Since the earliest development of liquid rocket engines, there has always been the need for tools capable of predicting the heat flux peaks impacting the engine material, in order to control the thermal behaviour of the engine walls as well as all the other components. To prevent thermal failures, the engine is generally cooled by a coolant flowing through pas- sages machined or manufactured inside the hottest regions of the engine (combustion chamber and nozzle). The coolant extracts heat from the wall similarly to the cooling liquid used in automotive radiators, thus limiting the wall temperature. This configuration is referred to as a fluid-cooling system. Intheaerospacefield, itisnotpossibletoemployaclosed-loopcoolingcircuit, asinautomotive applications, because there is no external air available to dissipate the heat absorbed by the coolant. For this reason, after flowing through the cooling channels, the coolant must either be discarded by ejecting it into the atmosphere or space, or reused. Given the limited onboard resources and strict mass constraints, the most common and ef- fective solution adopted in rocket engines is coolant reutilization. In particular, the propellants themselves can be used as coolants and subsequently injected into the combustion chamber. In this way, besides cooling the chamber, the thermal energy absorbed during the cooling process contributes to improving engine performance. More advanced solutions employ the Expander Cycle configuration. In this case, the heated coolant is used to drive the turbopumps, further increasing the overall system efficiency. This configuration is referred to as a Regenerative Cooling System, whereas the case in which the coolant is externally discharged is known as a Dump Cooling System. The operating conditions of the cooling circuit strongly depend on the engine type and on the working fluid. In the case of high-performance cryogenic engines, such as LO2/CH4 engines, the coolant may operate under subcritical, two-phase, or supercritical conditions, meaning that its behaviour significantly differs from that of an ideal liquid or gas. In both academia and industry, it is common practice to analyse the thermal behaviour of both the fluid and the wall through simplified approaches based on empirical or semi-empirical correlations. These methods are empirical because they are generally calibrated using exper- imental data obtained from tests performed on engines with architectures similar to the one being simulated. Although simplified approaches cannot accurately describe all the physical phenomena occur- ring within the fluid, especially when three-dimensional effects are involved, they still provide reliable results and are therefore extremely useful for obtaining a general overview of the engine behaviour and performance. However, they are less accurate when describing local phenomena in detail. For this reason, regenerative cooling circuits are often oversized. Moreover, empirical cor- relations frequently depend on the specific engine architecture and coolant employed, making recalibration or the adoption of new correlations sometimes necessary. In recent years, the development of additive manufacturing technologies has enabled the pro- duction of new geometric configurations optimized for specific coolants and operating conditions. This evolution has increased the need for more accurate analysis tools, such as three-dimensional Navier–Stokes solvers (Computational Fluid Dynamics (CFD) solvers), capable of accurately de- scribing three-dimensional fluid behaviour as well as multiphase flows. Nevertheless, simplified approaches are still essential because of their limited computational requirements, which make them ideal for rapidly iterating engine designs. Furthermore, detailed 3D analyses can be integrated into the design process and used to validate the simplified models whenever necessary. The purpose of this thesis is to present the work carried out during an internship period at PANGEA PROPULSION, focused on the development of reduced order models used for the design of rocket engine cooling circuits, with particular attention to supercritical coolant flows inside cooling channels. The work was divided into the following steps: • Cooling code validation: the in-house code used to simulate both the internal gas com- bustion and the cooling circuit was analysed in all its components and partially rewritten in order to improve efficiency, accuracy, and flexibility. During this process, new heat transfer models (based on thermal resistance theory) were val- idated and the computational logic was restructured to reduce computational cost. Finally, new models were developed, including both 1D and 2D approaches. • Cooling channel design: finally, the codes developed and improved throughout this work were used to generate and iterate new cooling channel designs for the company products, in- cluding rocket nozzles and aerospike engines. For this purpose a new optimization algorithm computing the Regenerative Cooling Circuit (RCC) optimal design was coded.
