Compare Small-Turbojet Design Proposals
Weigh competing small-turbojet proposals: normalize the design point, pin down TSFC basis, read the off-design deck, and rank evidence over confidence.
Two proposals land on your desk. One promises 12.8 kN, the other 13.3. Choosing on that line alone teaches you nothing: thrust quoted without a flight condition, a mass basis and a fuel figure beside it will not survive the first design review. The proposals that differ in the headline often differ much more in the fine print, and the fine print is where the aircraft lives.
This is the order of operations for comparing them like an engineer.
Normalize the comparison first
Before comparing any two numbers, force them onto common ground. Same flight condition: altitude and Mach, not "takeoff" for one bidder and "cruise" for the other. Same atmosphere: a standard day, unless you have a declared hot-day requirement. Same installation basis: uninstalled (bench) thrust and installed (net of inlet and nozzle losses) thrust are different numbers, and both are legitimate as long as you know which one you are holding. Same throttle rating: maximum continuous and a short-duration rating are not interchangeable.
A bidder working from a generated performance model can restate their figures at your common point on request. A bidder working from a slide cannot. The request itself is a filter.
Pin down TSFC: units and basis
Specific fuel consumption is the figure most easily gamed by accident. Two unit systems circulate (kg/(kN·h) and lb/(lbf·h)) and the same engine scores roughly a hundred times higher in the first than in the second, which makes any unitless comparison meaningless before physics is even discussed; NASA Glenn's specific fuel consumption reference uses the imperial form, which is worth knowing when a proposal's number looks suspiciously good in the wrong units. Beyond units, pin three things: at which flight condition the figure was evaluated, whether it is installed or uninstalled, and whether it is a design-point value or a mission-weighted one.
The same engine can honestly carry very different numbers. Orion's TSFC is 99.3 kg/(kN·h) at its sea-level static design point and 123.7 at 10 km and Mach 0.70. Neither is deceptive. Quoting only the first to a buyer whose mission lives at the second condition would be.
Read the off-design deck, not the design point
The design point is where the engine was optimized; the mission is where it will spend its life. Ask every bidder for an altitude × Mach deck (solved points, not scaled curves) and compare the proposals at your mission's critical corner, not at theirs.
Thrust lapse with altitude is steep for any small turbojet: Orion's 13.34 kN at sea level becomes 7.93 kN at 5 km and Mach 0.40, and 5.01 kN at 10 km and Mach 0.70. A proposal that wins on the bench can lose at altitude if its cycle was tuned for the bench. If a bidder cannot show you off-design points, the supported reading is narrower: the proposal has not substantiated performance at your mission conditions.
Grade the mass estimate's maturity
Not all mass figures are estimates of the same thing. There is a maturity ladder: a single round number; a figure scaled from an existing engine; a component-level budget with declared material bases; and a budget floored by explicit subsystem allowances with a fallback construction alongside it. Each rung down the ladder is a place weight growth can hide.
The most useful single question you can ask: what does this engine weigh in conventional construction? Orion's answer is public (165.2 kg in advanced materials, 209.0 kg as an all-metal fallback) and the 44 kg between those numbers is a quantified statement of how much of the design's competitiveness depends on the materials actually being deliverable. A proposal with no fallback construction has not told you its growth risk; it has declined to.
Check the mechanical anchors
Cycle numbers are only credible if the hardware they imply stays inside defensible bounds. Four anchors catch most overreach: rotor tip speed, relative compressor tip Mach, per-stage pressure ratio, and turbine inlet temperature. Orion's release declares its own check bounds for these: tip speed 435.4 m/s against a declared 455 ceiling, relative tip Mach 1.381 against 1.40, per-stage pressure ratio 1.491 inside a declared 1.20-1.70 band. Read those as Orion's declared checks for that release, not as universal demonstrated precedent, and require any bidder, us included, to state the source and rationale behind the bounds they check against.
A proposal outside its declared bounds is not automatically wrong; materials and aerodynamics do advance. But the burden of proof moves to the bidder, and "our analysis says so" restates the claim rather than supporting it.
Interrogate the assumptions
Under every cycle deck sits a stack of assumptions: compressor and turbine efficiencies, combustor pressure loss, bleed and cooling flows, fuel properties, installation losses. Two proposals with identical headlines can rest on efficiency assumptions several points apart, and the assumptions (not the headlines) decide which design survives contact with hardware. Ask for the model basis in writing, the way Orion's is published: conservation laws, efficiency relations, choking behavior, what is solved and what is held fixed. If the assumptions are not written down, you are being asked to underwrite them blind.
Rank the evidence, not the confidence
Confidence is free; evidence has a cost structure. The hierarchy, strongest first: declared thresholds with measured values against them; rejected alternatives on the record; a package that re-executes deterministically; and, far below all of these, conclusions, pedigree and adjectives. The reasoning behind that ordering is laid out in what published verification actually means, and the artefact-by-artefact view of what you should be holding is in the package deliverables guide. When two proposals are close on physics, buy the one you can audit.
The comparison worksheet
| Dimension | Ask each bidder | A strong answer looks like |
|---|---|---|
| Design point | At what condition is the headline thrust quoted? | One declared altitude/Mach/throttle point, restatable at yours |
| TSFC | Units, condition, installed or uninstalled? | Units stated, basis stated, both design-point and mission-corner values |
| Off-design | Show the altitude × Mach deck | Solved points across your envelope, thrust lapse visible |
| Mass | What is the estimate's basis, and the conventional fallback? | Component budget with material bases and a declared fallback |
| Mechanical anchors | Tip speed, tip Mach, stage loading, TIT versus declared bounds | Inside declared bounds, or outside with evidence attached |
| Assumptions | Publish the model basis | Written statement of efficiencies, losses and what is solved |
| Evidence | What re-runs, and what was rejected? | Thresholds with measured values; rejected points on the record |
Score the proposals on the worksheet rows, and the headline thrust row will usually stop being the deciding one. That is the point of the exercise.
When one proposal wins, the next problem is the airframe it has to live in, the integration checklist covers that handoff. And if the comparison exposed that your own requirements were the weak link, fix the input first: the design-brief checklist is the place.
Reference point: the Orion cycle-selection trail is what one row of this worksheet looks like fully public: baseline, rejected point, selected cycle, thresholds and all.