direct e feature based modeling

[Trapy]

a greeting to all, I am going and while being already enrolled for a while at the forum I had not yet had a chance to write anything.
I am at the third year of industrial engineering (mechanical address) and we were introduced to the use of cad 3d parametric (inventor) at the first year.
concluded the semester and given the drawing examination of machines, I put aside inventor and started to use quite intensely solidworks for about a year and a half. I recently started learning and using creo parametric 3.0, which I would like to support solidworks.
So I've been using cad 3d for about a couple of years, it's not exactly like it was yesterday, but it still remains the certainty of having avalanches of things to learn.
I don't know if it can apply as a presentation, or I have to post in a specific topic of presentation in some subforum.


if there were already topic in this regard, I apologize, but I couldn't find them with the search function of the forum.

my question would be this, addressed to more experienced users of me on the topic (not that it takes much, on a similar forum).
I have recently come across this spaceclaim promotional video, which compares direct modeling (is synchronous technology of solid edge?) with that based on the features applied to the model.
[video=youtube;yTzlPjk93Ds]https://www.youtube.com/watch?v=ytzlpjk93ds[/video]In short, from this video it seems almost that the "classic" cads are such because of the technical limitations of the computers on which they were initially conceived and continue to exist only, or almost, for a fact of tradition, when their use represents almost a loss of time and energy compared to using more modern cads that make use of direct modeling more immediate and less problematic.

The many "problems" of the features and history of classical cads are highlighted, but nothing is said about what could be a difficulty of operating only in direct mode.

At the same time, it seems to me that almost no one, minimum percentages, uses these most modern technologies, the most used cads continue to be the usual.


from that pretty drastic video, it seems that the classic cads are so obsolete and destined to disappear in the turn of not much.
I would be curious to have some opinion of more experienced cad users on the subject.

Thank you for any answers :smile:

 

[marcof]

my question would be this, addressed to more experienced users of me on the topic (not that it takes much, on a similar forum).
I have recently come across this spaceclaim promotional video, which compares direct modeling (is synchronous technology of solid edge?) with that based on the features applied to the model.[video=youtube;yTzlPjk93Ds]https://www.youtube.com/watch?v=ytzlpjk93ds[/video]In short, from this video it seems almost that the "classic" cads are such because of the technical limitations of the computers on which they were initially conceived and continue to exist only, or almost, for a fact of tradition, when their use represents almost a loss of time and energy compared to using more modern cads that make use of direct modeling more immediate and less problematic.

here I believe that the only one who uses and really knows in depth is a parametric cad (solidedge) that contextual (cocreate) is hunter, so we wait for his considerations. when I tried cocreate I felt lost, how to travel downhill with broken brakes:smile:
by gaining the video would be to be answered that those of spaceclaim are not able to use a parametric cad, and instead they know perfectly how they work therefore the other answer is that it is an advert with the usual common places then with technical value near zero.
there are advantages and disadvantages in both systems, so much so that both nx and solidedge have introduced direct modeling in their cad and a historical context cad as cocreate has a parametric module to buy aside (for which I think it serves to sell a kidney and probably will be more complex than anyone should be forced to endure). the tendency to mix the two environments something will want to say.

The many "problems" of the features and history of classical cads are highlighted, but nothing is said about what could be a difficulty of operating only in direct mode.

and call them stools!... hardly show you the strong points of the parametric cads that are then the weak points of the contextual cads:smile:

I would be curious to have some opinion of more experienced cad users on the subject.

you have a discreet experience in using parametric cad. download the demo of spaceclaim or cocreate and prove them.

 

[Trapy]

Thank you for the answer.
the fact that it was marketing quite "painted" was suspectable.
I had no idea, however, the fact that the parametric module can be purchased separately for cocreate.

my experience, never having "worked" with a cad is limited in the sense that at the moment I have trained almost only from the point of view of modeling 3d, I have just tried in a few cases to create from scratch some small set in top-down but I do not think I own a real strategy (and experience) of design that to identify agility the difference between a parametric and a contextual in design optical. Surely then I can try to use demos as suggested.

anyway I thank you very much for the information.

