LandSpace Completes IPO Tutoring, Zhuque-3 Deputy Chief Engineer Discusses Bottlenecks in Commercial Space Recovery Technology | Xinxing PORTFOLIO

Recently, the China Securities Regulatory Commission's online service platform disclosed that LandSpace, a Heart Capital portfolio company, has officially completed its IPO tutoring process. LandSpace is poised to make a run at becoming the "first commercial space stock."

Founded in 2015, LandSpace was among China's earliest private rocket companies. It is dedicated to building a full industrial chain of "R&D, manufacturing, testing, and launch" centered on medium-to-large liquid oxygen-methane launch vehicles.

In July 2023, LandSpace achieved orbit with the second flight of the Zhuque-2 (ZQ-2 Y2), the world's first liquid oxygen-methane launch vehicle.

On December 3 this year, LandSpace's self-developed Zhuque-3 test rocket (code name: ZQ-3 Y1) successfully completed its maiden flight, with the second stage precisely entering its designated orbit. The mission also conducted first-stage recovery verification, marking a notable breakthrough for China in the reusable launch vehicle sector.

Recently, Dong Kai, deputy chief designer of the Zhuque-3 overall system at LandSpace, sat down with Tech Buzz China for an interview. From the perspective of a frontline engineer, he systematically broke down the real wins and losses of the Zhuque-3 maiden flight, delved into why China's recoverable rockets have broadly converged on the "Falcon 9 configuration," where exactly the recovery attempt proved most difficult, and what reusable rockets mean for the future of China's commercial space industry. Audio link below:

This article is adapted from the Tech Buzz China podcast.

Reading time: 18 minutes.

What Kind of Rocket Is Zhuque-3?

Yaxian: Welcome back to the show, Mr. Dong. Your last visit was in October last year, right when SpaceX was doing its "chopsticks catching rocket" and Zhuque-3 had just completed its 10 km vertical takeoff and landing test. At the time, you previewed a test flight in 2025—who knew it would happen by year-end. As China's first rocket to attempt recovery on its very first launch, we'd love to hear about this launch and the development of reusable rockets.

Dong Kai: Zhuque-3 is our liquid oxygen-methane, medium-to-large reusable rocket. The maiden flight vehicle launched on December 3 was slightly shorter than the "full version"—netizens nicknamed it the "youth edition." The full version stands 76 meters tall, with nine engines on the first stage and one on the second, and the body is 80% stainless steel. The maiden flight had the same configuration, but used the TQ-12A engines with 83 tons of thrust; the full version will use the TQ-12B with over 100 tons, so the body was about 10 meters shorter. For safety on the maiden flight, the recovery range was set at 390 kilometers; at maximum capacity, it could reach 550–600 kilometers.

Is Zhuque-3 a "Domestic Falcon 9"?

Yaxian: Many netizens call Zhuque-3 the "domestic Falcon 9." What's your take on this comparison?

Dong Kai: The Falcon 9 is a proven configuration validated by engineering. We studied it and recognize its rationality—this is learning, not copying. I consider "domestic Falcon 9" high praise. It's the world's only stably operating heavy reusable rocket, a target we aspire to ("see the worthy and seek to equal them"). It launches about 150 times a year, with a record of 32 reuses for a single booster—that's the direction we're working toward.

Yaxian: The design also borrows from Starship, like stainless steel body and liquid oxygen-methane engines. Elon Musk said Zhuque-3 might surpass Falcon 9 within five years. Is that realistic?

Dong Kai: Engineering borrowing requires understanding the why. As for surpassing it in five years, I'm actually not that optimistic. Falcon 9 alone launches about 150 times a year, roughly three times a week. All of China's space launches across all models this year total about 100. For one company to reach that frequency is extremely difficult—it takes an entire ecosystem. I prefer to see his comment as encouragement and motivation.

Yaxian: Without reusability, could existing technology achieve three launches a week?

Dong Kai: Very difficult. This isn't an either-or. Manufacturing capacity expands the base; reusable technology is a multiplier. We focus more on "delivery capacity"—the weight of payloads put into orbit per unit time. If a recovered rocket can relaunch in ten days, one vehicle can fly multiple times a year. Reusability isn't the only path, but it's currently the fastest and most effective.

Yaxian: This would also dramatically reduce costs.

Dong Kai: Yes, but more precious than material cost is time cost. In engineering, time is the scarcest resource. If I could trade twice the money for half the time, I'd do it without hesitation. Reusable technology is precisely about winning us time.

