SolidWorks 2013 Update
There are two things I want to discuss about SolidWorks 2013. The first is the Conic sketch element. Remember that when 2013 originally released, you were not able to make a spline c2 to a conic. There was a discussion on the Conic in the SW forums a while back, where Mark Biasotti, who is usually correct, and usually fair-minded got caught up in saying that it was mathematically impossible to give the conic c2 capabilities. There is a point of view from which that statement might be correct. I’m not sure that you can drive a conic by the first derivative, although the derivative of second order stuff is easy enough, you might not be able to drive by it, although you’d have to show me why.
When you investigate a little, it turns out that ellipses and parabolas are also unable to have anything c2 to them in SolidWorks. (Ok, I wrote that about 2013 SP0, but in SP3 you are able to make the c2 relation between partial ellipse and a spline. I checked 2011, and 2011 does not allow c2 between a partial ellipse and a spline). It doesn’t make sense to make two arcs c2 to one another unless they are “co-radial”. But you can make a spline c2 to a circle. Making anything c2 to a straight line except a spline doesn’t make any sense. I recognize all of this and understand the reason why. But to say it is mathematically impossible to make something c2 to a conic is simply incorrect.
Although the items within sketches are not history based, I’m sure there is some sort of order to getting things done. Whether you can drive a conic by some other element’s curvature, it is undeniable that you can drive a spline by some other element’s curvature. And if you can’t drive a conic by a spline, you should at least be able to drive a spline by a conic. I just didn’t see why that wouldn’t work. People were making stupid claims, and frankly I think claims that they didn’t understand.
In any case, reseller employee Alin Vargatu pointed out a while back that 2013 SP2, you were suddenly able to add c2 between a conic and a spline. And now that I check it, you can also now add c2 between a parabola and a spline, and a partial ellipse and a spline. (But not between a partial ellipse and a conic, partial ellipse and a parabola, or a parabola and a conic.) I can buy this situation. So there was obviously someone somewhere who had a better idea of what was going on than the people pedantically stamping their feet saying it couldn’t be done.
Plus, I just want to point out right here that I called this one, in this blog post in the paragraph starting “of course”. I said that it was incompletely implemented, but would eventually get the rest of the functionality in a couple of years. I was wrong, it didn’t take them a couple of years, just a couple of service packs. Well, thank you very much to whoever fixed this. And to Mark B, sorry to be so hard on you, you’re normally on the mark about this stuff, but this time you missed.
So. I pounded that one into the ground hard enough, I think.
The second thing I want to talk about is the Intersect tool. Is anyone using it? Anyone completely in love? I’m personally bored. I spend more time trying to figure out what’s happening when it fails than I will ever save in actually using it. I wish it could use sketches so it could replace the Split feature (and the interfaces are almost identical). This is a great example of different egos running different areas of the software. If everyone were just on one big team, or one person were leading the development in specific areas, things like Intersect and Split wouldn’t duplicate so much functionality. Intersect does almost exactly what Split does, except Intersect can’t use sketches and Split can’t use solids as tools. Neither of them can result in surface bodies. We still don’t have a tool that can split a single surface body into two. And then there are the features that save bodies out as parts. 4 of them, each subtly different.
Also about the Intersect tool, has anyone ever used it without using the Invert Selection option at the end? The interface seems to be asking you exactly the opposite question that you want to answer. I always have to invert the selection, because I’m always selecting the stuff I want to keep. Must be an incorrigible pessimist at SW putting together the interface.
Here’s a comparison of the two interfaces. Notice that the Intersect doesn’t have the Message at the top, so they aren’t even trying to explain to people what to do, what it’s supposed to do, or how to get started. Fortunately, it does have the interactive what’s new symbol, but that will be gone next release.
Even though these look so similar, there are some serious differences. In Intersect, you select the target body along with all the tool bodies. In Split, just select the tools. Plus, notice that Split only seems to indicate that you can split with surfaces, doesn’t say anything about planes or sketches. Intersect at least indicates that planes can be used.
The guys that I know at SW that do the interface and usability stuff are probably twice as smart as I am. Which makes me wonder what in the hell has to happen at that company to cause smart guys like that to eff up an interface like this.
