Topic: Edit solid models

[SDETERS]

I have not Used Pro_e.  I have nothing but hear say on this software.  Actually to say it correctly Wildfire.   

What I meant by crossestions is you have to tell the software where the section is needed and then it will drop the section in.  In I-Deas it creates the cross section 90% of the time.

When you make a change on the 3D model poof the geometry changes and any 2D dimension lines that are attached to that will also move.  (In I-Deas)

[CADaver]

What I meant by crossestions is you have to tell the software where the section is needed and then it will drop the section in.  In I-Deas it creates the cross section 90% of the time.

In AutoCAD, I tell it where to cut the slice and place it in the desired viewport ALL the time.

When you make a change on the 3D model poof the geometry changes and any 2D dimension lines that are attached to that will also move.  (In I-Deas)

In AutoCAD with DIMASSOC=2 the same is true.

[SDETERS]

OOpps got a little of the subject here sorry

Autocad is a 3D option.  It is very hard for someone like me to see Autocad and 3D in the same sentence. 

Learn Pro-E Learn 3D autocad it will make your designs easier in the long run.  Just do not be afraid to try new software.  Always give it a chance and it might suprise ya.

Thanks

PS

 me mouse pad rocks

[Serge J. Gianolla]

Get this you make a change on the 3D model poof all of your views upate to that mod.

Same is true for AutoCAD.

Same is true for AutoCAD
? ? ?? Never saw that in AutoCAD, when a user is changing the 3D model and the 2D views generated via SOLVIEW, SOLDRAW... are updating automatically!

What's the trick CADaver cause that'd save me time and I'd be happy to learn it 

[Jim Yadon]

Anybody know efficient ways to:
*select a set of features, and translate them.
*select a set of edges and chamfer or fillet them.  
*scale a model along one or two axes only.
Currently, know how to extrude faces just fine, also slice, section & boolean.  Appreciate input.

I'm not sure what you mean by translate.
I've never had a problem using the fillet or chamfer command with solids in any version of ACAD 2k4 and above.
Scaling along independant axis... hmmm. Sounds like you need to look into 2k7. You have to see some of the features that are there to believe them. Picture using one solid to subtract from another. Oops, you placed the solid you subtracted a little bit off? Well just select feature that was created by the subtraction and move it! Seriously, you gotta see it to believe it. It sounds too good to be true but it works. You can also align faces of objects by selecting the face to align with and moving the object accross the faces to choose what you align to. For those of us using solid modeling on a regular basis, 2k7 has sevral tools that make our jobs a dream. Again, the only feature that still needs help is dimensioning. Maybe in 2k8?

[CADaver]

Get this you make a change on the 3D model poof all of your views upate to that mod.

Same is true for AutoCAD.

Same is true for AutoCAD
? ? ?? Never saw that in AutoCAD, when a user is changing the 3D model and the 2D views generated via SOLVIEW, SOLDRAW... are updating automatically!

What's the trick CADaver cause that'd save me time and I'd be happy to learn it 

Simple, just avoid using the idiotic SOLxxxx commands.  They were obsolete after R12 and removed from R13, but because of users who were incapable of understanding DVIEW CLIP, they were adding back to the program with R13C4.  The MUCH better option is to use DVIEW CLIPping to produce the views required so that as the model is modified the views of the model update.

[Serge J. Gianolla]

Fair enough about DVIEW.
But a drawing - normally - is to tell the builders or makers how to construct the thing(s). There are some drafting conventions to respect! How does the CLIP command behave when you have different lineweights to show or different linetypes for what is hidden for 3rd angle projection say? I mean maybe in your field or state you can get away without drafting standards or QA?
What about dimensioning complex objects never in proper UCS... Trying to understand how it is viable, for Architects for instance where the thickness of a line is related to what they want to stand out, the more prominent features?
Can you clip a section not linearly with DVIEW, as opposed to having a "kink" with the SOLxxx?
What about adding hatching? Isn't it gonna appear on all other views because it is on the model, or you have to have a stack of layers to manipulate?

[CADaver]

Lineweights are contolled in AutoCAD like they've always been; CTB/STB. 

For hidden 3D elements use OBSCUREDLTYPE and OBSCUREDCOLOR. 

If you "really" "need" a kinked section, use multiple aligned viewports.  That way your "section" can bend in any direction desired.

Hatching is annotation, and as such resides in PS, along with dims and notes.

