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

Community Tip - You can Bookmark boards, posts or articles that you'd like to access again easily! X

Help! Equation Solution Request.

TomD.inPDX
17-Peridot

Help! Equation Solution Request.

How do I "define" a semi-circle by an equation when Y is linear?

I know the cartesian formula:

x = 4 * cos ( t * 180 )

y = 4 * sin ( t * 180 )

z = 0

But I need to compensate for the trajectory (Y) which has a linear value.

I need the formula applied only to X given Y is linear 0-1.

I want to sweep a VSS to make this shape. The trajpar is the given value (t as Y):

equation_question.PNG


This thread is inactive and closed by the PTC Community Management Team. If you would like to provide a reply and re-open this thread, please notify the moderator and reference the thread. You may also use "Start a topic" button to ask a new question. Please be sure to include what version of the PTC product you are using so another community member knowledgeable about your version may be able to assist.
1 ACCEPTED SOLUTION

Accepted Solutions

Dear Antonius

I have tried to create SEMICIRCLE BY SWEEP feature - it is working (you can investigate my model)

Note: I have used for sweep feature the Relation (for UPPER_HALF_OF_THE_SEMICIRCLE) sd4=(r^2 - (trajpar*r)^2) ^ (1/2)

Regards,

Vladimir

Best Regards,
Vladimir Palffy

View solution in original post

30 REPLIES 30

Dear Antonius,

I think that you can combine Sweep feature with Graph feature - the values of graph will be inputs for sweep trajpar relations.

Regards,

Vladimir

Best Regards,
Vladimir Palffy

...or that The graph works nicely but there has to be a way to do the math.

Hi Antonius,

Maybe "Mathcad" community able to help you: http://communities.ptc.com/community/mathcad

I think the Mathcad community considers us an odd bunch and I rarely get replies... go figure

Patriot_1776
22-Sapphire II
(To:TomD.inPDX)

I'd love to help, but I suck at math! I'd be graphing that baby!

I think that is why we are "visual" people... Geometry is so much simpler to "grasp" - literally

Patriot_1776
22-Sapphire II
(To:TomD.inPDX)

Exactly! Same result, just visual!

Dear Antonius

I have tried to create SEMICIRCLE BY SWEEP feature - it is working (you can investigate my model)

Note: I have used for sweep feature the Relation (for UPPER_HALF_OF_THE_SEMICIRCLE) sd4=(r^2 - (trajpar*r)^2) ^ (1/2)

Regards,

Vladimir

Best Regards,
Vladimir Palffy

That's the one! I suspected I would have to divide it into quarters.

Here is the simplified "on the fly" version: sd4=sqrt((5^2 - (trajpar*5)^2))

And as a bonus, it covers ellipses

Thanks Vladimir!

arc_traj_eq.PNG

Gopal, you didn't have to remove your reply. It is definitely worthy of this discussion:

That's the formula of a circle x^2+y^2=radius^2

with value of Y varying from with trajectory parameter.

I was hesitant to reply,i am surprised how did you recovered my reply,i have to be careful now.

...email notifications. Never hold back, my friend

You are Welcome Antonius.

It was challenge for me too - I like Relations and Sweep feature too - it is powerfull combination.

Here is some old example

wallpaper_creo_logo_wire.jpg

Regards,

Vladimir

Best Regards,
Vladimir Palffy

Okay, I finally got exactly what I was after; a full semicircle.

By converting the trajpar to a range from -1 to 1; and making the initial value more than 0, it works:

sd3=5-sqrt((5^2 - (((2*trajpar)-1)*5)^2))

...((2 * trajpar) - 1)... => converts the trajpar to range from -1 to 1 and the Origin for the sweep is "10" (2 * R)

So we start off with "5" - Zero (5^2=25; 2*trajpar-1=2*0-1=-1 *5=-5 ^2=25; 25-25=0 ^2=0; sqrt(0)=0; 5-0=5)

When trajpar=0 (halfway) sd3=0 ; (5^2=25; 2*trajpar-1=2*.5-1=1-1=0; 0*5=0 ^2=0;25-0=25; sqrt(25)=5; 5-5=0)

All the "5"s are the radius value and can therefore be substituted wit relations or parameters.

arc_traj_eq2.PNG

I find it comforting that you can have a line go to zero within the range... but not so much when you cannot define it at the beginning of the range.

