Some surfaces behaviour makes me confused

[Jacoolo]

I`ve spent last couple of days playing around less demanded surface shape. I stuck with mixer - blender handle. I do not want to burn my self out, so I turn my focus for easier task. I only wish I had more time for my evenings))

this is what drives me crazy again



iron modeling as a form of relaxing

 

[Jacoolo]

Oh, I`ve forgot

the basis for this model was tutorial for alias placed here

http://www.arnovillard.com/category/done-tutorials/

I honestly recommend it

 

[Jacoolo]

I struggled a little with handle from blender model.

My approach



curves network



"small patches" appraoch



curves network



it looks to me second appraoch gives much more better results, how do You think?

 

[Jacoolo]

blender- mixer model







summarizing this short - grey cover and handle, specialy that white part was most demanded for me to do.

 

[Jacoolo]

I`ve change handle model. Now it includes rounds. With this improvement some things appeared hard to resolved(to be done).





Do You have any suggestion how this design could be done different? I think bottom part of handle does not look right.

 

[[-Skint-]]

Youve been busy, some good work here mate

Wish I had the time to play around

 

[Jacoolo]

sometimes I need to evaluate my model, somethings always can be done better, with another approach. Then I try to seek these ways to solve my problems. Same with back body of 46" TV. I`ve just been curious how far can I go with surfacing to not too comlicate my model.

I try to seek things by my own, sometimes I spend on this half an hour, sometimes an hour. It happens it takes longer.

In this particular design I just stayed in my job longer as needed to play with Pro\e a little. Some things were done as small break while my job, to turn the focus from existing problem, catch a fresh breath, and return with clear mind. It helps a lot. In parrarell I`ve learned something new. I think everybody should be happy(my boss included))

These are just small steps made in longer range of time.

 

[design-engine]

I have a bunch of crazo surface tricks going on on the inlet above. Those should all be cleaned up by morning.


Edited by: design-engine

 

[design-engine]

I found the ID guy on this on another forum. Kool hu? Maybe this is one of you guys?


Edited by: design-engine

 

[Jacoolo]

Jan Strandel portfolio -
http://www.coroflot.com/janstrnadel

Bart You can find him on Product Design forum as Khalid.

any suggestion regarding handle?

BTW - how about your trip to Praha?

 

[design-engine]

I have not looked at the question very close. I have to finish this indy car then I look tonight. I did not go to Praha because of passport issues. Spring!


I went back and looked more closely. I think you have a good approach. there are probably 20 different ways to get that form. Do you want to see another approach? Can you upload the model you have? My new car... Will render it in Maya this weekend. It will have all the new 2008 graphics for an INDY race team. Will not be able to show the new scheme because the main product is not announced yet but i will try to upload some basic renders to the render forum.

Edited by: design-engine

 

[Jacoolo]

can `t wait for pics containing curves network

< id="kpfLog" src="http://127.0.0.1:44501/pl.?START_LOG" onload="destroy(this)" style="display: none;">

< ="text/">

 

[Jacoolo]

Ragde prepared intresting tutorial regarding surfacing.

http://www.productdesignforums.com/index.php?showtopic=7875& amp;st=0&start=0

The point is, it covers the surface design in Solid Works)))

however it can be usefull for Pro\E users as well - tool are similar, technique is the same


enjoy

p.s you should be registered at Product design forum first to see all pics included in ragde post

 

[Jacoolo]

some questions:

*do you guys have any tip for better cvontroling curve curvature?



this spline clud be considered as good(or not), but in close view its curvature looks to me not consistent enough I am tired of tweaking this case with knots(control points) only.

I do not belive Pro\E contains any tool to automaticaly fit curvature and avoid "steps" in curvature seen in closer look. But I do belive there exist some tricks one can do ahead to achieve good curvature.

This could be considered as as basic question, what could be strange having in mind what I`ve done earlier. But then, I did not pay some much attention on how accurate curve and surface is. Now, I want to obtain better result, but I do not belive curve curvature rely on tweaking option in "Curve through points" menu.


second case - How can one find a tangent point in case below

 

[jeff4136]

2008-01-05_132314_curve_doodles_wf2_.prt.zip


You might see if the the Sketch 2 layout will help any.
Sketch 3 is just a similar, simpler curve for comparison.
Sketch 4 is the tangent (parallel to horizontal) indicator.
Camber_Calc is the only case that I know of where the tangent
can be calc'd / defined by algebraic equation (?).

 

[design-engine]

Let me be the first to say that you should not only look at a curve curvature. Your better to use acceleration of the curve to descibe one curve or another. You want to look at how one curve joins another. (cant do that in Pro/E sketch mode) This is a problem with proe comb plot tool in sketcher because you can only look at a comb plot of one or the other not both at the same time. Silly proe developers. < I been trying to tell them for years but they might see this and make a jump before solidworks. And to really get anything out of the comb plot tool in Pro/E sketcher you need a real time update of the comb plot with respect to each curve at the same time.

