Model Drivetrain Noise

This example shows how to inject a fault into a drivetrain using a Torque Noise Source block. Injecting noise into your model allows you to predict how your actual physical system responds when it experiences environmental or internal disturbances. It also allows you to test the robustness and responsiveness of your control system.

In the example, the noise is injected into one of the sprockets on a metal sheet feeder that is moving a piece of stock. Although you can perform most of the steps in this example using tools that the Simulink^® and Simscape™ Driveline™ user interfaces provide, programmatic commands are supplied. You can combine the programmatic commands to create a script for parameter sweeps.

1. Open the model. At the MATLAB^® command prompt, enter:

   openExample('sdl/SheetMetalFeederExample')
   model = 'SheetMetalFeeder';

   

2. Simulate the model and plot the results.

    Script for Generating and Plotting Simulation Results

   %% Simulate
   sim(model)
   
   %% Define the data
   simlog1 = simlog_SheetMetalFeeder;
   time = simlog1.Sensor_Stock.Motion_Sensor.x.series.time;
   sheetPosition = simlog1.Sensor_Stock.Motion_Sensor.x.series.values;
   sprocketTorque = simlog1.Sprocket_to_Belt_R.t.series.values;
   
   %% Create figure
   figure1 = figure('NumberTitle','off',...
       'Name','Simscape Results: SheetMetalFeeder_noise',...
       'OuterPosition',[565 52 733 822]);
   
   %% Create subplot 1: Sprocket to Belt L Tourque versus Time
   % Create axes for 
   axes1 = axes('Parent',figure1,...
       'Position',[0.13 0.709 0.775 0.216]);
   hold(axes1,'on')
   box(axes1,'on')
   grid(axes1,'on')
   % Create title & axis labels
   title('Stock Position versus Time')
   % xlabel('Time (s)')
   ylabel('m')
   % Create plot 
   plot(time,sheetPosition,'Parent',axes1) 
   
   %% Create subplot 2: Sprocket to Belt R Tourque versus Time
   % Create axes
   axes2 = axes('Parent',figure1,...
       'Position',[0.13 0.409 0.775 0.2156]);
   hold(axes2,'on')
   box(axes2,'on')
   grid(axes2,'on')
   % Create title & axis labels
   title('Sprocket-to-Right-Belt Torque versus Time')
   xlabel('Time (s)')
   ylabel('N*m')
   % Create plot with linked axes
   plot(time,sprocketTorque,'Parent',axes2)
   linkaxes([axes1,axes2],'x')

   

3. Add, configure, connect, and arrange these blocks as shown:

   • Step block — Specify 7.5 for the Step time parameter and 300 for the Final value parameter.

   • Simulink-PS Converter block

   • Mechanical Rotational Reference block

   • Torque Noise Source block — For the Sample time, specify 1e-1, for Repeatability, select Specify seed, and for Seed specify 0.

    Script for Adding, Configuring, Connecting, and Arranging Blocks

   %% Expand the model window to make room for new blocks
   set_param(model,'Location',[108    73   881   682])
   
   %% Add and configure Step block
   % Define block
   stepPath = [model,'/Step'];
   stepLib = 'simulink/Sources/Step';
   stepPosition = [-195   205  -165   235];
   stepTime = '7.5';
   stepValue = '300';
   % Add block
   add_block(stepLib,stepPath,'Position',stepPosition)
   % Configure Step block
   set_param(stepPath,'Time',stepTime,...
       'After',stepValue)
   % Check block configuration
   open_system(stepPath)
   pause(3);
   close_system(stepPath)
   % Get output port
   stepPortHandle = get_param(stepPath,'PortHandles');
   stepOutport = stepPortHandle.Outport;
   
   %% Add and configure S-PS Converter block
   % Define block
   sPSConvPath = [model,'/Simulink-PS Converter'];
   sPSConvLib = ...
      'nesl_utility/Simulink-PS Converter';
   sPSConvPathPosition = [-115   205   -85   235];
   % Add block
   add_block(sPSConvLib,sPSConvPath,'Position',sPSConvPathPosition)
   % Get output port
   sPSConvPortHandle = get_param(sPSConvPath,'PortHandles');
   sPSConvInport = sPSConvPortHandle.Inport;
   sPSConvConPort = sPSConvPortHandle.RConn(1,1);
   
