrier aircraft development suggests their new engines will reach somewhere in the 60 to 70 kilonewton dry range." He looked at Vardhan. "We are at 97."
"We are at 97," Vardhan said. "Yes."
"Afterburner projection from yesterday's partial run," Karan said.
"137 kilonewtons per engine," Vardhan said. "The revised spray bar geometry gave us the remaining margin. Full augmentor section now shows uniform fuel distribution — no lean zones, no energy loss." He paused. "Two engines. 274 kilonewtons total afterburner thrust."
He said the total the way you say something that has been a long time coming.
"TBO," Karan said.
"Current projection — 1,100 hours," Vardhan said. He said this one differently — with the satisfaction of someone who has built the answer into the design rather than discovering it. "The thermal barrier coatings on the turbine blades, the active clearance control, the FADEC managing the thermal cycling precisely — the engine is not accumulating fatigue the way a conventional high-performance engine does. We're running hotter than anyone, but we're managing the heat better than anyone." He paused. "1,100 hours is a conservative figure. We may see better in extended testing."
Karan looked at the engine through the observation port. "The J79 runs 1,400 hours but at 52 kilonewtons dry," he said. "We are running 1,100 hours at 97 kilonewtons dry." He let that sit for a moment. "That is not a compromise. That is an achievement."
Vardhan looked at his engine.
"Yes," he said quietly. "It is but not enough for me."
Hangar Three at seven in the morning had a specific quality — the flat hard light of the fluorescents, the organised sound of sixty-three people doing sixty-three different things that were all aimed at the same outcome, the smell of machining oil and metal and the faint chemical sharpness of high-temperature materials.
P-1 stood at the far end — the aircraft that had flown eleven times, that wore its hours in the discoloration along the exhaust nozzles and the faint scoring on the port wing leading edge from the third sortie when Rathore had pushed it to Mach 1.4 at low altitude to find the edge and had found it.
P-2 occupied the center — both Kaveri Mk 1.5 engines now seated and connected, installed two days ago, the aircraft looking purposeful in a way it hadn't before. The variable geometry mechanism housing gleamed under the work lights. The leading edge root extensions — the strakes running forward from the wing roots along the fuselage — gave the nose a particular intent that you noticed without being able to name.
Karan walked the floor before anything else.
This was not procedure. It was habit — the thirty seconds of a room before it knows you're watching that tells you things no briefing does.
He saw Srinivasa Ramanathan at the drafting table in the corner. This was not a good sign in the specific way that a symptom is not a good sign — not catastrophic, but indicating something that needed attention. Ramanathan used the drafting table when the computational analysis wasn't giving him what he needed and he wanted to think with his hands. He'd been there since before Karan arrived. The papers around him had the look of an argument with physics that was going badly for the person, not the physics.
He saw Major Vikram Rathore beneath P-1's port wing. Not working — standing, arms folded, looking at the leading edge scoring with an expression Karan had learned to read over eleven sorties. It was the expression of a man running a private calculation between what he'd done to the aircraft and what the aircraft had done in return.
He saw Dr. Padma Krishnan at the structural testing station with Neeraj Bhatia, both looking at a fatigue data printout that clearly contained something Krishnan had found and Bhatia was in the process of verifying. Krishnan had a quality of focused stillness when she was working through something significant — a particular economy of movement that meant her full attention was elsewhere. She had found two things in P-1's fatigue data that nobody else had found in the first three weeks after joining. Both had been minor. The fact that she found them had not been.
He saw Lt. Col. Amarjit Dhaliwal at the electronic warfare station reviewing the EW suite integration logs — a man who had spent twenty years in the army learning exactly how aircraft got killed electronically, and who brought to every EW question a professional intimacy with the problem that pure engineers couldn't replicate.
He saw Commander Subramaniam and Lieutenant Commander Pillai at the side table where they had been sitting since seven. Naval liaison officers sent by Nanda's office six weeks ago. They had the particular stillness of men who have been told to observe and not interfere and have decided the best way to do that is to take up as little space as possible while missing nothing.
Karan looked at them for a moment longer than the others.
