• Non ci sono risultati.

Dynamic Model of a Full Toroidal Variator

N/A
N/A
Protected

Academic year: 2021

Condividi "Dynamic Model of a Full Toroidal Variator"

Copied!
11
0
0

Testo completo

(1)

Dynamic Model of a Full Toroidal Variator

• The Full Toroidal Variator (FTV) is the core element in many mechanical hybrid systems for vehicle applications like Kinetic Energy Recovery System and Infinitely Variable Transmission (IVT) because it allows to manage the power transfer with continuous variation of the speed-ratio.

• In KERS applications the FTV manages the power flows between the ve- hicle and a flywheel (the kinetic energy is stored in the flywheel during braking and then reused under acceleration of the vehicle). A schematic of the considered FTV is the following:

r2θ

ω2 J2, b2 K2b

r3θ ω3

J3, b3

Kb3

τ3 ra

Jb,bb

ωb

rb

θ

r2

K12

r1 ω1

J1, b1

τ1

(2)

• Meaning of parameters and variables of the system:

J1 shaft moment of inertia

b1 shaft linear friction coefficient ω1 shaft angular velocity

τ1 torque acting on the shaft r1 gear radius

K12 stiffness coefficient between shaft and inner disk J2, J3 inner and outer disks moments of inertia

b2, b3 inner and outer disks linear friction coefficients ω2, ω3 inner and outer disks angular velocities

τ3 torque acting on the outer disk r2 inner disk radius

θ tilt angle of the rollers r2θ, r3θ contact radial positions K2b, K3b contact stiffness coefficients

Jb rollers moment of inertia

bb rollers linear friction coefficient ωb rollers angular velocity

rb rollers radius

ra toroid centerline radius

• The two toroidal cavities contain six rollers (three for each cavity) that transfer the power from the inner disc to the outer disc and viceversa.

The variator speed-ratio is a function of the rollers inclination: a change of the tilt angle θ of the rollers causes a speed-ratio variation.

• The contact radial positions r2θ and r3θ are related to radii ra and rb as follows:

r = ra + rbsin θ, r = ra − rb sin θ.

• The gear ratio R(θ) relating the angular speed ω2 of the inner disk to the angular speed ω3 of the outer disk is function of the tilt angle θ as follows:

R(θ) = ω3

ω2

= r2θ

r3θ

= ra + rb sin θ

ra − rbsin θ = 1 + α sin θ 1 − α sin θ where

α = rb ra

(3)

• Plot of the gear ratio R(θ) for different values of α ∈ [0.1 : 0.1 : 0.9]:

−100 −80 −60 −40 −20 0 20 40 60 80 100

10−1 100 101

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Gear ratio R(θ) for different values of α

R)

θ [degree]

• The FTV is a non-linear and time-variant system and writing its lumped parameter dynamic model is non trivial problem.

• The POG model of this physical system can be easily obtained introducing

“fictitious” elastic elements between the inertias of the system. The elastic elements are the only dynamic element that can be put between two masses or inertias respecting the integral causality of the system.

• The presence of these elastic elements simplify the writing down of a dynamic model of the system. The obtained POG model is characterized only by integral causality element suitable for simulation.

• The elastic elements can be afterwards eliminated applying a proper con- gruent transformation to the original system and so obtaining the POG dynamic model of the reduced system.

(4)

• The POG block scheme of the considered FTV system is the following:

τ1

ω1

1 Shaft

- 

?

1 s

?

J11

?

 -

 

b1

6

6

- -

2 Stiffness and gear radii between shaft and inner disk

- r1 -

 r1  -

6

1 s

6

K12

6

f12

-   r2  - r2 -

3 Inner disk

- 

?

1 s

?

J21 ω?2

 -

 

b2

6

6

- -

4 Stiffness and gear radii between inner disk and roller

- r2θ -

 r2θ  -

6

1 s

6

K2b

6

- 

f2b

˙xb

· · ·

· · · f2b

˙xb  rb 

- rb -

5 Rollers

- 

?

1 s

?

Jb1 ω?b

 -

 

bb

6

6

- -

6 Stiffness and gear radii between rollers and outer disk

- rb -

 rb  -

6

1 s

6

Kb3

6

fb3

-   r3θ  - r -

7 Outer disk

- 

?

1 s

?

J3−1

?