BTW, just because my drafting standards are "different" doesn't mean I "can get away without drafting standards or QA".  Many drafting standards and conventions were created to communicate a 3D concept on an orthographic 2D drawing.  The 3D model can now be easily portrayed and displayed on a drawing from any viewing angle needed for clarity, thus alleviating the need for some of the drawing conventions we come to know and love over the decades.

[Chuck Gabriel]

For hidden 3D elements use OBSCUREDLTYPE and OBSCUREDCOLOR. 

At the risk of revealing my ignorance, please enlighten us as to what these are.  I vaguely remember reading about some undocumented method for causing hidden edges to show in a different linetype/color, but that was a loooong time ago.

[edit]
Oops.  Never mind.  I see now that those are system variables that were added in R2002.  Must remember to Google before asking.
[/edit]

[CADaver]

you can also explore the use of HALOGAP

[Serge J. Gianolla]

OK CADaver, tried "your" method, and I have to disagree with some of your remarks [I know, who would have thought! ]
When using OBSCUREDLTYPE it helps only to show an object in dashed in one view while continuous in another, but the continuous should be 0.5mm and the dashed one 0.35mm [well I am metric].

Lineweights are contolled in AutoCAD like they've always been; CTB/STB.

"always been" Careful with this type of remark, I do not recall version 9 having CTB/STB!!
Because it is convenient to you, do not assume that it is the only way, I and countless of others like to see the lineweight as we draw, that's why there is a toggle LWT, so we do not have to go to print preview.
The tessalations lines are still showing even when tinkering with DISPSILH! Unsightly, cluttering the dwg, not proper std.
Even if using the MVSETUP, it takes some mucking around on kinky section planes, let alone curved ones. Yep, a customer in the past asked for a clipped area to be curved!
You have a 3D model in ModelSpace, and you have to show the cut section with hatching; so you told me this goes in PaperSpace, pray tell me how you define the outline you want to fill in being in a different environment.

[CADaver]

OK CADaver, tried "your" method, and I have to disagree with some of your remarks [I know, who would have thought! ]
When using OBSCUREDLTYPE it helps only to show an object in dashed in one view while continuous in another, but the continuous should be 0.5mm and the dashed one 0.35mm [well I am metric].

the viewports must be set to hideplot or they won't hide, and therefore won't show the hidden ltype and weight.

Lineweights are contolled in AutoCAD like they've always been; CTB/STB.

"always been" Careful with this type of remark, I do not recall version 9 having CTB/STB!!

very true, I just added the CTB/STB comment for clarification sorry if it confused you.

Because it is convenient to you, do not assume that it is the only way, I and countless of others like to see the lineweight as we draw, that's why there is a toggle LWT, so we do not have to go to print preview.

being very comfortable with the software and what it produces, I neither need to see the lineweights nor preview plots.  But I don't see my method as being a deterant to either.

The tessalations lines are still showing even when tinkering with DISPSILH! Unsightly, cluttering the dwg, not proper std.

Do you mean while you're drawing, or when plotted.  What you're seeing while drawing is probably a function of SHADEMODE that always shows the tessellations lines.  DISPSILH will produce a clean plot, regardless of the display of SHADEMADE.

Even if using the MVSETUP, it takes some mucking around on kinky section planes, let alone curved ones. Yep, a customer in the past asked for a clipped area to be curved!

It only takes mucking around because you're not used to the interface.  Try this for quick DVIEW slicing (I'm sure the lisp gurus around here can make this a lot slicker).

Code: [Select]

