Maximising digital printing performance

Digital printing has quietly edged its way into every part of our lives. The parcels we receive almost daily have uniquely printed, on-demand labels; the trainers we wear have their brands printed on them; and you can even print car parts using three-dimensional (3D) printers. The digital printing industry has many branches and technologies within it, but all are driven by the need to deliver repeatable, high-quality results and to reduce consumable waste.

As well as being able to control all parameters that bear on the quality of its output, a well-designed digital printer should be able to detect output-quality issues and, when required, be capable of aborting lengthy print jobs to prevent wasting valuable production time and materials.

This may seem like a blindingly obvious point to make. But the fact is, there are hundreds of thousands of digital printing systems out there – inject, drop-on-demand, thermoplastic, continuous, laser printers and a vast array of multi-axis 3D printing technologies –that lack any feedback or closed-loop control. These products are normally built to be low-cost to the customer, so some printer designs have never even considered feeding data back into the printer or to the manufacturer to learn how their printers fail or degrade with time. Without feedback, there is no possibility of introducing decision-making based on quality and repeatability over the lifetime of the product. Consequently, long-term production risk and material inefficiencies are transferred to the end user.

Thankfully, the digital printing industry is shifting its awareness to account for production yield and consumable waste. This shift is linked to more businesses attempting to reduce their carbon footprint and assess the environmental impact of their supplier and logistics networks. Environmental standards, such as ISO14001, encourage measuring and improving the production efficiency of your business. As a result, businesses now expect newer digital printers to report yield statistics, failures and consumables used. Doing this properly requires the printers to audit their own work.

Even though so many printers are not self-aware, designing a printer with real-time qualitative measurements is possible for both the start-ups wanting to enter the market and for established businesses that already have printers in widespread use. Thankfully this doesn’t require the wholesale re-engineering of established product lines, but it does require a willingness to better understand your printers even more and an eagerness to measure what quality means to the customer!

In my experience, there are two distinct but necessary steps to achieve this objective. Firstly, there’s a need to capture and understand all calibration parameters that govern the quality of the printer itself. This first step ensures you know as much as possible about your printer, and its assembly, before it leaves your factory floor. As you identify the parameters that affect your printer it will become clear how certain parameters are related and how some influence quality more than you might expect. Secondly, you can look beyond the printer itself and focus on the quality of what it produces (your paper print or 3D model). In this step you want to recognise what represents a quality failure: this is what you’ll use to close the loop on your system.

# Capturing the calibration parameters

This first step aims to get a full understanding of your printer and to remove any assumptions about the actual tolerances of its construction. Once done properly, you will end up with a highly configurable printing system (great for auditing against) and a well-proven production process. Key parameters worth capturing include:

## Consumable sensitivity

Consumables are often tightly linked to specific printer parameters that get loaded into a printer before or at the time of use. Example settings for ink might include a different drive frequency linked to differences in pigmentation loading, for resins it might be pump speeds or cure speeds linked to the liquid’s viscosity, or in the case of toner/powder-based printers you might adjust laser intensity based on particle size. These parameters are usually arrived at in ideal lab conditions. It is therefore vital to adjust the parameters of the printer according to your end customers’ environment and results collected in the lab. Only with both can the same high-quality results be guaranteed. The consumables themselves also need to be tracked on a per batch basis and a per job basis, to allow a system to reject a consumable based on contamination or expiry dates.

## Electromechanical tolerances

The tolerance of mechanical and electrical components of are subject to subtle variations. Measuring and having control of these in production is key to delivering consistent results. Building systems and software with integrated calibration routines helps to ensure you can deliver consistent results on a per unit basis.

## Environmental conditions

It is likely that environmental conditions will influence the quality and success rate of a printing system. The environment affects both the consumables and the highly sensitive electromechanical parts of the system. Factors like temperature and humidity are the most common conditions to control, but for some systems, this can go as far as air pressure and oxygen levels and surrounding vibration levels. If factors affect your systems, they can be sensed and managed as part of per job environmental controls.

## Production tolerances

During the production and assembly of the printers, it is vital to make assembly and measurement equipment to measure the actual variation between your units and assembly batches. These variations need to be captured on a per batch and per unit basis.

## Incorporate findings from life testing

Qualitative measurements taken during long-life testing of a printer are a key source of unaccounted quality errors as the printer and its components age, wear, loosen, stretch and degrade. Identifying and matching those variables that connect to quality is key to making a printer that also re-tunes itself as it, and its sub-systems, age from continuous use. This should be done using accelerated life tests on a per model basis.

# Sensing Changes in Quality

As a second step, the focus should shift away from the printer, to validating the connection between the mechanical and electrical steps it carried out and the resulting print quality. This can be done as a series of experiments which can later be engineered back into the printer to perform these assessments automatically. I have found it best to carry out this testing without any consumables (a ‘dry test’) during the product life test. Monitoring the behaviour of the printer over an extended time, gives a ‘baseline’ or ‘golden reference’. I have also used this as an opportunity to work closely with customers who have identified specific quality issues and get them involved with the study. In general, it is customers that report the outliers once a product is launched, and they can be very helpful when connecting the dots around the circumstances of the quality failure.
For the experiments themselves, I have integrated third-party data acquisition systems into the printer, before designing a custom test system that can be bolted onto a printer solution. The testing aims to correlate the times that quality issues arise (surface deformities, dosing issues, spatter) during the operation of the printer, and match this to the conditions to which the printer and its consumables were subjected. For me, the following steps have been most useful when trying to identify quality failures:

## Measure drive signals

In the case of a piezo driver, monitoring drive voltage might show you are not reaching the ideal drive frequency or voltage you would expect. This might correlate to an unexpected ink viscosity caused by a particular batch of ink and a lower-than-normal air temperature.

## Validate time-critical functions

If a printer requires exact exposure times, connect a sensor to monitor the actual exposure time rather than trust the internal timer of your system.

## Monitor consumables

Always log the consumables. Monitor ink temperature, monitor the volume used, and monitor the dosing system that supplies your consumables.

## Environmental monitoring

Monitor the external environment. For example, failures can arise from fluctuations in air temperature at different times of the day.

## Consider secondary indicators

If your control systems are designed to hit target voltages and frequencies, consider monitoring the current of the drive circuit to see if a current spike or drop matches the time of a quality failure. It’s a good indication that your control system is working well but other downstream electronics are working at capacity. If your printer has many moving parts, measure the forces applied by motors. It may be that an unusually high force profile is a precursor to a failure later down the line.

With some data insights gathered from sensors and output quality, focus can move to provide the missing inputs back into your printer. This will allow the printer to recognise a quality failure and abort poor-quality jobs early.

# Conclusion

We live in such a fast-paced world where production continues 24/7 and we need to meet our customers’ expectations all the time. This means there is not much room for error, even if our customers are pushing the boundaries of the printer’s capabilities. To achieve this, we need to design products that can detect and act when quality errors occur, to avoid wasting valuable production time and consumables.

If you are a printer designer looking to increase yields and reduce consumable waste for your customers, I hope you’ve found food for thought here. Going through this process proved very effective. It promoted direct customer engagement and prompted a reassessment how printers were tested, assembled, and calibrated, resulting in the resolution of critical quality issues. Insights gleaned were used by aftersales support and also helped set the requirements for the next generation of printers.