The study focuses on the implementation of a linear regression model applied to a 2k factorial design of experiments for rocket applications. The mathematical method allowed the detailed investigation of a heterogeneous solid propellant able to comply a selected performance cost function. The work has been divided in three main parts: preliminary composition research, selection of a precise formulation and relative experimental characterization. This work contributes to a wider project aiming at the experimentation of innovative extinguishable propellants in missiles. Among all the speci cations the propellant has to comply high pressure de agration limit and speci c impulse. The propellant choice focused on a dual-oxidizer AP/PSAN. Being the speci c impulse the only parameter known a priori, the rst preliminary research was based on nding a combination of HTPB, AP and PSAN able to grant Is = 240s through a 2k approach. The general formulations investigated have oxidizer to fuel ratios of 80/20, 85/15 and 90/10 and their relative speci c impulse produced within the missile system was calculated through CEA NASA-Glenn program. From this rst analysis the only feasible blend is the 85/15 because the 80/20 does not reach the requested Is, whereas the 90/15 was not practically manufacturable at SPLab, in Politecnico di Milano. The implemented regression model was then applied to the combinations found and the speci c impulse was recalculated through the mathematical method. The model evaluates each ingredient in uences on Is, including their interactions, and traces contour plots of their behaviour. The optimization of the 85/15 combinations is determined by analysing the trends and the nal propellant is selected: 14% HTPB, 16% PSAN and 70% AP. These percentages grant the requested impulse maximizing the PSAN amount, ensuring the highest PDL. The manufacture of the propellant grain follows. Its production procedure involved high temperature issues and it has been slightly modi ed when binder degradation was detected. Once produced, the grains were cut into several specimens and they have been used to test burning rate and pressure de agration limit. The rst one has been evaluated experimentally at 15, 30 and 50 bar, the latter through a quasi-static depressurisation of 5 mbar=s circa because the process was manually driven. After post-processing analysis, the burning rate is equal to 8.4 mm=s at 70 bar, with the pressure coe cient n of the Vieille's law equal to 0.589, while the PDL is 230 mbar.
Il presente studio si concentra sull'implementazione di un modello a regressione lineare applicato ad un metodo fattoriale 2k di progettazione sperimentale per applicazioni spaziali. Il modello matematico permette l'investigazione dettagliata di un propellente solido eterogeneo capace di soddisfare una determinata funzione di costo delle prestazioni. Il lavoro è stato suddiviso in tre parti principali: ricerca di una composizione preliminare, selezione di una formulazione precisa e relativa caratterizzazione sperimentale. Questo studio contribuisce ad un più grande progetto che punta alla sperimentazione di propellenti estinguibili innovativi in missilistica. Tra tutte le speci che il propellente deve soddisfare in particolare un'alta pressione limite di de agrazione e impulso speci co. La scelta ha volto verso un doppio-ossidante AP/PSAN. Essendo l'impulso speci co l'unico parametro scibile a priori, la ricerca preliminare è stata basata sulla identi cazione di una formulazione di HTPB, AP e PSAN in grado di garantire Is = 240s attraverso l'approccio 2k. I rapporti ossidante/combustibile investigati sono stati 80/20, 85/15 e 90/10 e il loro relativo impulso prodotto all'interno del sistema missile è stato calcolato attraverso il programma CEA NASAGlenn. Da questa analisi l'unica mistura fattibile è l' 85/15; l' 80/20 infatti non raggiunge Is richiesto mentre la 90/10 non è manifatturabile all'interno dell'SPLab, nel Politecnico di Milano Il modello a regressione lineare implementato è stato successivamente applicato alle combinazioni trovate e l'impulso speci co è stato ricalcolato attraverso il metodo matematico. Il codice valuta quantitativamente l'in uenza di ogni ingrediente su Is, comprese le loro interazioni, e traccia i pro li del loro comportamento. L'ottimizzazione delle combinazioni 85/15 è stata determinata con l'analisi dei trend e il propellente nale scelto è così composto: 14% HTPB, 16% PSAN e 70% AP. Queste percentuali assicurano l'impulso richiesto massimizzando la quantit à di PSAN, garantendo la più alta PDL. Segue la manifattura del grano. La sua procedura di produzione è stata caratterizzata da problematiche relative all'alta temperatura generatasi, è stato infatti notato del degradamento del legante. una volta prodotti, i propellenti sono stati tagliati in diversi provini e sono stati utilizzati per testare la velocità di combustione e pressione limite di de agrazione. La prima è stata valutata sperimentalmente a 15, 30 e 50 bar, la seconda tramite depressurizzazione quasi stazionario di 5 mbar=s circa a causa del controllo manuale del processo. A seguito dell'analisi di postelaborazione, la velocità di combustione risulta essere di 8.4 mm=s a 70 bar, con l'esponente di pressione n della legge di Vieille pari a 0.589, mentre la PDL è di 230 mbar.
Formulation of dual-oxidizer propellant using design of experiment method
FERAZZI, FABRIZIO
2018/2019
Abstract
The study focuses on the implementation of a linear regression model applied to a 2k factorial design of experiments for rocket applications. The mathematical method allowed the detailed investigation of a heterogeneous solid propellant able to comply a selected performance cost function. The work has been divided in three main parts: preliminary composition research, selection of a precise formulation and relative experimental characterization. This work contributes to a wider project aiming at the experimentation of innovative extinguishable propellants in missiles. Among all the speci cations the propellant has to comply high pressure de agration limit and speci c impulse. The propellant choice focused on a dual-oxidizer AP/PSAN. Being the speci c impulse the only parameter known a priori, the rst preliminary research was based on nding a combination of HTPB, AP and PSAN able to grant Is = 240s through a 2k approach. The general formulations investigated have oxidizer to fuel ratios of 80/20, 85/15 and 90/10 and their relative speci c impulse produced within the missile system was calculated through CEA NASA-Glenn program. From this rst analysis the only feasible blend is the 85/15 because the 80/20 does not reach the requested Is, whereas the 90/15 was not practically manufacturable at SPLab, in Politecnico di Milano. The implemented regression model was then applied to the combinations found and the speci c impulse was recalculated through the mathematical method. The model evaluates each ingredient in uences on Is, including their interactions, and traces contour plots of their behaviour. The optimization of the 85/15 combinations is determined by analysing the trends and the nal propellant is selected: 14% HTPB, 16% PSAN and 70% AP. These percentages grant the requested impulse maximizing the PSAN amount, ensuring the highest PDL. The manufacture of the propellant grain follows. Its production procedure involved high temperature issues and it has been slightly modi ed when binder degradation was detected. Once produced, the grains were cut into several specimens and they have been used to test burning rate and pressure de agration limit. The rst one has been evaluated experimentally at 15, 30 and 50 bar, the latter through a quasi-static depressurisation of 5 mbar=s circa because the process was manually driven. After post-processing analysis, the burning rate is equal to 8.4 mm=s at 70 bar, with the pressure coe cient n of the Vieille's law equal to 0.589, while the PDL is 230 mbar.File | Dimensione | Formato | |
---|---|---|---|
Thesis Ferazzi.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
21.12 MB
Formato
Adobe PDF
|
21.12 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/10589/146769