Again, thanks all for helping me picture this in my mind. Somehow I was thinking I needed the sin/cos functions to get there. Had I remembered the equation Gopal originally posted, I might have been able to work it backwards from there.

Some things just don't transfer to Creo easily. I went through 4 iterations to make this work, only to go back to the original to have it come out properly.

Since you cannot start out a VSS at zero, I have a tail on the swept line. The long edge is the semicircle we've been discussing. The sweep also has an angle change. Without the angle change, I could easily trim the surface with the initial sweep arc (origin), but once the angle was introduced, it failed the trim near the ends. Okay, merge? Merge failed! Revolve-remove? Failed! Merge flat sheet? Failed!

I went through several other scenarios to make this work including a half section mirror/merge and projecting the edge of a copy. They all created a small non-tangent deformation at the end.

Finally, I ended up extruding-removing a circle (arc didn't work either!) straight up through the sweep sketch. That finally did it.

Give it a try...

cowl.PNG

How can there be fewer mouse clicks when you have to do everything a half dozen times?

Patriot_1776
22-Sapphire II
(To:TomD.inPDX)

Cool stuff! I'll definately have to look at the equations, thanks!

It is the bill of a ball cap?

Part of a cowling...

An interesting thing happened on the way to creating this part. One that forced me to put in a CS request. When I said I had trouble trimming the "tail" from the sweep (to keep a line from starting at 0), it only fails with my config.pro. The CS didn't have an issue with my version and when we opened it with the default config.pro, it was fine. I didn't think I had anything that could cause this kind of issue with geometry integrity so when I find out, I will post it here. But I did confirm that when I imported my config.pro into the "good" session, it went stupid.

I tried a couple of setting but I didn't stumble across the offending setting. I pay PTC to resolve these so I will get back to you. It is however significant to learn this can happened.

Definition of "Stupid":

trim_problem.PNG

Patriot_1776
22-Sapphire II
(To:TomD.inPDX)

I've seen weird things like that in boundary blends near corners. IMO, Boundary blend was better and more robust before WF.

So far we've ruled it down to the display quality. Medium and low are fine and High and above it fails. No SPR yet, but I think one is coming since display quality really shouldn't change geometry. It is highly reproducable on this part.

Patriot_1776
22-Sapphire II
(To:TomD.inPDX)

Hmmmm, so the geometry ISN'T all weird like that? Interesting.....

Does it fail to create the surface on the high quality settings if you actually confirm creation of the sweep? Or will it just create distorded surface?

This is like really really odd. The boundary of that surface shouldn't change at all, while the mesh of the surface could of course with the different display quality.

Jakub Fojtik wrote:

Does it fail to create the surface on the high quality settings if you actually confirm creation of the sweep? Or will it just create distorded surface?

...

When you confirm, it stays that way. You can use measure on it and it shows values at the failed location. I can pick up vertexes. Oddly enough, I cannot save the measure feature.

Jakub Fojtik wrote:

...

This is like really really odd. The boundary of that surface shouldn't change at all, while the mesh of the surface could of course with the different display quality.

That is pretty much what I am saying to CS. I think I just found the perfect storm to cause this, but it explains a lot of the weird things everyone is seeing at times when working with surfaces. It really needs to be looked at carefully by PTC engineers.

I'll post up a quick video sometime today.

That's nasty.

Tom, could you also please show the curvature analysis of the curves in your video? Just so one can see these are not some crazy oscilating splines, but regular curves.

Good idea. I will show that.

Of course it is crazy, however It has two trajpar statements. But the sweep is a simple arc.

trim_problem_section.PNG

trim_problem_fin_analysis.PNG

trim_problem_heatmap_analysis.PNG

I will add the short version...

The lesson here is pay attention to the preview. It is less likely to lie than the geometry. Regardless, this feature should work and internal rounding has really corrupted the edge. Oddly enough, there is a method where this works perfectly, but it is a lot more involved. And it uses exactly the same geometry.

Video Link : 4113

This is the solution video. Basically, I reduced the trajpar to only 1 and used the very same feature with both trajpar's for a -remove material- sweep.

Jakub, thanks for the tip on the analysis. That helped my understanding a lot.

Video Link : 4114

You are welcome, I'm glad you got it working, and with so few features, really interesting approach.

Actually, I never knew portion of a surface can be cut out with another on-laying surface in Creo.

So, just to say again, learn something new everyday.

True Frank. This is an eye-opening moment!

Top Tags