 -

 

b3

6

6

- -

8

τ3

ω3

• The state space model L ˙x = A x + Bu of the POG system is:











J1 ˙ω1 1 K12

12

J2 ˙ω2 1 K2b

2b

Jb ˙ωb 1 Kb3b3

J3 ˙ω3











| {z }

L ˙x

=











−b1−r1

r1 0 −r2

r2 −b2 −r r 0 −rb

rb −bb −rb

rb 0 −r

r3θ −b3











| {z }

A









 ω1

f12

ω2

f2b ωb fb3

ω3











| {z }

x

+











1 0 0 0 0 0 0 0 0 0 0 0 0 −1











| {z }

B

 τ1

τ3



| {z }

u

• The matrices and vectors L, A, B, x, u can be directly read from the POG block scheme.

(5)

• When K12 → ∞, K2b → ∞ and Kb3 → ∞ the state variables are linked by the following constraint:









0 = r1ω1 −r2ω2

0 = rω2 − rbωb 0 = rbωb − r3θ ω3.









ω2 = r1

r2

ω1

ωb = r

rb ω2 = r r1

rb r2

ω1

ω3 = rb

r ωb = r2θ r1

r r2

ω1.

• The following state space congruent transformation can be used:

x = T(θ) ω1









 ω1

f12

ω2

f2b

ωb fb3

ω3











| {z } x

=









 1 0 R2

0 Rb(θ)

0 R3(θ)











| {z } T(θ)

ω1

where gear ratios R2, Rb(θ) and R3(θ) have the following structure:

R2 = r1

r2

, Rb(θ) = r1r

r2rb = r1(ra + rbsin θ) r2rb , R3(θ) = r1r2θ

r2r = r1(ra + rbsin θ) r2(ra − rbsin θ).

• Since the transformation matrix T(θ) is a column matrix, the transformed and reduced model is a first order dynamic model that can be written in the following explicit form:

J(θ) ˙ω1 + N (θ, ˙θ) ω1 = −b(θ) ω1 +

1 −R3(θ)

| {z }

B¯

1

τ3



(1)

where ¯B = TTB, coefficients J(θ) and b(θ) are

J(θ) = TTLT = J1 + R22J2 + R2b(θ) Jb + R23(θ) J3

b(θ) = −TTAT = b1 + R22b2 + R2b(θ) bb + R23(θ) b3

(6)

and

N(θ, ˙θ) = TTL ˙T = r rb ˙θ cos θ R22

Jb

rb2 + 2 raJ3

r33θ



= J˙(θ) 2 .

• The transformed and reduced system (1) can be graphically represented using the following POG block scheme:

τ1

ω1

1

- 

J1(θ)

1 s

?

?

?

 -

 

N(θ, ˙θ)

6

6

- -

 

b(θ)

6

6

- - -R3(θ) -

 R3(θ)

8

τ3

ω3

where

J(θ) = J1 + R22J2 + R2b(θ) Jb + R32(θ) J3

b(θ) = b1 + R22b2 + R2b(θ) bb + R23(θ) b3

N(θ, ˙θ) = r rb ˙θ cos θ R22

Jb

rb2 + 2rra3J3

3θ

 and

R2 = r1

r2

Rb(θ) = r1(ra + rb sin θ) r2rb

R3(θ) = r1(ra + rbsin θ) r2(ra − rbsin θ).

(7)

• The system can be graphically represented by the following POG scheme:

...

...

T1 J1

w1 b1

T3 T4

v5

T = r1

K12 F6

v7

T8

T = r2

J2

w2 b2

T10 T11

v12

T = r2th

K2b

F13

v14

T15 Jb

wb bb

T17 T18

v19

T = rb

T = rb

Kb3

F20

v21

T22

T = r3th

J3

w3 b3

T24 1

a

3

c

4 5

e 7

7

g

g

8 9

i

11

m

12 13

o 2

b d

6

f

h

10

l n

• This scheme has been obtained using the following source code:

**, Gr, Si, Sn, Si, As, Si, POG, Si, EQN, Si, SLX, Si

**, pSplit, 42, Sr, No, pSr, No, pEsu, No, Split, 7 rT, 1, a, An,-90, Ln, 1.2

rJ, 1, a, Sh, 0.5, Kn, J 1, En, w 1, Kn0, 13, Qn0, 2 - -, 1, 2

- -, a, b

rB, 2, b, Kn, b 1

CB, [2;3],[b;c], Kn, E2=r 1*E1, Sh, [0.6;-0.2]

mK, 3, 4, Ln, 0.8, Kn, K 1 2 - -, c, d, Ln, 0.8

CB, [4;5],[d;e], Kn, E1=r 2*E2, Sh, [0.2;-0.6]

