Tie-Bar Strain in Injection Molding: Measuring the Clamp Force You Actually Have

What the clamp setpoint does not tell you

On a hydraulic machine the operator enters a clamp pressure or a percentage of maximum tonnage. The controller regulates oil pressure. Neither number is the force that actually holds the two mould halves together, because the conversion from oil pressure to clamp force runs through a toggle geometry or a hydraulic cylinder, a moving platen, four tie bars and a stack of steel that is never perfectly flat.

The gap between the setpoint and the real force is not academic. A toggle clamp delivers almost no force until the linkage approaches straight, so a small change in mould height or a few hundredths of a worn pivot changes the force at the tool by a large fraction. A hydraulic clamp is closer to linear, but the pressure gauge still sits upstream of the seals and the friction in the platen guides.

Then there is distribution. Even when the total force is correct, it does not have to arrive evenly. The projected area of a part is rarely centred on the platen, the cavity pressure is higher near the gate, and a mould that is mounted a few millimetres off centre loads two tie bars more than the other two. The tool sees a tilted clamp, not a uniform squeeze.

This is why flash at one corner is such a common and frustrating symptom. The machine reports the same clamp tonnage it reported last week, the process sheet has not changed, and yet one corner of the parting line opens. The total force may be unchanged while the local force at that corner has fallen by enough to let the cavity pressure win.

Tie-bar strain measurement closes the loop. Each tie bar stretches elastically as the clamp closes, and that stretch is proportional to the load in that bar. Four sensors turn a hydraulic setpoint into four real force readings, which is the only form of the number that can explain a one-sided defect.

How tie-bar strain becomes a force reading

A tie bar under clamp load behaves as a spring. Stress is force divided by the bar cross-section, strain is stress divided by the elastic modulus of the steel, and the measured elongation is strain multiplied by the gauge length. For a typical bar of 80 to 120 mm diameter on a 160 tonne machine, full clamp load produces an elongation of roughly 0.1 to 0.3 mm over a 1 m gauge length.

That is small but well within the range of a bonded strain gauge, a magnetostrictive sensor or a dial-type extensometer. Commercial clamp force systems usually mount one sensor per bar, wired to a display that shows each bar and the sum. The sum is the real clamp tonnage; the spread between bars is the number that diagnoses tooling and machine condition.

Calibration matters. A strain sensor reads elongation, not force, so the system needs the bar diameter, the modulus and the effective free length, and it needs a zero taken with the clamp fully open and the tool at temperature. Warm steel and cold steel give different zero offsets, which is one reason a check done at start-up and a check done after two hours can disagree.

Temperature compensation is not optional on a hot mould. Heat conducts into the platens and the inner ends of the tie bars, producing thermal growth that a strain gauge cannot distinguish from load unless the system compensates for it. Take the zero at the same tool temperature you intend to run, and log that temperature with the reading.

A portable kit is enough for a periodic check. Four magnetic extensometers clamped to the bars, a slow clamp close onto a guarded tool, and a written record of the four values will tell you more than a permanently installed system that nobody reads. Permanent sensors earn their cost on high-cavitation tools and on jobs where flash is expensive.

Why four bars never read the same

A perfectly centred mould on a perfectly aligned machine would load all four bars equally. Real machines do not. Small differences in tie-bar preload, platen parallelism, toggle pivot wear and mould height setting leave a residual pattern that is stable enough to be a fingerprint of the machine and sensitive enough to change when something wears.

The useful measure is the spread, usually stated as the difference between the highest and lowest bar divided by the mean. A well-maintained machine can sit within about 5 to 10 percent. Once the spread climbs past roughly 15 to 20 percent, flash at one corner, uneven venting and inconsistent part thickness start to appear even though the total tonnage looks correct.

The pattern tells you where to look. One high bar with three low ones usually means a preload or height setting problem. Two adjacent bars high and the opposite pair low points to an off-centre mould or a platen that is not parallel. A diagonal split, with opposite corners high, is more often a twisted platen or a machine that is not level on its foundations.

Table 1 - Reading the four-bar strain pattern before changing the process.

Cavity pressure makes the same spread more damaging. At higher injection and hold pressure, the opening force at the parting line rises, so a bar that is 20 percent light reaches the point where it can no longer hold the tool shut. That is why a job that runs cleanly on one material lot can flash on a lot that needs a little more pressure to fill.