I like Alain’s videos. Solidworks is full of hidden features. I would prefer a two page written presentation with some screen shots that can be easily downloaded for future refrence. It would also be nice to know what kind of shapes will cause trouble. I use intersection curve because SW does not know how to project sketches.
Good one, Neil. 🙂 Nope, I was actually on-topic (read the article).
I think maybe Alin is trying for SW salesperson of the year….
These little video promotions pop up everywhere don’t they? 😉
Ah the dedicated SW VAR, a dying breed…
One more application for using the new Intersect Tool:
http://www.youtube.com/watch?feature=player_detailpage&v=vFh6Q_T874s
Note how fast is the second technique shown in the video.
Thinking about conics some more I think SW need to distinguish 3 situations where they might be used and where the user has a slightly different objective and therefore there should be loft/boundary sheet algorithms tuned to those purposes. At the moment loft/boundary may not be as ‘conic aware’ as they ought to be.
1. Where you have a path that makes large excursions or changes of direction and you want the conic to flow well along it as in a fillet – the proverbial ID rollercoaster. Here you want SW to generate a clean surface but one that holds up well to any path deviation and will not have too restrictive rules that prevent it generating or fail on rebuilds or when impinged by further features or dimension changes. It builds easily and holds together well.
2. A patch situation like for a shoe instep where it does have some localised restraints and edge conditions but the internal outcome doesn’t need to be too specific a shape but rather have a pleasing swoop/smoothness in several directions concurrently and blend with adjacent surfaces reasonably well but over a relatively small distance. Perhaps the patch has a max x/y ratio of 5:1. The excusion out of plane is only moderate. It doesn’t matter if the overall geometry changes a little as a result of the construction to satisfy the objective in the manner knit can be tolerant of a gap.
3. Where you specifically want to create a quality fair surface between carefully constructed/controlled profiles that mirrors well like a glider fuselage, a HPV shell, or wings and the like. You dont want SW to modify the geometry to do a best fit but rather want to strive for maximum engineered precision. In this case the flow is along an axis ie parallel planes/profiles and the surface fairness and faithfulness to the conic is a priority.
Perhaps you could identify these algorithms as compliant, semi-adaptive and conformal.
Dunno, hope that helps someone at SW make some progress with conics 😉
I guess we’ll have to wait until SW2015 for a further tweak… if SW hasn’t gone extinct by then…
Of course having been identified as something potentially useful for otherwise nearly development starved SW surfacers its probably going to be bought out by Autodesk…no identifiable future as a SW addin and all that.
DS had better get out their check book promptly. We’ve been here before 🙁
Why doesn’t Dassault just license this technology from cadcamcomponents? As a third-party add-in it is definitely weird, because it just doesn’t DO enough to justify being a third-party add-in. The problem here is the uptake, not very many companies use it.
From what Rick has shown here it would be very well welcomed to have by many surfacing users. Which is still not many users, I do get that 😉
Matt, Mark, in the example from your video, there should not be an equal curvature relation available to the user.
I was under the impression that you should be able to force a spline to match the curvature of a conic, but not the other way around.
@matt
You wrote:
“The second thing I want to talk about is the Intersect tool. Is anyone using it?”
I promised you a video showing one great reason for using the new Intersect Tool and it is about simplifying geometry.
One gets these large models from his/her suppliers or customers and just needs solids without all the internal faces in order to 3DPrint the models or something similar.
The idea for the technique shown in the video came from my customers, so I can tell you that this tool is being used quite a bit.
Enjoy:
[youtube http://www.youtube.com/watch?v=7pI9xIgJXqY?feature=player_detailpage&w=640&h=360%5D
If the embedded code does not work, please use this link for watching the video:
http://www.youtube.com/watch?v=7pI9xIgJXqY
@Neil
Dear Neil, that is a very good answer. Thanks for taking your time for explaining this topic. I will continue visiting this forum, is really useful to me. Best wishes
Jose,
People have been interested to get conics in SW for years. Conics produce sweet surfaces for styling purposes and for more ‘technically correct’ things like aeroplane fuselages. The trouble with conics as they arrived is that they couldn’t blend in seemlessly with adjacent surfaces. It was a bit like being given a new pair of Nike trainers without the laces. It was sorta cool but also a bit useless. With sp2 and equal curvature that part is fixed however people have further issues with constructing surfaces from conic profiles in that those aren’t as sweet along the loft or boundary surface as we would like. That is to use the shoe analogy – upon closer inspection the soles of the shoes were a bit malformed/warped in the mold.
