Making the transition from using a job shop to owning a 3D printer is a major event for an engineering department. Having access to a printer at your own facility brings many benefits to a company, but among the most noticeable is speed. Prototypes that used to take weeks to send out for can now be had in days, or sometimes even hours. This in turn frees up time in the design cycle that can be used to facilitate additional iterations or even early delivery, at least at first.
Eventually expectations begin to shift, schedules tighten, and before you know it things are back to the status quo of nail biting and hand wringing over whether parts will be ready in time for the big design review. The only difference is now instead of calling the job shop three times a day for status updates, you are staring at your workstation trying to find the part orientation or toolpath option that shaves enough time off the build to make the deadline.
While printer settings are important, when every day (or even every hour) counts, there are also ways to adjust part design to increase print speed. Below is one example of how making temporary changes to the CAD file combined with some manual post processing, can save substantial build time on an FDM part.
How Small Design Changes Can Alter 3D Printed Part Build Time
Consider the example part on the right. It’s six inches in diameter and eight inches tall and has a ⅛” wall thickness. It has a flat bottom and curved walls, but because the angle of the walls never fall below 60 degrees to the horizontal, it can be printed without the use of any support material. When processed in Catalyst we find it takes roughly 12 hours 30 minutes to print at 0.013” step height on a Dimension 1200es. Unfortunately this part is incomplete; in addition to the flat bottom and curved walls, it also needs a few holes which is where the problems start.
The final design requires three holes, each ¾” in diameter and six inches above the bottom of the part. By most reasonable definitions this is a very minor change from the original version. The mass of the part has been altered by less than 1% and has technically been reduced. At first glance it would seem that the impact to print time would be small and maybe even positive. A reasonable estimate might be a reduction in print time of five or six minutes because there is now less material that needs to be deposited on the tray. This however is not what ends up happening. Adding those three holes does not decrease the print time. It actually increases it from 12 hours 30 minutes to 19 hours!
Why Did The Time Increase?
A quick look at the toolpath shows why we now need all this extra time. The new part requires support material that extends nearly to the top of the model. Catalyst has identified the holes as being unable to self-support, and created support structures to prevent them from collapsing during the build. While each of the support structures is not very big (they only consume a combined total of 4 in3 of material), there is a disproportionate effect on print time because adding support material to a part adds both time to print the material as well as the time necessary to switch between the model and support toolpaths in every layer where support is present.
In our example adding the holes requires this switching to take place in 490 out of 619 layers. In this part, it is actually the number of layers that contain support toolpaths, rather than the total volume of material, that drives the build time.
Option One: Manually Edit Your Model And Support Toolpaths
Now that we know a full third of the build time for this part is tied up in those three holes, is there anything we can do to get that 19 hours back down to the 12 or 13 hours that this part would have taken without the holes? If this were being printed on a Fortus printer using the Insight toolpath software the answer would be “Yes”.
Insight allows users to manually edit model and support toolpaths. Here a user could elect to simply delete the support toolpath entirely. There would be some risk of a build failure but the most likely bad outcome would be that the tops of the holes droop under the weight of the layers above it. The holes would still be there, but may be misshapen, however this can easily be fixed post print by drilling or reaming the holes back to size. In this example, we are using Catalyst software and a Dimension printer, where those options are not available. However if you are able to make changes to the original CAD file, there is another option, that works just as well if not better.
Option Two: Pilot Cones
One of the best things about Stratasys FDM plastics is their relatively high strength compared to other 3D printed materials. The parts are strong enough that under most circumstances they can even stand up to the forces involved in machining operations. This means that if there are features that you need in an FDM part, but are inconvenient to print (like our three holes), it is possible to exclude them from the print, and then add them back later with via post-machining. This can be done on a milling machine if you have one, but all that is needed here is a drill and some forethought. The problem is not drilling through the plastic, but accurately locating the hole centers on a part that lacks edges and flat surfaces to mark and measure from.
The solution here is to replace each hole with a small “pilot cone”. The cone should have the same 118o angle as the tip of a drill bit and should stop just short of breaking through the inside wall of the part. This results in a feature that can accurately locate the drill bit, yet does not require support material to print. Because there is no support material the print time drops to 12 hours 30 minutes. Also because you have created an inherently self-supporting part rather than over-riding the software and deleting the support structure as would be done in Insight, you are not introducing additional risk of a build failure.
Once the modified part is printed, all that is left to do is secure it in a vice and drill the holes with whatever tools are on hand. Using the pilot cones to align the drill bit will allow for locational accuracy on par with other features on an FDM part. The entire process shouldn’t take more than 30 minutes including cleanup, which makes the total time to build the part about 13 hours. This represents a savings of six hours over printing the part with the holes already in it.
Admittedly this is an extreme example. After accounting for post-printing steps, time savings for real parts are likely to be closer to one hour than the six shown here. That being said, little savings add up. Just like the minute or less it takes to switch between model and support material during a print job can add hours to the build time, savings of an hour per job can add up to days or weeks over the course of an iterative design project. In addition to saving time, these techniques of post-machining features can also have other benefits such as improving surface finish and accuracy; but those are the topics of another post.