INS Vikrant — India's sole aircraft carrier. Flight deck 213 metres long, 34 metres wide, ski-jump ramp at 12 degrees. Operating in the Arabian Sea and Bay of Bengal in conditions that regularly reached sea state four — wave heights of 1.25 to 2.5 metres, deck pitching, crosswinds that arrived without asking. The men who would eventually fly this aircraft from that deck were currently on MiGs somewhere, waiting for something they didn't know yet was coming.
Everything in this hangar existed because of that ship and those conditions and those men.
He walked to Ramanathan's drafting table.
"Twenty-five degrees,did you test that?," Karan said, pulling a stool.
Ramanathan looked up. He had the expression of a man who has been in an argument with himself for three days and is not entirely sure which side he's on. "I've been running the crosswind case."
"Show me."
Ramanathan pushed the sheet across. It was covered in airflow calculations — vortex lift coefficients at different forward sweep angles, lift-to-drag ratios through the launch speed range, the crosswind correction factors for Vikrant's deck dimensions. At the bottom, two columns of numbers under the headings 20° Forward* and *25° Forward, each with sub-columns for nominal launch conditions and crosswind launch conditions.
Karan read the crosswind column.
At 20 degrees forward sweep, the vortex generated by the leading edge root extensions combined with the wing's forward position gave a lift coefficient that was, in the crosswind case with full weapons load, adequate. The aircraft launched. It had margins.
At 25 degrees, the same vortex was stronger — the wing presented more leading edge area to the oncoming air, the vortex was larger and more energetic, the lift coefficient in the crosswind case was measurably better.
The difference between adequate and comfortable, in a 15-knot crosswind off Vikrant's deck with 2 Kaumodaki missiles and 4 air-to-air missiles and full fuel, was not a small difference.
"The transition loads?," Karan said.
"Yes," Ramanathan said. "That's where I'm stuck." He pulled a second sheet. "When the wing sweeps from 25 degrees forward to 68 degrees aft at transonic speed, the aerodynamic loads during the transition are higher than at 20 degrees forward. The mechanism works harder."
"Venkatesh's doing structural analysis on the 25-degree transition case," Karan said. "Have you run it jointly?"
"That's what I need to do today." Ramanathan looked at the drafting table. "I've been working the aerodynamics in isolation. I need his load numbers before I can give you a definitive answer."
"Joint answer by end of day," Karan said. "Not two separate reports. One number, with both signatures on it."
"If the structural analysis doesn't support 25 degrees—"
"Then I want to know exactly why it doesn't and what it would take to make it support 25 degrees," Karan said Seriously. "The answer is not 20 degrees. I've seen the Vikrant crosswind case numbers. I'm not accepting adequate when the aircraft is capable of comfortable."
Ramanathan looked at him. The look of a man who has been circling something and has just been told to land on it. "You've already worked this through, right?."
"I've worked through it," Karan confirmed. "I want your analysis to confirm it independently. Because if your numbers say something different from what I expect, I want to know that more than I want to be confirmed."
Ramanathan pulled a fresh sheet and began. He didn't say anything else. He was already working.
The main review happened at ten.
Karan walked to the center of the floor and stood. The group came to him — not called, pulled, the way a decision point pulls people who have been working toward it.
The full assembly: Vishwakarma, Ramanathan, Arvind Pratap Singh, Rathore, Vardhan, Venkatesh, Zorawar Singh, Kaushik, Deshpande, Bhutia, Prasad, Rawat, Ranbir Singh, Misra, Tyagi, Mukherjee, Gupta, Tripathi, Somnath Iyer, Tiwari, Reddy, Sharma, Krishnan, Dhaliwal, Meera Rao, Suresh Pillai, Kamala Iyer, Bhatia, Anand, Kulkarni, Menon — and others, the full complement of a program that had grown from a set of ambitions into a room full of people who had made those ambitions structural.
Commander Subramaniam and Lieutenant Commander Pillai had moved from their side table. They stood at the back of the group. Subramaniam's arms were folded. He was looking at the data board with the expression of a man who has been waiting six weeks for a specific conversation and is now about to have it.
Karan looked at the board. Then at the room.
"P-2 with Kaveri Mk 1.5," he said. "Dry thrust — 97 kilonewtons per engine. Afterburner — 137 kilonewtons per engine. Two engines — 194 kilonewtons dry total, 274 kilonewtons afterburner total."