Really I suppose this issue isn’t to do with the conic sketches/profiles but how the loft and boundary algorithms interpolate between them. Possibly these need a tighter conic specific algoithm to be added now as well rather than have slightly relaxed ones that attempt to satisfy for a wide range of conditions . Rick was pointing out that presently from a precision viewpoint there are small unwanted deviations in the constructed surfaces. For instance although they are conic at the profiles they may not be so much in between or might have slight curls and waves along the edges. Also if you mirror the surface the final result may not maintain a normal to the plane continuity across the mirror. The whole attraction of conics though is the lack of inflection and a superior fairness so this is disappointing or frustating. You would kind of hope that this situation arises from a lack of time to get it done entirely rather than a fundamental inadequacy in the approach or something being provided that looks like conics but isnt. SW do have a history of only doing things by half when they are introduced but the conics implemetation seems to have fallen short of even that standard in the usefulness category.
As I said I think this is a matter of needing an option to maintain precision rather than produce a solution for as many cases as possible. Perhaps we should have a ‘interpolation for conics’ option in the UI even though we might be greeted with a message like ‘conic precision could not be obtained in this case please try…’ Perhaps it might only be available for parallel profiles or straight paths or something. I guess Mark would be able to tell us if that was possible. Its sort of like having a variable tolerance for knit I suppose.
Neil,
A conic sketch item is useful. Conic curve makes such a beautiful corner that can have the “sharpness” adjusted. The conic curve can replace many two point splines with no inflection points. You might be able to get a symmetrical shape using mirrored conics for each section of a boundary surface. The edges will not be perfect but they might look ok.
The GW3D conic curves are much more versatile that SW conic, as they can be controlled in at least a dozen ways. Unfortunately they are curves and not a sketch entity. The conic surface is a wonderful versatile surface. The spine may be a general curve. I have only used that one time. GW3D add in from cadcamcomponents.com has many nice surfaces. I think that they added the conic surface in the 2009 version. Price is about $1300 + $200 subscription. I bought it direct. It is available through your favorite SW VAR. They have a 30 day demo. It takes some practice to use it. Support is excellent.
Hi there. I have been following this post for the conics topic. It is Very interesting even when i do not use it, but i am trying to understand it. My question would look like a little noob, but after the update of SW SP3, do conics really work as Matt wants? i mean, on the book he said : “The Conic sketch entity is a new feature in SolidWorks 2013. While this is used primarily to create smooth shapes, it is not possible to add curvature continuous relations to it, only tangent. This is a welcome addition, but it still lacks important functionality ” But with the new updates is working fine right? Thanks
>The basic problem is that a SW boundary surface with all conic profiles does not preserve the conic shape between profiles
Right, so Rick, this is what I am interested to know as someone lagging behind with SW upgrades. Tell me more. How errant are they? Are conics as they are in SW presently all but practically useless except for revolves and simple fillets? (and I guess even for those conic sketches probably weren’t much use before sp2…) btw what’s the price of GW3D in round numbers? Thanks 🙂 Nice plane too. 😉
Edit: your conic sections would naturally be parallel along the axis of the airframe, does GW3D also preserve the conic profile between sections normal to a 3d spline or at least a very fair transition? Perhaps I need to see the manual (good?) or demo it…
Arthur,
Solidworks boundary surfaces will mostly make a symmetrical shape if all of the sections and guides are symmetrical. It can still do a rounded butt crack which shows as a reverse in curvature in the curvature combs. The basic problem is that a SW boundary surface with all conic profiles does not preserve the conic shape between profiles. The most controllable guide curves for this kind of shape are a plan sketch and a profile sketch projected to make a 3d curve. How do you mirror a curve in Solidworks? Just to irritate me the boundary surface consumes the guide curve that is needed for other surfaces.
