This blog is based on the reading public number "Robot study notes," the study notes from ~
First of all, to sum up the previous contents;
1. stress singularity stress calculation results can not be determined accurately;
displacement result is no singular points, only the stress singularity.
2. The method of determining non-stress and stress singularity grid resolution values;
first calculation result and then look at the size of the red area coverage region.
3. The results of correlation displacement relatively low accuracy of the grid.
The above is to ensure that the simulation results and experimental results can be compared basic requirements, with more awareness of the problem, we can compare a range of analytical and experimental results and theoretical calculation results.
Example: L-bracket, the arrow direction is applied to the red end surface 900N load, both end surface of the fixed rail, and the support rail is zero is assumed fitted state, stress and displacement structures investigated.
Materials: Structure Steel
Then stroked over procedure:
1 . Import model
2 . Because the assembly, so to change the contact (to frictional, 0.2)
. 3 mesh generation. Because we do not know what kind of automatically generates a grid, you may wish to nothing about the set to generate grid. Found crossbar hexahedron mesh, L-shaped bracket is a tetrahedral mesh. So we select the above mesh control this body will be changed to the crossbar tetrahedral mesh.
Because L-bracket there appeared rounded, so need to set the size. Also choose the fillet surface mesh control, and then change the element size 0.5mm.
Finally, the overall grid settings. Click on the mesh, then the following element size in there to 3mm, the transition was changed to slow.
Then generates grid ~
4 . Constraints and add a fixed load.
5 . Solving. Click the static structural recommended solve.
6 . After solving is complete, we should add a calculated result.
emmmm, made this first pass, resulting in an error : an internal solution magnitude limit was exceeded.Node Number 15403, Body Part 2 ...
after Baidu should find the solution process the resulting stress exceeds the strength of our material as defined in the limit, resulting in failure to solve.
Need to check the material properties we set the load applied or if there are problems, such as whether an error value, whether the unit is wrong. If the model, provided no problems and the load material, load value at this time should be reduced.
Investigated again from start to finish, there was found to forget to change the contact frictional provided a friction coefficient of 0.2! ! ! ! Do not forget to have finished hit Enter to confirm! ! !
Thinking ~~:
The results are :
general results of the examinations, the most is the examination of the value of the displacement, stress very few companies will do.
This means we get a message experiment: experimental deformation amount is 1mm. But we do out of the theoretical value of almost about 2.5mm.
The contrast is easy to overlook the problem analysis and experimental results:
Position 1. experimental test is the actual maximum position it?
In fact, we do not know where exactly the location where the experimentally measured.
2. The results show the finite element and is the same type of results?
Experimental test sensor, the sensor will change after deformation, it is the degree of deformation in the direction indicated thimble. Therefore, the results of experimental tests are directional!
The simulation results are combined deformation. So we can choose what deformation direction.
We can look at the changes in the Y direction. Then you can probably find experimental measured position.
In general, we can measure the deformation is deformation direction; can be measured stress is stress
when normal, and then select the fillet surface (equivalent stress posted a paste), and then to select the Z axis, this is the test could sensed.
On resistance strain gauges, need remember the following points:
1. The strain values read are specified position and orientation;
2. stress is not directly obtain the amount (strain amount is a direct access; stress = modulus of elasticity × strain ) in the test system must be defined that the modulus of elasticity , thus apply the experimental and simulation results are compared, the set value of the modulus of elasticity must be consistent. ****
Add, for Display Option selection of Normal Stress:
stress display settings,
select unaveraged displayed without averaging the results of stress, such as stress on the common node will show FIG 20MPa, 22Mpa in each cell region , 18Mpa, 18Mpa.
(Cloud will be displayed when the color difference, a difference may be large around sharp corners.)
Select the average stress results averaged displayed, it is displayed (20 + 22 + 18 the common node such stress value + on FIG. 18) /4=19.5Mpa.
PS. This setting only for stress !
unaveraged stress is real value! In the project is the best display. However, in the calculation accuracy accurate picture, not too much difference between the two.
It recommended unaveraged.
Remember to above comments exclude stress singularity ! Stress singularity is not within the scope of our discussion.
