@Kitonum 

One thing I did not like about my trials was that they effectivley used the operator form of the plot command but showing a variable in the first argument of the plot command. Your variant is much cleaner also in functional notation

plot('r(t)',t= 0..3, size=[1000,400]);

It seems that functional notation inevitably goes hand in hand with quotation marks when using plot

Thanks!

@nm 

I basically have the same setup appart of having less memory (32 Gb) and the fact that your machine is much faster. I could check for the "experiences pack" if you tell me where you did the screenshot.

60 seconds did not have an effect. Maybe you could try shorter 

I will try later during the day on a faster machine.

@TechnicalSupport 

On windows 10 with the latest physics update

@acer 

Maybe the server also needs a restart: I see here more than than 10 deleted comments.

@Carl Love 

Interesting!

Thank you

Select the output in acers answer and drag&drop it to a new input line. This will give you the Maple statement you a looking for. Note that the first argument of hypergeom contains two elements

-r^2*((1 - beta)/gamma)^n*hypergeom([n, n/(a - beta)], [(-1 - n - beta)/(-1 + beta)], -r^((3 - 3*beta)/n)*chi^2/gamma)/3;

@acer 

There is a misunderstanding in terms of numbering.

We both talk about the same best way: the one with the nested seq commands.

I have updated my first reply to you.

@acer

For the intended purpose your way (the second in sequence in this post) is the best. My attempt of using a function might be better for a color function.

The comparison of both approaches is in any case instructive.

Thank you

@janhardo 

Thank you for a commented the code variant. It's simpler than mine!

rmod is even not needed.

@acer 

A comparision of the two methods is not only interesting in terms of speed.

Acers code:

Modified (by acer) attemept

I have two questions:

Why should I use undefined instead of NULL?

Why is the shading not continuous? (I do not need it for stl export but for renderings in Maple it would be desirable)

@mmcdara 

No problem, I do not necessarily expect replys.

I can add to your story the propasal to use Simulink instead of MapleSim. Simulink is more expensive and less user friendly but only more powerfull in some instances. I my case it is digital signal processing with built in blocks and system identification & spectrum analyser component which would have been handy in the case above to probe for high order frequency components in the vertical oscillation. But with the latest update, MapleSim how complies with Modelica 4.0 which brings it closer to process Modelica code developed for Dymola (less adjustments to be done for those who know).

Annother proposal to me was to use Matlab instead of Maple...

@nm 

I have updated my answer with a modifed version. Removing by setting the powers to zero works now for sums and products. There is probably also a way with subsindets.

@mmcdara 

I am impressed. I did not consider this possible only using Maple. The case you demonstrated would probably have interested MapleSoft. Although skilled designers with modern CAD tools can do animation of a mechanism they cannot (?!?) provide the calculation results you were after.

MapleSoft provides a CAD Toolbox to import complex geometries for simulation. That is a big help. However, the tool does not allow generation of contact ports  on surfaces (probably, I tested it 3 years ago). This has to be done manually in MapleSim which might not be straight forward because of limited visibility of hidden model features.

From my point of view it makes totally sense to expand MapleSim functionalities in this direction: Definition of contact surfaces inside the CAD Toolbox.

What happens when you use floats in the sine expression? I can't try it myself at the moment.