;;-------------------------------------------------------
;; DVIEW Clipping aid
;; original from Falcon Design Services with a lot of help from Mark Thomas
;; modifications and wierd little viewport change function by Ruul Morawetz
;; Enter DVP to execute
;;-------------------------------------------------------
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
(vl-load-com)
(vl-load-reactors)
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
(defun c:dvpmk ( / *error*)
   (defun *error* (msg)
      (cond
         (msg (command)(command)(princ (strcat "\nError: " msg)))
         (T nil)
      )
      (dvt:restore)
      (if msg (command "_u"))
   )
   (dvt:save)
   (dvp:sectiontop)
   (*error* nil)
      (prin1)
)
;------------------------------------------------------------------
(defun dvp:sectiontop ();/ ptfrt pt2 ang Ndvpdeep)
   (setvar "EXPERT" 5)
   (setvar "CMDECHO" 0)
   (setvar "orthomode" 1)
   (setq curvp (getvar "cvport"))
   (command "_ucs" "_s" "__TMP")
   (command "_ucs" "_v")
   (cond
      ((null (setq ptfrt (getpoint "\nSelect Front Clipping Plane of Section: "))))
      ((null (setq ptbck (getpoint "\n Select Back Clipping Plane of Section: " ptfrt))))
      ((null (setq ang (angle ptbck ptfrt))))
      ((null (setq dist (distance ptbck ptfrt))))
       (T
         (setq ptfrt (trans ptfrt 1 0) ptbck (trans ptbck 1 0));;;;;;; pt2 (trans pt2 1 0))
         (prompt "\n Select Viewport for Section Display (ENTER if necessary then) ")
         (dvt:rm:cvpc)     ; vp-change function (possibly unreliable)
         (setvar "WORLDVIEW" 1)
         (setvar "UCSFOLLOW" 0)
         (command "_ucs" "_w") ; moved here (only necessary after viewport change)
         (command "_.dview" "" "_po" "_non" ptbck "_non" ptfrt "_cl" "_f" dist "_cl" "_b" "0" "")
         (command "_.zoom" "_e" "_.zoom" ".8x")
         (command "_ucs" "_r" "__TMP" "_ucs" "_d" "__TMP")
      )
   )
)
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;------------------------------------------------------------------
(defun dvt:rm:cvpc ( / grrd weida p pold viewsize trigger)
   (setq weida T viewsize (getvar "VIEWSIZE")
         trigger (* viewsize 0.4)) ; adjust factor according to mouse speed etc
   (while weida
      (setq grrd (grread T )) ;(+ 1 2 4 8) 2
      (cond
         ((= 5 (car grrd))
            (setq p (cadr grrd))
            (cond
               ((and pold (> (distance p pold) trigger))
                  (setq weida nil)
               )
            )
            (setq pold p)
         )
         ((= 2 (car grrd))(setq weida nil))
         (T nil
         )
      )
   )
)
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
   (defun dvt:save ( / n)
      (setq *dvt* nil)
      (command "_undo" "_begin")
       (osmget)(setvar "osmode" 0)
      (foreach n '("CMDECHO" "NOMUTT" "CMDECHO" "WORLDVIEW" "UCSFOLLOW" "EXPERT")
         (setq *dvt* (cons (cons n (getvar n)) *dvt*))
      )
   )
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
   (defun dvt:restore ()
      (foreach n *dvt*
         (setvar (car n) (cdr n))
      )
       (osmret)
      (command "_undo" "_end")
   )
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
(defun c:dvp ()
(c:dvpmk)
(setvar "cvport" curvp)
)
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
(defun c:VPSwap ()
(command ".undo" "BEGIN")
(if
  (= (getvar "cvport") 1)
   (command "_mspace")
)
(command "_view" "_d" "vpswap*")            ;;;;Delete any existing vport views

(prompt "\nSelect a DIFFERENT viewport")    ;;;;Get the first viewport
(dvt:rm:cvpc)                               ;;;;Viewport select tool;; subroutine below
(setq vprt01 (getvar "cvport"))             ;;;;Get viewport setvar
(command "_view" "_S" "vpswap01")           ;;;;Save the view of that viewport

(prompt "\nSelect another viewport")        ;;;;Get the second viewport
(dvt:rm:cvpc)                               ;;;;Viewport select tool;; subroutine below
(setq vprt02 (getvar "cvport"))             ;;;;Get viewport setvar
(command "_view" "_S" "vpswap02")           ;;;;Save the view of that viewport

(setvar "cvport" vprt01)                    ;;;;Set first viewport
(command "_view" "r" "vpswap02")            ;;;;Restore Second View

(setvar "cvport" vprt02)                    ;;;;Set second viewport
(command "_view" "r" "vpswap01")            ;;;;Restore first View

(command "_view" "_d" "vpswap*")            ;;;;Delete any existing vport views
(IF (= nil id:lispold) (setq id:lispold "x")(setq id:lispold id:lisp))
(setq id:lisp "vpswap")
(c:counter)
(command ".undo" "END")
(princ)
)
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;


You have a 3D model in ModelSpace, and you have to show the cut section with hatching; so you told me this goes in PaperSpace, pray tell me how you define the outline you want to fill in being in a different environment.

First, we only hatch what's necessary for clarity and not according to some arbitrary standard.  But that aside, there are several ways to place hatching besides selecting a zone to hatch. True most are a little more tedious, but that modicum of tedium is more than outweighed by the advantages of viewing the real model instead of a facsimile of the model produced by the tedium of the SOLxxxx commands.

