cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - Did you get called away in the middle of writing a post? Don't worry you can find your unfinished post later in the Drafts section of your profile page. X

Generating a hemispherical dome surface plot using a cosine function

bobg100x
8-Gravel

Generating a hemispherical dome surface plot using a cosine function

I want to use Mathcad 6.0 to create a color-coded mesh surface plot of a hemispherical dome whose height ranges between 0 and 1.0 as defined by the function cos(πx/2a), 0 ≤ x ≤ a, with a = 300. The surface is “swept out” by taking this curve [plotted along the x axis] and rotating it a full 360 degrees about x= 0.  

1 ACCEPTED SOLUTION

Accepted Solutions

Here is a quick utility function which creates the number of slices demanded and also takes care that they all have equal height.

It will work OK only if the function provided is monotone. Otherwise there might be different regions of x values with the same z-values which would make creating the slices much more difficult. So the function assumes that the min and max values for z are found at either end of the x-range provided.

Werner_E_3-1642769733645.png

Werner_E_2-1642769110756.png

 

 

 

View solution in original post

22 REPLIES 22
StuartBruff
23-Emerald II
(To:bobg100x)


@bobg100x wrote:

I want to use Mathcad 6.0 to create a color-coded mesh surface plot of a hemispherical dome whose height ranges between 0 and 1.0 as defined by the function cos(πx/2a), 0 ≤ x ≤ a, with a = 300. The surface is “swept out” by taking this curve [plotted along the x axis] and rotating it a full 360 degrees about x= 0.  


 

I don't have much experience with Prime 3D plots, but AFAIA, you can't apply a colormap to a 3D surface.  I'm glad to be corrected on this.

 

However, it is possible to (partially) code a surface by building it up in strips.  Use the CreateMesh function to create the strips and then individually type the name of each strip into the 3DPlot placeholders and set the characteristics you want..  I've got Prime 7, so posting the worksheet won't help you.

 

2022 01 18 b.png

 

It would be straightforward in Mathcad 15.

 

 

2022 01 18 a.png

 

Stuart

LucMeekes
23-Emerald III
(To:StuartBruff)

Here's Stuarts file for Prime 6

 

Success!
Luc

Luc and Stuart,

Yes, I am using Mathcad 6.0 Prime. Because I have had very limited experience using this software, I'm totally flummoxed by the solution provided. For instance, I don't understand how "Curve(a, b)" defined in the worksheet relates to the cosine function I'm working with; also what does "K" denote? Frankly, its quite obvious I bit off much more than I can chew and should not have bothered those in the Mathcad community who responded to my question. My only recourse is to familiarize myself with the CreateMesh feature before continuing this discussion.

Bob


For instance, I don't understand how "Curve(a, b)" defined in the worksheet relates to the cosine function I'm working with;

It doesn't relate to your cosine function at all. Stuart demonstrated the cumbersome method of creating a couple of colored stripes using a normal sphere with radius 1.

 


also what does "K" denote?

You asked for a "color-coded mesh surface plot" and it was interpreted as if you would like it to be colored coded by height. Prime does not offer the ability to coloring a surface using a (more or less) continuous color map like Mathcad 15 and below do.

So Stuart suggested a workaround by putting together the sphere by a couple of slices of equal height. Each of theses slices must be plotted separately and can be colored differently from the other.

K simply is the number of slices you want to use. If you change K, you also must change the arguments in the plot and color them the way you like.

Here is the hemisphere with your cosine function. To slice the dome I simply divided the x-range into K equal parts and I added a slice by changing K from 4 to 5. Be aware that by simply dividing the x-range into equal part  the heights of the slices are not equal as you can clearly see in the plot!

Werner_E_0-1642594556522.png

I guess you are not expecting equal scale on all axis. Otherwise you would have to manually change the axis limits and the dome with height 1 would look like this 😈:

Werner_E_1-1642594808406.png

(the bottom brown slice is hardly visible)

 

Prime 6 sheet attached

OK, here is a variant with slices of equal height

Werner_E_4-1642596026451.png

 

Prime 6 sheet attached

 

 

The variant with slices of equal height is exactly what I wanted- -thank you very much!

 

Would it be possible to change the slice colors to [all bright colors, from top to bottom] red, yellow, green, blue, purple? 

 

Also, can a contour plot be created using the same colors I requested? 

 

Bob

 


@bobg100x wrote:

The variant with slices of equal height is exactly what I wanted- -thank you very much!


You are welcome

 


Would it be possible to change the slice colors to [all bright colors, from top to bottom] red, yellow, green, blue, purple? 


Yes, you can color each of the slices individually using the normal Plot menu. But as PTC decided not to implement a decent color picker, you are limited to the 60 predefined colors.

Werner_E_0-1642634674591.png

 


Also, can a contour plot be created using the same colors I requested? 


Nope. Using Primes contour plot limits you to one of six predefined color schemes. But at least you are allowed to edit the first, second and last value in the color legend. The rest is equally spaced corresponding to equal heights (which fortunately is what you are looking for anyway). In the pic below I set the limits to 0 and 1 and the second value to 0.1 which means a total of 10 slices.

Werner_E_1-1642635273934.png

If you find a color scheme which fits you needs you can try to simulate the same colors in the 3D plot.

Another option might be to use the normal 3D plot instead of the contour plot and fiddle around until you have the final view - quite difficult and you also loose the color legend.

