Basic usage¶

Installation¶

Download the source code from GitHub and place it in your working directory.

Initialization¶

Add the library to the MATLAB path by running the ./src/initToolbox.m script.

run('./src/initToolbox.m')

Note

The script has to be executed every time a new instance of MATLAB is started.

Introduction to Jelly¶

Suppose we want to model a 3D continuum soft robot of two bodies modeled under the Piecewise Constant Curvature (PCC) hypothesis. Each body has rest length $L_{0} = 0.3 [\mathrm{m}]$ , base and tip radius $R_{\mathrm{base}} = 0.01 [\mathrm{m}]$ and $R_{\mathrm{tip}} = 0.01 [\mathrm{m}]$ , respectively, and mass density $\rho = 1062 [\mathrm{kg/m^{3}}]$ . Furthermore, $E = 6.66 [\mathrm{MPa}]$ is the Young modulus, $\nu = 0.5$ the Poisson ratio and $\eta = 0.1 [\mathrm{s}]$ the material damping coefficient. Finally, use $N_{\mathrm{Gauss}} = 10$ Gaussian for the numerical computation of the kinematics and the integrals.

To create such robot two main ingredients are needed:

Two joints that connect the robot bodies;
Two 3D PCC bodies.

Create two fixed joints to connect the first body to the robot base and second body to the first.

%Allocate the joints and store them in a cell array
Joints = {FixedJoint(); FixedJoint()};

Now, create two 3D PCC bodies by using the PCC3D class.

%Rest length
L              = 0.3;
%Base radius
R_base         = 0.01;
%Tip radius
R_tip          = 0.01;
%Mass density
rho            = 1062;
%Young modulus
E              = 6.66*1e6;
%Poisson ratio
nu             = 0.5;
%Material damping
eta            = 0.1;
%Number of Gaussian points
N_Gauss        = 10;
%Create a vector of parameters
BodyParameters = [L; R_base; R_tip; rho; E; nu; eta; N_Gauss];
%Allocate the bodies and store them in a cell array
Bodies         = {PCC3D(BodyParameters); PCC3D(BodyParameters)};

In Jelly, a robot is modeled by the BodyTree class, which is a serial assembly of $N$ joints and bodies. To build the robot create a BodyTree instance.

Robot = BodyTree(Joints, Bodies);

Now, use the Robot object to compute some dynamic terms.

%Configuration variables and time derivatives
q   = [0; -pi; 0; pi/3; pi/4; -0.01];
dq  = zeros(Robot.n, 1);
ddq = zeros(Robot.n, 1);
%Generalized actuation force
tau = ones(Robot.n, 1);

%Evaluate the inverse dynamics
Robot.InverseDynamics(q, dq, ddq)

%Evaluate the forward dynamics
Robot.ForwardDynamics(q, dq, tau)

%Evaluate the mass matrix
Robot.MassMatrix(q)