Hypersonic Vehicles, Modeled in Days: How Specter Aerospace Accelerates Hypersonic Design with nTop

Hypersonic vehicle development is stuck in a loop most people don't see.

You start with customer requirements.

What is the mission? That drives requirements for speed, range, payload, and everything else.

Then you need to generate vehicle concepts, including outer mold line, propulsion integration, and thermal management. What is the shape? Will it package? Will it survive?

Here is where traditional CAD and low-fidelity methods can’t keep up.

Change a sweep angle and the feature tree breaks. Resize the fuselage, and downstream features fail. You're not exploring design space, you're debugging geometry. By the time you've hand-crafted 3-4 configurations in NX, CATIA, or SolidWorks, weeks have passed. The program moves to Excel and PowerPoint because nobody can move geometry fast enough to test real hardware.

Then, if you actually pick a concept and move to detailed design, the same problem repeats at the component level. Regeneratively cooled combustors, integrated ducts, complex internal structures—every design change touches multiple features. One iteration takes days or weeks. Testing gets pushed out. Competitors move faster.

The bottleneck isn't analysis. CFD tools are good. The bottleneck isn't manufacturing—additive and composite tooling can build complex geometries. The bottleneck is geometry creation itself.

Specter Aerospace needed a workflow that works from requirements to testing hardware, not just faster CAD. A fundamentally different approach.

The solution:

nTop Across the Program Lifecycle

The breakthrough isn't just that nTop is faster–it's that conceptual and detailed design can happen simultaneously.

While the OML is being refined, component-level engineering is already underway. Traditional programs wait months between these phases. Specter runs them in parallel.

The key insight: the same parametric modeling approach that works for vehicle-level concept exploration also works for detailed component design. One platform, one workflow, requirements-to-hardware.

Track 1: Conceptual Design

Building the OML in 3 Days

In a 3-day workshop, Specter and nTop built a parametric hypersonic vehicle model: combustor, duct, and outer mold line. Fully tied to performance parameters, ready for simulation.

I have never seen something as useful as nTop for conceptual design. I've been in concept design for hypersonic vehicles for 25 years.

Joel Malo-Molina

Technical Director

Specter Aerospace

That's 25 years of defense aerospace experience (programs you've heard of, programs you haven't) saying this is the best tool he's seen for this phase.

How the OML Model Works

The model is driven by performance parameters:

• Fuselage diameter and length
• Wing sweep angle and aspect ratio
• Inlet position and geometry
• Combustor-to-airframe integration points
• Duct routing and cross-section

Change any of these and the model updates instantly. Every time.

This isn't a static geometry file. It's a design space generator. The team can explore inlet configurations, evaluate different combustor integration strategies, and resize the vehicle for different payload requirements all while maintaining simulation-ready geometry.

The model feeds directly into their MDAO (Multidisciplinary Design Analysis and Optimization) workflows. No geometry cleanup, no manual rework, no CAD repair sessions.

If you can increase the fidelity of your concept design, it means the work you need to do to get to final production starts dropping like crazy. If you don't have those go-backs and you already understand your sensitivities, you save so much time

Arun Chundru

CTO

Specter Aerospace

This is the compounding effect. When you can explore the design space thoroughly at the concept phase, you don't end up rebuilding hardware in detailed design. The savings multiply downstream.

The Metrics That Matter

This work goes from about two to three weeks of someone's work to one or two days to get a concept design.

Arun Chundru

CTO

Specter Aerospace

Track 2: Detailed Design

The Regeneratively Cooled Combustor

Once the OML is locked, Specter moves to detailed component design. This is where most teams hit another wall.

Regeneratively cooled combustors use fuel flowing through small channels in the combustor wall to cool the structure before burning. The channel geometry is complex: hundreds of parallel channels, varying cross-sections, precise wall thicknesses. It's a heat exchanger that also has to survive hypersonic flight loads.

We typically do this in SolidWorks. Although it's great for doing detailed design, things get really hard when we want to make changes. When we partnered with nTop, one of the first things we went after was making a generalized model for a part like this.

Jake Brown

Propulsion Engineer

Specter Aerospace

The Traditional CAD Problem

In SolidWorks, this combustor is built with:

• Base geometry (combustor body)
• Pattern features (cooling channels)
• Boolean operations (subtracting channels from the body)
• Downstream features (mounting interfaces, fuel inlets)

Want to change the number of channels from 80 to 100? The pattern regenerates, the Boolean operations recalculate, and something breaks. Maybe the mounting interface. Maybe the fuel inlet intersection. You spend hours debugging the feature tree.