Fin dai primi sviluppi dei motori a razzo a propellente liquido, è sempre stata necessaria la disponibilità di strumenti in grado di prevedere i picchi di flusso termico che investono i materiali del motore, al fine di controllare il comportamento termico delle pareti del motore e di tutti gli altri componenti. Per prevenire cedimenti dovuti al surriscaldamento, il motore viene generalmente raffreddato mediante un fluido refrigerante che scorre attraverso condotti ricavati o realizzati all’interno delle regioni più calde del motore (camera di combustione e ugello). Il refrigerante estrae calore dalla parete in modo analogo al liquido di raffreddamento utilizzato nei radiatori automobilistici, limitando così la temperatura delle pareti. Questa configurazione è definita sistema di raffreddamento a fluido. Nel settore aerospaziale non è possibile utilizzare un circuito di raffreddamento a ciclo chiuso, come avviene nelle applicazioni automobilistiche, poiché non è disponibile aria esterna in grado di dissipare il calore assorbito dal refrigerante. Per questo motivo, dopo aver attraversato i canali di raffreddamento, il fluido refrigerante deve essere espulso all’esterno (nell’atmosfera o nello spazio) oppure riutilizzato. Considerando le limitate risorse disponibili a bordo e i severi vincoli di massa, la soluzione più comune ed efficace adottata nei motori a razzo consiste nel riutilizzo del refrigerante. In particolare, gli stessi propellenti possono essere impiegati come fluidi di raffreddamento e successivamente iniettati nella camera di combustione. In questo modo, oltre a raffreddare la camera, l’energia termica assorbita durante il processo di raffreddamento contribuisce a migliorare le prestazioni del motore. Soluzioni più avanzate adottano la configurazione a ciclo expander. In questo caso, il refrigerante riscaldato viene utilizzato per azionare le turbopompe, aumentando ulteriormente l’efficienza complessiva del sistema. Questa configurazione prende il nome di sistema di raffreddamento rigenerativo, mentre il caso in cui il refrigerante venga scaricato all’esterno è noto come sistema di raffreddamento a scarico (dump cooling). Le condizioni operative del circuito di raffreddamento dipendono fortemente dal tipo di motore e dal fluido di lavoro utilizzato. Nel caso di motori criogenici ad alte prestazioni, come i motori LO₂/CH₄, il refrigerante può operare in condizioni subcritiche, bifase oppure supercritiche; ciò significa che il suo comportamento si discosta significativamente da quello di un liquido o di un gas ideale. Sia in ambito accademico sia industriale, è pratica comune analizzare il comportamento termico del fluido e delle pareti mediante approcci semplificati basati su correlazioni empiriche o semi-empiriche. Questi metodi sono definiti empirici perché vengono generalmente calibrati utilizzando dati sperimentali ottenuti da prove eseguite su motori con architetture simili a quella oggetto della simulazione. Sebbene gli approcci semplificati non siano in grado di descrivere accuratamente tutti i fenomeni fisici che avvengono all’interno del fluido, soprattutto quando entrano in gioco effetti tridimensionali, essi forniscono comunque risultati affidabili e risultano quindi estremamente utili per ottenere una visione generale del comportamento e delle prestazioni del motore. Tuttavia, la loro accuratezza diminuisce quando si tratta di descrivere nel dettaglio fenomeni locali. Per questo motivo, i circuiti di raffreddamento rigenerativo vengono spesso sovradimensionati. Inoltre, le correlazioni empiriche dipendono frequentemente dalla specifica architettura del motore e dal refrigerante impiegato, rendendo talvolta necessaria una ricalibrazione o l’adozione di nuove correlazioni. Negli ultimi anni, lo sviluppo delle tecnologie di manifattura additiva ha reso possibile la realizzazione di nuove configurazioni geometriche ottimizzate per specifici refrigeranti e condizioni operative. Questa evoluzione ha aumentato la necessità di strumenti di analisi più accurati, come i risolutori tridimensionali delle equazioni di Navier–Stokes (Computational Fluid Dynamics, CFD), capaci di descrivere con precisione il comportamento tridimensionale dei fluidi e i flussi multifase. Nonostante ciò, gli approcci semplificati restano fondamentali grazie ai loro ridotti requisiti computazionali, che li rendono ideali per iterare rapidamente il progetto di un motore. Inoltre, analisi tridimensionali dettagliate possono essere integrate nel processo di progettazione e utilizzate per validare i modelli semplificati quando necessario. Lo scopo di questa tesi è presentare il lavoro svolto durante un periodo di tirocinio presso Pangea Aerospace, focalizzato sullo sviluppo di modelli a ordine ridotto utilizzati per la progettazione dei circuiti di raffreddamento dei motori a razzo, con particolare attenzione ai flussi di refrigerante in condizioni supercritiche all’interno dei canali di raffreddamento. Il lavoro è stato suddiviso nelle seguenti fasi: • Validazione del codice di raffreddamento Il codice interno utilizzato per simulare sia la combustione interna dei gas sia il circuito di raffreddamento è stato analizzato in tutte le sue componenti e parzialmente riscritto al fine di migliorarne efficienza, accuratezza e flessibilità. Durante questo processo, sono stati validati nuovi modelli di trasferimento termico (basati sulla teoria delle resistenze termiche) e la logica computazionale è stata ristrutturata per ridurre il costo computazionale. Infine, sono stati sviluppati nuovi modelli, sia monodimensionali (1D) sia bidimensionali (2D). • Progettazione dei canali di raffreddamento Infine, i codici sviluppati e migliorati nel corso di questo lavoro sono stati utilizzati per generare e iterare nuove configurazioni dei canali di raffreddamento per i prodotti aziendali, inclusi ugelli per motori a razzo e motori aerospike. A tale scopo è stato sviluppato un nuovo algoritmo di ottimizzazione in grado di determinare il progetto ottimale del Circuito di Raffreddamento Rigenerativo (Regenerative Cooling Circuit, RCC).