He let the afterburner number sit for a moment.
"Our assessment of where foreign carrier aviation development currently stands," he continued, "based on what intelligence and published sources give us — suggests the most advanced carrier aircraft programs are targeting engines in the 85 to 100 kilonewton afterburner range. I believe we are significantly above that. I cannot give you their exact numbers because I don't have their exact numbers. I can tell you the direction and I can tell you we are ahead of it."
Wing Commander Prabhat Anand, second test pilot, spoke from where he stood. "By how much, in your estimate?"
"Significantly," Karan said. "More than a marginal advantage. The technology combination we're running — single-crystal blades, film cooling, thermal barrier coatings, mature FADEC, 3D compressor blade design — I don't believe any program in the world is running all of these simultaneously. We may be running some of them ahead of schedule by years."
"TBO," Vardhan said. He said it the way a man introduces something he's proud of.
"1,100 hours," Karan said. "Which is the number that matters for Vikrant operations,it is not enough but what we have achieved at that 97KN is nothing short of generational ahead. High sortie rate, sustained deployment, maintenance burden that a carrier air wing can actually manage." He paused. "That number is a direct result of the thermal management program. The engine runs hotter than anything comparable. It lasts longer than most engines that run cooler. That is not an accident — it was designed."
Vardhan, quietly: "It was."
"Thrust-to-weight ratio at combat weight, 50% fuel," Karan continued. "1.24 with afterburner. The practical meaning — the aircraft can accelerate in any direction, including vertically, when the afterburner is lit. In a combat engagement, energy advantage is decisive. We have more energy available than anything we expect to face."
Subramaniam, from the back: "Maximum speed."
"Mach 2.2 in clean configuration," Karan said. "Limited by airframe heating at that speed — the kinetic heating of supersonic flight creates surface temperatures that constrain what the structure can tolerate — not by thrust. We have more engine than we need for Mach 2.2. Mohan Deshpande's thermal coating work on the airframe surface is what sets the limit, not the Kaveri."
Deshpande, from where he stood: "We're working on extending it. The next generation coating spec adds approximately 0.1 Mach to the sustainable limit."
"Combat radius," Karan said. "945 kilometres on internal fuel, standard profile. For Vikrant's operational context in the Arabian Sea and Bay of Bengal, this reaches targets that no current carrier aircraft can threaten. It changes what Vikrant can do and where she needs to be to do it."
Subramaniam nodded once. A controlled nod — not impressed, absorbing.
"Turn performance," Rathore said. He said it before Karan did, which was itself a kind of statement. "Say the numbers but say what they mean in the air."
"Sustained turn rate — 20 degrees per second," Karan said. "Instantaneous — 25 degrees per second." He looked at Rathore. "You've felt both."
"Say what they mean," Rathore said again.
"Sustained turn rate is how fast the aircraft can continuously change direction without losing speed," Karan said. "20 degrees per second means in a turning engagement — a dogfight — this aircraft points its nose at the target before the target can point at us. The instantaneous rate — 25 degrees — is the snap, the immediate direction change that creates the moment of advantage." He paused. "Based on our assessment of current and near-term foreign carrier aircraft performance, we are ahead on both measures. Not marginally. Clearly."
Lieutenant Commander Pillai had been writing steadily. He looked up. "G-limit."
"Structural limit — positive 9G, negative 3.5G," Karan said. "The FBW enforces the structural limit regardless of pilot input. The pilot cannot overstress the aircraft. Carefree handling in high-G manoeuvring." He looked at Dr. Alok Misra. "Pilot survivability at 9G."
"The pressure suit and the G-onset limiting in the FBW work together," Misra said. "The suit inflates to prevent blood pooling in the lower body. The FBW manages G-onset rate to give the suit and the pilot's physiology time to compensate. At 9G sustained, the pilot remains conscious and functional. This was validated in the simulator and will be confirmed in P-2's flight test program."
"The simulator data," Rathore said, with a quality in his voice that was not quite skepticism and not quite trust. "The simulator is not 9G."
"No," Misra said. "Which is why I said it will be confirmed in the flight test program."