Want to change channel width to improve heat transfer? Same problem.

Each iteration: over a day of engineering time.

The nTop Approach

The Specter team built a fully parametric combustor model in nTop. Channel count, channel dimensions, wall thickness, and combustor length that all update in real time with changes.

We can change the amount of channels, how big they are, different dimensions, all in real time. And it's allowed us to take this fluid volume, export it to a CFD software, to our own codes, and really figure out what the optimum sizing of that heat exchanger can be for any given design point

Specter Engineer

Zoom image

The model exports clean fluid volumes directly to their CFD solver. They run thermal analysis, evaluate heat transfer performance, adjust parameters, and re-export. The iteration cycle: 20 minutes instead of a day.

That's a 10x compression in iteration time that enables better decisions earlier in the program.

Why Traditional Parametric CAD Breaks

Traditional parametric CAD forces a choice: hand-craft every variant (too slow) or build parametric models that collapse under variation.

The problem is boundary representation (B-Rep). Every union, subtraction, or intersection requires perfect topological stitching. Surfaces have to trim exactly. Edges have to meet precisely. When you change a dimension, the topology changes, and the stitching fails.

A Real Example: Inlet Reconfiguration

What happens in traditional CAD:

1. Inlet surface intersects fuselage at a different location
2. Duct path changes to accommodate new inlet angle
3. Combustor mounting interface shifts
4. Structural load path updates
5. Model breaks at one or more intersection points
   

An engineer spends hours (or days) manually:

• Repairing surface intersections
• Re-trimming boundaries
• Recreating downstream features that depended on the old geometry

What happens in nTop:

1. Change inlet angle parameter to 15°
2. Implicit field recalculates
3. Geometry updates
4. Export to CFD

No repair. No rework. Ready for the next analysis iteration

Zoom image

Requirements to Engine Test: 4 Weeks

Here's Specter's actual timeline using nTop for a new combustor configuration:

• 1 day: Design in nTop (parametric model updates + CFD export)
• 15 days: Manufacturing
• 4 days: Assembly, final machining, test prep
• Total: 4 weeks from requirements to engine test fire

Zoom image

Compare that to traditional aerospace development timelines measured in years. Even "fast" programs take 6-12 months from design freeze to first hardware test.

Specter isn't just incrementally faster. They're operating at a different cadence entirely.

Why This Matters for Defense Programs

DoW is pushing for reconfigurable, modular hypersonic systems. The requirement isn't subtle: they want platforms that can be adapted for different missions, different payloads, different ranges.

The main thing is, you need to iterate fast to win programs. You need to show you can be 100x or 1000x faster. The DoW wants reconfigurable, modular hypersonic systems, and most primes can't move at that speed. We're really impressed with the Specter team's leadership in building these next-gen engineering workflows.

Brad Rothenberg

CEO

nTop

When you're constrained by traditional CAD, "reconfigurable" means years of re-engineering. When you have parametric models that actually work, reconfigurable means parameter updates and a few weeks of testing.

If I can get 20% of the range for 10% of the cost, that changes the game. And with nTop, we can show the full cost-performance menu.

Arun Chundru

CTO

Specter Aerospace

This is what DoW actually buys: options backed by analysis. Not theoretical trades in a PowerPoint. Real configurations with simulated performance and cost breakdowns.

"By being able to rapidly go from designs to performances, we can explore the design space and meet or exceed current system ranges at a fraction of the cost, enabling affordable mass without compromising performance."

"Affordable mass" is the DoW term for "we can build enough of these to matter in a real conflict." Not 10 exquisite systems that cost $200M each. 1,000 systems at $2M each that you can actually field.

What's Next: Hypersonic Vehicle in a Day

Specter is connecting their parametric nTop models to automated CFD and MDAO workflows.

The goal: generate a complete hypersonic vehicle concept—OML, propulsion integration, structural design—with performance estimates and manufacturing consideration, in a single day.

Not a sketch. Not a PowerPoint concept. A simulation-ready model with CFD results, structural analysis, and cost estimates.

Long-term vision: requirements to test hardware in under two weeks.

See how a 3-5 day Design Sprint can deliver tangible results on your active programs with parametric models, performance insights, and toolchain integration. This is co-engineering on problems that matter.