Conic surfaces do not present this problem. There is no possibility of an undesired inflection point since a second order curve cannot produce one. The conic shape mirrors perfectly and knits. There are no tits or curls to mess up thicken or offset surfaces.
@Rick… Very nice work!.. I can only imagine what you may have gone through to get what you have using surfaces.
I do remember GW3D being very forgiving and liked the surfaces it created!
So, is this a direct or subcontract for a company.. like, Icon?
Mirroring has never been a “good” idea no matter what program you’re in. Alias, Rhino, SW, SE, ST (SolidThinking)…the list goes on and on. Why leave it to “chance” to let any software figure it out for you? Does it take more time, sure, of that there is no question.
Try making a 3 (or more) point spline that goes around to create 1/2 a “circle”, make it horizontal at the top and bottom, surface extrude, then mirror…No bueno
Do the same as above but only this time complete the “circle” by going all the way around. Notice how the spline “snaps” back to itself. Now surface extrude and notice just how much nicer it is.
Work around…maybe, better overall approach definitely. I’m an “overbuilder” and proud of it…..anyone else wanting to join, membership is free (ok…I did laugh a bit at that last part)
Paul,
The pictured amphibian is a new project. My first design with almost all conic surfaces. The canopy from nose to end of pylon is one conic surface. It has a nose singularity, convex in the windows and reverses to a non-convex region and ends in a planar conic that is perpendicular to the spine. The ambient reflections flow smoothy.
@Rick… what aircraft company do you work for?.. what is this aircraft model?… and, is it in production?
btw, removed the previous sldprt,… here’s the updated file.. (I guess we can not update or add files?)
[file]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/sweep-boundary-w-loft-zxys-1.zip[/file]
GW3D conic surface makes beautiful and versitile surfaces. The “conic surface” is a function surface where everywhere the surface is on conic curves that are normal to a spine.
I most often use the simplest variation; conic surface is defined by edge curves, slope constraints and rho. Usually one edge curve is on the right plane. The other edge curve is usually a projection from a curve on the right plane and a curve on the top plane. The slope constraints are then normal to right pland and normal to top plane. Start with rho= 0.5 and adjust to taste. Surfaces with 4 edges are easiest, but often the conic surface will work smoothly right to a singular point.
The attached shape is mostly conic surfaces.
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/M7closewheels.jpg[/img]
I’ve seen that difference as well between normal to profile and tangent to an extruded construction surface. Run a deviation analysis on the knitted edges and there is a difference…
10 years or so ago I completed a CDRS course and there was a discussion about mirroring critical surfaces, like a car bonnet. I believe some surfaces like these are built full width as it is very hard to get a completely smooth curvature plot across a mirrored surface.
@Rick.. (GW3D.. it was interesting years ago.. don’t know how much it matured?) although I understand the issues with symmetry or inconsistencies with mirrored curves and surfaces… this particular shape looks pretty simple.. done this many times using different approaches or breaking this up into more patches for more control.
Anyhow,.. don’t have your part but I opened Marks part in edrawings and copy/pasted/measured the profiles.. and remodeled in SW2006 to see if anything “may” pop up using a loft (maybe SW2013 lofts suk more??)… it works fine imho. The curves being clean and using tangency with the surface.. it works most of the time in SW2006.
@Mark.. nice Carrera work… and Power Surfacing add-in (SudD to NURBS)… looks a bit like Freestyle in Creo.
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/sweep-boundary-w-conic-zxys.png[/img]
I second that. We have been evaluating GW3D, and to be honest, it is not an easy thing to evaluate. At the moment we were thinking that it didn’t quite have enough value to us since it looks like only a few of the tools would be of any use to us, and even then only rarely. But this conversation seems to be demonstrating a better example of how it can be used that we haven’t seen for ourselves yet.
We often mirror surfaces that we want to be nice and smooth across the mirror plane. I have noticed that I sometimes get a better result using tangency at the mirror plane instead of normal to profile. This requires first extruding a construction surface from the curves that’s on the mirror plane. It doesn’t seem logical at first that there should be any difference between normal-to vs. tangent, but often there is a noticeable difference.
Another wacky thing we’ll sometimes do if there is an offensively visible seam along the mirror plane, is split out an area along the seam and then do a fill surface with curvature continuous as a patch-job. Not an ideal solution but it can often improve the look and smoothness along the seam.