 

[cacciatorino]

The question is simple, but the answer is complex!

said in two words: Parametric cad is smarter but harder to use. The context is stupider but easier!

from what is my experience:
advantages of parametric
1) possibility of more simulation of cinematisms and assemblies
2) best association model-table
3) possibility to edit complex functions with a few clicks (for example loft, helicoids, etc etc.).
4) faster editing (if the models are well made in origin)
5) realization of families of parts
6) inter-part relations (imported)

Context benefits:
1) Faster learning
2) increased ease of editing files from other cad
3) simpler top-down modeling (but not associative)
4) possibility to work with great assemblies (but this is probably specific to cads I'm using, more than methodology).

a context seller showed me an interesting slide, in which this concept was expressed: a power-user definitely makes better with the parametric, but as the world is made by normal-men, the aggregate value of the productivity of a ut is perhaps better with a contextual, because it allows the standard designer to reach a higher level than the parametric. It's true he's a seller of a certain product, but I don't feel like I'm wrong.

For example, sometimes I found myself having to change solid edge files so bad that I preferred to remodel them from scratch. the contextual obviously does not suffer from this problem, and if you think of the world of current work, made of contracts to term and interinal work in which the time of training is reduced to zero, perhaps it is better to have a 3d tecnigraph (the context) rather than a very powerful spaceship but which takes a long time before being able to be used profitably.

 

[Trapy]

Thank you very much, the answer was very exhausting.

the power-user speech is what I was asking myself at the bottom of this topic.

I was not sure finally of point 4) of the parametric, but then if one has structured well in the beginning the relationships still manages to manage it well or possibly better than the contextual.

Thanks again.

 

[Er Presidente]

practically gave me some "normal-user" to me?!? ! !

life is not interesting without taking any risks, but you don't have to abuse luck... .

:

 

[ceschi1959]

Sorry if I intervene now but I hadn't seen the post. There are a lot of inaccuracies and common places that create a lot of confusion. The only cad that owns all three environments is nx. I speak of three environments and not of two because it is necessary to distinguish between parametric modeling, direct and 'sincrona' modeling. on the first we all agree, the second consists in shifting and acting parametricly on the so-called b-rep (presented in ug-nx from the night of time), the third (on which not even the programmers of the various cads have clear ideas) consists in acting on b-rep in a direct and without history. The example I do when I train is that, I hope it is clear, given in figure. the speech would be much longer and articulated, of course it is that, at least in the nx field, you are moving towards forms of modeling very different from the previous ones. to explain it is not enough a post.
in the figure you can see:first column parametric modeling, to move the d point you need to move the straight and recalculate the systemSecond column direct modeling, to move the point you need to calculate the system and apply the shiftThird column modeling 'sincrona', do not need calculations to move the point, at a certain time it is located in x1;y1 then in x2;y2

 

[cacciatorino]

I'm sorry to interrupt now, but I didn't see the post. there are a lot of inaccuracies and common places that create a lot of confusion. The only cad that owns all three environments is nx.

thanks for the words of esteem. :smile:

I think that the user has sufficiently understood the difference between the various modeling methods. The question was how this is reflected in the everyday practice of design, and it is to what has been tried to respond.

if I'm not mistaken also solid edge offers the three methods you describe: features, handles with quotas, handles without quotas.

 

[maxopus]

in truth I also create parametric but, the world is done so.

 

[marcof]

Sorry if I intervene now but I hadn't seen the post. There are a lot of inaccuracies and common places that create a lot of confusion.

Since you have identified them, I think you would do what grateful to all if I list all the inaccuracies and common places of the thread on time by writing the appropriate corrections, this to prevent someone reading then you will learn wrong things or you should keep the doubt of what they are wrong and what not by reading what you have written.

first column parametric modeling, to move the d point you need to move the straight and recalculate the systemSecond column direct modeling, to move the point you need to calculate the system and apply the shiftThird column modeling 'sincrona', do not need calculations to move the point, at a certain time it is located in x1;y1 then in x2;y2

I would like to know in how many years you understand what this implies from the practical point of view in using the cad if you are working with reference to the first column, the second or the third.
+1

would it not be more useful to explain when it is better to use one of three types of modeling or a combination of the three rather than to know in which case the cad performs certain calculations?

