Torque Wrench Calibration Explained

Torque wrenches are one of the most commonly calibrated tools in any Singapore workshop — and one of the most commonly misunderstood. The difference between "calibration" and "indication check," the confusion between ASME and ISO 6789, and the silent drift that happens to click-type wrenches long before they look worn all combine to create a situation where a lot of so-called calibrated torque wrenches aren't actually fit for use. Here's what to know.

Click-Type vs Indicating Torque Wrenches

Before the standards, know your tool. Most torque wrenches fall into one of two categories:

• Click-type (setting) wrenches — you set the desired torque on a scale, and the wrench "clicks" when that value is reached. Simple, popular on assembly lines, but the internal spring fatigues over time and the scale drifts before anything visible happens.
• Indicating wrenches — dial or digital read-out showing the actual applied torque in real time. More expensive, more accurate, and easier to calibrate because you read the value directly rather than relying on a mechanism.

Click-type wrenches require more frequent calibration because their mechanism is what drifts. Indicating wrenches are more stable but still need periodic checks.

The Two Standards You'll Hear About

ISO 6789:2017 — The international standard

ISO 6789 is the internationally recognised standard for hand torque tools. It was updated in 2017 and is now split into two parts:

• ISO 6789-1:2017 — specifies the requirements that torque tools must meet: accuracy, design classes, marking requirements. This is the "conformance testing" part that covers new tools and verification.
• ISO 6789-2:2017 — specifies the calibration method for torque tools, including test points, number of measurements, uncertainty calculation, and what must appear on the certificate.

The critical distinction: ISO 6789-1 is the specification, ISO 6789-2 is the calibration. An accredited calibration should explicitly reference ISO 6789-2 and follow its methodology — not just do a quick indication check and call it "calibrated."

ASME B107.300 — The US equivalent

ASME B107.300 (formerly B107.14) is the American standard for hand torque tools. It defines similar requirements to ISO 6789-1, with slightly different tolerances and test conditions. In most Singapore workshops, ISO 6789 is the default reference, but some US-owned facilities or US-market work may specify ASME B107.300 on tool procurement specs or calibration requests.

Practically speaking, a calibration lab accredited to perform torque calibration under ISO/IEC 17025 can typically calibrate to either standard on request. If you have US-sourced tools or work under US contracts, ask specifically for ASME compliance — don't assume.

ISO 6789 Tool Types and Classes

ISO 6789-1:2017 splits hand torque tools into two fundamental types and several sub-classes based on construction:

• Type I — Indicating torque tools. These display the actual applied torque in real time, either on a mechanical dial/scale (Classes A–C) or electronically (Classes D and E). An indicating wrench or a digital torque meter falls here.
• Type II — Setting torque tools. These are pre-set to a target torque and signal when it's reached — the familiar click-type wrench. Classes A through G cover different sub-designs, from adjustable graduated scale (Class A) to fixed-value pre-set (Class D) to various scale/display combinations.

The permissible deviation under ISO 6789-1:2017 depends on both the tool's class and its nominal torque range — typically ±4% or ±6%, with the tighter figure usually applying to higher-torque ranges. The exact values are set out in Tables 3 and 4 of the standard. What the class actually determines is the construction of the tool and, by extension, the number of measurements and uncertainty components the lab must evaluate during calibration.

The practical implication: any tool rated to ±4% should be calibrated with a lab uncertainty small enough to give a Test Uncertainty Ratio of at least 4:1. A lab with a published CMC that's marginal for your tolerance will produce technically compliant certificates that are useless for making pass/fail decisions close to the limit.

What a Proper ISO 6789-2 Calibration Includes

1. Visual inspection and exercise cycle. Inspect for mechanical damage, then apply at least five loads at maximum capacity in each direction to condition the mechanism before measurement begins.
2. Measurement at three defined test points. ISO 6789-2 specifies measurement at the lowest value of the marked torque range, 60% of maximum, and 100% of maximum — not at a single full-scale point.
3. Multiple readings per test point. The standard requires enough repeat measurements to evaluate repeatability and reproducibility; the exact count depends on the tool type and class.
4. Uncertainty budget per the GUM. The lab calculates the relative measurement error and builds a full uncertainty budget covering the reference device, geometric effects, repeatability, and reproducibility.
5. Comparison with permissible deviation. The deviation at each point is compared against the maximum permissible value from ISO 6789-1 (±4% or ±6% depending on the nominal torque range).
6. Certificate issue. Showing as-found values, any adjustment performed, as-left values, expanded uncertainty with coverage factor, and a clear statement of conformance or non-conformance.

How Often to Calibrate a Torque Wrench

ISO 6789-2:2017 states that if the user doesn't operate a control procedure, a default recalibration interval of 12 months or 5,000 cycles (whichever comes first) may be used, starting from first operation of the tool. In high-cycle Singapore production environments — engine assembly lines, automotive shops, structural fastener work — 5,000 cycles can come in a matter of months, which is why many aerospace and precision manufacturing facilities here run 6-month intervals on production tools rather than the annual default.

The standard also requires immediate recalibration if the tool has been subjected to an overload above the limit specified in ISO 6789-1, after repair, or after any improper handling that could affect torque accuracy. Damage from drops, using a torque wrench as a breaker bar, or loading it beyond its rated range all fall in this category — and all are far more common than most workshops admit.

Storage and Handling That Affects Calibration

• Click-type wrenches should be wound down to their lowest setting before storage. Leaving them loaded at high torque accelerates spring fatigue.
• Never exceed the tool's rated range — a 50 Nm wrench used at 80 Nm is damaged, not "stressed."
• Don't use a torque wrench as a breaker bar to loosen fasteners. The shock loads permanently deform the mechanism.
• Store in original case where possible. Singapore humidity corrodes exposed steel parts — a wrench left on a bench rusts faster than you'd expect.

UT Metrology's Torque Calibration Scope

Our SAC-SINGLAS accredited scope covers torque calibration across several instrument types: torque wrenches per ASME and ISO 6789 (0.4 to 1500 Nm), torque drivers (0.1 to 14.1 Nm), and torque meters and indicating devices. All calibrations follow ISO 6789-2:2017 methodology — multiple test points, full uncertainty calculation, and a certificate that stands up to AS9100, ISO 9001, and regulatory audit. View our mechanical calibration page for the full scope or send us your wrench list for a quote.