The method we use most is direct hatching.  Of course we have several lisp functions to place the desired hatch, but below is the base AutoCAD Command sequence.

Code: [Select]

Command: HATCH

Enter a pattern name or [?/Solid/User defined] <ANSI31>:

Specify a scale for the pattern <1.0000>:1.5

Specify an angle for the pattern <0.00000000>: 45
Select objects to define hatch boundary or <direct hatch>, <-Press ENTER Here

Select objects:

Retain polyline boundary? [Yes/No] <N>:N

Specify start point:  <-Pick Point
Specify next point or [Arc/Length/Undo]:  <-Pick Point
Specify next point or [Arc/Close/Length/Undo]:  <-Pick Point
Specify next point or [Arc/Close/Length/Undo]: C  <-“C” for close
Specify start point for new boundary or <apply hatch>:<-Press ENTER


You disagree after attempting the procedure for only a few minutes.  I'm reasonably sure you gave AutoCAD a lot longer than that to get accustomed to it, otherwise you wouldn't be using it now.  Understand, I did it your way for several years, always fiddling with the SOLxxxx commands. Something moved?, either re-run the SOLxxxx commands or fix both the model and the drawing. Something changed?, either re-run the SOLxxxx commands or fix both the model and the drawing. Something deleted?, either re-run the SOLxxxx commands or fix both the model and the drawing. Something added?, either re-run the SOLxxxx commands or fix both the model and the drawing.  Now that was true "mucking around".  R14 gave use OBSCREDLTYYPE and OBSCUREDCOLOR, R2000 gave us multiple layout tabs, and the little DVP routine above filled in the rest.  We are now considerably faster that we were using the SOLxxxx commands, and the drawings are not divorced from the actual model.

BTW, it seems the R2K7 has added a feature very much like the DVP functions above

[Serge J. Gianolla]

Hi CADaver,

very true, I just added the CTB/STB comment for clarification sorry if it confused you.

It did not confuse me for been using AutoCAD for a long time, but had the potential to confuse others!

being very comfortable with the software and what it produces, I neither need to see the lineweights nor preview plots.  But I don't see my method as being a deterant to either.

It does not matter if one is being comfortable, cause I do not draw just for myself. Sometimes customers want to see what they pay for on the screen.

Do you mean while you're drawing, or when plotted.  What you're seeing while drawing is probably a function of SHADEMODE that always shows the tessellations lines.  DISPSILH will produce a clean plot, regardless of the display of SHADEMADE.

Yeah, meant while drawing. Well, you have to be in HIDDEN mode to have the OBSCURELTYPE working, so the tessalations will show up and on complex design clutter dwg unnecessarily.

It only takes mucking around because you're not used to the interface.  Try this for quick DVIEW slicing (I'm sure the lisp gurus around here can make this a lot slicker).

When we were comparing the benefits of one vs other, I was under the impression that we were comparing oranges! I agree with you that the potential of AutoCAD is not so much what it does, but what it allows one to do. But in this case, you were considering add-ons to AutoCAD [whether in-house development]. Not fair.
I am sure that when you were comparing AutoCAD 3D capabilities against the other software, other users could have said OK mine does not do that out-the-box, but I can make it do!

First, we only hatch what's necessary for clarity and not according to some arbitrary standard.  But that aside, there are several ways to place hatching besides selecting a zone to hatch. True most are a little more tedious, but that modicum of tedium is more than outweighed by the advantages of viewing the real model instead of a facsimile of the model produced by the tedium of the SOLxxxx commands.

The method we use most is direct hatching.  Of course we have several lisp functions to place the desired hatch, but below is the base AutoCAD Command sequence.

Right. Picking point is only gonna work on curveless objects.

You disagree after attempting the procedure for only a few minutes.  I'm reasonably sure you gave AutoCAD a lot longer than that to get accustomed to it, otherwise you wouldn't be using it now.  Understand, I did it your way for several years, always fiddling with the SOLxxxx commands. Something moved?, either re-run the SOLxxxx commands or fix both the model and the drawing. Something changed?, either re-run the SOLxxxx commands or fix both the model and the drawing. Something deleted?, either re-run the SOLxxxx commands or fix both the model and the drawing. Something added?, either re-run the SOLxxxx commands or fix both the model and the drawing.  Now that was true "mucking around".  R14 gave use OBSCREDLTYYPE and OBSCUREDCOLOR, R2000 gave us multiple layout tabs, and the little DVP routine above filled in the rest.  We are now considerably faster that we were using the SOLxxxx commands, and the drawings are not divorced from the actual model.