Rathore looked at him. "I'll be doing the confirming."
"I know," Misra said. "That's why I'm confident."
A few people in the room responded to that. Rathore made a sound that was not quite a laugh.
Karan looked at the board. "The EW suite," he said. "Dhaliwal Sahab."
Dhaliwal stepped forward. He moved with the particular economy of someone who spent twenty years in environments where unnecessary movement drew fire — literal or figurative. "The electronic warfare suite integration is complete," he said. "Radar warning receiver — tuned to the frequency bands of systems we currently know about, updated quarterly as our intelligence improves. The jamming capability — we have active jamming in the forward and aft sectors. Coverage in the side sectors is partial and is being extended." He paused. "The suite is at what I would characterise as a solid operational standard. Not the ceiling of what's possible. The ceiling of what we've had time to build."
"What's missing," Karan said.
"Chaff and flare dispensers need the automated sequencing finalized," Dhaliwal said. "Currently the pilot manually activates each countermeasure. The automated threat-response sequencing — where the system recognizes the type of incoming threat and selects the appropriate countermeasure without pilot input — is in development. Probably three months from operational standard."
"Three months," Karan said. "It's a priority."
"Understood," Dhaliwal said.
"The Netra-1," Karan said, looking at Rawat.
Siddharth Rawat had the expression of a man who has been working on something for long enough that his relationship with it has become complex. "Operational," he said. "Pulse Doppler, transistor-based. Tracking performance in ground clutter is significantly better than older radar types — the Doppler processing separates moving targets from stationary background, which older radar cannot do reliably." He paused. "Integration with the fire control system is complete. The targeting solution goes from radar acquisition through the fire control computer to the weapons release system. Pilot confirmation required before release. The full sequence from acquisition to weapons-ready — approximately 4 seconds."
"Against a manoeuvring target,"Subramaniam said.
"Longer," Rawat said, without hedging. "A manoeuvring target changes the solution continuously. The fire control computer updates continuously. The pilot releases when the solution quality is acceptable. Against an aggressively manoeuvring target — 8 to 12 seconds for a stable solution."
"That's honest," Subramaniam said.
"It's accurate," Rawat said.
Capt. Ranbir Singh set the Kaumodaki Mk 2 data sheet on the table.
He did it without preamble — just placed it, the way you place something that has earned its position.
"Air-launched from the Makara at altitude," he said. "INS midcourse guidance for the first 170 kilometres — the missile navigates to the target area using its internal navigation system, without transmitting any signal the target ship can detect. Active radar terminal homing engages at 15 to 17 kilometres. Sea-skimming profile at 3 to 5 metres above the surface in the terminal phase." He looked at the naval officers. "Terminal speed — Mach 1.8."
Lieutenant Commander Pillai stopped writing.
He had been writing steadily since the review started. He stopped and looked at the sheet
.
"Total range air-launched — 187 kilometres," Singh continued. "Warhead — 178 kilograms, shaped charge with delayed fuze. The fuze is set to detonate after hull penetration — the missile enters the ship before the warhead activates. Below-waterline detonation."
Pillai looked at Subramaniam.
Subramaniam's expression had not changed. But something in how he was standing was different from before.
"At Mach 1.8 from 15 kilometres," Singh said, "the missile covers the distance to impact in approximately 28 seconds." He paused. "I'll let the naval officers speak to the significance of that timeline."
Subramaniam spoke. "A ship's close-in defence system requires, at minimum, radar acquisition, threat classification, weapons assignment, and first round on target." He said it in the flat tone of someone reciting something they know completely. "Against a fast low-level target in sea clutter, the acquisition step alone is not reliable under 20 seconds for current systems." He paused. "28 seconds from 15 kilometres means the window between detection and impact, if detection occurs at acquisition range, is approximately 8 seconds. That is not a manageable engagement timeline."
"For current systems," Karan said.
For current systems," Subramaniam confirmed. He looked at Karan directly. "You understand that this changes the threat equation for every surface combatant in the Indian Ocean."
"Yes," Karan said. "Including ours. Which is why the next development stage includes understanding how to defend against a missile of this type as well as how to employ it."
Subramaniam held his gaze for a moment. Then nodded.