Yaxian: Let's talk about this launch. From project initiation to now, it's been just over two years. What goals were set before the maiden flight, and how were they met?

Dong Kai: Project initiation was in August 2023, with the plan formally announced in December. Last year the company said there would be three launches in 2025. That target wasn't particularly stressful for our team—it was based on actual production capacity at the time, and the hope was to complete first-stage recovery within those three flights. The company actually had greater tolerance for the maiden flight goals.

Of course, as a launch mission, we had clear evaluation criteria. When the rocket passed its factory review in September, the flight outline separated orbit insertion and recovery into independent evaluations. For orbit insertion, based on perigee altitude, there were three grades: "complete success," "success," and "failure." For recovery, it was simple: standing on the landing site for one minute without falling over was success; otherwise, failure. Only both together constituted mission success.

Before launch, we set four standards: successful orbit insertion with perigee above 170 km and successful recovery would be "complete success"; perigee above 130 km would be "success"; achieving only one of orbit insertion or recovery would be "partial success." In the end, the commander announced "complete launch success," referring to orbit insertion fully meeting standards; the overall mission was "partial success" because recovery wasn't achieved.

From the team's internal perspective, this result is acceptable. But all our preparations and inner expectations before launch were aimed at "complete success"—hoping for an undisputed victory. So when we saw the recovery segment proceed to 4 km from the ground, about 2 km before landing ignition, we really almost thought we would succeed. At that final moment, some colleagues on site were so moved they shed tears—it was still somewhat regrettable.

But overall, it met expectations. Before launch, I told everyone that as long as we could hold out to reentry ignition (around 80 km altitude on the return leg), we would have already crossed a major threshold—this was expectation management. But we definitely aimed for the best and prepared for the worst. In fact, we had contingency plans for even worse scenarios.

Yaxian: Watching the recovery video, the landing point was just a hair from the designated site. Online, some said quite a few cameras nearby were destroyed in the explosion.

Dong Kai: The landing point was indeed extremely close. From a guidance perspective, once we passed 40 km altitude and shut down reentry ignition, the subsequent sequence was basically what we'd already validated in our 10 km flight test. At landing ignition, we were already very close to the recovery ship—colleagues could see it with the naked eye, someone was shouting "It's coming, it's coming." At that moment, we were truly very close to success. But engineering is like this: success is success, failure is failure—there's no intermediate state.

How Would You Score This Maiden Flight Engineering-Wise?

Yaxian: If you had to give this mission a score from the heart, what would it be?

Dong Kai: I've seen evaluations saying orbit insertion was a required question, full marks; recovery was a bonus question. Some leaders encouraged us, saying this was already 95 points. But the charm of aerospace engineering is that it doesn't need self-promotion or subjective scoring. As our program chief commander Director Dai used to say: for rocket launches, success and failure are 0 and 1.

So currently, orbit insertion is 1, recovery is 0. Although from a time perspective, the second stage completed its full 3,000-plus-second process, and the first stage's actions were normal for the first 450 seconds (about 90% of the time), falling just short at the end. But quantitative scoring is subjective—netizens or outsiders can comment freely. As the person responsible for this model, I won't engage in such controversial quantification: without success, it's 0.

Yaxian: You mentioned earlier that three launches were expected this year, but aerospace is often called the "pigeon industry"—delays are common. Does this create pressure in your work? Do you consider it normal? This launch was indeed postponed several times due to various factors. Were those days just waiting around?

Dong Kai: Honestly, from a work perspective, colleagues definitely feel pressure facing such goals. But this pressure differs from typical fatigue. Personally, engaging with natural science and engineering problems, I even somewhat enjoy it. This pressure actually brings a sense of fulfillment in life, alleviating anxiety. My most stressful days were precisely when the launch date was set and then postponed again and again. When there are engineering problems to solve, I have things to do and don't feel much pressure; waiting for dawn, waiting for launch—that's when all kinds of restlessness and pressure emerge.

In aerospace work, there's a method called "dual thinking"—anticipation and retrospection, essentially what The Art of War describes as "consider defeat before considering victory." Once the rocket is on the launch pad, there's very limited room for salvation—at most some operational adjustments to mitigate risks. If it's a design issue, you can only think about it. But slow work produces fine results; this slowness isn't during those launch days. From the moment the rocket transferred to the launch site on October 20, I told colleagues: the upper limit of success or failure is already set. What we must do is avoid deducting further points through our own work.