@Rick
This topic is hot, Rick. Would you consider creating a presentation on how you create your geometry? I am sure you would have a lot of attendees!
Or even better, if you find the time, start your own blog. I would be a frequent visitor, for sure!
So you still need to define the boundary condition in GW3D along the mirror plane?
I don’t see how specifying normal to profile means you have set up the construction in a dodgy fashion. It just tells the software what the tangent direction is between curves. Another way of controlling the input curve influence.
I can set up a boundary surface in ProE and it will show a crease along the mirror line, unless I specify normal to/tangent to etc (dependant on construction). I don’t see this as a failing in the geometry creation of the software, it is just a way of specifiying design intent. I might actually want a crease.
I agree 100% with you about SW boundary surfs though, sometimes it creates a thin wiggle right along a boundary when logically it should not be there as the input construction is clean as a whistle. Loft does not normally exhibit this.
The normal to profile surface will mostly band aid a bad Boundary surface. it is so much nicer to have all positive curvature. Very carefully placed sections will almost avoid the hogbacks and butt cracks. Use curvature combs when reviewing a boundary surface. The GW3D conic surface has a perfect edge condition at the mirror plane. It is amazing to have such a good looking surface that is accurately dimensionable. I rarely use a boundary surface now.
Rick, so do you think a mirrored surface that uses normal to profile along the mirrored edge as a defect? How often do you succeed in getting a surface that is smooth across the mirror plane without using normal to profile? Do you have to add alot of extra sections to control this?
Solidworks surfacing will make better aerodynamic shapes than lofts. I have given some of the common defects names. When a surface should mirror perfectly but does not produce a shape that matches slope at the mirror plane, a concave dihedral is a butt crack, a convex dihedral is a hogback. Setting the edge condition to normal to curve will round the defect but it is still a defect. Using more than 3 curves in a direction can result in solidworks finding a surface fit that has ripples or wrinkles rather than the smooth surface intended. You must look at curvature colors, curvature combs and zebra stripes to see these. When a surface has a very tight curvature at the edge that was not desired is a curl. Tits are tiny peaks usually in a corner but sometimes can result from a curve wrinkle. For some reason, I feel the occasional need to practice vocabulary when Solidworks fails to make the desired geometry.
Unfortunately these geometric defects are real and will be translated into a problem by CAM software. Solidworks Lofts are particularly untrustworthy. Sometimes profiles are ignored, and occasionally it will genertate a brocolli shaped surface from sensible sections. I have seen Solidworks loft an airplane wing which should have resulted in a ruled surface, produce a sag of 1/16 inch. I have seen lofts totally destroy an airfoil profile and produce a blob twice as thick. Which was made into a Sla model that sits on a shalf as a nasty reminder. Lofts that have only two profiles and two guide curves work the best.
Fit spline is terrible for airfoils, it adds many inflection points and ripples to smooth data. I make an airfoil shape by a manual curve fit. I start with a fit spline, activate curvature combs, in a new sketch Pick a nose point, and trailing edge upper corner, and two other points where the curvature is changing, set the slopes of the spline handles, and adjust the lengths to get a good looking curvature comb that has the character of the airfoil. Fix all of the handles by setting dimensions.
I now use an add in to do all of my aerodynamic surfaces. GW3D from cadcamcomponents.com. The user interface is not fully integrated into solidworks and a bit clunky. It is fussy about the edge curves being continuous. Solidworks convert entity will not always join the ends of curves. The features that I use the most are conic surfaces, blend surface, conic curve, helical sweep, and offset curve. The surfaces tend to be more relaible than SW surfaces. They will occasionally produce a defect a tiny defect near a singular point that I call a ribbon fold, not pretty. It can be fixed by changing some parameter in the conic surface. GW3D features do not consume geometry so the same curve can be used over again.
@David Paulson
Heck, I would be happy with a definition from Rick as to what each meant. Like you, I’m no surfacing wiz and I’m wondering what it all means.