 

[maxopus]

Dear Marco, the modeling brep instead of feature based is at the end of the fair something that we all "elder" know. is the same mode used by autocad 3d, throw or rotate a face instead of changing the odds of a certain function.
on the one hand this lightens the model because it does not force the cad to keep in mind the logical tree of the work, on the other is a mode that in case of changes can give rise to considerable problems. It is not a better feature based system, otherwise all the global manifacturing colossi would have adopted it for a long time, it is simply very useful for certain contexts. for example, management and model modification with thousands of features: some "strurm und drang"-cad own nun glia do manage heavy models, then they export a step and on that they carry out the work. other example, the third party/supplier does not have your cad and by this method you can import a neutral file and "slip" for the necessary changes. all this however has a limit, forget the power of control of a cad feature based.

 

[maxopus]

at the end of the fair the best thing, in my opinion, is to have both modes. I work in feature based ... if once a year I need the mode brep use. Therefore the road of ptc and siemens under this point of view seems to me drawn, it will not be a question of exile ... is a tool in more work than in certain contexts it comes useful.

 

[ceschi1959]

Thank you for the opportunity given me to try to explain how much I wrote in the hope of being clear.

to hunter: Please don't mention...:smile::smile::smile:

attach two images, the first relative to the direct modeling commands present in nx5 (they are those present more or less always in nx) the other relative to the commands present in nx 9. The latter can be used in history or histoty free fashion (concept introduced by nx6 and which takes the name of synchronous modeling) . As you see, they are not simple 'handle' commands and I don't think you find them all in itself or in creo.
some of these commands are usable in various ways, such as the delete face can be used with the heal option activated or not, i.e. can reconstruct or not what is below. each command would require explanation pages. However, we come to the basic question, i.e. when to use some options and when they may be inopportune, taking into account that it is not irrevocable choices but that can be reviewed during the execution of the project or the particular. nx has this of beautiful and that is that you can walk different paths when and how you want.

direct modeling

first example banalissimo (real and direct personal experience): a door of a molded cast iron stove (from another designer) with about 300 features. you must move a rib connected to a sketch that commands the whole door; first option to understand that caspita is connected to that sketch and try to repair everything; second option apply a direct modeling command and have the change immediately and parameterically.
second example, even this banal; in the execution of a series of fittings you are forced to make them in a certain sequence with the result you see in the first figure. applying the reorder blend command you get the correct effect without going to reorder features with an uncertain outcome.
examples of this type I can do at dozens and knowledge of certain commands leads to shape a little differently.

modeling 'free' or contextual or synchronous or as cabbage you want to call, also this direct experience:

you must inspect the 'hairs' made with other cad to obtain a solid to be made by laser sintering to obtain a support on which to lay the carbon fiber to realize the famine of an Italian racing bike with which it ran vr years ago. It was also necessary to carry out the run-offs of these skins (cases at about 90° length of about 20 mm on the outer perimeter). It was a job where parameterization was even in the way.

from these few examples I think we can say that the various types of modeling can be used in an integrated way.

I ask you to look at the footage on the tube at this link. https://www.youtube.com/watch?v=5q2iijyfy7e and tell me what you understand about synchronous siemens and in particular to hunter to compare it to concepts in itself.

I'm sorry if I'm not ready in the answers, but I'd have more to do on Sunday. as I believe and I hope you too.

kisses

 

[ceschi1959]

Dear Marco, the modeling brep instead of feature based is at the end of the fair something that we all "elder" know. is the same mode used by autocad 3d, throw or rotate a face instead of changing the odds of a certain function.
on the one hand this lightens the model because it does not force the cad to keep in mind the logical tree of the work, on the other is a mode that in case of changes can give rise to considerable problems.

This is one of the inaccuracies that turn, because if it is related to the second mode I have described, it even weighs the model because it compels to add calculations to the calculations. and that's what I want to emphasize in my post: it makes confusion between commands and modes. it is absolutely not true that siemens and creo have marked the way: from what I find siemens has two different paths for nx and if, for the latter has changed from st1 to st7 (ask confirms to hunter) manifesting lack of objectives.

