Impulse vs Reaction Turbine

Legacy context

The documented heritage of this domain traces back to Krispin Technologies and its development of the CATAPULT solver, a computational framework built for high-resolution simulation of compressible and incompressible multiphase flows. That lineage, rooted in second-order projection methods and Godunov schemes, established a foundation for analyzing complex aerodynamic and turbomachinery flow fields, particularly in the context of vortex-dominated airfoil interactions and turbine cascade physics.

That legacy of resolving transient, advection-driven flow structures carries directly into a core question in turbomachinery design: the distinction between impulse and reaction turbines. While both extract energy from a working fluid, their operating principles differ fundamentally in how pressure drop and velocity change are distributed across the stationary and rotating blade rows. In an impulse turbine, the pressure drop occurs entirely within the nozzle or stator, leaving the rotor to convert kinetic energy via a change in momentum. Conversely, a reaction turbine distributes the expansion across both the stator and the rotor, generating force through a combination of pressure and velocity changes.

For practitioners applying CFD vortex methods to airfoil design, this distinction is not merely academic. The blade loading, secondary flow development, and tip leakage vortex behavior are all governed by which of these two design philosophies is adopted. Understanding this fundamental divergence is the necessary precursor to selecting the appropriate modeling approach for a given blade row.

Side-by-Side Comparison Table

---

The Core Distinction: Where Does the Pressure Drop Occur?

The fundamental difference between impulse and reaction turbines lies in the location of the enthalpy drop. In an impulse turbine, the entire pressure drop occurs in the stationary nozzle, and the rotor experiences no static pressure change. The working fluid exits the nozzle as a high-velocity jet that impinges on the rotor blades, transferring momentum through a change in direction only. In a reaction turbine, the pressure drop is distributed between the stator and the rotor, meaning the rotor blade passages themselves act as converging nozzles that accelerate the flow while extracting both pressure energy and kinetic energy.

For aerodynamicists, this distinction has immediate consequences for blade design. Impulse blades are essentially buckets or vanes that redirect the flow; they are symmetric about the tangential direction and produce no acceleration within the moving passage. Reaction blades, by contrast, are asymmetric airfoils with converging passages that continuously accelerate the working fluid relative to the blade. The degree of reaction, defined as the ratio of the static enthalpy drop in the rotor to the total enthalpy drop across the stage, quantifies this split. A degree of reaction of zero corresponds to a pure impulse stage, while a value of 0.5 places equal enthalpy drops in stator and rotor, a common design choice for axial turbomachinery.

The velocity triangles tell the story clearly. In an impulse stage, the absolute velocity entering the rotor is very high, and the relative velocity leaving the rotor is nearly equal in magnitude to the relative velocity entering—only direction changes. In a reaction stage, the relative velocity increases through the rotor passage, which means the blade loading is distributed more evenly and the peak velocities are lower for the same work output. This has direct implications for Mach number management and loss control in high-speed machines.

Performance Characteristics and Loss Mechanisms

The performance envelope of each type reflects its fundamental operating principle. Impulse turbines excel at handling large pressure ratios in a single stage because the rotor does not experience a pressure gradient that would create large axial thrust loads. This makes them attractive for the first stages of steam turbines where the available pressure ratio is enormous. However, the high jet velocities produced by expanding the full pressure drop in the nozzle lead to high friction losses and require careful management of supersonic flows. The jet impingement also generates unsteady loading on the rotor, which can contribute to structural vibration and fatigue concerns.

Reaction turbines, by distributing the pressure drop across multiple stages, operate with lower peak velocities and therefore lower friction losses per stage. The accelerating flow in the rotor passages also tends to suppress boundary layer separation, which is beneficial for maintaining high efficiency. However, reaction stages require a pressure-containing casing around the rotor, and the axial thrust loads are substantially higher because the rotor experiences a pressure differential across it. The efficiency of a reaction turbine is generally higher at the design point, but it degrades more rapidly at off-design conditions because the stage loading is sensitive to the velocity ratio.