@Rick McWilliams
Rick, I follow your posts because I respect your use and practical knowledge of surfacing. Surfacing is a skill that doesn’t seem to apply greatly to my practice, but I always wonder if I am missing something. About half of what I do is Flow, and creating fan blades and turbine wheels may be better with a defined surface. My lofts are somewhat primitive, and surfacing done right would make a lot of sense.
But when you refer to hog backs and butt cracks that are created by SW, are these features that show up in the manufactured product or just in your model. I can see where this would be a problem with machined parts that are just replicated on a CNC machining center. But would it be relevant how these imperfections show up on an actual aircraft fuselage?
Why do software companies insist on using words such as ‘freeform’ and ‘organic’, especially when relating their software to industrial design? Go to an ID studio and no one will be using those terms, apart from organic when describing a potato…
@Rick McWilliams
Rick – have you looked into the Power Surfacing add-in for SW?
Mark
Matt, “Intersect” is very usefull.
Once I have a bit of time, I will demonstrate how, for some applications, Intersect saves the day.
Mark,
I have found that almost all boundary surfaces with variable sections will produce hogbacks and butt cracks. Some are hard to see. Your sweep will satisfy the edge conditions but the rho of the conic is variable.
I insist on control of the surface shapes. I will not accept what a CAD program likes to generate. A conic surface is easily defined and dimensioned. It will be the same shape if constructed by hand or by software.
The SW boundary surface will mostly contact the defining curves and then does something between those surfaces. Edge conditions are rather loosely followed and often make an ugly ripple, wrinkle, tit or curl. These defects are often very small. I once saw a 1/4 inch gap in a 20 inch part. Solidworks is very careful to not say what a feature is supposed to do, rather left to user discoverability. I would much prefer to work with untrimmed surfaces bounded by cubic splines or projections of cubic splines and Coons Patches. Maybe we need to get past nurbs to include some nice function surfaces.
Hi Rick,
Take a look at this model -it isn’t the exact console from the Porsche but it does show using conics with boundary. You’re right, that for this example, holding the “Next Curve” does not give as good as solution as “Global” but the better way is to sweep a conic (See second configuration in file.)
[file]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/sweep-boundary-w-conic.zip[/file]
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/boundary-w-conics.png[/img]
Mark its another snippet I remember you posting on the forum. I do remember these small details you know…. even from 4-5 years ago 😉 Sounds like she doing really well for herself. I bet she has proud parents.
What’s satisfying about recalling this is that although senility beckons I didn’t need a search engine (apologies to Jim) and that we were able to share a ‘social industry experience’ without us both having to open SW. (apologies to DS marketing).
@Rick McWilliams
Hi rick
post your model and I can take a look.
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/bsconic.jpg[/img]
Zebra stripes show hogback and butt crack. Adding normal to curve constrant will help hide it, a smooth butt crack is still a butt crack.
Mark,
I modeled a boundary surface much like your console. The center curve was a spline, and the edge was a curve consisting of a spline projected on a spline. All sections were conics, with influence to next curve. The result looked pretty good. I mirrored it and turned on zebra stripes and the reuslt looked not so good. Two hogbacks and two butt cracks. Not a pretty animal.
Boundary surface is very powerful, and in general better than a loft. As long as you are happy with a butt crack right down the canopy of an otherwise attractive airplane it is marvelous.
Neil,
I suppose your daughter is though college now?
How did you know about that? – anyway, yes, she has her masters now and is working full time in So CA in a Pysc Hospital as admission/intake counselor.
thanks for asking…
Mark, I do. Some day I’m going to hunt down that snippet 😉
Now I don’t want you to take this personally because I know you endeavour to do your best for your employer and users but DS have to realise they are going one on one with SE for my and others attention come ST6/SW2014 release time. Its a fork in the CAD road we will only linger at for so long. SW2014 is going to have very useful enhancements and be a very convincing buy for people to persist with knowing in a larger sense its soon to be a lost cause. It is in the interests of DS to create some advanced interest in this release because SE will be out first and they have a lot to gain and SW a lot to lose. Again I’m not getting at SW staff but its a reflection of the big reality. If DS don’t publicise the best of what’s new in SW ahead of their usual launch or chart a clear direction for SW development or still won’t provide Catia Lite info then they are not going to have a credible market presence to compete with. I think DS should put aside the usual non disclosure/confidentiality game in the interests of retaining customers. Do I think they are smart enough to do that? Um, no. How a successful company can be progressively buggered by such inept management is hard to understand but here it is manifest in SW and the poor old users and the loyal staff are the ones left with the mess.
