The subject of this study is the hydrostatic screw used to drive axes of very large horizontal milling machines. This mechanism is used for the horizontal motion whose travel is up to 30 meters and as well for the vertical motion whose travel is up to 8 meters. The moving machine column has a mass of 120 ton which is causing huge inertial loads. This machine is equipped with a high torque spindle motor that require an axial load, due to cutting forces, up to 40000-50000 N. The current system has an axial force (60000 N) just enough to cover the maximum requested even if from the theoretical design the axial force was supposed to be around 90000 N. Due to the gap between the design and the performances of the real system it has been decided to study deeper the hydrostatic and mechanical system in order to understand what is not letting achieving the target as per the original specification. The various factors that may significantly affect the greatnesses of interest have been analyzed; they are: • Pressure loss at the output of capillary restrictor at the input of pockets. • Viscosity decrease inside the fluid. • Fluid compressibility. • Number of opposed hydrostatic bearings simultaneously engaged. • Errors due to geometrical tolerances and assembly errors that may cause unwanted variation of the clearance. Measurements have been taken on the machine to monitor the pressure value during the axis motion and here below is what has been observed: • The pressure in changing within a large range (from 32 to 18 bar) and dependent from the angular position of the nut. This is surely highlighting geometrical errors between different racks. • The pressure fluctuation has a periodical behavior and strongly linked to the characteristic frequencies (1XScrewRevolution and 2XScrewRevolution). This behavior is caused by the screw run out, not enough stiffness from the structure and by the too large bearing backlash of the rotating hydrostatic lead screw. • The frequency content of the pressure signals is changing according the axis speed; precisely the contribution (10XScrewRevolution) is significantly increasing. This contribution depends from the oil refill system of the screw channels that is clearly not performing as expected. Once the measurements on the machine were analyzed and studied the factors that are influencing the hydrostatic greatnesses (pressure and flow) it has been decided to develop an integrated model inside Matlab - Simulink® - SimScape™ environment. The implemented model developed with a multi tasks approach is necessary to understand the real machine behavior and it contains the following parts: hydraulic (hydraulic power unit, capillary restrictor, distributor and screw), hydrostatic (opposed hydrostatic bearing), mechanical (all components of the kinematic chain) and the portion relevant to the axis driver (speed loop and position loop). The results out from the model are reliable enough to validate the model and to estimate the axis performance: • Loss of pressure of the hydraulic system: distributor input, distributor and screw: 7 bar real – 6.5 bar in the integrated model. • Pressure out of the capillary restrictor: 32-18 bar real – 25 bar in the integrated model. • Flow inside the channels on the teeth side: 9-12 l/minute real – 11 l/minute in the integrated model, the flow is referred to one single tooth side and in the nominal condition. • Pressure fluctuation between teeth side under a inertial load of 10000 N: 6.5 bar real, 6 bar in the integrated model. • Static nut stiffness: 1500-2000 N/μm real, 1800 N/μm in the integrated model (both evaluated under the maximum load condition). • Maximum axial load: 50000-60000 N real, 60000 N in the integrated model (if a clearance of 10 μm is accepted). The main results produced by the developed model are: • Analysis and understanding of the hydrostatic axis functioning. • Development of software tools within Matlab - Simulink® - SimScape™ environment to support the axis design. • Proposal for design modification.
La trasmissione a chiocciola idrostatica in oggetto di studio serve per la movimentazione di una fresalesatrice di grandi dimensioni. Le corse che devono essere garantite sono molto grandi, circa 30 metri per l’asse orizzontale (X) ed 8 metri per quello verticale (Y). La massa della struttura che ne consegue può raggiungere ed oltrepassare le 120 tonnellate, causando notevoli carichi inerziali. Anche le spinte in fase di lavorazione sono molto elevate, addirittura 40000-50000 N. La spinta che attualmente l’asse garantisce è circa pari a quella massima richiesta (60000 N), nonostante dalla progettazione la spinta massima dovrebbe essere circa 90000 N. Date la discrepanza tra progetto e realtà si è deciso di studiare a fondo il sistema idrostatico e meccanico per capire quali sono i fattori che maggiormente influiscono negativamente sulle prestazioni della trasmissione. I fattori analizzati sono: - Perdite di carico dall’uscita degli strozzatori all’ingresso del pozzetto. Diminuzione della viscosità nel fluido. - Comprimibilità del fluido. - Numero di pattini idrostatici contrapposti effettivamente attivi. - Errori sulle tolleranze geometriche e sul montaggio che provocano una variazione non voluta dei meati. Si è eseguita una campagna sperimentale rilevando l’andamento della pressione a valle degli strozzatori durante la movimentazione dell’asse. Le analisi svolte hanno permesso di identificare le seguenti risultati: - La pressione varia in modo estremamente significativo (da 32 a 18 bar) al variare della posizione della chiocciola. Ciò è indice di errori geometrici sulla cremagliera molto diversi tra una cremagliera e l’altra. - L’andamento della pressione presenta un andamento fortemente periodico con frequenze caratteristiche (1xGiroVite e 2xGiroVite). Tale andamento dipende dall’eccentricità della vite, dalla cedevolezza della struttura e dai giochi dei cuscinetti che sostengono la chiocciola idrostatica. - Il contenuto in frequenza dei segnali di pressione varia in funzione della velocità dell’asse, precisamente il contributo (10xGiroVite) aumenta in modo significativo. Tale contributo dipende dal sistema di pre-riempimento dei condotti della