I found that I could apply some probes myself, and then have a graph of stress vs. # probes.
You can also create a plot along the curve. For example, to create a von-Mises stress along the curve, you'd right click on on the von-Mises stress definition (Stress1 by default), go to 'Probe', select 'On selected entities' under 'Option" at the top, selected the curve in the 'Results' box, hit 'Update' below, then hit the 'Plot' button under 'Results Options'.
However, what I wish to do is have specific values.. what I mean is I would like to obtain the stress at specific points.
The plot points are the values at nodes along the curve, so if you wanted results every 0.1 mm or less, then you'd need to define your mesh such that you have nodes spaced 0.1 mm or less along the curve (such as a mesh control along the edge with the corresponding element size definition).
From 0mm (the center of the notch) to Xmm (the tip) by steps of 0.1mm (or less).
To only plot along half of the curve, you'd first need to create a Split Line to sub-divide the surface, thereby sub-dividing the curved edge. You could skip this part and just plot all of the values along the edge and then export them to Excel if you wanted to avoid the Split Line for whatever reason.
Hi Shaun and Jared,
Thank you for the reply. However, I can't seem to select the curve. I can only select the entire face of the notch. Maybe it's due to how I sketched the part in the first place?
I sketched the outline of a rectangle with a notch and then used the revolution function to have a notched cylinder.
Anyway, when I select the surface and then click on Update, it gives me about 50 000 different points.. I don't know if I can extrapolate the points of a single curve from this data.
And, Jared, I am working on my master's thesis and I need to base my first values on the ones found by the previous master's student. He used Abaqus and determined the stress applied in the notch at the locations I previously mentioned.
I just need to check them to make sure I am in the same range before I move on to other mechanical tests.
Also, I am basing my experimental values on papers that use that same method.
And, finally, I have never used vertices with splitlines.. if you could briefly explain how to do it?
Thank you so much to both of you for taking the time to help me out!
screenshots would really help here.
you can't use curves for plotting, it has to be real geometry. ie create a curve and project it with a splitline.
regarding splitlines with vertices, draw a square, split line. you have 4 vertices. use the concept with a different type of sketch that lets you create vertices wherey ou want to measure.
however based on your description of why you need those particular values, i'm surprised that a probe that is saved as a sensor isn't enough. it should give you the numbers to check ify ou are in the ballpark without being exactly the same locatino as the other report.
Thanks for the post, but I still don't see how to add those splitlines. I've tried adding points all along the curve in a sketch but they stop appearing as soon as the probe is on.
Also, I need to test about 100 specimens and have an FEA for all of them. Not only would this precise probing allow me to check the values from other papers but it would give me some consistency as to where I'm obtaining my stresses. For example, if I use the normal probing technique and have the measurements at points more or less on the curve at regular intervals, I'll find myself with a graph plotting stress vs. probe# (cf screenshot). And since I would need to show what these probe# correspond to in the notch, I'd have to have the screenshot of the curve with the 80-100 probes on it. Multiply this times 100 for all of my samples, it comes to a lot of screenshots. I just wish there were a simpler method where I can just have some kind of stress vs. distance plot, it would be more accurate.
Thanks in advance
Can you post the part file of the geometry? That'll help is determining how to divide up the surface.
That being said, the general process is as follows:
1. Start with the base part.
2. Define a closed sketch to divide the area up (and thereby divide the curve up).
3. Create a Split Line.
4. Use the sketch and respective surface for the Split Line definition.
5. Define a mesh control along the curve such that your nodes are spaced how you need them.
6. Define a probe for the results you'd like to plot.
7. Select the sub-divided curve that was created by the Split Line.
8. Press "Update" and the press "Plot" under the "Report Options" section.
You can check the "Flip edge plot" box to change which end of the curve defines the starting point.
If, for whatever reason, you can't sub-divide your geometry using a Split Line, then you can always plot along the entire curve, export the values to Excel, and then create a plot of the subset of values in there.
Wow thank you for the detailed explanation! Just one question, can I apply this to a cylindrical part? I attached one of my files with this reply, I hope there won't be any problem opening it (Solidworks 2013). I wasn't able to check it since the computer I am on doesn't have Solidworks.
Another question, may be a bit silly but, what is a closed sketch? Maybe I can do this on the plane that passes through the middle of my notched cylinder?
Again thank you.
Closed sketch means not just a line.
And the technique can be applied to any type of geometry. A plane could be used with the intersection split line type.
There several ways you can probably do this, but here is what I did.
1. Start with your base geometry.
2. Go to "Insert" -> "Curve" -> "Split Line"
3. Select "Intersection" for "Type of Split", select "Plane 1" and "Plane 2" for "Splitting Bodies/Faces/Planes" and select the curved/notched surface for "Faces/Bodies to Split".
4. You now have a curved edge to define a mesh control and plot along. First define a 0.1 mm mesh control.
5. Mesh, solve, and then plot your quaintly of interest along this curve.
Please keep in mind that you'll need to apply additional mesh controls and go through additional steps to get converged results (you'll notice some rather large discontinuities in the stress field).
I followed your steps and managed to get the split lines done, as well as the meshing (I'll have to work on that to have it finer but it's a start!)
I can't thank you enough for your kind help.
shaun's really got it here
you can also create points where you want to measure using vertices with splitlines
but why do you need the data at such specific locations?