"Combined rail curves exceed Max Surf Spans tolerance"

[davide-ita]

Hi all,

this is my first post on the forum.......i've done a surface between two curve on surface using b-rail;  i've created 2 generation lines ( degree 4 and 5 cv), and when i've used the command b-rail, alias don't make me the surface... in the prompt line there was "Combined rail curves exceed Max Surf Spans tolerance"  ...

i've seen in the constructions options that the value is for default (100), but if in the options on the b-rail, click rebuild at the rails (the curves on surfaces), alias make the surface but whit irregular span.....

my question is :

- the max surf spans value, is in relation at the knots that the curve on surface's have??

- i've read on internet this "

COS is always 3rd degree and adds spans to achieve the shape. when ever you build a rail or square, you have to turn on rebuild for atleast one of the edges if not both.".......this definition it's true?

-is there a method for don't exceed the max surf? for example i need to rebuild the original curve that i project on surface?

sorry for my english  🙂

[ravenzep]

Hi,

As you can imagine, a surface with 100 spans, is far from being acceptable. And alias is not like rhino, regarding these spans stuff. Basically, alias doesnt like spans, which makes sense for what it was build for, because spans are bad news. You should gett used to work in explicit mode, so that you have total control over what you are doing, and yes, it takes skill and a lot of knowledge to work this way.

You are  getting these many spans, because you are building surface from curves on surface. That is a bad practice, generally. You should have curves with good geometry (no spans) when you build surfaces from them.

Same applies to trimmed edges, they are bad geometry too.

I will give you a hint:

Delete those COS, and re-project from main curves, but this time, dont project as COS, but rather project as CURVES. It s the other option. Check the deviation and do it. It probably will have some deviation, which is normal. Build again the surfaces from those good curves, and then you align the surface edges to your need.

This would be a good example of a good workflow. I bet you got a lot better surfaces, right?

Garbage IN garbage OUT.....Your input geometry should have quality, or everything will get complicated downstream.

Good luck

[davide-ita]

Spoiler

hi ravenzep; thank you for reply 🙂

i've try to project like curves, and i've seen that the result is better....the problem is that i can't cut the surface whit the curves projected in that mode;  so i've try to use  duplicate curve,  and  i've try to rebuild the Cos, and after i've used b-rail, and i've worked on explicit control and in the same time i was looking the hull for to see the distribution on the cv... the result it was a nice surface whit 1 span and a regular distribuction of the cv; the problem after it was to have the tangency whit the surfaces cutted..

questions :

- this tipe of workflow that i've done can be correct?

- i've learn that in alias is important for the continuity of the surface, the position of the cv's.....my question is... for to have continuity between 2 surfaces, they must to have  the same degree?? can 2 surface have continuity whit different degree??  maybe is a stupid question, but in the tutorials and guide that i've found, i've never see the answers at this question.... for me need the same degree in the direction that i search continuity between 2 surface, but i'm not sure ..so i ask

thank you so much 

have a good day

[ravenzep]

Hi there.

You seem to have understood.... And before u trim, in the end, those surfaces, they should be G2, and a good G2, because if things are not done well, even with G2 you can have surface artifacts, quite visible is a render, for example, let alone in manufacturing.

By the way, set your modelling tolerances to CATIA V5.

Regarding your second question, CV layout (distribution) is the most important thing really. It s the great advantage of such free surface modellers, as opposed to parametric modellers.

WHat you ar used to hearing and seeing, is that people  use degree 5 surfaces, when they need G2 on both sides, because such surface will have the minimum necessary number of CVs fo rit being possible.

In theory, you can even have a G3 alignment with degree 2 surfaces, because G depends on shape, not on number of CVs..... But never mind this last bit i said, it s not important now.

It s a very extensive subject, and is hard really to explain things thorougly here.

Read the manual, try to understand things, and you ll gradually start understanding why things are done the way they are. I m sure you will get it with some work.

I will send you a little PM. Check it out.

Good luck