Introduction: Why Printing Blocks on Ember Is Difficult
Ember, a SLA 3d printer, is excellent at printing complex, filigreed objects. It does not excel at printing parts that have regions of large cross-sectional area, like large blocks, prisms, spheres, or pyramids.
Do not print blocks on Ember if they can be avoided! If blocks are unavoidable, there are excellent strategies(see Step 4)which lower their cross-sectional area or apply corrective settings.
Large cross-sectional area causes deflections in the printer, which in turn lead to solidified layers that are not the correct thickness. This effect can result in a number of failures like jamming, floating layers, or misplaced layers. Parts with low cross-sectional area, or many small areas that are disconnected, print extremely well. Such lattice-like or filigreed objects are well suited to Ember.
The following instructable will explain this in detail.
Step 1: Understand the Print Cycle
Ember’s print cycle is as follows:
1) The build plate or already-adhered print moves down into the resin.
2) The new layer is solidified by shining blue light on the resin through a window in the bottom of the resin-tank.
3) The resin-tank/window rotates away from the print.
4) The print lifts up.
5) The resin-tank rotates back into position for the next cycle.
Note: the first cured layer adheres to the build-plate and each subsequent solidified layer adheres to the previous layer. The model grows "upside-down," with its base adhered to the underside of the build-head.
Step 2: Squeezed Resin
Each time the build-head moves down (step 1) the resin between the window and the print is “squeezed” into an extremely thin layer. The thickness of that layer is, in fact, the layer-thickness you have chosen. The standard is 25 microns (0.025 mm), which is about the thickness of rice paper. The squeezing action to form such a thin layer of resin can generate significant forces, which cause the entire resin-tank to move downwards. The larger the cross-sectional area* of the print, the larger the forces generated during squeeze, and thus the larger the deflection.
The phenomena is similar to the experience of picking up a wet glass or cup from a table. It takes surprisingly more force to pick the glass straight up from the table than to slide it off the edge. A glass with a large base is going to be harder to pick straight up than a similar glass with a small base (assuming all other attributes equal, like mass). This also explains why the rotation-mechanism to separate a cured layer is necessary; if the print lifted straight-up, the forces involved could destroy the print or even harm the machine, especially for prints with large cross-sectional area.
* The forces generated during squeezing an arbitrary shape vertically down into a fluid are not only related to cross-sectional area, but also the outflow distance that fluid must flow to escape from the interior of the shape to the edges.
Step 3: Failure: Jamming
The effect of the resin tank deflection can cause failure in a number of ways.
1) The layer is too thick, and in the next cycle causes an even greater displacement. The cycle of ever-thicker layers continues until the print is much larger than it should be and it pushes down into the window with so much force that the rotation mechanism jams. Once jamming ensues, the print is likely to fail.
2) The gap caused by the squeeze motion is so large that the cured layer does not adhere to the previously-adhered layers. This floating layer will likely hang-out on the window, and start growing into a big massive clump.
3) Similar to (2), the resin layer is too thick, but in this case the solidified layer only just adheres. Upon rotation this layer becomes displaced, like a deck of cards sliding.
Even without resin tank deflection, a large cross-sectional area requires more torque from the rotation motor, and can thus cause jamming!
This is why printing large blocks and prisms is not a good idea: these shapes have large-cross sectional area and tend to jam the printer.
Step 4: Solutions
So far the best solution to this problem is to not print solid blocks! Even as test objects. There are a host of other manufacturing methods for creating blocks that work quite well. However, it is sometimes unavoidable to have geometry with significant cross-sectional area. Here are some ways to deal with such cases:
So far the best solution for large cross-section layers is adding an “overpress” motion, where 1) the printer moves down to the target layer-thickness, then 2) pushes down even further, squeezing the layer down to the correct thickness despite the deflection of the resin-tank and 3) returns to the target-position and cures. Figuring out how to assign the correct overpress motion based on the geometry of the print has turned out to be quite complex and is currently being investigated. Automatic methods of determining overpress are also being explored.
Another option is making the block-regions in a model hollow, with a lattice infill or ribs to add mechanical stability.
In the case of functional parts, like microfluidic devices or small mechanical parts, there are often ways of achieving the same functionality as a large block, but with less cross-sectional area. For example a block riddled with channels could instead be a network of channels connect by thin "webbing" or ribs. A hefty bracket could instead be a composite of thin plates intersecting at various angles, etc.