recreating a rounded pyramid in moi. 1-20  21-39

 From: telnoi 14 Jul 2008  (1 of 39)
 I'm trying to recreate this shape in moi, however I am having issues rounding the different faces. Any hints, tips etc are very welcome. Cinema 4d Moi attempt Basically, I have no idea how to get the bottom of the pyramid rounded like the c4d example, along with the rounding of all the different "faces". I have tried: incorperating the rounding of the bottom in the curves prior to boolean merge them, however this leaves an angular result and destroys the capability to boolean the entire object. Attachments:

 From: Frenchy Pilou (PILOU) 14 Jul 2008  (2 of 39)
 1782.2 In reply to 1782.1 Maybe Try with a profil + a rail =base bottom view by the top and the function Rail Revolve Here with a simple square so take a rounded square :) Attachments:

 From: telnoi 14 Jul 2008  (3 of 39)
 1782.3 In reply to 1782.2 darn, now why didn't I think of that. Thanks, that should partially solve my issue.

 From: Michael Gibson 14 Jul 2008  (4 of 39)
 1782.4 In reply to 1782.1 Hi telnoi, I think you'll want to create a kind of rounded surface shape for each side, probably sweep is easiest to use for this. If you take 2 curves that cross each other like so: You can then select one of them, run Construct/Sweep, then pick the other curve as the rail path, and that will create this kind of surface: I think that one is a little too rounded, you want something a little flatter than that. But that should give you the idea hopefully - you can control how round it is by the shape of the original curves. If you want curves on all sides, you could take that sheet and then copy it around to all 5 sides and then use Construct/Boolean/Merge to fuse it all together and extract out the solid in the middle of them all. So that's one method, I think another way would be to use the Rail revolve command, I'll try that alternate way in a minute here. (EDIT: looks like Pilou has already shown the rail revolve method, thanks Pilou!) Also I guess you could just use Revolve instead of sweep to build a rounded cap surface similar to the above. Once you have a rounded cap surface like that, you can use Boolean difference to slice a solid with it to leave the rounded imprint in the solid. - Michael Attachments:

 From: telnoi 14 Jul 2008  (5 of 39)
 thank you both, learned yet another trick. I'll post the result and source file when I'm done (it's 1:19 over here, time to sleep).

 From: Brian (BWTR) 14 Jul 2008  (6 of 39)
 1782.6 In reply to 1782.5 My effort usimg Pilous way. Brian EDITED: 31 Dec 2008 by BWTR

 From: telnoi 14 Jul 2008  (7 of 39)
 1782.7 In reply to 1782.6 one remaining issue. How do I boolean the remaining sharp edges (the ones selected in the image below)? I presume i have to simplify the object somehow. Source file has been attached. Attachments:

 From: Michael Gibson 14 Jul 2008  (8 of 39)
 1782.8 In reply to 1782.7 Hi telnoi - normally you would use Fillet for that, but in this case the structure of the object is interfering with the filleting. Filleting is a somewhat cranky function, the more unnecessary edges and pieces that you have in a model will tend to make fillet failures more likely. In this case you've ended up with each face being made up of 4 different fragments instead of one big surface - that fragmentation is not good for filleting. It looks like this was made with the sweeping method? When you set that up, I would recommend mirroring the curves and then joining the 2 halves together and actually delete the point where they join to make one single long smooth segment, instead of 2 segments. That will help to make larger single sheets in the final result. I'll see if I can make an example. - Michael

 From: Michael Gibson 14 Jul 2008  (9 of 39)
 1782.9 In reply to 1782.8 Basically in the surfaces there you are inheriting the segmentation of the original curves, so if you can fuse those into single segments, you'll instead get single big surface sheets which are better for filleting at the end.

 From: telnoi 14 Jul 2008  (10 of 39)
 1782.10 In reply to 1782.8 Hey, Yes, I used the sweep method since that was the only way to reproduce the shape I wanted. I didn't use two joined curves however. What I did was draw a single line, add a point in the center and drag it down dead center to create the curvature of the base spline. I copied and rotated that and constructed the solid with sweep. An example would be very helpful. Thank you for your time.

 From: telnoi 14 Jul 2008  (11 of 39)
 1782.11 In reply to 1782.10 here's the spline curve I used. In this case i'm not sure how I would maintain the curvature while deleting center points. Attachments:

 From: Michael Gibson 14 Jul 2008  (12 of 39)
 1782.12 In reply to 1782.10 Hi telnoi, I've attached a sample here, with this set up as larger surface sheets it will make the filleter a lot happier, you should then be able to select the edges of this one and fillet it. - Michael Attachments:

 From: telnoi 14 Jul 2008  (13 of 39)
 1782.13 In reply to 1782.12 thank you, I still have to get used to the idea that points in certain places will create more edges :(

 From: Michael Gibson 14 Jul 2008  (14 of 39)

 From: Michael Gibson 14 Jul 2008  (15 of 39)
 1782.15 In reply to 1782.13 Hi telnoi, > thank you, I still have to get used to the idea that points in certain > places will create more edges :( It's not so much the points themselves (unless you used Add Pt with the "Make corner point" option), it is when you use Edit/Join to glue 2 segments together into one composite curve object. With that curve being made up of 2 segments, many things will process it in a segment-by-segment basis, same thing as if you had a rectangle made up of 4 segments and extruded it, you will get edges at the ends of the line segments making up the rectangle. The same basic thing is happening here due to the segmentation even though the shapes are smooth to one another. It could be possible for me to automatically fuse segments together in cases like this that have smooth pieces to one another, but unfortunately although that would have been better for this particular case, it tends to cause problems in a lot of other cases for mechanical objects, I was describing that some in this previous post here. So to avoid those problems it tends to make things lean more towards segment-by-segment processing. I definitely want to make some better tools for making single segment symmetrical stuff like this easier though. - Michael

 From: telnoi 14 Jul 2008  (16 of 39)
 1782.16 In reply to 1782.15 > I definitely want to make some better tools for making single segment symmetrical stuff like this easier though That would most certainly be welcome. However, learning to construct models correctly in the first place seems to be a better learning method ;). One last question related to this shape. Would it be possible to construct a base shape which would allow the construction of a 4 sided pyramid? In this case, the base needs 3 edges instead of 4, complicating the matter somewhat. The curvature of the base shape is correct, but the amount of edges is not. EDIT: never mind, managed to figure it out by intersecting a sphere with the old shape and using that base to construct the entire pyramid. EDITED: 15 Jul 2008 by TELNOI Attachments: Image Attachments:

 From: manz 15 Jul 2008  (17 of 39)
 1782.17 In reply to 1782.16 Hi telnoi, You look like you are making it difficult for yourself. Pilou put you on the right track by showing you to use the rail revolve, but, I do not think the intention was for you to produce the shape shown and then attempt to fillet. If you progress from what was shown, you can produce both pyramids with a rail revolve without a need to fillet etc. EDITED: 3 Aug 2009 by MANZ