Mike Allen

It is fully documented in the standard Help file - go to the index & type in "SmartCursor." It can even be disabled (if anyone really wants to work without it) - from the Snap dialog box (right-click when you are in cross-section mode.) It really is a pretty elegant way of cutting down on the "noise" that comes from too many SmartSnap guides.

Hey, that works pretty well! I didn't notice before that you could turn them off by pausing over the x. They are really quite useful.

I played with this for quite some time this morning & it really seems like the x shows up wherever your cursor happens to cross a curve the first time. Each time you draw a new curve, all of the x's are reset. I'd be surprised if it were anything other than edge detection. The confusion comes in when you are looking for a specific relationship to the existing curves. For instance, if you want to draw a line from the midpoint of a curve, you move your cursor near where you expect to find the midpoint. IronCAD detects that your cursor is over the curve & puts an x there - it will be close to the midpoint, but has no real relation, other than that's where your cursor crossed over the curve. My guess is that one of the IC programmers was using the x while debugging a section of code & forgot to comment it out.

Well, I did an uninstall & restarted. Then I closed all system tray processes & disabled my virus protection (no other programs were running) and ran the IronCAD setup (I closed Windows Explorer as soon as I started setup - then once the actual installation program started, I closed the IronCAD Flash window.) IronCAD finished installing & I restarted again. Then I ran the IronCAD9Hotfix1.exe (after closing all system tray processes, programs & virus protection) and started IronCAD. After setting up Tools>Options and all my hotkeys, I closed & reopened IronCAD (don't want to lose all the settings, in case I crash.) A few more minutes of work to restore the file I was working on to where it was when I crashed (saving after each operation) - and eventually got the same Runtime Error & crashed during a Save. I have the file just at the point where inserting one more object from a custom 2D catalog and saving will cause the crash, so I will upload the Scene, Drawing & Catalog to IronCAD's FTP (Skyline_Save_Crash.zip). You should be able to open the Drawing file, insert the object from the catalog (it's the only one in there) & click the Save button. Strangely enough, inserting the note from the catalog into a blank drawing & saving does not cause a Runtime Error & crash. Also, replacing the catalog note with a simple text note works just fine - so I can finish the job without recreating the drawing.

I totally agree. No need for 3 sets of symbols to describe the same vectors. XYZ is the least ambiguous - most everyone has attended a geometry class in which X&Y describe a 2D plane and Z is the extrusion into the 3rd dimension. I wouldn't mind having the extra terminology to relate IronCAD to other modeling/animation software, as long as XYZ is always referenced at the same time - so Height (z) would be acceptable.

Elder, do you tend to get this error when saving? I've seen that a couple of times here - very annoying. I'm close what Windows identifies as "Programs" during IC9 installation - but don't mess with the "Processes" that Windows usually has running in the background (including the system tray.) As far as virus protection goes, we can disable the "On-access scan", but not the actual program (it's running on our network.) I'm going to try a reinstall with as much turned off or disabled as possible & see if this makes a difference.

IronCAD appears to be detecting the first time the cursor passes over an existing curve. It only seems to happen once per curve segment & if you pass over the segment too quickly, the x doesn't show up. Draw a line segment and then slowly pass your cursor across that segment (with the line tool still active) and you'll see what I mean.

I experience a slightly different behavior from my colleagues. The delay occurs only when I have my rendering set to OpenGL. When set to "Software", there is no delay. I went to the nVidia site & downloaded the latest driver for WinXP, set the Video Overlay to "On". Then in IronCAD, I changed the rendering to OpenGL. When I opened a medium sized scene (about 30 feet x 30 feet), there was a definite delay of more than one second on the box select. Switching to Software rendering got rid of the delay. What is strange is that we have at least one user who experiences the delay on Box Select regardless of the rendering setting. Personally, I switch back and forth between OpenGL and Software, depending on what I am doing. I like some of the features that OpenGL provides, but there are also some annoying behaviors that come with it (for instance, when moving parts with the TriBall, the original part location isn't highlighted - and the above mentioned delay with box select.)

So, do you have any idea as to how we can get ours to work like yours? I know it isn't our templates, because I have tested this with a blank drawing and a new view. The Change Source button always takes us to our working directory.

I've been playing around with using the TriBall to position 3D curve points on the surface of a solid object & it does work with the 3D arc - but only when the "Create Smoothly Connected Curves" button is not depressed.

I started looking at this from another angle & came up with another solution. This time, I kept the entry & exit angles of the lines & the exact radius of the arc (or the cylinder that the arc fits around.) I realized that a curve wrapping around a cylinder at an angle is going to be elliptical. I measured the angles of the lines relative to the cylinder & to each other, then used surface reshaping to taper the end of the cylinder to match the first line. I arranged the original 3D curve so that the first line would be in the same plane as the tapered end of the cylinder. Then I copied the cylinder & tapered the other end, & arranged it so that it would match the other line as closely as possible (I had to fudge, since that line isn't tangent.) Then I created two 2D shapes - one that projects the first line and the top taper of the cylinder, the other that projects the second line and the bottom taper of the cylinder (on the second shape, I projected the line & then moved it so it would be tangent to the ellipse.) The two projected ellipses don't intersect at a tangent point, but that's necessary if a single piece of rebar is going to wrap around the cylinder. So I moved one of the 2D shapes (and the corresponding cylinder, for reference) so that the two ellipses intersected as close as possible to a single point (this is where I really had to fudge, because the facets of the edges don't show the actual curve & I can't quite figure out how to find the point of intersection geometrically.) Finally, I created a 3D curve by tracing over the two 2D shapes. I used two arcs - one from the end of the first line to the point of intersection of the ellipses - again this took several tries, because I couldn't get the intersection closer than a few thousandths of an inch - and a second arc from the end of the first arc to the end of the second line. The arcs don't perfectly follow the cylinder, but they are extremely close & there is no 3D elliptical arc tool. I also tried using the TriBall to place the points of the arc segments directly on the surface of the cylinder (hitting "p" for each point), the way you can with a spline (my workaround for Carlo's magnetic curve), but the arcs didn't want to cooperate. It's funny, because it works without the cylinder. Anyway, here's my second attempt (actually, about the 50th - but the others were pretty unsuccessful.) -Mike fullrad3.ics