rJ, 5, e, Kn, J 2, En, w 2 - -, 5, 6, Ln, 0.5

- -, e, f, Ln, 0.5 rB, 6, f, Kn, b 2

CB, [6;7],[f;g], Kn, E2=r 2 t h*E1, Sh, [0.6;-0.2]

mK, 7, 8, Ln, 0.8, Kn, K 2 b - -, g, h, Ln, 0.8

CB, [8;9],[h;i], Kn, F2=r b*F1, Sh, [0.2;-0.6]

rJ, 9, i, Kn, J b, En, w b - -, 9, 10, Ln, 0.5

- -, i, l, Ln, 0.5

(8)

rB, 10, l, Kn, b b

CB, [10;11],[l;m], Kn, E2=r b*E1, Sh, [0.6;-0.2]

mK, 11, 12, Ln, 0.8, Kn, K b 3 - -, m, n, Ln, 0.8

CB, [12;13],[n;o], Kn, F2=r 3 t h*F1, Sh, [0.2;-0.6]

rJ, 13, o, Kn, J 3, En, w 3 rB, 13, o, Sh, 0.5, Kn, b 3

• The POG Modeler provides the following POG block scheme:

T1

1 J1

w1 w1

b1 T3

r1

r1

K12 F6

r2 r2

1 J2

w2

b2 T10

r2th

r2th

...

...

K2b

F13

1 Jb

wb

bb

T17

rb

rb

rb

rb

Kb3

F20

r3th

r3th

1 s

1 s 1

s

1 s

1 s

1 s 1

s

1 J3

w3

b3 T24

• The POG state space equations L ˙x = Ax + Bu and y = Cx + Du are:

L ˙x =

b1 r1 0 0 0 0 0

r1 0 r2 0 0 0 0

0 r2 b2 r2th 0 0 0

0 0 r2th 0 rb 0 0

0 0 0 rb bb rb 0

0 0 0 0 rb 0 r3th

0 0 0 0 0 r3th b3

w1

F6 w2 F13

wb

F20

w3

+

1 0 0 0 0 0 0

T1

(9)

where

L =

J1 0 0 0 0 0 0

0 1

K1 2

0 0 0 0 0

0 0 J2 0 0 0 0

0 0 0 1

K2 b

0 0 0

0 0 0 0 Jb 0 0

0 0 0 0 0 1

Kb 3

0

0 0 0 0 0 0 J3

• System parameters:

-- Matlab commands --- (POG_KERS_Par_SLX_m) ---

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

syms J_1 b_1 r_a r_b r_1 r_2 J_2 b_2 r_b J_b b_b J_3 b_3 th R2=r_1/r_2;

r2th=r_a+r_b*sin(th); r3th=r_a-r_b*sin(th);

Rbth=r_1*r2th/(r_2*r_b);

R3th=r_1*r2th/(r_2*r3th);

Jth=J_1+J_2*R2^2+J_b*Rbth^2+J_3*R3th^2;

bth=b_1+b_2*R2^2+b_b*Rbth^2+b_3*R3th^2;

dJ_dth=diff(Jth,’th’)/2;

%%%%%%%%%%%% SYSTEM PARAMETERS %%%%%%%%%%%%%%%%%%%%%%%%%%

J_1=0.026*kg*meters^2; % 2. Inertia. Internal parameter.

b_1=1.06*Nm/(4000*rpm); % 3. Angular friction. Internal parameter.

r_1=17.2*mm; % 5. Parameter. Transformer/Gyrator.

K_1_2=150*Newton/(0.2*mm); % 6. Stiffness. Internal parameter.

r_2=51.5*mm; % 12. Parameter. Transformer/Gyrator.

J_2=0.675*kg*r_2^2; % 9. Inertia. Internal parameter.

b_2=2.2*Nm/(4000*rpm); % 10. Angular friction. Internal parameter.

K_2_b=80*Newton/(0.2*mm); % 13. Stiffness. Internal parameter.

r_a=34.3*mm; r_b=27.5*mm;

J_b=1.0*kg*r_b^2; % 16. Inertia. Internal parameter.

b_b=1.9*Nm/(4000*rpm); % 17. Angular friction. Internal parameter.

K_b_3=80*Newton/(0.2*mm); % 20. Stiffness. Internal parameter.

J_3=100*J_2; % 23. Inertia. Internal parameter.

b_3=1.3*Nm/(4000*rpm); % 24. Angular friction. Internal parameter.

