How to Make (Almost) Anything | Anders Häggman

WEEK 3 // 3D-Printing

I have several years experience with different CAD software such as Pro/E, Catia, and I-DEAS, although I have not spent significant amounts of time using CAD in the last few years. My all-time favourite CAD program is Pro/E. The last time I used it extensively was when Wildfire 3.0 was the newest version , sometime back in 2007 (before it was rebranded Creo). Although I am not as comfortable with SolidWorks yet, I felt it did not hinder me significantly as the shapes and geometry I created were simple.

I have previously printed parts with the MakerBot Replicator 2, UP! Mini, Stratasys Objet30, and once with the Stratasys Dimension. Based on my previous experience, it was clear to me that there would potentially be problems with clearing out the support material from my part, but I decided to push the limits a bit, and see how it would turn out.

As I had worked with a Dimension once before, I had a general sense for what it was capable of, and what the output quality would be. I also had a cursory look at the material datasheet for the material being printed, and asked Charles Guan for advice on tolerances. However, given more time I would have created a couple of smaller test parts, to test different clearance gaps to find the optimal clearance.

* The School of Architecture at McGill University in Canada has an interesting 3D ice printer, which is probably quite unique in the world. Unfortunately there is only little information on their website, at The Laboratory for Integrated Prototyping and Hybrid Environments . Here is an example of a thin ice shell .

Level of previous experience:
Time taken:
Tools used:
SolidWorks // Stratasys Dimension SST 1200es 3D-printer
Got help from:
-
Main learnings:
Clearance limits for support removal
Rendering with SolidWorks

This is the final 3D-printed part – a battery holder for rechargeable AA and AAA batteries. The support material has not yet fully dissolved, and is still visible in the holes. The part was taken out for a photo but was then put back in the NaOH bath. A few days later, the remaining support material had dissolved on the outer shell, so the AA battery compartments are functional, although as of writing this, the AAA inner shell does not rotate. Also, the tolerances for the battery holes are a bit tight, so in a future iteration, I would increase the diameter of the holes to make battery insertion/removal easier. Ideally, I would have created a few smaller test pieces to test tolerances, but due to lack of time I had to try and get it right the first time.

The rechargeable battery holder I designed. This is a rendering made with the PhotoView 360 rendering tool in SolidWorks. It was incredibly frustrating to use, and google was unable to help me find what I needed. I finally figured it out by reading somewhat unrelated tutorials, and just brute force trial and error. Perhaps the hardest part for me was to figure out how to scale my model so that it looked appropriate in size compared to the background. The way that I finally solved the problem was to 'insert a camera' and then change the camera's position. Although the size of background did not change when moving the camera further, nonintuitively, the size of the part changed (although it's relative position in the background did not). So in essence moving the camera forward and backward scaled the part. Also, I was not able to create this effect with the default camera, but had to first insert a new camera. Moving the part relative to the background was also a ridiculous endeavour and consisted of zooming in and out until the part was roughly where I wanted it to be.

This is an earlier rendering I created. I was not happy with the scale of my product, although the lighting of the product was better. (Although the lack of a shadow makes it seem ungrounded.) Frustratingly, the render preview does not give a realistic representation of the final lighting. It seems to be more of a 'trial-and-error' approach that one needs to take. Also, the surface finish of the product in this rendering I felt was better to the image above. To give a sense of scale, the product holds 9 AAA batteries, and 13 AA batteries. There are three independently rotating parts, the inner cylinder, and two outer covers that rotate independently (you can see the parting line between the outer AA ring, and the inner AAA ring).

When creating the model, I tried to be very cognisant of what dimensions I was creating in the file, so that they reflected the actual dimensions I was interested in, rather than simply trying to define the geometry with dimensions that seemed most convenient at the time. For the inner ring of AAA battery holders, for example, I was interested in being able to easily adjust the inner wall thickness later on, so I defined that dimension, but other dimensions were not defined as numbers, but rather through relations and geometric constraints. That way, at the end, I only had a couple of measurements that I controlled, and the whole model would automatically up-date based on those dimensions. The parameters that I controlled are wall thickness, number of battery holders, and clearance gap between the three different parts of the model. All other dimensions would up-date based on these main parameters.

When creating the rotating top plates for example, the distance of the seam was not defined as a unit of length, but rather as the midpoint (taking into account the clearance gap that would be formed between the two rotating pieces) between the outer edges of the two different rings of battery containers (holes).

A cross-section of the part,with the outer shell missing, and the inner shell not yet having the through hole. The 'outer shell' sits over the outer edge, creating a 'C' shape where the top acts as the lid, and the bottom part goes underneath the main body to hold the outer shell in place.

The final SolidWorks part. The larger holes are for inserting/extracting batteries, and the smaller holes are 'indicators' so that one can see if a battery is in a hole or not (ie. which holes are still empty). They also serve the dual-function of hopefully letting some of the solvent bath into the part to remove the structural support, although I fear this will be minimal, and I anticipate I will have to do some post-cleaning to clear all the holes. As long as the outer and inner shells spin I am willing to call it a success. As mentioned earlier, I have some experience 3D-printing, and in my experience the Makerbot Replicator was decent for what it was, but it is not particularly high resolution or incredibly strong, and although the Objet creates beautiful parts, they tend to be quite brittle. I therefore wanted to print on the Dimension, to get accurate, durable parts. I also did not want to take any chances with the UP! Mini, as our lab has had a terrible experience with our printer that we bought brand new.

Here are some examples of prints that we have made in our lab. The bottoms of these parts (left side, in the picture) are supposed to be totally flat. Look at the second part from the left, for example. A particularly bad print.

An almost complete part.

Part of the problem is certainly that the print does not always stick to the raft. Issues with thermal stresses.

The bottom here is supposed to be flat. The corner on the right should touch the cutting mat, and is clearly incredibly warped.

The actual part being printed in the Dimension printer. It's a little bit exiting to see it grow. (My part is the one under the print head, being printed on as the picture was taken.)

The part stayed overnight in the NaOH bath, but the support structures inside the part had not yet fully dissolved, so I put it back. After a few more days in the bath, most of the remaining support material had dissolved, and the outer shell of the part actually rotates! I was somewhat surprised that the AA battery compartments were as clean as they were considering they only had a very small hole through which to interact with the NaOH solution. The AAA battery compartments also seemed mostly cleared from support material, but the inner shell does not rotate. My guess is that it is held in place by support material around the inner cylinder (with the through-hole in it), as exposure to the NaOH solution would have been very limited for this section. Future design improvements would include making cut-outs on the inner cylinder, similar to those seen here for the outer shell.