Those who manufacture pumps and valves are familiar with a recurring problem when marking: the part, especially if large or with an articulated geometry, rarely offers a flat surface parallel to the laser’s focal plane. Cast valve bodies, flanged couplings, fittings with pronounced curvatures, steps between a machined section and a rough one, hydraulic blocks with cavities and reliefs: in all these cases, the area to be marked can vary in height by as much as several millimeters from the focus reference. With a conventional two-axis scanning head, the beam remains in focus only within a limited depth of field, generally between 1 and 4 mm depending on the focal length used. Outside this range, the marking loses sharpness, contrast drops and, in the worst cases, the Datamatrix code is neither readable nor conforms to the required grading classes.
The 3D head, or three-axis head, was created to meet this need: to uniformly mark surfaces that are not flat while maintaining proper focus on sloping, curved, stair-step, or three-dimensional geometries in the full sense of the word.
How a three-axis scanning head works
A conventional scanning head moves the laser beam via two galvanometric mirrors, which deflect it along the X and Y axes within a marking area defined by the focus (typically Ø100, Ø160, Ø254 or Ø330 mm in our most popular configurations). The focus is fixed: it is set mechanically at the beginning of the process and remains unchanged throughout the marking.
A 3D head adds, before the two mirrors, a third moving optical element, usually a motorized divergent lens positioned on the optical Z axis. This element dynamically varies the divergence of the beam before it is deflected, modulating the position of the focal plane along the vertical axis in real time. The marking software controls the movement of the lens synchronously with the two X-Y galvanometers, so that the beam remains in focus even on surfaces that diverge from the nominal plane, without the need to mechanically move the part or the entire head.
The same principle also makes it possible to extend the useful depth of field with respect to a 2D head, working on volumes as large as 50 to 100 mm of Z travel within the same marking area, depending on the optical configuration.
Why it is needed on valve bodies and pumps
In hydraulics and fluid power components, the typical cases are different. On large valve bodies, serial number and Datamatrix marking is often done on a cast or machined pad that is not always perfectly parallel to the reference flange, especially when marking is done on blanks. On distribution blocks, the combination of threaded holes, milling operations and rough surfaces creates highly variable Z dimensions in the same marking field. On pumps, cylindrical bodies are frequently machined, where the curvature generates a non-negligible height difference between the center and edges of the marking.
In all these scenarios, the 3D head allows the entire graphic to be marked with uniform quality, avoiding both the typical artifacts of out-of-focus marking (characters that blur, DMC codes with unresolved cells) and the need to fragment the marking into multiple passes with mechanical repositioning of the part.
Even for small valves and fittings, where at first glance a standard head would suffice, the use of the 3D head makes sense when marking over a wide circumferential arc is needed. In this case, the curvature of the workpiece quickly takes the marking at the two ends out of focus: with a 3D head, the entire curved marking is handled directly without resorting to additional rotary axes, with an obvious positive impact on cycle time.
In-line and robotic cell integration
The most remarkable aspect of the 3D head is its flexibility of integration. The fact that focus is handled optically and not mechanically means that the head can be installed in fixed positions, even mounted at the tip of an anthropomorphic robot, and operate on parts of different shapes without requiring complex movements.
A concrete example: a cell for marking large pump bodies may consist of a loading table, a vision system for part recognition and localization, and a six-axis robot that brings the 3-D head to the marking area. Once the robot is positioned, it is the head itself, through driving the three optical axes, that tracks the geometry of the part. Configurations of this type are in operation on castings intended for the automotive and hydraulic industries, with 50 W fiber lasers, integrated vision systems, and cycles that include both marking and code grading according to AIM-DPM and ISO/IEC 15415 standards.
On lines where the dimensional tolerance of the part is significant-which is common on foundry blanks, where differences in height of tenths and sometimes millimeters can occur between one part and another-the 3D head is often combined with a sensor-based autofocus (typically a laser distance sensor) that measures the actual dimension and recalibrates the work plane before marking. It is the combination of autofocus for absolute calibration and 3D head for point-to-point modulation that ensures process repeatability even with dimensionally uneven batches.
When the 3D head is not the most suitable choice
It is worth stating clearly: a 3D head is not always the right answer. If the part is flat, small, and with minimal Z variability, a standard head coupled with a mechanical Z axis is generally simpler and more robust. Even for marking nameplates-a solution that many manufacturers of large valves and pumps prefer instead of marking directly on the body-the 3D head does not benefit, because the surface is flat and the additional optical investment does not translate into a quality gain.
Thus, the 3-D head makes sense when at least one of these conditions is met: effectively three-dimensional surface, local elevation differences greater than the focal depth of field, parts loaded in positions that are not perfectly repeatable, integration on robots with multiple poses, or the need to mark large areas on cylindrical bodies.
Mistakes to avoid in the design phase
Three practical pointers for those considering a 3D head system. First: define from the start the dimension map of the marking area, because it is from that that the useful optical Z stroke is sized and the correct focal length is chosen-short focal lengths (FFL100, FFL160) offer higher energy density but reduced marking area, long focal lengths (FFL254, FFL330) allow covering large pitches but should be checked on the power density required by the material, particularly on brass. Second, carefully evaluate the repeatability of part placement. If the position varies significantly from component to component-as is typically the case on foundry blanks or on bodies loosely loaded on the plane-the 3-D head alone is not sufficient: it must be complemented by an autofocus or vision system that recalibrates the reference before each marking. Third, test in the lab on real parts, not simulations. The response of a stainless steel, cast brass or cast aluminum with the same laser parameters is very different, and an evaluation conducted on actual materials and geometries remains the most reliable way to correctly size power, focus and marking cycle.
In summary
If you are working with valve bodies, pumps or hydraulic blocks characterized by non-planar surfaces, significant elevation changes in the marking area or extensive cylindrical geometries, the 3D head is the technology that allows you to maintain marking uniformity, code readability and low cycle times while avoiding additional mechanical handling. On flat, repeatable parts, a standard head remains the most rational choice.