A thermal-fluid model for the preliminary design of regeneratively cooled liquid rocket engines
De Carli, Matteo
2025/2026
Abstract
Since the earliest development of liquid rocket engines, there has always been the need for tools capable of predicting the heat flux peaks impacting the engine material, in order to control the thermal behaviour of the engine walls as well as all the other components. To prevent thermal failures, the engine is generally cooled by a coolant flowing through pas- sages machined or manufactured inside the hottest regions of the engine (combustion chamber and nozzle). The coolant extracts heat from the wall similarly to the cooling liquid used in automotive radiators, thus limiting the wall temperature. This configuration is referred to as a fluid-cooling system. Intheaerospacefield, itisnotpossibletoemployaclosed-loopcoolingcircuit, asinautomotive applications, because there is no external air available to dissipate the heat absorbed by the coolant. For this reason, after flowing through the cooling channels, the coolant must either be discarded by ejecting it into the atmosphere or space, or reused. Given the limited onboard resources and strict mass constraints, the most common and ef- fective solution adopted in rocket engines is coolant reutilization. In particular, the propellants themselves can be used as coolants and subsequently injected into the combustion chamber. In this way, besides cooling the chamber, the thermal energy absorbed during the cooling process contributes to improving engine performance. More advanced solutions employ the Expander Cycle configuration. In this case, the heated coolant is used to drive the turbopumps, further increasing the overall system efficiency. This configuration is referred to as a Regenerative Cooling System, whereas the case in which the coolant is externally discharged is known as a Dump Cooling System. The operating conditions of the cooling circuit strongly depend on the engine type and on the working fluid. In the case of high-performance cryogenic engines, such as LO2/CH4 engines, the coolant may operate under subcritical, two-phase, or supercritical conditions, meaning that its behaviour significantly differs from that of an ideal liquid or gas. In both academia and industry, it is common practice to analyse the thermal behaviour of both the fluid and the wall through simplified approaches based on empirical or semi-empirical correlations. These methods are empirical because they are generally calibrated using exper- imental data obtained from tests performed on engines with architectures similar to the one being simulated. Although simplified approaches cannot accurately describe all the physical phenomena occur- ring within the fluid, especially when three-dimensional effects are involved, they still provide reliable results and are therefore extremely useful for obtaining a general overview of the engine behaviour and performance. However, they are less accurate when describing local phenomena in detail. For this reason, regenerative cooling circuits are often oversized. Moreover, empirical cor- relations frequently depend on the specific engine architecture and coolant employed, making recalibration or the adoption of new correlations sometimes necessary. In recent years, the development of additive manufacturing technologies has enabled the pro- duction of new geometric configurations optimized for specific coolants and operating conditions. This evolution has increased the need for more accurate analysis tools, such as three-dimensional Navier–Stokes solvers (Computational Fluid Dynamics (CFD) solvers), capable of accurately de- scribing three-dimensional fluid behaviour as well as multiphase flows. Nevertheless, simplified approaches are still essential because of their limited computational requirements, which make them ideal for rapidly iterating engine designs. Furthermore, detailed 3D analyses can be integrated into the design process and used to validate the simplified models whenever necessary. The purpose of this thesis is to present the work carried out during an internship period at PANGEA PROPULSION, focused on the development of reduced order models used for the design of rocket engine cooling circuits, with particular attention to supercritical coolant flows inside cooling channels. The work was divided into the following steps: • Cooling code validation: the in-house code used to simulate both the internal gas com- bustion and the cooling circuit was analysed in all its components and partially rewritten in order to improve efficiency, accuracy, and flexibility. During this process, new heat transfer models (based on thermal resistance theory) were val- idated and the computational logic was restructured to reduce computational cost. Finally, new models were developed, including both 1D and 2D approaches. • Cooling channel design: finally, the codes developed and improved throughout this work were used to generate and iterate new cooling channel designs for the company products, in- cluding rocket nozzles and aerospike engines. For this purpose a new optimization algorithm computing the Regenerative Cooling Circuit (RCC) optimal design was coded.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_07_De_Carli.pdf
accessibile in internet per tutti a partire dal 25/06/2027
Dimensione
7.48 MB
Formato
Adobe PDF
|
7.48 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/10589/261427