Stuck On You

The most important part of any 3D print is the first layer. If the first layer doesn’t stick to the build plate well, you will never get a good print. If you actually make through a print with a dodgy first layer, you usually won’t be happy with the look of the print and it can be structurally compromised as well.

So, how to get it to stick? Well, the most important method is to have the print head the correct height above the print bed at all points. This is known as bed leveling. In the traditional method for Cartesian printers, you put a piece of paper under the print head and lower the head just until the head just grabs the paper and there is a slight bit of resistance when you tug on the paper as seen below:

If you were to cut a well-printed 3D printed piece in half and looked at each layer, you would notice that rather than round filaments stacked on top of each other, you would see flat oval shapes:


Looking at the picture above, it is obvious that the flat oval shape allows more surface area of each layer to touch, resulting in much better adhesion not only of each layer to the one before it, but of the first layer to the bed. Since more surface area is a good thing, most 3D prints are set to extrude more material on the first layer to have as much contact as possible between the filament and the bed – often 150% of the set layer height.

Once you have the Z-home height set in one corner, you repeat the process in all four corners and adjust the bed height screws until it is at the same height in all places. There are ways to automate this process using Force Sensing Resistors, microswitches mounted to the print head or proximity sensors where the controller takes measurements before the print. It uses these measurements to take the minute height variations into account and adjusts the z-height on the fly during the print. I have not chosen to incorporate that yet, but there are modifications I can make at a later date should I choose to.

So once you have the bed leveled and the extruder pushing out the right amount of plastic, you still might have adhesion issues. The next step would be to lay down more plastic around the perimeter or even lay down a few solid layers. The first is called a skirt (if the perimeter does not touch the actual model) or brim (if the perimeter does touch the model) and the latter is called a raft.

Another method to combat layer adhesion issues is to heat the bed surface. Plastic tends to be tacky when it is slightly heated and will grip the build surface better. Most 8″ and smaller printers use what is essentially a circuit board that is just one long trace of copper. For this printer I am using a silicone heating mat. The mat makes direct contact with the build surface, is flexible and heats up faster, especially when run at 24v like it will be with FrankenCore.

Sometimes even with good heating and a well-adjusted printer, you just cannot get the object to stick to the print bed. At that point you will generally turn to an adhesive to make the bed surface more tacky. I have used everything on the print bed from glue sticks to Aqua Net Extra Super Hold hair spray to slurries such as “ABS Juice” (a combination of acetone and pieces of ABS) or “PVA Juice” (standard white glue thinned with water) to blue painter’s tape. All offered varying degrees of success, but I couldn’t get anything consistent across multiple materials.

Earlier this year I was introduced to Polyetherimide or PEI. A thermoplastic sheet permanently attached to the build surface with a 3M 468P adhesive sheet, it is an excellent surface. I have been able to print both PLA and ABS with almost no warping or lifting of the print and no surface preparation, and when the bed is cool, I have had no problem removing the prints from the bed. I am using it on FrankenPrusa (that is the amber color visible in the video above) and it will definitely be going on FrankenCore’s build surface.

Build Log

Not much progress this week due to another error on my part. When counting up the number of bearings I would need for the belt paths, I left out two idler arms that are used to adjust the belt tension, so wound up 4 bearings short. Unfortunately, no one local carries these particular bearings, so further building will be delayed until afte Monday the 2nd. However, I was able to assemble the two hotends so there is still some movement forward.

Left to do:

  • Insert the idler arms
  • Run the X/Y belts
  • Modify the RAMPS board to take 24v
  • Mount the Extruder motors and hotend to the carriage
  • Run the wiring to the RAMPS board and hook up the 24v power supply
  • Upload controller code to the RAMPS board (known as flashing the firmware)
  • Fix any issues with the wiring or firmware then try to start printing!