Detuning and Proving Resonance - Fatigue Cracks and Failure Due to Resonance

Practical Solutions to Machinery and Maintenance Vibration Problems

Chapter 3, Detuning and Proving Resonance

Section 4, Fatigue Cracks and Failure Due to Resonance

Whenever cracking is reported, resonance should be suspected. Cracking almost always occurs at one of the mode shape nodes. The material does not break due to excessive stresses but due to the many millions of reversed stresses concentrated about the nodes, causing fatigue. The phenomenon is similar to bending a wire back and forth until it breaks. The break has the characteristics of a pure fatigue failure which is often mistakenly called "crystallization." The break usually shows a crystalline appearance and sharp edges.

This accounts for so many familiar situations whereby pump shafts "mysteriously" break within a few weeks to a few years, even though the calculated stresses indicate the shaft should last a lifetime without breaking. Premature foundation or floor cracking, pipe weld breakage (often repetitively), gear cracking or failure (usually in equal-spaced, radial lines), gear driven shafts broken with the characteristic 45° torsion break, have all occurred as the result of metal or concrete fatigue at one of the nodes due to resonance.

Sometimes high speed, high output major rotating machinery does not show excessive vibration at the usual points of measurement, such as at the bearings. However, by feeling with fingers along the small diameter lubrication tubing, or small diameter cooling water piping, it can be determined if there are telltale large "loops" (antinodes) and nodes, characteristic of a resonant condition. Most often the vibration frequency is high enough and the tube or pipe rigidity low enough to allow a higher resonance with several nodes. It may take many months or even a couple of years for the tube or pipe to crack at one of the nodes, but when it does, lost lubricant can wreck an expensive machine or a flammable product can be released. The same phenomenon is very common for pressure gauges cantilevered on a pipe. If non-resonant, there is no problem. But sometimes the pipe, with a gauge acting as a weight, is resonant to a specific frequency of machine vibration. Eventually it will break, usually at its point of connection.

At one gas pipeline pumping station, a low speed reciprocating compressor was plagued with expensive shutdowns due to gas leaking from pipe nipples that cracked at the threaded point of connection with the main case. The series of short, stubby nipples (about four inches in diameter by ten inches in length) were all connected to the same headers. The nipples acted as resonant cantilevered beams vibrating at their first resonance frequency. This caused the reversible stresses at the node, which coincided with the point of connection. The low speed of the compressor did not resonate the pipe nipples. Instead, they were resonated by a small turbo-supercharger mounted on top of the compressor. Its operating speed of 22,000 rpm resulted in acceptable vibration levels on the supercharger itself, but that same vibration traveled along the compressor's main frame to the nipples, causing them to resonate. The cracking problem was cured by simply replacing the nipples with others of the same pipe diameter but about three inches longer in length.

Textbook Index

This textbook contains only part of the information in our Practical Solutions seminar.

Link to Seminar Schedule. Order a print version of this entire textbook.