Aircraft Propulsion: Dimensional Analysis and Unit Consistency in Compressor Design

Abstract

Dimensional analysis and unit consistency are foundational to rigorous compressor design in aircraft propulsion systems. This article examines how dimensional reasoning underpins the aerodynamic and mechanical analysis of multistage compressors, from pressure ratio requirements through blade element design. We demonstrate that consistent application of dimensional principles—particularly in velocity diagrams, incidence and deviation angles, and stage matching—ensures that empirical correlations and computational predictions remain valid across different operating conditions and engine scales.

Background

Modern aircraft engines demand core compressors capable of achieving pressure ratios of 32:1 or higher ^{[core-compressor-pressure-ratio-requirements]}, a requirement that drives the design of multistage axial compressors with tightly coordinated aerodynamic behavior. The design process relies on a hierarchy of analytical methods: meridional flow analysis, blade element theory, and three-dimensional computational fluid dynamics. Each method operates within a specific dimensional framework, and consistency across these frameworks is essential for reliable predictions.

The fundamental challenge is that compressor design involves multiple physical domains—aerodynamics, thermodynamics, and mechanics—each with its own natural variables and scales. Dimensional analysis provides the discipline to ensure that empirical corrections, computational results, and experimental measurements remain coherent across these domains and across different operating points.

Key Results

Dimensional Consistency in Velocity Diagrams and Incidence Angles

The foundation of blade element theory ^{[blade-element-theory]} rests on velocity diagrams constructed from meridional flow analysis ^{[meridional-flow-analysis]}. In these diagrams, absolute and relative flow angles must be dimensionally consistent with the blade geometry they interact with.

The incidence angle ^{[incidence-angle]} is defined as:

$i = \beta_{\text{relative}} - \beta_{\text{blade inlet}}$

Both terms on the right are angles (dimensionless in radians), so the incidence angle is itself dimensionless. This dimensional homogeneity is critical: empirical correlations for incidence-angle losses or flow-separation limits are developed from experimental data where incidence is measured in the same way. When a designer applies an incidence correlation to a new blade design, the dimensional consistency ensures the correlation remains valid provided the underlying physics (Reynolds number, blade geometry family) is similar.

Similarly, the deviation angle ^{[deviation-angle]} is defined as:

$\delta = \beta_{\text{relative, exit}} - \beta_{\text{blade outlet}}$

Again, both terms are angles, preserving dimensional consistency. Deviation correlations, typically expressed as functions of blade geometry and Reynolds number, depend on this dimensional structure. A deviation correlation developed for a particular blade family at a particular Reynolds number regime will produce incorrect results if applied to a fundamentally different flow regime without dimensional re-scaling or re-validation.

Stage Matching and Pressure Ratio Distribution

Stage matching ^{[stage-matching-in-compressor-design]} requires that each stage in a multistage compressor receives properly conditioned flow from upstream. The pressure ratio achieved by each stage must be dimensionally consistent with the overall engine pressure ratio requirement. For a compressor with $n$ stages, if the overall pressure ratio is $\Pi_{\text{total}}$, then:

$\Pi_{\text{total}} = \prod_{j=1}^{n} \Pi_j$

where $\Pi_j$ is the pressure ratio of stage $j$. Each $\Pi_j$ is dimensionless (a ratio of pressures), ensuring that the product is also dimensionless and represents a valid pressure ratio. This dimensional structure is not merely formal; it reflects the physical fact that pressure ratios are multiplicative, not additive. Designers must distribute the overall pressure ratio across stages in a way that respects this multiplicative structure while maintaining aerodynamic efficiency at each stage.

Inlet Guide Vane Optimization and Operating Envelope

Inlet guide vanes ^{[inlet-guide-vanes]} are optimized across the compressor operating envelope using an IGV stagger angle schedule ^{[inlet-guide-vane-optimization]}. The schedule maps compressor operating speed (or equivalently, pressure ratio) to the optimal IGV angle. Both the independent variable (speed or pressure ratio) and the dependent variable (IGV angle) must be dimensionally consistent with the physical mechanisms they represent.

