This shows you the differences between two versions of the page.
Both sides previous revision Previous revision Next revision | Previous revision Last revision Both sides next revision | ||
faq [2021/06/06 17:22] marijn.nijenhuis |
faq [2022/11/04 12:08] jan.dejong |
||
---|---|---|---|
Line 11: | Line 11: | ||
This means that there’s no general answer to the number of beams that are needed per element. The approach to deciding on the number is to investigate if the quantity of interest (displacement, | This means that there’s no general answer to the number of beams that are needed per element. The approach to deciding on the number is to investigate if the quantity of interest (displacement, | ||
- | The number of beams that are used per element can be set using the '' | + | The number of beams that are used per element can be set using the '' |
For an entirely rigid element, there is no benefit to using more than one beam. | For an entirely rigid element, there is no benefit to using more than one beam. | ||
Line 18: | Line 18: | ||
* For motion in the DOF (the actuation direction of the mechanism), it is likely that the deformation, | * For motion in the DOF (the actuation direction of the mechanism), it is likely that the deformation, | ||
* For higher mode shapes, associated with motion in a constrained direction of the mechanism, it is likely that multiple beams are needed per flexible element. A good starting point would be to use '' | * For higher mode shapes, associated with motion in a constrained direction of the mechanism, it is likely that multiple beams are needed per flexible element. A good starting point would be to use '' | ||
- | * By default, Spavisual only shows the first 10 mode shapes of the mechanism. To increase this number, see the use of the '' | + | * By default, Spavisual only shows the first 10 mode shapes of the mechanism. To increase this number, see the use of the '' |
* Increasing the number of beams leads to more degrees of freedom and more eigenmodes (with generally higher eigenfrequencies). The state space and transfer function representations will contain many states. If possible, avoid unnecessary conversions from state space to transfer functions. | * Increasing the number of beams leads to more degrees of freedom and more eigenmodes (with generally higher eigenfrequencies). The state space and transfer function representations will contain many states. If possible, avoid unnecessary conversions from state space to transfer functions. | ||
* To use such a detailed model in Simulink, pre-process the state space representation to discard eigenfrequencies higher than the Nyquist frequency. See the project lectures for details. | * To use such a detailed model in Simulink, pre-process the state space representation to discard eigenfrequencies higher than the Nyquist frequency. See the project lectures for details. | ||
- | * Add some modal damping (relative damping of 0.001 to 0.01) to the modes for more representative Bode plots. See the use of the '' | + | * Add some modal damping (a relative damping of 0.001 to 0.01 for steel materials) to the modes for more representative Bode plots. See the use of the '' |
===== 2. Overconstraints could not be solved automatically ===== | ===== 2. Overconstraints could not be solved automatically ===== | ||
Line 46: | Line 45: | ||
**Manual releases**: An alternative solution is to provide appropriate releases for the deformations in the '' | **Manual releases**: An alternative solution is to provide appropriate releases for the deformations in the '' | ||
- | ===== 3. Animated GIFs of mode shapes ===== | + | ===== 3. Changing units ===== |
+ | When the system parameters are expressed with very small numbers, you may receive the warning "Mass coefficients turn out to be very small" or the related error " | ||
+ | |||
+ | The example file that comes with SPACAR Light assumes meters as the default unit for length. To change to other units, your script should be adjusted in several places, depending on how each physical quantity scales with the length L: | ||
+ | * coordinates of the nodes ('' | ||
+ | * input displacements ('' | ||
+ | * (rotations in radians are independent of length) | ||
+ | * moment of inertia ('' | ||
+ | * forces ('' | ||
+ | * moments ('' | ||
+ | * material stiffness properties ('' | ||
+ | * material density ('' | ||
+ | * cross-sectional dimensions ('' | ||
+ | * gravitational acceleration ('' | ||
+ | |||
+ | In other words, if your script assumes meters as the unit for length, and you want to go to millimeters, | ||
+ | |||
+ | Note that output values, so numbers you see in the '' | ||
+ | ===== 4. Bode plots ===== | ||
+ | When making Bode plots of the system, in particular with a lot degrees of freedom (e.g. when '' | ||
+ | |||
+ | {{ :: | ||
+ | |||
+ | It can be helpful to go to Properties (after right-clicking) and the Options tab, and: | ||
+ | * //uncheck// "Wrap phase" | ||
+ | * //check// " | ||
+ | |||
+ | {{ : | ||
+ | |||
+ | This way, multiples of 360 degrees are subtracted from the phase such that the phase at zero frequency is closest to 0 degrees. After zooming in on the phase: | ||
+ | |||
+ | {{ :: | ||
+ | |||
+ | Also see the [[parallel_flexure_guide_transfer|example on producing transfer functions]] for tips on making Bode plots that look nicer. | ||
+ | ===== 5. Animated GIFs of mode shapes ===== | ||
For presentation purposes, it can be helpful to create an animated GIF of a visualization in Spavisual. You can make a GIF for a particular mode shape or for the general motion of the system. In the Spavisual window, deselect the ' | For presentation purposes, it can be helpful to create an animated GIF of a visualization in Spavisual. You can make a GIF for a particular mode shape or for the general motion of the system. In the Spavisual window, deselect the ' | ||