Commissioning Dual Closed Loop Drive Systems on Roll Welders

Installing a roll welder is a mechanical task; commissioning one is a measurement task. The machine arrives with a transformer, a pneumatic force system, a water circuit, a pair of

Nov 16, 2026 - johnmin ren

Installing a roll welder is a mechanical task; commissioning one is a measurement task. The machine arrives with a transformer, a pneumatic force system, a water circuit, a pair of wheel electrodes and a digital controller, and each of those subsystems can be individually correct while the combined process is still wrong. Dual closed-loop feedback control — one loop for motor speed and one for the welding conditions it governs — is the feature that makes long seams repeatable, but only if it is verified under the same load, part geometry and thermal state it will see in production. A structured commissioning sequence turns a new Seam Welding Machine from a collection of settings into a qualified process with a documented baseline that can be returned to after any future service.

Before energising anything, confirm the incoming supply matches the nameplate: single-phase 380 V, 50 Hz, with feeder capacity appropriate to the model's 197 to 420 A input current. Verify protective device ratings, cable cross-section, earthing and the absence of other large intermittent loads on the same branch. Check that compressed air is clean, dry and regulated, and that a gauge is installed at the cylinder so force can be read where it is applied. Confirm cooling water flow at the machine inlet, not merely at the pump; a nominal 120 litres per hour should be measured with the actual hose run and wheel connections in place. Finally, inspect the wheel electrodes for correct profile and secure mounting, and verify that the roughly 20 mm electrode travel is free through its full stroke with no binding or hose interference.

Verifying the Pneumatic Force Chain

With the wheels brought together on a gauge or load cell, step the pressure through the intended operating range and record commanded versus measured force in both directions. Hysteresis in the regulator, sticky cylinder seals or a misaligned electrode holder will show up as different readings on the way up and the way down. Repeat the measurement after a few hundred actuations, because a new cylinder often beds in during the first shifts. High-quality cylinders and pneumatic components are specified for fast response, so also observe how quickly force settles after a step change: a sluggish response leaves the first few centimetres of every seam under-welded. Record the settling time as part of the commissioning baseline; it is a useful early indicator of regulator or valve degradation months later.

Dry Runs Before Current

Run the drive without welding current using a representative part, or a dummy of equivalent mass and inertia. A dual-wheel drive is intended to move large structural components smoothly, so the test should include the heaviest part and the most awkward rotation, not just a flat coupon. Observe the wheels for slip, marking and tracking, and confirm that the part is transported rather than dragged. Where the seam is circumferential, rotate the part through a full revolution and watch for a point where the drive slows or the fixture binds. Any speed disturbance at a specific angular position will later appear as a local burn or a weak spot at exactly that point on every part.

Setting Up and Checking the Speed Loop

The published welding speed range of 0.8 to 3.2 metres per minute should be confirmed by measurement rather than accepted from the display. Mark a known distance on the part or use a tachometer on the wheel, and compare actual travel against the commanded value at the low, middle and high ends of the range. Then repeat the middle setting under load, with the part clamped and the force applied. Dual closed-loop feedback control exists to hold constant motor speed and stable welding speed through these load changes; a properly tuned loop will show almost no difference between the unloaded and loaded readings. If it does, check encoder mounting, drive belt tension, wheel diameter entered in the controller, and any acceleration or deceleration ramps that may be trimming the effective speed on short seams.

First Welds and the Heat Input Ladder

Begin with coupon stock from the production material lot, cleaned the way production parts are cleaned. Choose a mid-range speed, a moderate force and a current well below the expected final value, then increase current in small steps while keeping speed and force fixed. For each step, produce a short seam, section it, etch it, and measure nugget width and penetration. The goal is to find the centre of the usable window — the point where nuggets overlap reliably without expulsion — rather than the first setting that happens to hold together. Because the machine family covers 2+2 mm through 3.5+3.5 mm low-carbon steel, repeat the ladder for each thickness that will be produced; a single universal recipe is rarely the right answer. Keep the rated 50 percent duty cycle in mind while running repeated coupons, and allow the transformer to cool between groups so that thermal state does not quietly shift the results.

Testing the Loop Under Disturbance

A speed loop that only works on a flat coupon is not yet commissioned. Introduce deliberate, realistic disturbances: a change in part mass, a formed edge that raises the wheel, a section where the operator has to guide the part, and a long uninterrupted seam that lets the wheels heat up. Log actual speed and, where the controller permits, actual current during each run. Then compare the etched sections from the disturbed zones with those from undisturbed zones. Consistent nugget geometry across those zones is the practical proof that the closed-loop control is doing its job. If penetration grows in the slow sections, the loop is compensating too late or the current regulator is not coordinated with the speed command.

Cooling Circuit Verification Under Load

Water cooling stabilises wheel geometry and contact resistance, so commissioning should include a thermal soak. Run a production-length seam sequence and measure coolant inlet and outlet temperatures and flow at the start, middle and end. A rising outlet temperature that stabilises is normal; a continuing upward trend or a falling flow points to a restriction, an air lock or an undersized supply. Inspect the wheel faces after the soak for discoloration or softening, and re-measure force after the machine has reached its steady thermal state. Machines are sometimes set up cold and then produce different welds after twenty minutes of production; catching that during commissioning saves a great deal of later troubleshooting.

Controller Recipes, Fault Logs and Safety

Once the window is established, store a named recipe for each part number that records current, speed, force, wheel diameter, dressing state, coolant flow and any pulse or modulation pattern. Back the recipes up outside the machine. Verify that fault logging captures overcurrent, low flow, low air pressure and drive faults, and that each alarm stops the weld in a safe state. Test every guard, emergency stop and two-hand or foot-switch interlock with the machine in a realistic configuration. Confirm that the new-generation digital controller reports the values an operator actually needs during a shift, and that changing a recipe requires deliberate action rather than an accidental keypress.

Acceptance Documentation and the Baseline

The commissioning deliverable is a signed record: service measurements, force calibration curve, speed verification at several setpoints, the heat input ladder with etched section photographs, leak or pressure test results on production-representative parts, thermal soak data, recipe list, and the names of the people who witnessed the tests. Include the wheel diameter and dressing interval at the time of qualification, because both will change. This baseline is what makes future troubleshooting rational. When a seam starts leaking six months later, the team can compare current, force, speed and coolant readings against a known-good state instead of adjusting settings by trial and error. A properly commissioned machine with stable closed-loop drive control should need only small, documented corrections as wheels wear and material lots change.

More Posts