GENESIS VENTURES
BUILD STORYAMINREZA KHOSHBAHARAUGUST 13, 20265 MIN READ

Gas Turbine Design: The One-Shift Drum

One shift of gas turbine design: the GT1 compressor drum re-architected, 12 of 12 spinning stress cases passed, every shaft seated to zero interference.


Re-architecting the rotating heart of a small turbojet sounds like a month of work. The GT1 compressor drum just did it in one shift: re-pitched to the aerodynamic design, screened under spin, re-joined to both stub shafts, and rendered for the record on the GPU. This is what gas turbine design looks like when the pipeline does the heavy lifting: measured at every step, assumed at none.

The drum is the part the whole compressor hangs on. Every rotor row bolts to it, both stub shafts locate on it, and it carries the full spinning load of the front end of the engine. Touching it means touching everything.

GT1 fuel manifold ring with its injector stems, rendered headlessly on the GPU, gas turbine design verification render

Six stages, each on its own step

The previous drum spaced its rotor joints uniformly, because uniform is easy to draw. The aerodynamic design does not care about easy. Each rotor row lives where the gas path says it lives, so we re-pitched the drum: six stages re-landed, every designed rotor row now sitting on its own drum stage, exactly where the flowpath puts it.

Then we measured the result from the re-imported solids rather than trusting the builder's intent. Six rotor-to-drum pairs seated. Twelve stator and vane pairs clear. Zero interference anywhere in the stack.

The new drum also came out lighter than the one it replaces. Mass was never a target; the weight went away because material goes only where the stages actually need it.

Twelve spin cases, twelve passes

A re-pitched drum is a claim until it survives the stress screen. We ran the screen across all six rotor rows at both ends of the speed envelope, rated and trip, twelve cases in all.

Twelve for twelve. Every case closed under its allowable with margin in hand, and even the tightest row kept comfortable room between itself and the line.

A word on what that means. This is a screening-tier result: a fast, conservative force-balance screen run directly on the promoted solids, not a finite-element release. Its job is to kill weak geometry before it costs anyone a week, and it just waved the new drum through to the next gate.

The shafts seat with nothing left over

A drum is only as good as what joins to it. With the stages re-pitched, the front and rear stub shafts were re-seated onto the drum's lands through their curvic couplings, the toothed rings that locate a shaft to a drum with no slop and no ambiguity.

Front stub shaft curvic coupling seated on the GT1 compressor drum, zero-gap turbomachinery CAD interface render

The contact scan came back the way you want every scan to come back: both interfaces seated, gap zero, shared volume zero. No interference, and no daylight either. The shaft line is closed end to end.

The casing gets its one new hole

The fuel manifold from the 18-for-18 build story needs exactly one way out of the engine: a single inlet boss passing through the outer casing to meet the fuel line. That penetration is now cut into the casing and verified: the boss passes through its declared hole, and the structure around it was proven clear by boolean measurement, not by eyeballing a section view.

GT1 fuel manifold inlet boss passing through the re-cut combustor casing penetration

One hole. That count matters, because every penetration in a pressure casing is structure, sealing and future inspection burden. The design buys its entire external fuel connection with exactly one.

Gas turbine design at GPU speed

Every render in this article came off a headless pipeline that draws straight on the GPU, no workstation screen involved. The same shift, the verification factory worked through a full hundred-blade row in batches: one hundred blades in, one hundred through every gate, while the host machine stayed responsive enough to keep designing on.

Renders that used to queue behind the CAD work now arrive while the measurements are still warm. That convenience is what makes a one-shift re-architecture possible, because the evidence keeps pace with the geometry instead of trailing it by a day.

That is the real story here. Speed in gas turbine design does not come from heroic sessions; it comes from a pipeline where the solids, the stress screen, the interface scans and the pictures all regenerate from the same source, so a re-architecture measured in hours still ships with receipts.

Follow the whole engine on the Orion page, and the pipeline that makes one-shift turns possible on the Method page.

Next step

Your geometry, checked the same way.

Threshold study, $3,800, two weeks
Engagements

A threshold study writes your check matrix in two weeks, from $3,800.

[email protected]