Modeling Challenge: Silver Spoon
This one is a little easier than the shoe, or you can make a big study of it if you want. One thing I have learned about complex modeling is that as you add detail, the modeling time grows exponentially. You can do the main shape of the spoon in a single sweep feature using 2 sketches for a projected curve, a path sketch and a profile sketch. Or you can get really involved and put a couple of hours into adding some fluting, thickness, fillets and troubleshooting a degeneracy.
How would you go about creating this part? I used a little flourish at the handle end that Mark Biasotti used on a similar model, but it shows some nice techniques and introduces some awkward complexities. I encourage you to send in models, and I’ll post them here for other folks to learn from. I’ll do the same as in other challenges, where I have a look at models and make comments.
The basic shape of the spoon follows a bit of a formula, which it turns out can serve you for a wide range of modeling. If you need a hand getting started, here’s how I made the inital basic shape:
- I started with a side profile. This is the shape at the mid plane, the shape you’d get if you cut the spoon in half along the length.
- Next I created one side of the outer edge. This is a 3D edge, but it is easy to create from a projected curve. Draw the edge as you see it from the right, then draw it as you see it from the top, and make the curve. The next steps are easier if you plan for a spoon with a square ended handle.
- The main trick here is this next
step. You have the center plane and one side. Curve features cannot be mirrored, so you’ve got to get symmetry some other way. Theway I did it was with a 3 point spline, where the midpoint is pierced by the center plane sketch, one end is pierced by the projected curve, and a centerline and Symmetry sketch relation control the other end. - Create a sweep. Use “keep normal constant”. The center plane sketch is the path, the projected curve is the guide curve. The 3 pt spline is the profile.
Ok, there’s more to it than that. In particular, getting the projected curve right means using some controls on the splines used for the sketches. I’ll hand out some more hints as we go. Although the main shape can be done in a single feature, this part is actually more difficult than you might think. There are a lot of ways to create this, but in the sample part I used a sweep. Think about how to use a sweep to create the flip in convexity (the bowl is concave up, the handle is concave down).
Click on the picture of the spoon above to download my SW09 version of the spoon. Remember, simplify if you don’t want to get involved in the messy details. Let’s see what you come up with.
=====================================
Entry #1 is from James Canney. James produced the projected curve by making two extruded surfaces and using the intersection curve between the extrudes. Then he drew the profile at the centerplane and made a boundary between the intersection curve and the profile, thickened, filleted and mirrored.
The truth is I like the intersection curve better than the projected curve. First, it is a sketch instead of a curve so it gives you more options. Second, the curve features are said to be less accurate than edges. With that, maybe doing a trim between the surfaces and using the edge would be even better than the sketch. Not sure about the accuracy (tolerance) of intersection curves (3D sketches) compared to other 3D curve types.
Also, I should mention, nice use of reference surfaces here. Surfacing is not usually the end product, often it is just a tool to get the geometry you need.
=====================================
Entry #2 is from R. Paul Waddington, an AutoCAD reseller from Australia. I’m a bit surprised that Acad does this much surface work. With the new parametrics in AutoCAD and the Inventor name change (from Autodesk Inventor to AutoCAD Inventor), maybe we are going to see more 3D tools in an old 2D product. We don’t have a feature list here so we don’t know how he did this.
=====================================
Entry #3 is from Garrett Brooks. Garrett did a wooden spoon, curved on the top to add a bit of challenge. Garrett made good use of the “model the wireframe” concept, modeling curves in perpendicular direction for both Boundary and Fill surfaces.
=====================================
Entry #4 is from Mark Kaiser. It looks like Mark discovered that you can’t make a sweep go to a point on both ends, so he made 2 sweeps back to back. He mentioned that he could not thicken this model unless the end is trimmed off. This is because of the “degenerate” point caused when a feature goes down to a single point, and the normal 4 sided mesh collapses to 3 sides (one side is zero length).