vite che evidentemente non funziona a dovere. Una volta analizzati i dati sperimentali e studiati quelli che potrebbero essere i fattori che influenzano le grandezze idrostatiche (pressioni e portate), si decide di implementare un modello integrato in ambiente Matlab - Simulink® - SimScape™. Il modello, con un approccio multi-disciplinare necessario ad interpretare il reale funzionamento della macchina, contiene le seguenti parti: idraulica (centralina idraulica, strozzatori, distributore e vite), idrostatica (pattino idrostatico contrapposto), meccanica (componenti meccanici della catena cinematica) e l’azionamento (anello di velocità e posizione). I risultati ottenuti dal modello sono buoni sia in termini di validazione del modello che di stima delle prestazioni. - Perdita di pressione del sistema idraulico: ingresso distributore, distributore e vite: reale 7 bar – modello integrato 6,5 bar. - Pressione uscita strozzatori: reale 32-18 bar – modello integrato 25 bar. - Portata nei condotti ai fianchi dei denti: reale 9-12 l/min – modello integrato 11 l/min. Le portate sono per fianco, in condizioni nominali. - Delta pressione tra i fianchi dei denti sotto un carico inerziale di 10000 N: reale 6,5 bar – modello integrato 6 bar. - Rigidezza statica chiocciola: reale 1500-2000 N/μm – modello integrato 1800 N/μm (entrambe valutate nella condizione di massimo carico). - Spinta massima ammessa: reale 50000 – 60000 N – modello integrato 60000 N (se si accetta un meato minimo di 10 μm). I principali risultati prodotti sono: - Analisi ed interpretazione del funzionamento dell’asse idrostatico. - Sviluppo di strumenti in ambiente Matlab - Simulink® - SimScape™ a supporto della progettazione dell’asse. - Proposta di modifiche costruttive.
Analisi di una trasmissione a chiocciola idrostatica per fresalesatrici di grandi dimensioni
CASSINARI, ALBERTO
2010/2011
Abstract
The subject of this study is the hydrostatic screw used to drive axes of very large horizontal milling machines. This mechanism is used for the horizontal motion whose travel is up to 30 meters and as well for the vertical motion whose travel is up to 8 meters. The moving machine column has a mass of 120 ton which is causing huge inertial loads. This machine is equipped with a high torque spindle motor that require an axial load, due to cutting forces, up to 40000-50000 N. The current system has an axial force (60000 N) just enough to cover the maximum requested even if from the theoretical design the axial force was supposed to be around 90000 N. Due to the gap between the design and the performances of the real system it has been decided to study deeper the hydrostatic and mechanical system in order to understand what is not letting achieving the target as per the original specification. The various factors that may significantly affect the greatnesses of interest have been analyzed; they are: • Pressure loss at the output of capillary restrictor at the input of pockets. • Viscosity decrease inside the fluid. • Fluid compressibility. • Number of opposed hydrostatic bearings simultaneously engaged. • Errors due to geometrical tolerances and assembly errors that may cause unwanted variation of the clearance. Measurements have been taken on the machine to monitor the pressure value during the axis motion and here below is what has been observed: • The pressure in changing within a large range (from 32 to 18 bar) and dependent from the angular position of the nut. This is surely highlighting geometrical errors between different racks. • The pressure fluctuation has a periodical behavior and strongly linked to the characteristic frequencies (1XScrewRevolution and 2XScrewRevolution). This behavior is caused by the screw run out, not enough stiffness from the structure and by the too large bearing backlash of the rotating hydrostatic lead screw. • The frequency content of the pressure signals is changing according the axis speed; precisely the contribution (10XScrewRevolution) is significantly increasing. This contribution depends from the oil refill system of the screw channels that is clearly not performing as expected. Once the measurements on the machine were analyzed and studied the factors that are influencing the hydrostatic greatnesses (pressure and flow) it has been decided to develop an integrated model inside Matlab - Simulink® - SimScape™ environment. The implemented model developed with a multi tasks approach is necessary to understand the real machine behavior and it contains the following parts: hydraulic (hydraulic power unit, capillary restrictor, distributor and screw), hydrostatic (opposed hydrostatic bearing), mechanical (all components of the kinematic chain) and the portion relevant to the axis driver (speed loop and position loop). The results out from the model are reliable enough to validate the model and to estimate the axis performance: • Loss of pressure of the hydraulic system: distributor input, distributor and screw: 7 bar real – 6.5 bar in the integrated model. • Pressure out of the capillary restrictor: 32-18 bar real – 25 bar in the integrated model. • Flow inside the channels on the teeth side: 9-12 l/minute real – 11 l/minute in the integrated model, the flow is referred to one single tooth side and in the nominal condition. • Pressure fluctuation between teeth side under a inertial load of 10000 N: 6.5 bar real, 6 bar in the integrated model. • Static nut stiffness: 1500-2000 N/μm real, 1800 N/μm in the integrated model (both evaluated under the maximum load condition). • Maximum axial load: 50000-60000 N real, 60000 N in the integrated model (if a clearance of 10 μm is accepted). The main results produced by the developed model are: • Analysis and understanding of the hydrostatic axis functioning. • Development of software tools within Matlab - Simulink® - SimScape™ environment to support the axis design. • Proposal for design modification.| File | Dimensione | Formato | |
|---|---|---|---|
|
Analisi di una trasmissione a chiocciola idrostatica per fresalesatrici di grandi dimesioni.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
4.23 MB
Formato
Adobe PDF
|
4.23 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/10589/45961