Yaxian: As chief designer, being deeply involved in this process must have been quite an emotional rollercoaster?

Dong Kai: Not too bad. Definitely more anxious than when there's concrete work, but the swings weren't that large. After all, from the transfer onward, everyone had already adjusted their mindset: the upper limit is set, do our best.

Why Choose the Liquid Oxygen-Methane + Stainless Steel Route?

Yaxian: Let's return to Zhuque-3 itself. Though it's often said to benchmark Falcon 9, Falcon 9 uses an aluminum alloy body with liquid oxygen-kerosene, while Zhuque-3 uses stainless steel and liquid oxygen-methane. How did you make these technology choices at project initiation?

Dong Kai: We don't actually consider liquid oxygen-methane or stainless steel as inherently "more advanced" in engineering. There's no absolute advancement in engineering; all choices are about finding the optimal path for tactical execution within an unchanged grand strategy.

LandSpace's grand strategy is clear: solve the three core challenges of launch vehicles—heavy lift capacity, high frequency, and low cost. Expendable rockets can't achieve this through capacity expansion alone—it's an "impossible triangle." So only reusable rockets can realize this strategy.

On this foundation, one of my technical views is: reusable rockets must be designed for reuse from the very beginning, in every aspect of overall design. The features everyone notices about Zhuque-3—stainless steel, liquid oxygen-methane, grid fins, landing legs—are just the visible ones; there are many internal design elements. Overall, every design step prepares for increasing reuse frequency. Recovery is the foundation; only rapid reusability completes the technical closed loop. If recovery takes six months or a year of refurbishment, it loses meaning.

An important reason for choosing stainless steel is that it doesn't require external thermal insulation, greatly reducing post-recovery maintenance needs and cycles. Although Falcon 9's aluminum alloy body gets scorched black without repair, stainless steel is more heat-resistant and can return without insulation. Liquid oxygen-methane propulsion is similar. Kerosene engine coking isn't a core difficulty to address with current technology, but it always takes time to clean—sometimes ten days to half a month. Liquid oxygen-methane engines simply don't need this procedure. When reuse frequency is calculated in days or even hours, this advantage becomes enormously significant.

This is from the forward design perspective. Second, from LandSpace's own "tactical realizability." Talking strategy detached from implementation capability is meaningless. Our choices—flying the TQ-12A first, specific stainless steel production processes, designing the "youth edition," setting the maiden flight recovery site at 390 km—all serve to most reliably achieve technical validation and iteration within existing capabilities.

Yaxian: Mr. Dong, could you specifically explain how recovery range affects payload capacity?

Dong Kai: This depends on rocket scale. For Zhuque-3, the full version is designed for about 600 km recovery range. In that design, after first-stage separation, inertia alone carries it to the recovery site 600 km away, without needing extra fuel for lateral deceleration.

But for this maiden flight, for absolute safety, we set the recovery site at 390 km. This meant I had to separate the first stage earlier (from about 76 km down to 66 km), sacrificing some velocity increment and making the second stage work harder. Every 10 km increase in recovery range adds roughly 150 kg to Zhuque-3's payload capacity. On the maiden flight, we landed with about 10 tons of propellant remaining; if that energy were used to fly farther to the recovery site, payload capacity would immediately increase by 2–3 tons. So once recovery technology matures and safety is validated, we'll extend to the design range and fully release payload capacity.

Yaxian: Previously you expected a mid-year maiden flight, but it ended up being year-end. The original plan was three flights this year—does the current pace seem slower than expected?

Dong Kai: First, the three-flight annual target was hoping to achieve recovery within those three. When we knew there would only be one opportunity this year, naturally we hoped to "capture both kings in one battle." Regarding pace, aerospace engineering often follows "aim for the top, settle for the middle."

Zhuque-3 was formally initiated in August 2023, with scale approaching the Long March 5—classified as a large rocket—plus the extremely high difficulty of reusability. By conventional experience, having such a rocket achieve maiden flight orbit insertion by 2026 would already be decent. But we went from drawings to a successfully flying vehicle in two years—this itself is already a miracle.

We did strive for earlier targets. Without the impact of the August Zhuque-2 Y3 failure and its subsequent investigation, Zhuque-3's maiden flight could potentially have been earlier. Although the team had doubts about ambitious goals like "achieve recovery in 2025," doubt doesn't affect our full commitment to action. When you've done everything possible for a goal without regret, any outcome can be accepted calmly.