This part also makes nice use of reusing curve data. Mark Biasotti recently showed an example of being able to use handles to pull back the extent of a sweep similar to the way Boundary works. This will make Sweep more useful, and give users more options.
=====================================
Entry #5 is also from Mark Kaiser. Mark used a different approach on this one. This one was lofted as half of the spoon, mirrored, the filleted to smooth the non-tangency at the plane of symmetry, then thickened. Mark could have avoided the fillet if he had used a start/end constraint to make the loft go normal to the centerplane sketch profile.
Notice that in Marks two parts, he has edges that go in opposite directions. It is good to be able to think about parts in different directions. Often turning a problem 90 degrees in your mind will help you come up with different solutions that may get you out of a modeling tight spot.
=====================================
Entry #6 is from Mark Landsaat. Mark’s model uses a different approach, making the bowl and handle separately, making them solid, then taking them back to surface models with the Delete Face commands, smoothing the transition between them, and finally knitting them back to a solid again. Often features are easier in either solid or surface, so switching back and forth like this can become necessary. You need to be careful, though, because if you get in the habit of moving back and forth frequently, you may be doing work that is unnecessary, or at least increasing rebuild times unnecessarily.
=====================================
Entry #7 is from Kyle Stellpflug. Kyle has to get the award for “over the top” this time. Kyle used the spoon as an excuse to model a toy airplane. The entire model was 92 features, and the spoon was about 9 of that. Kyle used projected curves to set up a loft. He had to make 2 projected curves because this type of curve cannot be mirrored in SolidWorks. Nicely done.
You may also want to check out some of Kyle’s technique on the plane. An upcoming challenge will tackle some aircraft modeling, so maybe you can take some ideas from Kyle’s work here.
=====================================
Entry #8 is from SolidWorm. This should have been higher in the list, but I lost track of it while changing domain hosts. SolidWorm did a nice job on this part. In fact, I would go as far as to say that SolidWorm “gets it” when it comes to this kind of work. He starts with the bowl with a Boundary made with the X type of sketch arrangement, which is one of the coolest uses of Boundary. You just can’t do that with other surface types. Then he trims to shape the bowl, and uses a Fill to make the blend between the handle and the bowl. Very nicely done. I encourage you to download and study the model. I think it represents nice workflow and technique.
Surface modeling is very different from solid modeling. The way you go about building a part uses a completely different method of thinking.
=====================================
Entry #9 comes from Jason Knox. Jason did his entire spoon in a single feature. It’s nice to have that end to end continuity. But he said that he could only thicken one side, not the other, couldn’t get it to accept a fillet, and that the wrinkle in the middle was unintentional.
I like what Jason did with the boundary surface. He establishes the two outer curves and the mid plane curve (for direction 1), then sketches in several sections for direction 2. Nicely done. To avoid the wrinkle, I would have tried either more direction 2 curves in that area or try to add a connector in that area. Remember that connectors can be used like quick and dirty profiles, but mainly used to control the flow of the U-V grid rather than the shape.
As for the fillet and the thicken problems, I only have one thing to say – “degeneracy”.
See how all of the lines come down to points at the ends? These points are actually zero length sides, and cause a lot of local problems with shells, thickens, fillets, offsets, and other stuff like that. Everything coming down to a point makes little wrinkles that cannot be dealt with by offsets or other features. If you look at the part I did, you notice that the tip of the spoon was cut off and recreated with a Fill surface. That’s because the Fill surface makes a 4 sided patch and trims it to fit. I did this because of the same problem that Jason ran into here.
Degeneracies are a very common problem. Sometimes they don’t cause problems, and sometimes they do. They are best to avoid when you can, and you must know how to get rid of them when you need to. Even in simple prismatic parts with fillets, you know those 3 sided fillets where 3 filleted edges intersect? Those frequently fail on imported models.
Anyway, much more reading on this topic in the Surfacing Bible.