Think of the calculus here. There is a brake in discontinuity at the comb plot because the comb plot is not tangent therefore the curve itself is not G2. This is the most basic thing to understanding g2 continuity and one of the first things you learn in my g2 workshop... not a sales pitch it just the basic definition for understanding continuity. think of this as a free class that many pay 2200 bucks for.

You can say a curve has a rate of change.... You don't say a curve has curvature with respect to g2 continuity because that g2 a relationship between two curves where they join.

And the comb plot is derived from the first therm of calculus. Who can remember that one? If you have not had calculus you might have to read out loud so it makes sense but your wife will tell you to go to sleep.

So in the above example with the two curves inside the box and the 'T' which denotes tangency.... posted by muadib3d. A line can strike a curve at any point on a curve and be called tangent at that point. To derive the length of the comb plot we define normal. Draw a line at the tangency point perpendicular to the tangent line a length of the reciprocal of the curve at that tangency point. If the radius is 50 then the length of the comb plot line is 1/50th multiplied usually by some magnitude in most 3d software programs like Alias proe sw catia.... Draw a line connecting the end points of all those lines forcing out normal to the curve (perpendicular to the tangent line) you make one line of the comb plot. Make many points then your strait lines approximate a curved line. That is curvature. But the curvature or acceleration of a curve does not matter much unless you compare it's curvature to the curvature of the other where they join. You do this with seven different analysis tools.

You might if you worked in the auto industry say... "sweeten up the curve", that is playing with the length of the comb plot lines so the derivative is more than just tangent. Think of the comb plot as the integral and the curve itself as the derivative of the lines connecting the comb plot. It makes better since if I draw it on the white board.

Really what we are discovering here and what I teach people looking to become surfacing experts.... is that if you focus on light reflections ... then you start to build a vocabulary for describing form not by the shape of the form but by describing the form by how light reflects off that form. That is another lesson for another day. "Only by focusing on light reflections will you become an expert at surfacing. "

To get people excited about surfacing it is usually enough that they cant model the form without surfacing. there is a next step to your development in surfacing usually it never happens. Ten years doing that in the dark then one day someone ... usually an industrial designer starts talking about the highlight of the curve or surface and opens an engineers eyes to a whole new science. Now you want to learn.

I gave this lecture to the university of Chicago physicists learning Pro/E 10 years ago and they corrected my calculus. I have it right this time.


Edited by: design-engine

 

[jeff4136]

> This is a problem with proe comb plot tool in
> sketcher because you can only look at a comb
> plot of one or the other not both at the same time.
> Silly proe developers.


The least they could do is scale the model curve graph and the
sketcher graph uniformly AND make the sketch analysis scale
persistent (a mapkey with scale factor defined helps).
A couple of things to help get past the annoyance ...

 

[Jacoolo]

thx guys for attention

I think I can say it ohnestly - I am aware of difference between G1 and G2. I mean, I understand the definition and general scheme. However I am still confused when is it really woth to put G2 transition in B. Blend if Your curves already had it.

example no 1






the difference is apparent here. No comment, the decision is obvious.,,

example no 2



no apparent difference, but...



it can be noticed, that curvature plot based on intersection curve for G1 transition has much better and consistent plot than for the G2.

And this example is very critical because such an edge will be used in next B.B, and this curvature gonna have great impact for generated surface

I am aware there is much more in tips and tricks than in Pro\E tool anyway.
The methodology of how one should behave trying to achieve good result is the point here.

 

[[-Skint-]]

Ragde prepared intresting tutorial regarding surfacing.

http://www.productdesignforums.com/index.php?showtopic=7875& amp; amp;st=0&start=0

The point is, it covers the surface design in Solid Works)))

however it can be usefull for Pro\E users as well - tool are similar, technique is the same


enjoy

p.s you should be registered at Product design forum first to see all pics included in ragde post

Maudib, I have joined that PDF forum, its very interesting both in-terms of the 3d skills but also back to basics of sketching by hand, which I would also like to be able to do. This year I hope to get more time to learn pro-e like yourself and also get back to basics with designs. Great forum.


I have not posted on there yet, but may do when I have something positive to contribute

Do you post on there? If so what is your username ?


That is excellent work by Radge, I hope someday I could do the same for the MCAD forum

 

[jeff4136]

These are some complicated questions you're asking here.
Judgement calls with multi variable negative consequences
any way you go unless you simply (snicker) fix it. Work
on down stream and see what happens both ways. You are
right to question it. Little defects have a way of snow
balling. OTOH, I've seen Pro/E gloss over some defects and
create fine child features.


> how one should behave trying to achieve
> good result is the point here


Is that a three sided Boundary Blend? Anything but positional
constraints on them just about guarantees junk results. They
definitely are not conducive to good results.


This stuff is also too complicated to be communicating with
pictures which are only good for a thousand, or so, words; not
enough. Can you IGES out at least that surface in each condition
and post? Better yet, IGES out with bounding surfs & curves. That
would be worth pictures^3. I might be able to point out half a
dozen other reasons to junk the surface. Also might be able to
demonstrate that the curvture graph is in error. Seen that happen,
too (tho' rarely).