Werner_E_2-1642635586292.png

 

 

Thanks again for applying your obvious expertise to my question, Werner. Just wondering... Could you use the same methodology to create a color-coded mesh plot of an exponential decay-type of surface whose height ranges between (almost) 0 and 1.0 as defined by the function

bobg100x_0-1642751524251.png  , 0 ≤ x ≤ a, with a = 300 and c = 5.5. Visualize the surface as being swept out by taking this curve (plotted along the x axis) and rotating it a full 360 degrees about a vertical axis located at x = 0.

LucMeekes
23-Emerald III
(To:bobg100x)

Would this do it?

 

Success!
Luc


@bobg100x wrote:

Thanks again for applying your obvious expertise to my question, Werner. Just wondering... Could you use the same methodology to create a color-coded mesh plot of an exponential decay-type of surface whose height ranges between (almost) 0 and 1.0 as defined by the function

bobg100x_0-1642751524251.png  , 0 ≤ x ≤ a, with a = 300 and c = 5.5. Visualize the surface as being swept out by taking this curve (plotted along the x axis) and rotating it a full 360 degrees about a vertical axis located at x = 0.


Are yout talking about the built-in contour plot or about Stuarts approach of creating the 3D plot using a couple of slices?

In any way you simply have to switch the cosine function for the new exp function as Luc just had shown (you have to recalculate his sheet to see the results).

In case of the 3D plot with slices you may notice that the slices won't have equal heights - this would require additional changes.

Here is a quick utility function which creates the number of slices demanded and also takes care that they all have equal height.

It will work OK only if the function provided is monotone. Otherwise there might be different regions of x values with the same z-values which would make creating the slices much more difficult. So the function assumes that the min and max values for z are found at either end of the x-range provided.

Werner_E_3-1642769733645.png

Werner_E_2-1642769110756.png

 

 

 

The easiest way! Without cone!

spere.png


@ValeryOchkov wrote:

The easiest way!

 


Hmm, I guess, when the OP wrote "I want to use Mathcad 6.0" he meant Prime 6 (also named Mathcad Prime 6) and not real Mathcad 6. And I am sure that you know that Prime does not support (out of the box) other coordinate systems than cartesian in its 3D plots.

I don't know but are you sure that old Mathcad 6 (dating back to 1995) provided similar 3D plot capabilities like your Mathcad 15?

StuartBruff
23-Emerald II
(To:Werner_E)

Unless the message has leaked through from an alternate universe where Mathcad Prime 6 has an M15 3DPlot analogue ... I feel a transdimensional experiment is called for.   "Boris!  To the lab, my trusty minion."

LucMeekes
23-Emerald III
(To:Werner_E)

Here's the plot requested by the OP, though modified in the sense that I changed a=300 to a=3, in Mathcad 6 (Yes, I fired up an old machine with Mathcad 6 plus installed).

 

LucMeekes_0-1642604745637.png

Success!
Luc

 

Nice, but the OP request the cosine curve to be rotated not about the x-axis but about x=0, which means an axis orthogonal to the x-axis. As he described the surface as a "dome" I supposed that the axis of rotation should be vertical.

BTW, when I open your sheet in MC15, I see this

Werner_E_0-1642606216902.png

When I create a new plot and format appropriate, I can duplicate your plot

Werner_E_1-1642606567089.png

 

 

LucMeekes
23-Emerald III
(To:Werner_E)

Guess you're right. I misread the rotation.

Don't know why Mathcad 15 would interpret the plot differently. Or Mathcad 11.... They must have changed some internal administration in one of the intermediate versions.

 

Bottom line is though: Even Mathcad 6, some 20 years ago, was ahead of Prime 7 in certain aspects.

 

Luc


@LucMeekes wrote:

Guess you're right. I misread the rotation.

Don't know why Mathcad 15 would interpret the plot differently. Or Mathcad 11.... They must have changed some internal administration in one of the intermediate versions.

 

Bottom line is though: Even Mathcad 6, some 20 years ago, was ahead of Prime 7 in certain aspects.

 

Luc


On the other hand - its quite easy for a software to be ahead of Prime 7, especially when it comes to plotting and especially concerning 3D plots.


@LucMeekes wrote:

Yes, I fired up an old machine with Mathcad 6 plus installed


You are not the only one 😉!

 

LucMeekes
23-Emerald III
(To:LucMeekes)

Here's the corrected plot.

LucMeekes_0-1642705055562.png

The difference between Mathcad 6 versus 11 and higher is that mathcad 6 requires the three matrices X, Y and Z to be separately placed in the 3d graph, while Mathcad 11 and higher require them to be enclosed in parentheses.

 

Luc

 

Thank you for the effort you expended to answer my question, Luc. The apparent simplicity of generating my desired 3D plot using pre-Mathcad 6.0 Prime software makes me wish I had purchased the latter-type software.

 

Bob


@ValeryOchkov wrote:

The easiest way!

 


In Mathcad<=15, yes. However, when I plot it, I get this after applying a colormap with the default settings:

 

2022 01 18 c.png

 

Not only is the hemisphere in an unusual orientation for a hemisphere (unusual not incorrect, obviously), but the coloring also doesn't accord with the normal "vary by hemisphere height" display.  An easy way to fix the colouring problem is to change the colormap direction:

 

2022 01 18 d.png

 

Alternatively, change the coordinate ranges:

 

 

2022 01 18 e.png

 

 

 

(And just for completeness, here's the default cylinder)

 

2022 01 18 f.png

 

 

Stuart

 

I would suggest a change in the 

Top Tags