Operating speed is typically normalized as a fraction of design speed (dimensionless), or equivalently, as a percentage. Pressure ratio is also dimensionless. The IGV stagger angle is measured in degrees or radians. The optimization algorithm that produces the schedule must respect these dimensional distinctions: the schedule is a function from a dimensionless operating parameter to an angular parameter. This dimensional separation ensures that the schedule remains valid when applied to engines of different sizes or operating at different absolute speeds, provided the normalized operating point is the same.

Computational Validation and 3D Euler Analysis

Three-dimensional Euler codes ^{[three-dimensional-euler-code-for-compressor-flow-prediction]} solve the inviscid flow equations on a discretized domain. The outputs—velocity, pressure, density, temperature—all carry dimensional units (m/s, Pa, kg/m³, K). Validation against experimental measurements ^{[multistage-compressor-experimental-assessment]} requires dimensional consistency: predicted mass flow rate (kg/s) must be compared against measured mass flow rate in the same units; predicted pressure rise (Pa) against measured pressure rise.

A critical aspect of computational validation is that empirical corrections for viscous effects—such as loss coefficients or deviation angles—must be applied consistently. A loss coefficient, typically defined as a pressure loss normalized by dynamic pressure, is dimensionless:

$\omega = \frac{\Delta p_{\text{loss}}}{\frac{1}{2} \rho V^2}$

This dimensionless form ensures that loss correlations developed from one set of experiments can be applied to different flow conditions and blade sizes, provided the Reynolds number and flow geometry are appropriately matched. Dimensional inconsistency in applying such correlations—for example, using a loss coefficient developed at low Reynolds number in a high Reynolds number regime—is a common source of prediction error.

Worked Example: Pressure Ratio Distribution in a Five-Stage Core Compressor

Consider a core compressor required to achieve an overall pressure ratio of 32:1 ^{[core-compressor-pressure-ratio-requirements]}. A designer must distribute this pressure ratio across five stages while maintaining aerodynamic efficiency and stall margin at each stage.

A simple uniform distribution would assign each stage a pressure ratio of:

$\Pi_{\text{stage}} = 32^{1/5} \approx 2.01$

This is dimensionally correct (a ratio of pressures, dimensionless) and respects the multiplicative structure of stage matching. However, a uniform distribution is rarely optimal because inlet stages operate at lower pressures and temperatures than exit stages, allowing them to achieve higher pressure ratios without excessive losses.

A more realistic distribution might be:

$\Pi_1 = 1.8, \quad \Pi_2 = 1.9, \quad \Pi_3 = 2.0, \quad \Pi_4 = 2.1, \quad \Pi_5 = 2.2$

Verification: $1.8 \times 1.9 \times 2.0 \times 2.1 \times 2.2 \approx 32.0$ ✓

Each individual pressure ratio is dimensionless, and their product is also dimensionless. The designer then uses blade element theory and meridional analysis to design each stage to achieve its assigned pressure ratio. Inlet guide vane angles are optimized ^{[inlet-guide-vane-optimization]} to ensure that the first stage receives properly conditioned flow across the operating envelope. Experimental testing of representative stage groups ^{[multistage-compressor-experimental-assessment]} validates that the designed pressure ratios and efficiencies are achieved in a multistage environment.

References

• ^{[core-compressor-pressure-ratio-requirements]}
• ^{[multistage-compressor-experimental-assessment]}
• ^{[inlet-guide-vane-optimization]}
• ^{[three-dimensional-euler-code-for-compressor-flow-prediction]}
• ^{[stage-matching-in-compressor-design]}
• ^{[inlet-guide-vanes]}
• ^{[meridional-flow-analysis]}
• ^{[blade-element-theory]}
• ^{[incidence-angle]}
• ^{[deviation-angle]}

AI Disclosure

This article was drafted with the assistance of an AI language model based on personal class notes from an Aircraft Propulsion course. The AI was instructed to paraphrase note content, maintain dimensional rigor, and cite all factual claims. The author is responsible for technical accuracy and has reviewed all mathematical statements and physical reasoning.