Impact wrench marketing runs on one number, printed large: nut-busting torque. It is a real measurement, and it does not describe what happens when you tighten a bolt.

The two numbers, and what they mean

Nut-busting torque is the maximum torque the tool can apply to break a fastener loose. It is measured in a controlled setup, usually against a fastener specifically prepared for the test. Loosening a stuck fastener is a peak-force event and this number reflects it.

Fastening torque is what the tool can apply while tightening. It is always lower, sometimes substantially, and it is the number that matters for assembly work.

Both figures also depend on the battery. A tool that reaches its rated numbers with a high-output pack will not get there on a small compact battery, because the peak current is simply not available.

Drive size sets the ballpark

Before torque, look at drive size, because it defines the class of work.

  • Three-eighths drive. Compact and light. Suspension clips, small automotive work, tight spaces.
  • Half-inch drive. The general-purpose size. Lug nuts, most automotive and equipment work, general mechanical.
  • Three-quarter and one inch drive. Heavy equipment, structural bolts, agricultural work.

Within a brand's half-inch lineup there is usually a compact, a mid-torque, and a high-torque model. The mid-torque tool is the most useful single choice for most people: it handles lug nuts and suspension work without the weight and bulk of the high-torque model, which struggles to fit anywhere confined.

Why the real bolt behaves differently

Corrosion, thread condition, thread locker, prior overtightening, and joint stiffness all change the picture. Rust in the threads can require far more torque to break loose than the fastener was ever tightened to. Meanwhile, a clean fastener in a soft joint can be overtightened by a strong tool in a fraction of a second.

Penetrating oil and patience beat brute force on seized fasteners. Heat helps too, when the surrounding materials allow it. Snapping a stud off in a casting turns a ten minute job into a whole afternoon of extraction.

The socket question is a safety question

Use impact-rated sockets. Chrome hand sockets are hardened for a different kind of load and can crack or fragment under impact. Impact sockets are made from a tougher alloy with thicker walls, usually in a black oxide finish.

The same applies to extensions, adapters, and universal joints. Impact-rated versions exist for a reason, and a failed adapter under a high-torque tool sends pieces somewhere.

Wheel work has its own accessory: torque-limiting extensions, sized to a specific torque, which help avoid grossly overtightening lug nuts when you do not have a torque wrench in hand. They are a useful backup, not a replacement for a proper torque wrench.

Final torque needs a torque wrench

An impact wrench cannot deliver a repeatable, specified torque value. The impact mechanism is a series of blows, and the result varies with battery charge, joint condition, and how long you hold the trigger.

Anything with a published specification, and that includes wheels, engine components, structural connections, and most equipment fasteners, gets its final torque set with a calibrated torque wrench following the manufacturer's sequence and value. Overtightened wheel fasteners can distort brake components and stretch studs, and undertightened ones come loose.

Use the impact wrench for what it is good at: removing fasteners quickly and running new ones down to just short of seated. Then switch tools.

Choosing one honestly

Ask what you actually work on. Wheels and suspension: a half-inch mid-torque with a high-output pack. Engine bay and interior work: a three-eighths compact. Heavy equipment and stuck structural bolts: high torque, and expect the size and weight that comes with it.

The advertised nut-busting figure is fine as a rough class indicator across models within a brand. Comparing it across brands, or treating it as a promise about your specific rusted bolt, is where people get disappointed.