Where the load actually goes
The pipe body carries tension, torsion and internal pressure reasonably well. It is a uniform tube with a uniform section, and it has been quenched and tempered full length for exactly that job.
The tool joint is a different problem. It has to:
- carry the full string weight plus pullback in tension,
- transmit full rotary torque,
- seal against internal pressure,
- and be broken out and made up again several hundred times.
Every one of those loads concentrates at a change in section — the transition from tube to joint. That transition is a stress concentration by definition, and stress concentration is what fatigue cracks grow from.
The three things that start a crack
1. Stress concentration at the geometry change. The tube-to-joint transition is the sharpest change in stiffness in the whole string. Everything that reduces this helps: an upset-forged pipe end spreads the load over a longer zone, and a smooth forged transition zone does the same on an integral pipe.
2. Insufficient makeup torque. A connection made up loose works under cyclic bending. Every cycle pumps a tiny amount of fretting movement into the thread flanks, and that movement is a fatigue initiator. This is why the recommended makeup torque is a specification and not a suggestion — on a 127 mm / NC50 joint it is 65,255 N·m.
3. Fatigue from cyclic bending. On a directional bore every curve is one cycle. A string that has run a few hundred metres of curved bore has been through more bending cycles than a vertical string sees in a year. The crack grows quietly until the remaining section can no longer carry the steady load.
What the manufacturing process does about it
Each of the three has a process answer, and they stack:
| Failure driver | Process response | Effect |
|---|---|---|
| Stress concentration at the transition | Upset-forged pipe ends before welding; smooth forged transition zone on integral pipe | Load transfers over a longer zone instead of one plane |
| Weak or defective weld | Solid-state friction welding — no filler metal, no flux, no cast structure | Grain runs continuously through the joint |
| Insufficient joint strength | Custom heat treatment of the tool joint, separate from the body | Toughness and wear resistance tuned for the joint, not inherited from the tube |
| Thread galling on make-up | Nitriding the joint surface | Surface hardness up by roughly an order of magnitude |
| Undetected flaw | Dual inspection — machine flaw detection plus manual re-inspection | Catches what the machine alone would pass |
What you can control on site
- Make up to the recommended torque, every time. The most common self-inflicted failure on a directional bore is a loose connection.
- Do not exceed maximum torque. It is a ceiling. Yielding a thread flank at the tongs is damage you cannot see.
- Respect the bend radius. A pipe taken past its elastic limit has already been damaged, whether or not it breaks that day.
- Track cycles. A string that has run hard on curved bores is not the same asset it was when new, regardless of how it looks.
The joint is where the geometry changes, so the joint is where the stress concentrates, so the joint is where the crack starts. Everything in the manufacturing process — upset, friction welding, separate heat treatment, nitriding, dual inspection — exists to push that failure point as far away as possible.
Frequently asked questions
Why do drill pipes break at the tool joint?
Because the joint is where the section changes. The tube-to-joint transition is the sharpest change in stiffness in the string, so stress concentrates there, and fatigue cracks grow from stress concentrations. The pipe body is a uniform section and rarely the weakest point.
Can a friction-welded joint fail at the weld?
It can, but a solid-state friction weld has no filler metal, no flux and no cast structure, so the grain runs continuously through it. The more common failure is at the thread or the transition rather than in the weld itself — which is why the pipe ends are upset-forged to spread the load and the joint receives its own heat treatment.
Does higher makeup torque prevent joint failure?
Up to the recommended value, yes — a connection made up to specification resists the fretting movement that starts fatigue cracks. Beyond the maximum torque it becomes harmful, because you can yield the thread flanks or the pin nose before the string ever sees a load. Recommended torque is the target; maximum is the ceiling.
How do I know if a joint has already taken damage?
You usually cannot see it. That is the argument for dual inspection at manufacture and for tracking bending cycles in service. A string that has run hard on curved bores should be re-inspected rather than assumed sound because it looks sound.
Ask about inspection before you order
Every weld is inspected by machine flaw detection, then re-inspected manually by a veteran craftsman. Tell us your application and we will confirm the specification.
Request a quotationWhatsApp us