Shock waves are a major source of aerodynamic losses in non-ideal compressible fluid dynamics (NICFD) turbomachinery, where the working fluid can be both highly molecularly complex and far from ideal gas conditions. A dedicated numerical solver based on the steady one-dimensional Navier-Stokes equations is used to compute the internal structure of normal shocks in single-phase pure fluids. Resolving thermophysical property variations across the shock layer allows the definition of an entropy-based loss coefficient that (i) separates the two dissipation mechanisms within the shock, namely viscous dissipation and heat-transfer loss; and (ii) enables consistent comparisons of both the individual contributions and the total loss across different fluids and thermodynamic conditions. Results show that shock losses increase with molecular complexity and shock strength. Under non-ideal thermodynamic conditions, losses increase with the upstream compressibility factor and with decreasing average generalized isentropic pressure–volume exponent. Minimum losses occur for pre-shock states near the critical point. Regarding the loss mechanisms, viscous dissipation is generally the dominant contribution, but its relative importance decreases as the flow becomes more non-ideal. At comparable levels of thermodynamic non-ideality, differences in viscous losses across fluids mainly reflect changes in the velocity-gradient distribution within the shock layer; for sufficiently complex molecules, the velocity field approaches a self-similar structure and viscous losses become less sensitive to further increases in molecular complexity. Heat-transfer losses shift from being dominated by temperature-gradient distributions in near-ideal states to being driven by fluid property variations as non-ideality increases. Finally, comparison with the shock-structure solver shows that physically resolving the shock layer using CFD software would require grid spacings approximately three to four orders of magnitude smaller than those adopted here.
Le onde d’urto sono una delle principali cause di perdite aerodinamiche nelle turbomacchine che operano con fluidi in condizioni termodinamiche non ideali, specialmente se caratterizzati da elevata complessità molecolare. In questo lavoro si impiega un solutore numerico dedicato, basato sulle equazioni di Navier–Stokes monodimensionali stazionarie, per calcolare la struttura interna di urti normali in fluidi puri monofase. La caratterizzazione delle proprietà termofisiche nello strato d’urto permette di definire un coefficiente di perdita che (i) separa le perdite per dissipazione viscosa da quelle per scambio termico e (ii) consente un confronto sistematico e uniforme dei singoli contributi e delle perdite totali al variare del fluido e delle condizioni termodinamiche. Si osserva che le perdite d’urto aumentano con la complessità molecolare e con l’intensità dell’urto. In condizioni termodinamiche non ideali, inoltre, crescono con il fattore di comprimibilità a monte dell’urto e al diminuire del valore medio dell’esponente isentropico pressione-volume generalizzato. Le perdite minori si osservano per stati termodinamici a monte dell’urto prossimi al punto critico. Quanto ai meccanismi di perdita, la dissipazione viscosa è in generale il contributo dominante, ma la sua importanza relativa diminuisce con la non idealità termodinamica. A pari non idealità, le differenze nelle perdite viscose tra fluidi sono riconducibili a variazioni nella distribuzione del gradiente di velocità nello spessore dell’urto; con l’aumentare della complessità molecolare, il profilo di velocità adimensionale tende a una forma autosimile e le differenze nel contributo di perdita viscosa diminuiscono. Le perdite per scambio termico sono governate dai gradienti di temperatura in condizioni quasi ideali e, con l’aumentare della non idealità, dalle variazioni delle proprietà fisiche del fluido. Infine, il confronto con il solutore di struttura d’urto mostra che la risoluzione della struttura fisica dell’urto con software CFD richiederebbe una spaziatura di griglia circa tre o quattro ordini di grandezza più piccola di quella adottata in questo lavoro.
Dissipation in Non-Ideal Shock Waves. Model development and analysis of the effects of molecular complexity and non-ideal thermodynamics
Bergamaschi, Giandomenico
2024/2025
Abstract
Shock waves are a major source of aerodynamic losses in non-ideal compressible fluid dynamics (NICFD) turbomachinery, where the working fluid can be both highly molecularly complex and far from ideal gas conditions. A dedicated numerical solver based on the steady one-dimensional Navier-Stokes equations is used to compute the internal structure of normal shocks in single-phase pure fluids. Resolving thermophysical property variations across the shock layer allows the definition of an entropy-based loss coefficient that (i) separates the two dissipation mechanisms within the shock, namely viscous dissipation and heat-transfer loss; and (ii) enables consistent comparisons of both the individual contributions and the total loss across different fluids and thermodynamic conditions. Results show that shock losses increase with molecular complexity and shock strength. Under non-ideal thermodynamic conditions, losses increase with the upstream compressibility factor and with decreasing average generalized isentropic pressure–volume exponent. Minimum losses occur for pre-shock states near the critical point. Regarding the loss mechanisms, viscous dissipation is generally the dominant contribution, but its relative importance decreases as the flow becomes more non-ideal. At comparable levels of thermodynamic non-ideality, differences in viscous losses across fluids mainly reflect changes in the velocity-gradient distribution within the shock layer; for sufficiently complex molecules, the velocity field approaches a self-similar structure and viscous losses become less sensitive to further increases in molecular complexity. Heat-transfer losses shift from being dominated by temperature-gradient distributions in near-ideal states to being driven by fluid property variations as non-ideality increases. Finally, comparison with the shock-structure solver shows that physically resolving the shock layer using CFD software would require grid spacings approximately three to four orders of magnitude smaller than those adopted here.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_03_Bergamaschi_Thesis_01.pdf
non accessibile
Descrizione: Thesis
Dimensione
82.43 MB
Formato
Adobe PDF
|
82.43 MB | Adobe PDF | Visualizza/Apri |
|
2026_03_Bergamaschi_Executive Summary_02.pdf
non accessibile
Descrizione: Executive Summary
Dimensione
8.75 MB
Formato
Adobe PDF
|
8.75 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/251554