Singh continued. "The limitation." He said it without waiting for Karan to prompt him, which Karan noted. "The seeker acquisition range — 15 to 17 kilometres — is shorter than a subsonic missile's seeker range of 25 to 30 kilometres. The reason is aerodynamic heating at Mach 1.8 — the thermal boundary layer around the radome affects radar wave transmission. If the INS midcourse navigation places the missile outside the seeker acquisition basket — because the target has manoeuvred or because of navigation error — the missile does not find its target."
"Current INS accuracy at 170 kilometres," Karan said.
"Circular error probable of approximately 380 metres," Singh said. "The seeker basket at 15 kilometres active range is roughly 2.8 kilometres diameter. In nominal conditions, we have comfortable margin. In non-nominal conditions — target manoeuvring, navigation error compounding — the margin reduces." He looked at the sheet. "The Kaumodaki Mk 3 development path addresses both the seeker range and the INS accuracy. For now, the Mk 2 is what we have."
"And what you have," Subramaniam said, quietly, "is the most capable anti-ship missile in the world."
Singh looked at him. "Yes," he said. "It is and Expensive too."
Rathore caught Karan at the edge of the hangar during the midday break.
He had a single sheet of paper. Not the formal kind — handwritten, the notes of a man thinking something through in the margins of whatever was at hand.
"The eleven sorties," he said.
"The third sortie — Mach 1.4 at low level. I was looking for the edge." He paused. "I found it. Not a hard edge — a soft one. The aircraft told me where it was before it got there." He looked at P-1 at the far end of the hangar. "Most aircraft I've flown, you learn the edges by crossing them and recovering, if you're lucky. This one — the FBW, the way the control response changes quality before the limit, not after — it's communicative. It tells you."
"The departure resistance," Karan said.
"The departure resistance is part of it," Rathore said. "But it's more than that. At high alpha — high angle of attack, nose well above the horizon — the LERX vortices energise the wing flow and the aircraft stays controllable well past where a conventional design would be departing. The nose wants to come up and it stays up and you're still flying." He paused. "I've been in aircraft where high alpha is the beginning of an incident. This one, high alpha is a manoeuvre option."
"And the stick grip," Karan said.
Rathore looked at him. "I put that in the fifth sortie report."
"I read the fifth sortie report."
"Then you know the weapons release switch needs to move three centimetres forward on the grip," Rathore said. "At 9G, eight centimetres of thumb travel is a significant workload spike. Three centimetres is natural."
"Tyagi," Karan said, slightly louder.
Rajesh Tyagi Arrived from the direction of P-1 with the immediate responsiveness of someone who has learned that being
findable is a professional virtue.
"Stick grip redesign," Karan said. "Weapons release switch, three centimetres forward Asap. Rathore Sahab will mark the exact position."
"It will take One week," Tyagi said.
"One week," Karan confirmed.
Tyagi looked at Rathore. Rathore gave him a slight nod that meant: I'll come to you this afternoon with the measurement.
Tyagi left. Rathore looked back at P-1.
"The ninth sortie," he said. "I took it to 8.2G in a sustained turn. The pressure suit was working. The FBW was working. The aircraft was working." He was quiet for a moment. "I've been a pilot for twenty-two years. I've flown MiGs, Gnats, Hunters." He paused again. "This is a different category of aircraft."
He said it flatly. Not as a compliment — as an assessment. The way a professional states a conclusion after sufficient evidence.
Karan looked at P-1.
He thought about what this aircraft would do — not in December 1972, but in the years after. The hands it would be in. The situations it would be called into. The men who would climb into it at 0300 from Vikrant's pitching deck with something specific required of them. Whether those men would come back.
The numbers were right. He had checked them many times from many directions and they were right. The aircraft was what the numbers said it was. He had been more certain of this as P-1's flight hours accumulated — each sortie confirming or refining but not fundamentally contradicting.
What he felt standing here was not pride. Pride was too simple. It was something closer to the feeling of having made a promise to people he hadn't met yet and having kept it.
"One more thing," Rathore said.
Karan came back. "Go,on."