What uneven clamp load does to the part

The first visible sign is localised flash. It appears at the same corner or along the same side of the parting line, and it does not move when the operator raises hold pressure or lowers melt temperature, because those changes affect the whole cavity. Somewhere on that side the platen is being allowed to open by a few hundredths of a millimetre.

Venting follows. A parting line that is lightly loaded on one side is also a parting line that is not sealing the vent land. Gas escapes unevenly, burn marks appear at the end of flow on the tight side, and a setter often responds by deepening the vents. That makes the flash worse rather than better.

Dimensional scatter is the expensive consequence. The cavity depth depends on how far the two halves are pressed together at each point, so a tilted clamp produces a thickness gradient across the part and a small but repeatable difference between cavities in a family tool. A capability study then shows a spread that nobody can explain from the process settings.

Tool wear accelerates as well. The mould halves shift slightly relative to each other at every close, rubbing the leader pins, the interlocks and the shut-off faces. Shut-offs that are designed to seal on a small angle of contact are particularly sensitive, and the resulting galling is often mistaken for a lubrication problem.

There is also a machine cost. A toggle that is repeatedly asked for full tonnage while misaligned loads its pivots and its tie-bar nuts unevenly. A clamp balance check is therefore a maintenance task as much as a quality task, and it belongs in the same schedule as platen parallelism and tie-bar nut torque.

Running a clamp force balance check

Start with a clean, empty tool at its normal running temperature and the mould protection set to a safe low pressure. Install the four sensors at the same axial position on each bar, somewhere the bar is free of brackets and covers, and zero them with the clamp fully open. Write down which sensor is on which corner.

Close the clamp slowly and record the four values at 25, 50, 75 and 100 percent of the intended clamp setting. A machine that is healthy scales roughly linearly at all four corners. A corner that lags at low force and then jumps at high force is a sign of a mechanical bind or a toggle that is not extending evenly.

Repeat the measurement with the tool closed but with no injection, then again on a live shot if the system can log fast enough. The difference between the static and the dynamic reading shows how much of the clamp force is being consumed by cavity pressure, which is a useful cross-check on a cavity pressure sensor if one is fitted.

Move the mould, do not just move the settings. If two adjacent bars are high, loosen the clamps and re-centre the mould by the amount the strain pattern suggests, then re-measure. A 5 mm shift on a large tool can change the spread noticeably. Recheck the mould height at the same time, because the two adjustments interact.

If re-centring does not help, check platen parallelism with a dial indicator on the four corners of the closed platens, and check the machine level. Only after those are within the builder's tolerance should you adjust tie-bar preload, and then only following the machine manual, because preload changes the whole clamp geometry.

Table 2 - What belongs on a clamp balance record sheet.

Keep the record with the tool, not with the machine. The same mould on a different press will have a different pattern, and the history is what tells you whether the drift happened in the tool mounting or in the clamp.

Setting a limit and keeping it honest

Set the alarm on spread rather than on total tonnage alone. A total that is 5 percent low is usually harmless; a spread of 20 percent is not. Choose a limit from a baseline taken when the job is known to run cleanly, then tighten it only as far as normal machine-to-machine and temperature variation allow.

Do not use extra clamp tonnage to hide a spread. Raising the setpoint lifts all four bars, but the light corner stays proportionally light, and the extra load increases platen deflection and tool wear. It also pushes a toggle clamp closer to full lock, where a small mould height error does the most damage.

Tie the check to events rather than to the calendar. Re-measure after any mould change, after a platen or tie-bar job, after a crash or a mould protection stop, and whenever flash appears at a repeatable location. Those are the moments when the pattern actually moves.

Where cavity pressure sensing is fitted, use the two measurements together. A rising cavity pressure with a stable strain pattern is a process or material change; a stable cavity pressure with a drifting strain pattern is the machine. Separating those two saves a great deal of process tweaking.

Finally, put the four numbers on the setup sheet alongside mould height, cushion and transfer position. When a job comes back after six months, the baseline is what tells the setter whether the clamp, the tool or the process moved. Without it, the first response to one-sided flash is almost always to change something that was never wrong.