Can you jet upwards with inkjet printheads?

Most inkjet systems assume one thing about orientation: the printhead points down, and the drops fall onto the substrate below. A growing set of applications needs the opposite — jetting upward onto a surface above the head. Across several customer feasibility studies in our lab, we have found that upward jetting does work, but it exposes a few problems that downward jetting quietly hides. Here is what we have learned about making it reliable.

Gravity does a lot of unseen work in a conventional inkjet printing. It helps refill the nozzle after each drop, it helps the meniscus settle, and it pulls satellites and mist down toward the substrate and away from the nozzle plate. Turn the head over and every one of those behaviors changes at once.

We get asked whether a given printhead and fluid can jet upward more often than you might expect. Rather than guess, customers bring that question to our lab as a feasibility study, and we characterize what actually happens on the real hardware.

Why jet upward in the first place?

Some parts cannot be flipped. Some processes need material applied to an overhead or underside surface in place. In other cases the product or machine geometry leaves no room for a conventional top-down arrangement. Whatever the reason, the engineering question is the same: will this fluid, in this printhead, jet cleanly against gravity — and how much performance do you give up to do it?

We’ve tested this process and the video below shows what upward jetting looks like from an operator’s vantage point, as well as the JetXpert Dropwatcher view.

Getting stable jetting in the upward orientation

The first thing that changes when you invert a head is startup behavior. When a printhead sits idle and then begins jetting, it takes time for jetting to stabilize. In the upward orientation, we’ve found that latency is often longer than it is for the same fluid jetting downward. Presumably because gravity now is working against nozzle refill instead of helping it, requiring more actuations of the nozzle before a consistent rhythm is established. In our trials, the stabilization time went from essentially instant when jetting downward to as high as ten seconds when jetting upward, depending on the head and the fluid.

We found that recirculating printheads handle this noticeably better than purely gravity-fed designs. The likely reason is that the active recirculation, rather than gravity, governs how fluid refills the nozzle — so flipping the head matters less. Gravity-fed heads still jet upward successfully, but they are more sensitive, and bringing them to a stable state takes more care.

Either way, the levers are the same: meniscus pressure and recirculation settings have to be tuned for the inverted geometry. Settings that hold a stable meniscus pointing down will not necessarily hold it pointing up. Dialing this in is one of the tasks in an upward-jetting feasibility study.

Satellites and mist fall back toward the head

The second challenge is where the debris of drop formation ends up. Satellites and mist are a fact of life in inkjet; the only question is their destination. Jetting downward, most of it lands on the substrate, and only some gets pulled back toward the nozzle plate by turbulent airflow. Jetting upward, that picture changes. Satellites and mist that do not reach the target fall back toward the printhead at a significant rate, settling on the very surface you are trying to keep clean. Left unmanaged, that becomes a reliability problem.

The fix lives in the waveform. Tuning the drive waveform to suppress satellites and mist keeps the nozzle plate cleaner — but push too hard and you shorten the throw distance, which is exactly the dimension upward jetting can least afford to lose.

As a rough sense of the trade-off: if you rein in satellites solely by reducing drive voltage, we had to give up on the order of 1 – 2 m/s of drop velocity. Translating that into a throw-distance change is not straightforward, because it depends on drop size and on how tight your placement-accuracy tolerances are. Larger drops hold up better here, which is the regime we were working in. The practical upshot is that satellite control versus throw distance is specific enough to the fluid and head that it has to be measured, not assumed. This is where a dropwatcher earns its keep: you can watch satellite and ligament behavior directly as you change the waveform, instead of inferring it from print artifacts after the fact.

You’ll see this effect in the image below, which was taken in the downward jetting orientation. There is significant satellite formation and the drop velocity is between 6 – 6.4 m/s.

The below image was taken of the same material, in the upward jetting orientation. To reduce the number of satellites to make this sustainable, we had to reduce the jetting velocity down to 4.8 m/s.

Why this is a feasibility-study question

Upward jetting is a good example of a request where the honest answer is “it depends, let’s test it.” The interactions between fluid properties, printhead architecture, meniscus pressure, recirculation, and waveform are too coupled to predict on paper.

Our lab is set up to run exactly this kind of trial: take a customer’s fluid and a candidate printhead, characterize jetting in the orientation the application actually requires, and report back what is achievable and what the trade-offs are. Because we can prototype and measure in-house, customers get a grounded answer — and often a tuned starting point — that they can proceed forward with confidence in.

We keep the specifics of each project confidential, but the pattern across these studies is consistent. We have tested upward jetting with a broad range of printheads from different manufacturers — recirculating and non-recirculating alike — across different categories of fluids, and jetting performance held up as viable throughout.

Where to go from here

If you are weighing an application that needs to jet upward — or to jet in any non-standard orientation — it’s better to characterize it before you design around it. If that is on your roadmap, our feasibility and drop-analysis work is built for exactly this kind of question. Contact us with your new application ideas!