%%%%%%%%%%%% INPUT VALUES %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

th_A=70*degrees; % Maximum amplitude of theta

th_w=3; % Frequency of theta

tt=(0:0.001:10)’; % Time vector th=th_A*sin(th_w*tt); % theta(tt) dot_th=th_w*th_A*cos(th_w*tt); % dot_theta(tt)

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

th_0=th(1); R2=r_1/r_2;

r_2_t_h=r_a+r_b*sin(th_0);

r_3_t_h=r_a-r_b*sin(th_0);

%%%%%%%%%%%% INITIAL CONDITIONS %%%%%%%%%%%%%%%%%%%%%%%%%

T_1_0=0*Nm; % 1. Torque. Input value.

w_1_0=2000*rpm; % 2. Angular momentum. Initial condition.

w_2_0=w_1_0*r_1/r_2; % 9. Angular momentum. Initial condition.

w_b_0=w_2_0*r_2_t_h/r_b; % 16. Angular momentum. Initial condition.

w_3_0=w_b_0*r_b/r_3_t_h; % 23. Angular momentum. Initial condition.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

---

(10)

• Angular velocities ω1, ω2, ωb, ω3 when b1 = b2 = bb = b3 = 0:

0 1 2 3

0 500 1000 1500 2000 2500 3000

w_1 [rpm]

Angular Velocity w 1

0 1 2 3

0 200 400 600 800 1000

w_2 [rpm]

Angular Velocity w 2

0 1 2 3

0 200 400 600 800 1000

w_b [rpm]

Angular Velocity w b

Time [s]

0 1 2 3

0 200 400 600 800 1000 1200

w_3 [rpm]

Angular Velocity w 3

Time [s]

• Energy stored in the system when b1 = b2 = bb = b3 = 0:

0 0.5 1 1.5 2 2.5 3

0 0.2 0.4 0.6 0.8 1 1.2

Total Energy En, Energy En1 stored in in J_1, Energy En3 stored in in J_3

En, En1, En3 [kJ]

0 0.5 1 1.5 2 2.5 3

0 0.2 0.4 0.6 0.8 1 1.2 1.4

Total Inertia J_th

J_th [kg*m2 ]

Time [s]

Time [s]

(11)

• Angular velocities ω1, ω2, ωb, ω3 of the system with dissipations:

0 1 2 3

0 500 1000 1500 2000 2500

w_1 [rpm]

Angular Velocity w 1

0 1 2 3

0 200 400 600 800 1000

w_2 [rpm]

Angular Velocity w 2

0 1 2 3

0 200 400 600 800 1000

w_b [rpm]

Angular Velocity w b

Time [s]

0 1 2 3

0 200 400 600 800 1000

w_3 [rpm]

Angular Velocity w 3

Time [s]

• Stored energy of the system with dissipations:

0 0.5 1 1.5 2 2.5 3

0 0.2 0.4 0.6 0.8 1

Total Energy En, Energy En1 stored in in J_1, Energy En3 stored in in J_3

En, En1, En3 [kJ]

0 0.5 1 1.5 2 2.5 3

0 0.2 0.4 0.6 0.8 1 1.2 1.4

Total Inertia J_th

J_th [kg*m2 ]

Time [s]

Time [s]

Riferimenti

Documenti correlati

Feeding CL2006 worms at L1 larval stage with 50 m M OLE produced a protective effect significantly lower than that occurring when the same dose of drug was administered at L2 (45.8%

This kind of PVD process could be used when grinding operations were performed in order to achieve the right roughness to deposit a ceramic top coat without loosing the

Frequentemente nelle vasche o nelle botti di rinverdimento sono presenti odori di putrefazione oppure si riscontra sulla pelle eccessiva perdita di pelo,

the automated method in the present study were obtained by applying the boundary-effect model to ROIs of increasing volumes or by linearly extrapolating the number of

Different gene expression means between patients with radically resected colon cancer patients showing an unfavourable cancer stem cell gene profile, GROUP A, (red) and

Furthermore, cardiological involvement, leading to hypertrophic remodelling, occurs up to 40/50% of patients with MDs and the mtDNA A3243G tRNA Leu (UUR) mutation has been

the upstream regions of cases exons and less LTRs in the upstream regions of control exons, than in the downstream regions, were noticed; with regions of 230 nucleotides

Leoniceno inizia qui la sua opera di chiarificazione della posizione del medico di pergamo su questo particolare aspetto della sua teoria delle cause delle malattie,