"The carrier approach mode in the FBW," Rathore said. "The automatic glideslope and angle-of-attack hold when the gear is down and hook deployed. I've been running simulations." He paused. "It works. But the transition from normal flight mode to carrier approach mode — the moment the gear comes down and the control laws shift — there's a brief transient. Maybe one second where the feel changes. Not dangerous. But noticeable. A pilot who hasn't been briefed on it will compensate unnecessarily."
"Arvind," Karan said.
Arvind Pratap Singh had been within earshot, which was not accidental. "The mode transition logic," he said. "I can smooth it. It's a software rate limit on the control law transition. I'll have a revised version for ground test by the end of the month."
"Good," Karan said. He looked at Rathore.
"Anything else from the eleven sorties."
Rathore thought about it. Actually thought about it — ran through the sorties in whatever order he kept them. "No," he said.
"Everything else is in the reports."
"I've read the reports," Karan said.
"I know," Rathore said. "I can tell from the questions you don't ask."
Dr. Padma Krishnan found Karan at three in the afternoon.
She had the fatigue data printout under her arm and the expression of someone who has confirmed something and is deciding how to present it.
"P-1's main gear attachment," she said.
"Show me," Karan said.
She spread the printout on the nearest flat surface. "The fatigue accumulation at the main gear attachment points — I've been running the data from the landing simulations and the actual arrested landing tests. The attachment is performing within design limits." She pointed at a specific curve on the printout. "But the load distribution is not uniform across the attachment. This cluster of fasteners is carrying 34% more load than the design assumed. Over a high-sortie-rate carrier deployment — say 200 landings per year — this cluster reaches its fatigue limit at approximately year four."
"Which means," Karan said.
"Which means at year four you have an inspection requirement that you don't currently know about," Krishnan said. "If you know about it, you inspect and replace. If you don't know about it—" She paused. "You discover it the wrong way."
"The secondary load path," Karan said. "A second attachment structure running from the main gear beam to the secondary spar. Distributes the load, eliminates the concentration, adds redundancy."
Krishnan looked at him. "That's the solution I was going to propose."
"It changes the gear bay structure," Karan said.
"Moderately," she said. "Not extensively. And it also gives you redundancy — if the primary attachment is damaged in a heavy landing, the secondary path retains the gear long enough to matter." She paused. "It's worth the structural change."
"Do it," Karan said. "I want the revised design before the end of January."
"End of January," Krishnan said. She gathered the printout. Then, almost as an afterthought — the afterthought of someone who had been thinking about it since she found the data: "The team that designed the original attachment made a reasonable choice with the information they had. The data that reveals the issue only exists because P-1 has flown eleven times. You can't know what the aircraft will tell you until you fly it."
"No," Karan said. "You can't, that's why live testing is i trust"
She nodded and walked back toward the structural testing station.
He watched her go. The quality of a person who finds what others miss not because she's looking harder but because she's looking differently. The program had more of those than most programs did. He had tried to find them specifically.
Ramanathan and Venkatesh finished their joint analysis at six in the evening.
It had taken longer than expected — not because the aerodynamics were difficult but because the structural side had a complication that Venkatesh had introduced at two in the afternoon and spent three hours resolving.
He had found a sign error in his own moment calculation.
The moment he found it, he sat back from the drafting table and looked at it for a full thirty seconds without saying anything. Ramanathan, who had been working on the aerodynamics sheet beside him, noticed the silence and looked over.
"What," Ramanathan said.
"My moment calculation for the transition load at 25 degrees," Venkatesh said. "I had the sign wrong on the torsional component." He pointed at the line. "The load is in the other direction."
Ramanathan looked at it. He was quiet for a moment. "Which means."
"Which means the transition loads at 25 degrees forward sweep are lower than I calculated," Venkatesh said. "Not higher. The wing mechanism is working with the aerodynamic moment during the transition, not against it." He paused. "The structure supports 25 degrees. Comfortably."
Ramanathan looked at his aerodynamics sheet. Then at Venkatesh's corrected structural sheet. Then at the two of them together, which confirmed what Karan had said nine hours ago.
"How long have you had that sign error," Ramanathan said.
"Since October," Venkatesh said. He said it the way you say something when the honest answer is also the uncomfortable one.
Ramanathan was quiet. Then: "We've been arguing about this since October."
