Lug and terminal defects: cracks, plating, thread and crimp

Four defect families and where they start

A lug or terminal can fail in many ways, but most rejections in production fall into four groups. Cracks come from forming or from the material. Plating defects come from the surface preparation, the bath or the handling. Thread defects come from tapping, forming or gauging. Crimp defects come from the tool, the setup or the conductor.

The grouping matters because the detection method is different for each. A crack may need a bend test or a cross-section, not a glance under a lamp. A plating pore may need a corrosion test or a metallographic section, not an X-ray fluorescence reading. A crimp problem may need a crimp height gauge and a pull test, not a visual check of the barrel.

The other reason to group them is corrective action. Sorting the output removes the symptom for one lot. Changing the bend radius, the bath chemistry, the tap or the crimp tool removes the cause.

Cracks at bends, seams and staked joints

A crack starts where the material is strained beyond its ductility. In a stamped lug, the usual locations are the bend between the palm and the barrel, the seam where the barrel is closed, and the edge of a staked or riveted joint. In a turned lug, a crack can start at a sharp internal corner or a thread run-out.

The material matters. A hard temper, a high-strength alloy or a badly annealed strip will crack at a radius that a softer grade would form without trouble. Grain direction matters too: bending across the rolling direction is harder than bending along it. A strip layout that ignores grain direction can produce a part that passes the first article and cracks in the second coil.

Detection is not always visual. A fine crack at a bend can close up after forming and open only under load. A bend test on a sample, a dye penetrant check on a suspect area or a cross-section through the bend will find it. For a safety-related part, agree the inspection method and the acceptance limit before production, not after the first reject.

Plating: thickness, adhesion, porosity and coverage

Plating defects are often described as one problem, but they are several. Thickness can be low in a recess or high on an edge. Adhesion can fail after a bend or a crimp. Porosity can expose the base metal even when the thickness is correct. Coverage can be missing on an area that was masked or shielded in the bath.

Thickness is measured by X-ray fluorescence on a flat area, or by a coulometric method, or by a cross-section. Each method has a different spot size and a different sensitivity to the underplate. A reading on the palm may not represent the inside of the barrel, which is the area that matters for a crimp.

Adhesion is checked by a bend, a tape test or a thermal shock, depending on the product standard. Porosity is checked by a corrosion test or by a nitric acid exposure on a sample. Coverage is checked visually and by a measurement at the points named on the drawing.

Hydrogen embrittlement is a separate risk for high-strength steel parts and for some plated spring materials. It is not visible, and it may not appear until the part is under load. If the part is high strength and the process includes acid pickling or electroplating, the baking step and the test method belong in the specification.

Thread defects

A threaded lug or terminal can fail in ways that a visual check will not catch. The thread may be too shallow, cut at the wrong angle, stripped, crossed or filled with plating. A go/no-go gauge is the basic check, but it only proves that the gauge fits. It does not prove that the thread will hold the specified torque.

Thread depth is a common argument. A tapped hole in a thin palm may have only a few full threads, and the drawing may call for more than the material can support. A formed thread, rolled rather than cut, can give a stronger thread in the same thickness, but it needs a different hole size and a different gauge.

Plating inside a thread changes the fit. A thick tin deposit can close a fine thread enough to make a go gauge tight. Masking the thread avoids the problem but adds an operation and a risk of bare metal in a corrosive environment. The drawing should say whether the thread is plated, masked or plugged, and the inspection should match.

Crimp defects

A crimp is a controlled deformation, so a good crimp depends on the tool, the die, the setting and the conductor. A bad crimp can be too loose, too tight, out of position, missing strands, or cut into the conductor. The barrel may be bell-mouthed at the wrong end, or the insulation may be trapped in the crimp.

Crimp height and crimp width are the process variables that are usually measured. The target values come from the terminal maker and the tool maker, not from the drawing. A pull test at a defined rate confirms the result. A cross-section shows the strand compression and the voids, which a pull test alone may not reveal.

Strand damage is easy to miss. A few cut strands reduce the cross-section and the pull strength without changing the outside appearance of the crimp. A brush or a strand count at the cut end, plus a periodic cross-section, is a better control than a visual check of the barrel.

The tool is part of the specification. A crimp tool that is worn, mis-set or used with the wrong die will produce a joint that looks acceptable and fails later. Tool calibration, die identification and a first-off check at each setup are the controls that keep the process inside its window.

Building the inspection plan

Start with the drawing and list every characteristic that can affect function. Then choose a method that can actually detect a failure at that characteristic. A visual check is appropriate for a missing hole or a wrong label. It is not appropriate for a crack at a bend or a pore in a plating.

Set the sampling around the risk. A new tool, a new coil or a new plating batch deserves a first-off and a slightly larger sample. A stable process can move to a reduced frequency once the capability is demonstrated. The sampling plan should say what happens when a result is outside the limit, not only how many parts to check.

Record the result in a way that can be traced to the lot. A measurement without the part number, the lot, the gauge and the operator is difficult to use in a corrective action. For a safety-related part, keep the record for the period the customer or the standard requires.

Containment and corrective action

When a defect appears, contain the suspect lots first. Then find the process change that caused it: a new coil, a tool service, a bath top-up, a change of operator or a change of crimp tool. A defect that appears without a process change is often a sign that the process was never fully capable, and the first drift has exposed it.

Corrective action should change something measurable. Retraining alone is weak if the tool can be set incorrectly. Adding a gauge or a fixture is stronger. Changing the drawing to a radius the material can form is stronger still.

Close the loop with the next lot. Check the same characteristic, at the same frequency, and record the result. A corrective action that is not verified on production parts is only a plan.