Anyhow I haven’t been on maintenance for a while so I can’t participate in beta but thanks for the offer to give feedback. As we know however by beta time has rolled around its already too late to affect things for the better. If we are fortunate some kindly conscientious SW product manager might do a fix or two in a sp of late but that’s kind of an unusual occurance.
Keep up the good work. I suppose your daughter is though college now?
@Rick McWilliams
Creo is the king of conics, you can do this plus explicitly control Rho values with a relation or graph along the spine. Comes with the base package, no add-ins required 🙂
I noticed tonight that I can’t “reverse dimension sense” for a sketch offset in ’13. Is that new or specific to offset curves? It was really annoying.
@Mark Biasotti
opps – sorry about the format issue…
@Rick – There is an option in Boundary that will preserve the shape between conics and I did use it on the Porsche Console – it is in the Curve Influence option just above the selection box called “Next Curve” and it’s purpose is to hold influence from one curve to the next. Check out the file attached. I think it does a pretty good job of this.
[file]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/usingconicswithboundary.zip[/file]
@Neil – Sorry – I don’t recall ever saying we’d never have conics?, also, sorry, I am not at liberty to say any thing about functionality that is still in Alpha so sorry you’ll have to wait till June, but I do hope you try it then and give us some feed back.
If you want to know more about Conics, I wrote a blog post about them a while back at: https://forum.solidworks.com/docs/DOC-2285
Mark,
Your shape looks nice. It is my experience that boundary surface does not preserve the conic shape between sections. It may also produce an edge tangency problem such as butt crack or hog back. The outer edge can also develop microscopic curls.
Neil,
You could order the GW3D add in and get conics. The conic curves are curves not sketch elements. The interface is clunky, but the shapes are beautiful. A time limited demo is available from cadcamcomponents.com.
Mark,
I was happy conics arrived because it was my understanding from a comment you made previously that we wouldn’t be getting them at all. It was disappointing to find that the initial implementation of conics hadnt quite delivered enough to make them truly useful for ID purposes. Attribute that to product definition, hurried development or whatever. Thanks for putting in the extra work to add equal curvature in sp2. The only problem I have now is that as much as I’d like to upgrade my SW seat to take advantage of them DS HQ haven’t been able to satisfy me SW has a future and that I won’t be wasting $5290.
Have you anything you can share with users about the new shape spline functionality coming in SW2014? Does it offer something unique that splines can’t do today and is that a one off or are other tools based on them to follow along?? Sorry if you are put on the spot by that but you must realise these days SW dont have the same marketing wind at their backs as they used to. I have some sympathy for the good folks at Concord who have spent years pushing SW forward and consistently try their best. 😉
P. S. Tell Jim to make sure its all documented properly. Sorry no new search engines are allowed 🙂
Oh, and if you have some spare moments can you please get the UI tweaked as per Matts comments above…
Thanks
BTW,
It was really cool to use the Conic on the center console of the Porsche Carrera GT
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/interior-wide.png[/img]
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/conics.png[/img]
It sounded a lot like you or those defending you were telling me what I wanted was mathematically impossible. Apparently it wasn’t because I got what I wanted in sp2.
Hi Matt,
Then it was just a misunderstanding or miscommunication. We we’re wanting the same thing and I think we’re both happy it got in there in an SP.
Mark
A conic sketch entity is nice. A conic surface is wonderful. GW3D add in gives me conic surfaces.
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/conicsurfaceSnap1.jpg[/img]
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2013/04/conicsurfaceSnap2.jpg[/img]
These screen shots show a nice conic surface with very swoopy boundary curves. one boundary curve is the profile view edge, the other is a plan view edge projected on a profile of widest point. The conic tangency is to the top and front planes. Rho is 0.5, rho of 1.0 will have a square corner, rho of .3 will be an ellipse, rho of 0 approaches a straight line. I use this type of simple conic surface often. Occasionally I use alternative methods of defining the tangency. The rho can be constant or varying from end to end. joining two of these surfaces will only be C1 but it looks great.