 

[maxopus]

This is one of the inaccuracies that turn, because if it is related to the second mode I have described, it even weighs the model because it compels to add calculations to the calculations. and that's what I want to emphasize in my post: it makes confusion between commands and modes. it is absolutely not true that siemens and creo have marked the way: from what I find siemens has two different paths for nx and if, for the latter has changed from st1 to st7 (ask confirms to hunter) manifesting lack of objectives.

I am unconvinced baskets, although my (not only my) impact with nx was really disappointing. of all the phantasmic possibilities to achieve a result... 90% cylindrical, to understand how a normal radial series works there is to make the sign of the cross on the contrary with two crowns of garlic hanging on the neck.
in the world ptc a model without feature weighs less ... you tell me that in the world siemens weighs more ... there is something that does not come back.

 

[cacciatorino]

I am unconvinced baskets, although my (not only my) impact with nx was really disappointing. of all the phantasmic possibilities to achieve a result... 90% cylindrical, to understand how a normal radial series works there is to make the sign of the cross on the contrary with two crowns of garlic hanging on the neck.
in the world ptc a model without feature weighs less ... you tell me that in the world siemens weighs more ... there is something that does not come back.

I believe that the meaning of the statement is this: When I create a face, he actually adds a "face post" feature in the logical tree, which you don't see why he's hiding it. if you reopen the same model with create parametric instead you see your beautiful "face seat" in the model tree. I remember seeing this thing about the first creed, confirming it's still like that? If this is true, actually every operation of this kind is actually another feature: solid edge st (and all others working in contextual mode) instead do not hold memory of operations made with handles.

 

[ceschi1959]

I am unconvinced baskets, although my (not only my) impact with nx was really disappointing. of all the phantasmic possibilities to achieve a result... 90% cylindrical, to understand how a normal radial series works there is to make the sign of the cross on the contrary with two crowns of garlic hanging on the neck.
in the world ptc a model without feature weighs less ... you tell me that in the world siemens weighs more ... there is something that does not come back.

I see that it continues to not understand and this still brings me back to what has been said: There is great confusion. I did not say that a featureless model weighs less I said that direct modeling commands weigh the file and exactly these words: this is one of the inaccuracies that turn, because if it is related to the second mode that I have described, even weighting. This prompts me to think that nothing has been understood about direct modeling and modeling history free by confusing each other.
look at these two movies that I did on the fly, so excuse the banality, in one, you start from a file shot and you get through the modeling commands to create parameters. do you not answer the questions I have asked and that is, there is a vague likeness of commands with those I have illustrated of nx?https://www.youtube.com/watch?v=v8a6_zluyzuhttps://www.youtube.com/watch?v=j42-yjin2uu

 

[cacciatorino]

I ask you to look at the footage on the tube at this link. https://www.youtube.com/watch?v=5q2iijyfy7e and tell me what you understand about synchronous siemens and in particular to hunter to compare it to concepts in itself.

I'm sorry.

Sincerely it seems to me the same thing I have in solid edge, and in cocreate modeling. that I can take and move faces and/or add sena jobs having to follow an unexisting chronological tree of the work. Then the st of the siemens has nice that offers the possibility, but not the obligation to bind each other the faces, which does not exist in cocreate if not with specific advanced module.

However to answer your question I think that what you see in the video is equal in solid edge (I don't refer to the free form because it's not relevant to the discussion, and it's not part of my professional field so I don't know if nx succeeds if not).

 

[ceschi1959]

We're writing on each other....:smile:

 

[ceschi1959]

Sincerely it seems to me the same thing I have in solid edge, and in cocreate modeling. that I can take and move faces and/or add sena jobs having to follow an unexisting chronological tree of the work. Then the st of the siemens has nice that offers the possibility, but not the obligation to bind each other the faces, which does not exist in cocreate if not with specific advanced module.

However to answer your question I think that what you see in the video is equal in solid edge (I don't refer to the free form because it's not relevant to the discussion, and it's not part of my professional field so I don't know if nx succeeds if not).

My question was a trick question. in that video there is nothing st because it is all parametric, and here is confusion: in itself you could have done these things, as in nx, well before the introduction of the st?