   
   %% Add Mechanical Rotational Reference block
   % Define block
   mechRotRefPath = [model,'/Mechanical Rotational Reference'];
   mechRotRefLib = ...
      'fl_lib/Mechanical/Rotational Elements/Mechanical Rotational Reference';
   mechRotRefPosition = [-40   255   -20   275];
   % Add block
   add_block(mechRotRefLib,mechRotRefPath,'Position',mechRotRefPosition)
   %% Get output port
   mechRotRefPortHandle = get_param(mechRotRefPath,'PortHandles');
   mechRotRefConPort = mechRotRefPortHandle.LConn(1,1);
   
   %% Add and configure Torque Noise Source block
   % Define block
   torqueNoisePath = [model,'/Torque Noise Source'];
   torqueNoiseLib = 'sdl_lib/Sources/Torque Noise Source';
   torqueNoisePosition = [25   210    65   250];
   %% Add block
   add_block(torqueNoiseLib,torqueNoisePath,'Position',torqueNoisePosition)
   %% Get output port
   torqueNoisePortHandle = get_param(torqueNoisePath,'PortHandles');
   torqueNoiseConPort1 = torqueNoisePortHandle.LConn(1,1);
   torqueNoiseConPort2 = torqueNoisePortHandle.LConn(1,2);
   torqueNoiseConPort3 = torqueNoisePortHandle.RConn(1,1);
   %% Configure  block
   set_param(torqueNoisePath, 'sample_time','1e-1',...
       'repeatability','3')
   %% Check block configuration
   open_system(torqueNoisePath)
   pause(3);
   close_system(torqueNoisePath)
   
   %% Get block points for connecting as a branched line
   torqueNoisePortPoints = get_param(torqueNoisePath,'PortConnectivity');
   [torqueNoiseLConnPoints1,torqueNoiseLConnPoints2,torqueNoiseRConnPoints]...
       = torqueNoisePortPoints.Position;
   sprocketPath = [model,'/Sprocket to Belt R'];
   sprocketPortHandle = get_param(sprocketPath,'PortHandles');
   sprocketConPort1 = sprocketPortHandle.RConn(1,1);
   sproketPortPoints = get_param(sprocketPath,'PortConnectivity');
   [sproketLConnPoints,sproketRConnPoints] = sproketPortPoints.Position;
   
   %% Connect blocks
   add_line(model,stepOutport,sPSConvInport)
   add_line(model,sPSConvConPort,torqueNoiseConPort1)
   add_line(model,mechRotRefConPort,torqueNoiseConPort2)
   add_line(model,[sproketLConnPoints;torqueNoiseRConnPoints])
   
   %% Hold to check connections
   pause(5);

   

4. Simulate the model and plot the results.

    Script for Generating and Plotting Simulation Results

   %% Simulate model
   sim(model)
   
   %% Define the datalog variable
   simlog2 = simlog_SheetMetalFeeder;
   time = simlog2.Sensor_Stock.Motion_Sensor.x.series.time;
   sheetPosition = ...
       simlog2.Sensor_Stock.Motion_Sensor.x.series.values;
   sprocketTorque = simlog2.Sprocket_to_Belt_R.t.series.values;
   noiseTorque = simlog2.Torque_Noise_Source.t.series.values;
   
   %% Update plots 1 & 2
   figure1;
   hold on
   axes1;
   hold on
   plot(time, sheetPosition,'Parent',axes1,'Linestyle','--','Color','r')
   axes2;
   hold on 
   plot(time,sprocketTorque,'Parent',axes2','Linestyle','--','Color','r')
   
   
   %% Create subplot 3: Sprocket to Belt R Tourque versus Time
   % Create axes
   axes3 = axes('Parent',figure1,...
       'Position',[0.13 0.11 0.775 0.200]);
   hold(axes3,'on')
   box(axes3,'on')
   grid(axes3,'on')
   
   % Create title and axis labels
   title('Noise Source Torque versus Time')
   xlabel('Time (s)') 
   ylabel('N*m')
   linkaxes([axes1,axes2,axes3],'x')
   
   % Create plot
   plot(time,noiseTorque,'Parent',axes3,'Color','r')
   legend(axes1,'No Noise','Noise','Location','southeast');
   ylim(axes3,[-120 120]);

   

   The noise injected at simulation time, t = 7.5 seconds introduces torque disturbances.

5. Zoom in to see the effects of the disturbance.

    Script for Zooming In

   %% Zoom to examine data at peak noise
   xlim(axes3,[12.0 14.50]);
   ylim(axes3,[-120 120]);

   

   When the noise source torque exceeds ±50 N*m, it most significantly effects the torque applied by the sprocket and, therefore, the position of the stock.