Hi all,
I want to compare the deformation under vertical load of two similar structures, I’d like to do it with kangaroo 2, however I don’t know how the “load” goal node behave together with the other nodes and more importantly how the solver distribute the error on the points according to it’s weight and the wheight of other goal nodes.
anyone knows how does it work? or are there online resources that I could read?
You can take a look at K2engineering:
It is a plugin that contains a number of custom K2 goals, which makes it easier to calibrate K2 simulations to accurate structural behaviour and visualise forces/displacements. There are some example files on Github that can help you get started.
I wasn’t aware of this plugin! Thanks a lot! It might be exactly what I’m looking for!
R
Do you know if there are resources online to learn it? I can not find anything specific and it’s not clear to me how it works exaclty
The examples I posted in this thread might be of help:
1 Like
@riccardo.foschi2 You can find example files for K2E on Github as well to get started:
Thank you so much, I also tried to download the example files of kangaroo engineering from the food 4 rhino page but I get an error message when I try to pen them, yours works fine thanks.
So I’m not sure if I understand correctly how the strengths relate with physical values, you wrote that the forces at anchor points should be 1000 newtons, but how do they relate with the strengths values assigned to the other goal nodes? changing them will also change the amplitude of the strengths at anchor points is it right?
The strength value for an anchor is not actually a force in Newtons, but the strength of a spring attaching it to the target location. If you want to make an anchor absolute though, so it doesn’t move at all, just set its strength to 1e300.
But if I do so it’s weight won’t prevail too much on other weights making the simulation not moving at all?
if it’s not the case why it’s that??
is it maybe a special value that makes the anchor node behaving differently from its normal behaviour?(like a hard constrain maybe?)
Does this magic number also works with other goal nodes such as LengthLine??
It’s true that setting an extremely high strength like this for a goal will tend to make the points this goal acts on not move at all. However, since they are anchors this is exactly what we want here. Their high weight will not slow down the movement of other points, only the ones they are anchoring.
We can’t use this approach for other goals though - to make a moving line with absolutely constrained length for example, because of this fixing effect. For that we’d need to use a different approach - such as the soft/hard solver.
The ‘magic’ of 1e300 is just that it’s about as big as a double precision number is allowed to be. So it effectively completely overrides any other goals acting on that point.
On the subject of numerical error while modelling, for comparison try taking the length of a vector (1,1,0) and squaring it then comparing the resulting value to 2.0 (what the exact answer should be). You’ll see it goes wrong somewhere around the 15th decimal place. These minuscule errors are present even in such a simple case, and we don’t worry about them. The anchor errors we are talking about here are hundreds of decimal places smaller.
thanks a lot for the clarification.
About the soft hard solver I never used it, let’s say that I want to make moving lines with absolutely constrained lenght, and at the same time setting the anchor to be hard can I do it? 2 hard constrains.gh (19.5 KB)
does the soft\hard zombie component exists already?