I'm not clear on what "Full Radius" means. Does it mean that the radius is a critical dimension & can't be changed to meet the lines? Are the lines supposed to be tangent to the curved section, or are they meant to connect at an angle as you have shown? You definitely have one line that doesn't project tangent to the given arc. As I see it, the given radius won't connect tangent to both lines (unless you change the length or position of at least one of the lines), even with a twist in the middle, but it appears to be possible for another radius. Which is critical - the radius of the curve, or the length/position of the lines? Given the ability to change the length of one of the lines & also make the curve tangent to both lines, I came up with the following solution. I took a block (for a projection plane) and made it colinear to one of the lines (just to give myself a frame of reference.) Then I made a 2D shape & projected both lines & the curve onto the block. I made a copy of the resulting 2D shape & bisected the arc, creating 2 half shapes. Then I used the TriBall to twist one of the shapes from where the two halves of the arc met, until the line was parallel to the original line. You can see that they are not colinear & the arc has a different radius from your original arc. Finally, I created a 3D curve out of two line segments and two arcs. If I knew which elements were critical, I could probably get it much closer - assuming that what you are asking for doesn't require contradictory geometric conditions (for instance, a line can't be tangent to an arc if it would cross the arc at two points.) Here's my file - see if the concept works for you. fullrad2.ics

I wonder why it doesn't remember the last directory for me. I just changed the link for a drawing file to redirect the view to another scene (the Change Source button takes me to the working directory set in Options, not to the directory where I had just opened the scene.) Then I saved the drawing to the same directory as the new scene & did Ctrl/Shift+L again. The Change Source button still takes me to the working directory.

I think he's looking for a way to change the default behavior of the Links dialog (Ctrl/Shift +L), which always goes to the working directory. It would be useful if that dialog was consistent with the rest of IronCAD & opened in the last accessed folder, instead of always going to the working directory that is set in Tools>Options.

I must be missing something, because when I was beta testing, this worked perfectly. But with the release version of IronCAD 9 (build 7562), the TAB key doesn't seem to work to toggle between Polyline & Polyarc. Any ideas?

I modeled one just before that with a slightly smaller cross section that didn't have the facets. When I increased the cross section diameter, the facets showed up. I just checked some of the IntelliShapes & the torus has facets, but the part torus doesn't - until you change the spin value to 360º. The only hypothesis I can come up with is when I increased the diameter of the cross section, the "insides" of the turns get close to kinking, so IronCAD switches to a less-smooth surface to compensate.

See if this is what you want. I made a frame with 2D curves (much easier to manipulate than 3D curves), then traced around it with a 3D spline. Then I created a 2D circle on the spline (it seems to come in with the extrude direction tangent to the spline), right-clicked on the circle & selected Create->Sweep. I used the 3D curve as the path & voila! -Mike escher_knot.ics

Well, I guess we'll find out for outselves.

It would be a useful enhancement to have 2D & 3D curves treated as geometry, so they could be picked to be hidden in the drawing view. My current workaround is to switch to precise mode in order to hide the 2D shapes. Otherwise, I would have to create a new scene configuration with the 2D shapes suppressed - which becomes a problem if the default config needs to be revised (then it's a chore to hunt through all the configs to see which have been affected.) Otherwise, I'm not too crazy about the idea of "precise mode all the time." On large models, it takes considerably longer to create the views.

Can anyone tell me if the Granite Translator currently works with Wildfire 2.0 & 3.0? Our engineering department (which is using Pro/E) is going to be upgrading from 2001 to Wildfire and I'd like to know if we will still be able to import parts (we currently have a license for the Granite Translator & it works great.)

I don't know if this is exactly what Dave is referring to in point 4, but it is sometimes quite difficult to place a dimension on a small-radius corner. You have to be zoomed in far enough so you can pick the curve between the endpoint & midpoint of the arc, or you can't place the dimension. I've also noticed this on very small holes.

Hmm - I don't get the "invalid radius" message until I'm below .00005". (I can enter .00002 for the fillet, but when I dimension it, I get .00005). The message shows up when I try to enter .00001.

It works OK for me - both as Parasolid & ACIS.

Any parts that are suppressed using "Suppress Parts" rather than "Suppress" will show up in a config-based BOM. When you select an assembly along with either another part or another assembly and Suppress, it actually performs a "Suppress Parts" on all the parts of all selected assemblies - so they still show up in the config-based BOM. The workaround is to suppress each assembly separately (but parts can still be multiple-selected.) Here's my original bug report from last September: http://www.ironcad.com/support/community/i...?showtopic=2995

Right-click on the group, select "Properties..." & enter the desired values (or formulas) in the Anchor tab. Actual values apparently are based on 0,0 in the lower left corner of the group. By default, it appears that the anchor position is based on formulas - e.g. (0.5 * Sizebox\Length).