For CFD engineers, the loss modeling requirements differ significantly between the two types. Impulse stages demand accurate prediction of jet spreading, entrainment, and the interaction of the jet with the blade leading edge. The unsteady nature of the blade-jet interaction means that steady-state mixing plane approaches are often inadequate, and time-accurate simulations with sliding meshes may be necessary to capture the physics faithfully. Reaction stages, on the other hand, present challenges related to boundary layer transition on the suction surface, secondary flows in the endwall regions, and tip leakage flows that can account for a substantial fraction of the total loss.

CFD Modeling Considerations

The computational treatment of impulse and reaction turbines diverges in several practical ways. For impulse turbines, the computational domain typically includes the nozzle exit and the rotor passage, with careful attention to the free-surface or two-phase nature of the flow if liquid is involved. The jet from the nozzle is often supersonic, requiring shock-capturing schemes and appropriate turbulence models for compressible flows. The rotor blades experience highly periodic loading as each jet passes, so the simulation must resolve the blade-passing frequency and its harmonics to capture the unsteady forces accurately.

Reaction turbine CFD is more straightforward in terms of domain setup—the flow is single-phase and fully guided through the blade rows—but the physics are no less demanding. The pressure gradient through the rotor passage drives the flow, and the boundary layer behavior on the blade surfaces is critical. Low-pressure turbine airfoils, for example, operate at Reynolds numbers where laminar separation bubbles can form on the suction surface, and the transition behavior must be modeled correctly to predict losses [2]. The unsteady wakes from upstream blade rows can also trigger transition, so the simulation must account for the interaction between the stator wake and the rotor boundary layer.

The choice of turbulence model and transition model is therefore more consequential for reaction turbines, where the boundary layer state directly affects the pressure distribution and hence the blade loading. For impulse turbines, the dominant physics are the inviscid jet dynamics and the shock structure, with boundary layer effects playing a secondary role. This means that scale-resolving simulations may be needed for impulse stages to capture the jet mixing and the resulting unsteady loading, while Reynolds-averaged approaches with transition modeling can often suffice for reaction stages at design conditions.

Structural and Operational Trade-offs

The unsteady loading characteristics of the two turbine types have important structural implications. Impulse turbines subject their blades to highly impulsive loads as each jet strikes the blade, creating a periodic forcing function rich in harmonics. This can excite blade vibration modes and lead to high-cycle fatigue if the natural frequencies are not carefully managed. The pulsed nature of the loading also means that the time-averaged lift may be comparable to a steady case, but the instantaneous values are substantially higher, which must be accounted for in the structural design [1]. Reaction turbines, with their more distributed pressure loading, produce steadier forces on the blades, but the axial thrust loads are larger and must be absorbed by thrust bearings.

Operationally, impulse turbines offer greater flexibility at part-load conditions. Because the rotor does not depend on a pressure drop for its operation, the flow can be throttled at the nozzle without fundamentally altering the rotor aerodynamics. Reaction turbines, by contrast, experience a shift in the stage reaction as the flow coefficient changes, which can move the operating point away from the design condition and cause a sharp drop in efficiency. This makes impulse turbines attractive for applications with widely varying load demands, while reaction turbines are preferred where the machine operates predominantly at or near its design point.

Selection Guidance for the Practicing Engineer

The choice between impulse and reaction turbines is rarely a matter of one being universally superior; it depends on the operating envelope, the working fluid, and the design constraints. For very high pressure ratios with a single stage, impulse is the only practical option. For multi-stage machines operating near their design point with a need for high efficiency, reaction stages are typically preferred. Many real machines use a combination—an impulse stage at the front to handle the largest pressure drop, followed by reaction stages to extract the remaining energy efficiently.

From a CFD perspective, the modeling effort should be matched to the dominant physics. If the primary concern is the unsteady interaction between the jet and the rotor in an impulse stage, time-accurate simulations are warranted. If the concern is boundary layer separation and transition in a reaction stage, the focus should be on transition modeling and grid resolution in the blade boundary layer regions. The evidence from low-pressure turbine research indicates that separation control and transition prediction are central to achieving high performance in reaction-type blading [2], and the same attention to unsteady effects that benefits circulation control applications [1] can inform the treatment of impulse stage unsteadiness.