"Yes," Venkatesh said.l
"Because your calculation said the transition loads were too high."
"Yes."
"And they were never too high."
"No," Venkatesh said. "They were never too high."
Ramanathan set his pen down. He looked at the data for a long moment. When he spoke, his voice had a quality that was something between frustration and the specific kind of acceptance that comes from understanding something about how problems actually work.
"We should have run the crosswind case first," he said. "That would have shown us immediately that 25 degrees was right and we would have started from there instead of trying to talk ourselves out of it."
"We optimised for the nominal case," Venkatesh said. "It's what you do. It's what I always do."
"And the aircraft doesn't live in nominal cases," Ramanathan said.
"No," Venkatesh said. "It lives on Vikrant. In the Arabian Sea. In whatever conditions exist that day."
They gathered the sheets. Eight pages of work that had taken nine hours and had arrived, via a sign error and its correction, at a number that had been stated at eight in the morning.
"He knew," Ramanathan said.
"He usually knows," Venkatesh said.
"I find it consistently irritating," Ramanathan said.
"That's a consistent pattern in how you describe him," Venkatesh said.
"It's a consistent experience," Ramanathan said. Then, after a pause: "Also consistently useful. Both things. In equal measure and simultaneously."
By seven in the evening, most of the team had gone home or to the residential facility on site.
The hangar itself never fully emptied — there were always people with something that couldn't wait, something that made more sense to finish than to leave. Tiwari was probably still in the engine bay. Krishnan's structural analysis light was on. Dhaliwal had sent a message at five that said only EW suite sequencing — found something — staying.
The hangar lights stayed on. They always stayed on.
Karan stood in front of P-2 for a while.
Both engines in. The electrical and fluid connections complete. The data systems woken up three days ago, running their self-checks, confirming that the numbers the bench had produced were numbers the installed configuration would reproduce.
March.
He had said March to enough people now that it had acquired the quality of a commitment rather than an estimate. March was not far. Fourteen weeks. Fourteen weeks to complete the integration checks, run the ground vibration tests, complete the taxi trials, work through the pre-flight sequence until it was reliable enough that Rathore could climb in and trust it.
He thought about Rathore's words from earlier. A different category of aircraft.
He thought about what that category meant — not in terms of performance numbers, which were real and confirmed, but in terms of what it represented for the country that had built it. India had been buying its aircraft from Britain and the Soviet Union, taking what was offered, operating on someone else's development timeline, carrying capabilities that someone else had decided were appropriate. What was in this hangar had been decided here, by people here, for requirements specific to INS Vikrant and the Indian Ocean and the strategic situation that existed in this part of the world.
Nobody had given them permission to build it.
Nobody had needed to.
He looked at the aircraft for a long time.
The variable geometry wings were in the forward position — the position for slow flight, for launch, for the conditions of Vikrant's deck. The leading edge root extensions ran forward along the fuselage with their particular intent. The Netra-1 radome in the nose was smooth and slightly reflective under the hangar lights.
He thought about the men who would eventually fly this. Not Rathore — Rathore was the proof of the concept. The men who would take it into service. Young men, probably, with the specific combination of skill and aggression and judgment that carrier aviation required. Men who would fly from a pitching deck at three in the morning into whatever was waiting on the other side of the horizon, trusting an aircraft to do its part if they did theirs.
The aircraft would do its part.
He was certain of that in the way you're certain of things you've checked from many directions and found consistent.
He stayed a few minutes longer. Not because there was something more to see — because some things deserve the time to be properly acknowledged before you move on from them.
Then he picked up his coat and walked toward the door.
Behind him, P-2 stood in the hangar light. The engines were cool now, inactive, holding the heat of the day's calibration runs in the metal of their turbine housings. Waiting with the patience of machines that have been built correctly and know, in whatever sense machines know anything, that they are ready.
March was not far.
Outside, the December night was cold and clear, the stars sharp and close in the winter air above Gorakhpur. Somewhere in the northwest, INS Vikrant was at anchor. Her pilots were asleep, or on watch, or thinking about nothing in particular — unaware that fourteen weeks from now, an aircraft was going to exist that would change what they could do and where they could go and what the sea could ask of them.
Karan looked at the sky for a moment.
Then he went home.