Yaxian: Falcon 9 took nearly a decade from project initiation to achieving recovery. Zhuque-3 has only been two-plus years—do you consider that fast?

Dong Kai: Not just you—by my personal experience, it's extremely fast. Although I say it could be faster, internally I feel that for the team to accomplish this in two years not only exceeds my previous professional experience, but even approaches the legendary stories in Chinese aerospace history like the Long March 2E going "from drawings to maiden flight in 18 months." It feels like moving from hearing myths to participating in creating them—very gratifying.

Yaxian: Does time create great pressure? For example, from internal high targets, or external competition from other companies and the "national team"?

Dong Kai: Our team is relatively mature—we won't distort our technical route to compete for something. But on the other hand, "time and tide wait for no man"—if I can do it, there's no reason to waste time. External competitive pressure is somewhat like the "hardship" in "born in hardship"—it's a good thing. It drives us forward, but our engineering decisions aren't distorted by it. Healthy competition is an important driver of technological development.

Starship has been sharpening its sword for over a decade; its success is well-deserved. Rankings and "first place" vanity don't affect my value judgment and pride in my own work. Even if we're the fourth or fifth to succeed, as long as we achieve it ourselves, the inner satisfaction is the same.

How Was the First-Stage Recovery Process Designed? Where Were the Difficulties?

Yaxian: Alright, let's return to what everyone cares about most—the recovery. Mr. Dong, could you break down the first-stage recovery process and its difficulties?

Dong Kai: The process is roughly: after first-second stage separation, the first stage coasts to about 130 km apogee, simultaneously deploying grid fins and adjusting attitude. At about 80 km altitude, "reentry ignition" occurs—three engines fire for about 40 seconds, reducing velocity to safe range, then shut down. From 40 km to 4 km is the "aerodynamic control segment," relying on grid fins and strake wings for attitude control and deceleration. Finally, at about 4 km altitude, "landing ignition" occurs.

The difficulties and validation focuses this time were: first, whether "reentry ignition" is reliable in the environment at 80 km altitude—this was our first actual validation. Second, the actual effectiveness of the "aerodynamic deceleration segment"—real data was needed. For this, we configured all five engines in the landing segment to be capable of ignition, intending that if deceleration was insufficient, all five could "brake hard" together. Based on this flight's data, we now have confidence in aerodynamic deceleration effectiveness; going forward, we may not need so many engines firing simultaneously, which actually improves reliability.

Data indicates that in the end, the center engine ignited successfully, but the four outer engines may have had issues before thrust was fully established. The specific cause is still under rigorous analysis. Aerospace failure investigation is like criminal investigation—all evidence chains must be completely closed; you can't rely on partially matching speculation.

Yaxian: Using the 12A engine instead of 12B, and aluminum alloy for part of the body rather than all stainless steel—were these compromises to get to flight faster?

Dong Kai: Beyond the engine, there were no compromises. The stainless steel choice wasn't entirely for speed either. For the first stage, LandSpace's current stainless steel processing technology is quite mature; the maiden flight vehicle's tank wall thickness was under 2 mm, and in the future we can achieve 1.5 mm or even 1 mm—lighter than aluminum alloy. Our published payload capacity is calculated based on the stainless steel solution. For the second stage, this time using aluminum alloy was partly for maiden flight conservatism, and partly because this second stage's reentry into the atmosphere can obtain valuable real data on stainless steel tank reentry thermal environments for our next-generation fully reusable rocket. All our choices serve a longer-term technology roadmap.

Yaxian: SpaceX follows a rapid iteration, failure-tolerant model. Chinese aerospace tradition leans more toward stability. Are these two paths compatible?

Dong Kai: Rapid iteration and "blowing up" aren't necessarily linked. Pursuing high reliability and rapid iteration aren't contradictory—you can have both. What truly limits iteration speed is first and foremost production capacity. If you can only build one per year, you definitely can't afford to take risks. SpaceX also pursues extremely high reliability in ground testing; it doesn't trade reduced reliability for iteration speed.

For LandSpace, we have corresponding production capacity preparations. As mentioned before, the 2026 plan is ten rockets, with corresponding engine capacity (10 engines per rocket, 100 for ten rockets) laid out for this target. Meanwhile, our country's highly valued safety supervision—I don't consider this a limitation, but rather the prerequisite and guarantee for the industry's steady long-term development, and a responsibility enterprises must bear.

Commercialization and Next Steps: Where Does Zhuque-3 Go From Here?

Yaxian: Commercial space ultimately needs to make money. In your view, when and how do we make money?

