You look up the torque spec for a 3/4-10 Grade 8 bolt, find 382 ft-lbs, and tighten it to that number. The joint should be good, right? Maybe. That torque chart assumed a specific friction coefficient, a specific surface condition, and a specific K factor. If your bolt is dry, oiled, waxed, galvanized, or has anti-seize on it, the actual clamp force at that torque value can vary by 50 percent or more.
Bolt torque is really about clamp force. Torque is just the indirect method we use to achieve it. Only about 10 to 15 percent of the torque you apply actually stretches the bolt. The rest is consumed by friction. Any change in friction changes the clamp force at a given torque. This guide explains the relationship and describes practical techniques for getting consistent joints.
K Factor: The Number That Runs Everything
The torque-tension relationship is: T = K × D × F, where T is torque, K is the nut factor, D is the nominal bolt diameter, and F is the target clamp force. For a standard dry steel bolt on steel, K is typically 0.20. That is the number most torque charts assume.
But K varies with lubrication, surface finish, and plating. Dry steel on steel: K = 0.20. Oiled threads: K = 0.15 to 0.18. Cadmium plated: K = 0.12 to 0.16. Hot-dip galvanized: K = 0.20 to 0.25. Anti-seize compound: K = 0.12 to 0.15.
The impact is direct. If the chart says 382 ft-lbs for a dry bolt (K = 0.20) and you apply anti-seize (K = 0.13), the same torque produces 54 percent more clamp force than intended. You might yield the bolt or crush the gasket. Conversely, dry bolts on a galvanized spec give 25 to 35 percent less clamp force.
Torque specs without a stated lubrication condition are incomplete. A responsible spec says "382 ft-lbs, dry, K = 0.20" or states the lubrication condition. If the spec does not state the condition, you need to ask.
T = K × D × F
K = 0.20 (dry steel) is the default assumption.
K = 0.15 (oiled)
K = 0.12–0.15 (anti-seize)
K = 0.20–0.25 (hot-dip galvanized)
Bolt Torque Calculator
Calculate recommended torque values for bolts by size, grade, and lubrication. Covers SAE Grade 2/5/8, ASTM A325/A490, and Metric 8.8/10.9/12.9 with adjustable clamp load percentage and quick reference table.
Lubrication: Why the Same Torque Gives Different Results
About 50 percent of applied torque is consumed by friction under the nut face. Another 35 to 40 percent is thread friction. Only 10 to 15 percent stretches the bolt. Lubrication reduces the friction components, so more torque goes into bolt stretch.
A dry 3/4-10 Grade 8 bolt at 382 ft-lbs (K = 0.20) develops about 30,600 lbs of clamp force. The same bolt with anti-seize (K = 0.13) at 382 ft-lbs develops about 47,000 lbs, near the proof load of 47,450 lbs. You are at the edge of yielding the bolt.
The correct approach: T_new = T_chart × (K_actual / K_chart). If chart says 382 ft-lbs at K = 0.20 and your K is 0.13: T_new = 382 × (0.13 / 0.20) = 248 ft-lbs. Same clamp force, lower torque, because less is wasted on friction.
Consistency is more important than the absolute K value. If every bolt in a pattern has the same lubrication, the clamp force variation is +/- 10%. Mixed conditions can produce +/- 30% variation. Pick a lubrication condition and apply it to every bolt in the joint.
Bolt Torque Calculator
Calculate recommended torque values for bolts by size, grade, and lubrication. Covers SAE Grade 2/5/8, ASTM A325/A490, and Metric 8.8/10.9/12.9 with adjustable clamp load percentage and quick reference table.
Torque-Turn Method: When Torque Alone Is Not Enough
The torque-turn method uses torque to snug the bolt, then applies a specified additional rotation to stretch it a controlled amount. The snug torque brings joint surfaces into contact. The additional rotation elongates the bolt, developing clamp force determined by the bolt's stiffness rather than unpredictable friction.
In structural steel erection (AISC/RCSC), the method requires snugging all bolts, then applying 1/3 turn for bolts up to 4 diameters long, 1/2 turn for 4 to 8 diameters, and 2/3 turn for 8 to 12 diameters.
Torque-turn produces +/- 5 to 15 percent clamp force variation compared to +/- 25 to 30 percent for torque-only. The disadvantage is it takes longer and requires measuring rotation angle.
For critical joints (pressure vessels, engine head bolts), the torque-turn method is almost always preferred. For general assembly, torque-only is adequate with controlled lubrication.
Bolt length ≤ 4D: snug + 1/3 turn (120°)
Bolt length 4D to 8D: snug + 1/2 turn (180°)
Bolt length 8D to 12D: snug + 2/3 turn (240°)
Clamp force variation: +/- 10% vs +/- 25–30% for torque-only.
Joint Relaxation and Re-Torque: Why Bolts Loosen
Even properly torqued bolts lose clamp force over time. Embedment is the most common cause: high contact pressure crushes surface roughness peaks, reducing effective bolt stretch by a few thousandths. This causes 5 to 10 percent clamp force loss in the first few hours.
Gasket creep causes further relaxation. Soft gaskets compress under sustained load, reducing grip length and bolt tension. This is why gasketed flanges almost always require re-torque after initial pressurization and one thermal cycle.
Vibration-induced loosening is different: actual rotation of the nut. Prevailing-torque locknuts, nylon insert nuts, thread-locking adhesive, and safety wire are defenses against vibration loosening. A properly torqued bolt without a locking feature in a vibration environment will eventually come loose.
Re-torque is the simplest defense against relaxation. Torque the bolts, wait for the joint to seat, and re-torque to the original specification. Many industrial procedures specify a 24-hour re-torque for exactly this reason.
Tightening Patterns: Sequence Matters
On multi-bolt joints, tightening bolt #1 to full torque, then #2, then #3 in sequence causes elastic interaction: each subsequent bolt slightly relaxes the previous ones. The last bolt ends up at the highest tension.
The standard practice is a star or cross pattern, snugging all bolts to 30 percent of final torque, then 60 percent, then 100 percent. This distributes load evenly. For critical joints, codes specify the exact pattern and number of passes.
On gasket joints, uneven tightening causes gasket extrusion on the high-load side and leakage on the low-load side. A star pattern with 3 to 4 passes brings the gasket to uniform compression.
After the final pass, do one more check pass at full torque. If any bolt moves, the joint has not settled. Do another full pass. On gasketed joints, 3 to 5 passes before all bolts hold is normal.
Pass 1: 30% of final torque, star pattern
Pass 2: 60% of final torque, star pattern
Pass 3: 100% of final torque, star pattern
Pass 4: Check pass at 100%—any movement means another pass is needed.