The GW3D user interface is a bit clunky, and fussy about curves actually matching point and tangent at the common end points. These surfaces can be thickened and trimmed in the normal Solidworks ways. You do not need the addin to use the geometry, only to create a new feature. There are even two point singularities.
@Mark Biasotti
Mark,
No, I wasn’t asking for c2 on creation. Nothing else has that, why would I be upset it was missing? I was just asking for what was missing – the ability to apply c2 between a conic and a spline. There is no such thing as order from the end user point of view within sketches.
It sounded a lot like you or those defending you were telling me what I wanted was mathematically impossible. Apparently it wasn’t because I got what I wanted in sp2.
Hi Matt,
Glad that you circled back on this. I’d have to go back to the thread in the forum that you are referring to but I believe you misquoted me when you say “it was mathematically impossible to give the conic c2 capabilities.” I believe what I said was that ” it is impossible to Drive Conics to be G2 or C2″ – and I still hold to that. Those are two very different statements. Perhaps it was miscommunication on my part, but I have always wanted and supported having the ability to apply a C2 constraint between the conic and a spline, but that was a limitation at initial time of SP0 release (that we were aware of) and I and a colleague of mine worked hard to try to get that limitation lifted in an SP (and did this even while 2013 Beta) – that is why you see it suddenly becoming available in SP3. Correct me if I’m mistaken but I think you were wanting to have C2 at the point of creation of the Conic? If that is the case we just don’t have the current infrastructure to support that in our sketch constraints system – i.e. C2 needs to be applied after conic creation because the conic would have to drive the C2 between the conic and the spline (affect the spline at time of creation of the conic) – conic=master, spline= slave. I’m not saying that it couldn’t be done, but it would take some significant work to make it happen. I’m not backpedaling with that statement because it still would be impossible to drive the conic to be C2, rather it is a enhancement/change in behavior of our constraint system, not conic, in order to have it appear that you can create a Conic to be C2 to a spline. Just trying to be clear here…
Thanks
Mark
I use the GW3D add in to get nice smooth shapes. The conic curve in this add in is a curve defined in various ways, 2 points and 2 tangents and rho, 3 points and rho, 5 points, 4 points and 1 tangent, and all of the other combinations that I have yet to use.
The super useful feature is the conic surface which is everywhere a conic perpendicular to a spine with various nice ways to control the tangency and rho. Rho is basically the squareness of the transition between the end points. It is a very nice aesthetic control.
My airplane fuselage shapes have one conic surface for the top quarter and another for the bottom quarter. The side edge line junction has only C1 continuity but curvature is usually quite close. This surface joint has a very nice appearance when it is on the widest point curve. It preserves aerodynamic properties of rounded bodies very well. There are never butt cracks or hogbacks along the plane of symmetry. The resulting surface is defined by curves and it is not a trimmed nurbs surface so the edges do not have high order microscopic wiggles.
I recently started using the intersect feature. I noticed something else about it though that makes me re-think. I often use the indent tool to do similar things. So I recently tried using the intersect tool instead of indent, and what I noticed was that it took significantly longer to calculate. Feature statistics said the indent took only 0.1 sec, while intersect took over 5 whole seconds! End result was the same. Why such a huge difference in calculation time? If you have a long feature tree and pay attention to creating as quick rebuilding of a tree as you can, then something like this can really make a big difference.
Well apparently bumblebees can’t fly either…a mathematical impossibility I heard.
Perhaps you shouldn’t question divine providence, but quietly hope for a UI fix to arrive in sp4, basing your faith on the amazing good fortune that has befallen you with sp2, once the conic issue was widely known, thanks to this *humble* blog…
Failing that, I’ve heard that letters to Santa, asking for SW enhancements, are redirected to Concord, and that each one is read, and treated with the utmost respect and consideration. Appropriate action is then undertaken, only by the elffs (extremely lazy effer?) deemed most capable of these tasks, and shouldering the heavy responsibilities that go with them….Of course that could just be marketing BS.
OK seriously, are conics genuinely useful now? I think you said previously that lofts between them were prone to inflections?