Not really! It may appear that I tested it for a few minutes, but have been using AutoCAD for 20 years [since Version 2.18] So, was familiar with DVIEW and CLIP, but dropped it in the past for SOLxxx. I thought I'd give your method a try, thinking maybe he stumbled on a trick worth investigating. Still not convinced. You think the SOLxxx are too much mucking, and I know - for me - DVIEW... is too much involved for no much return.
As for the changes in the design, well I worked for a drafting bureau, so the designs were well tought out before we were commandited to draw! I do realise that in offices, especially with young generation of engineers, it is possibly not that easy. Old engineers were working the pros and cons of a design at 95% probability before giving it to draw. New ones, they have learnt that it is easy to modify on CAD, so why bother thinking ahead

I had a customer that had a curved covered walkway in past. Taking a long section centrally by aligning viewports would have been tedious and not accurate anyway!
Alright Monsieur Broussard, I'll keep struggling along my ol' way

[Kerry]

..>>>   Old engineers were working the pros and cons of a design at 95% probability before giving it to draw. New ones, they have learnt that it is easy to modify on CAD, so why bother thinking ahead
....

AMEN !

[CADaver]

It does not matter if one is being comfortable, cause I do not draw just for myself. Sometimes customers want to see what they pay for on the screen.

That's what 3dmodel reviews are for.

Yeah, meant while drawing. Well, you have to be in HIDDEN mode to have the OBSCURELTYPE working, so the tessalations will show up and on complex design clutter dwg unnecessarily.

I work with shademode set to 2d always, I built it, I should know where it is right?

When we were comparing the benefits of one vs other, I was under the impression that we were comparing oranges! I agree with you that the potential of AutoCAD is not so much what it does, but what it allows one to do. But in this case, you were considering add-ons to AutoCAD [whether in-house development]. Not fair.
I am sure that when you were comparing AutoCAD 3D capabilities against the other software, other users could have said OK mine does not do that out-the-box, but I can make it do!

I was offering a solution to the complaint about dview clipping.  But even without the aid, dview clipping maintains the connection between the model and the drawing that is destroyed by the SOLxxx commands

Right. Picking point is only gonna work on curveless objects.

sure it will as long as you not too "retentive" about the hatch line intersection the edges "exactly".

Not really! It may appear that I tested it for a few minutes, but have been using AutoCAD for 20 years [since Version 2.18] So, was familiar with DVIEW and CLIP, but dropped it in the past for SOLxxx.

and spent a lot of time learning and using that method no doubt.

I thought I'd give your method a try, thinking maybe he stumbled on a trick worth investigating. Still not convinced. You think the SOLxxx are too much mucking, and I know - for me - DVIEW... is too much involved for no much return.

You get the very same results in the view as you get with SOLLxxxx, so the return is the same for both except with DVIEW you're looking at the true model not some facsimile.  How is DVIEW "too much involved"?  You know where you want the slice, you know the current view target, you keyin how much in front or behind that target you want the slice.  You're done.

BTW, DVIEW works on ALL elements not just solids.  With the SOLxxx commands all the elements in the file must be 3dsolids, which prohibits the use of blocks for modeling.  Using blocks, our files run in the 10MB size, if all the elements were 3dsolids, the file sizes would be in the 200mb range.  Using blocks, we can attach intelligence to the model that can be extracted later (EATTEXT), something that can't be done with 3dsolids without some other application.

As for the changes in the design, well I worked for a drafting bureau, so the designs were well tought out before we were commandited to draw! I do realise that in offices, especially with young generation of engineers, it is possibly not that easy. Old engineers were working the pros and cons of a design at 95% probability before giving it to draw. New ones, they have learnt that it is easy to modify on CAD, so why bother thinking ahead

The problem now is not that it is easy, but that schedules have been compressed.  Projects that took 2 years to complete 30 years ago are being completed in 8 months now.  Products have to hit the market while the profitability window is open.  In these cases I have to get the steel into the fab shop long before the mechanical design is complete so that it can be erected prior to the pipe delivery, and before the steel is erected the foundations must be poured and prior to that the piles must be driven.  So very often we have issued the pile location plans well before even the process has been finalized.  That scheme leads to a multitude of revisions downline as the design is hammered out.

I had a customer that had a curved covered walkway in past. Taking a long section centrally by aligning viewports would have been tedious and not accurate anyway!

Might I suggest that what you finally gave him was not completely accurate either, and required no small amount of of "mucking about"?