Dong Kai: This somewhat exceeds an engineer's knowledge scope—there are many people who study these questions more thoroughly than I. But I believe a basic logic: as long as the launch service I provide has value—cheaper, faster, more stable than others—there will always be buyers. We're like the "road builders." SpaceX's truly profitable business now may be application services like Starlink—this is the economic effect of "after the road is built." My job is to build the "road" well and cheaply, laying the foundation for creating value. How to make money based on this road—I trust more professional people will consider that.

Yaxian: From maiden flight to achieving "low cost, heavy lift, high frequency," what iterations are still needed?

Dong Kai: The technical route already has benchmarks; the direction is clear. I may be slightly more conservative than Musk's "catch up to Falcon 9 in five years." But in the next two to three years, we hope to significantly close the gap. If our own "Long March 5-class" rocket can launch 50 times a year (equivalent to once a week), that would already be an order-of-magnitude leap. The arrival of the great space era definitely requires joint innovation in technology and management (including approval processes). I have great confidence in the country's pragmatism and wisdom in this regard—regulatory innovation will keep pace with technical iteration.

Yaxian: Finally, looking ahead at LandSpace's longer-term plans.

Dong Kai: We do have them. First, our 200-ton-class full-flow staged-combustion cycle liquid oxygen-methane engine (benchmarking "Raptor") completed full-engine testing in May this year and is currently iterating—this prepares for future larger fully reusable rockets.

For Zhuque-3, the maiden flight is just the beginning. We've adjusted plans, hoping to achieve successful recovery by mid-2026. Once successful, our three already-produced rockets will rapidly enter iteration, including structural weight reduction, recovery range extension, upgrading to TQ-12B engines, etc., with the goal of evolving to the "full version" around 2026. After that, the focus shifts to shortening launch cycles and increasing frequency.

If the second flight achieves successful recovery, we plan to use a reused first stage on the fourth flight—only then is the technical closed loop truly realized. We hope to shorten the reuse cycle to about 7 days, build additional launch pads, and construct recovery sites for different launch azimuths to meet future large-scale constellation launch demands. As for larger-scale fully reusable rockets (the so-called "Chinese Starship"), the company already has vision plans around 2030. We're developing the 200-ton-class engine precisely for this. Just as two years ago people thought Zhuque-3's 2025 maiden flight was very challenging, we've walked through step by step.

Yaxian: If building such a large rocket, would you also use the "chopsticks" method to recover it? In your view, how many years of gap remain between China's space industry and the United States or SpaceX?

Dong Kai: "Chopsticks" catching is one option. When rocket scale grows large enough, landing legs may go from "prohibitively expensive" to "technically difficult to implement." At that point, "chopsticks" or similar solutions may become the deterministic choice. Without better alternatives, engineers will choose known, feasible paths.

Regarding the gap with SpaceX, I personally find such objective quantification difficult. The gap has been continuously narrowing. From 1970 when we launched our first satellite, while the United States had already landed on the moon the year before—the gap was enormous then. But now, look at our space station's rapid construction, emergency space rescue capabilities—the catching-up speed is evident. I've never thought about precisely calculating the distance to the front-runner. I just run with all my might; the goal is to catch up and surpass. Perhaps when the day of surpassing truly comes, like our aviation industry's development—from chasing generational gaps to sixth-generation fighter maiden flights. At that point, what we need to surpass is only ourselves. As long as we can continuously surpass ourselves, that's a good thing.

Yaxian: Watching rocket launch videos is always stirring. Last year we left a thread—Zhuque-3 launching this year. This year let's leave a new thread—hoping to see Zhuque-3 successfully recover next year (2026)!

Heart Capital was founded in 2022 as an early-stage venture capital fund focused on technology and digitalization in China. The Heart Capital team is primarily composed of Yan Han, founding partner of Lightspeed; core investors; a CFO; and senior investors from industry. The team's past investments include Series A investments in Xpeng Motors (NYSE: XPEV, 09868.HK), Full Truck Alliance (NYSE: YMM), as well as FinVolution (NYSE: FINV), RoboSense (02498.HK), Baichuan, Manman Lengyun, Dedao, World Logistics, MinoSpace, LandSpace, Lanhu, Starfield, and others. Rooted in China with a global outlook, Heart Capital is committed to finding true value in non-consensus. Heart Capital respects the value of "people" and advocates the potential of the "heart," looking forward to accompanying more young Chinese entrepreneurs to strengthen China and go global.