Ask an Expert

ask-an-expertAll gearbox failure questions submitted here will be answered by one of our experts. We will not provide your email address to any other organization. We will only post details and answers to your questions with your permission, and will remove any sensitive information. Some past questions and answers are provided below. To submit your question, please use the form at the bottom of the page. We will respond directly to you with the answer to your question.

Question from D.L. : I found small cracks on the surface of gear teeth in my gearbox. Can you help me determine what caused the cracks?

subcase-fatigue

Answer from Bob Errichello of GEARTECH: The cracks in the gear tooth appear to be the result of subcase fatigue. Possible root causes for subcase fatigue include the following:

  1. Inadequate case depth
  2. Inadequate case hardness
  3. Inadequate core hardness
  4. The presence of an inclusion near the case/core boundary
  5. The presence of grind temper

Question from F.M. : Please see the attached picture of damage gears in my gearbox. The gearbox has been running for two years, and has recently shown an increase in vibration levels and acoustic noise. When we investigated the condition of the gearbox, we observed the axial lines on the gears shown in the photo. Can you help me identify the marks on the gear teeth?

Tiger-Stripe-Gear-Damage
Gear Damage

Answer from Bob Errichello of GEARTECH: The axial marks on the gear teeth are a form of electric discharge damage known as “Tiger Stripes”. The damage is caused by electrical currents passing through the gearbox, which cause the gears and bearings to weld to each other. See the entry on electric discharge in our failure modes section. The source of the electric current is often a VFD driving the gearbox motor, or the performance of a weld repair on or near the gearbox. All of the gears and bearings in the gearbox should be examined for signs of this damage.

Question from R.B. : Please see the attached pictures from a pump drive gearbox. After the gearbox failed, I disassembled it. I noticed that one of the bearings had seized, and found a piece of plastic caught in the bearing. Could this piece of plastic have caused the gearbox to fail? Why did all of the sun gear teeth fail, while only a few of the pinion teeth failed?

siezed-bearing
Seized Bearing
failed-gear-box
Failed Gearbox

Answer from Bob Errichello of GEARTECH: The debris found in the bearing was about 1/4 inch in size and certainly large enough to jam the bearing and prevent it from rotating. The bearing outer ring would then tend to rotate in the housing, or the bearing inner ring would tend to rotate on the planet shaft. Either way, the gear mesh load on the sun would be greatly increased and the sun teeth in mesh with the jammed planet may have failed. With spur gears, one tooth typically fails in the single pair zone. Once a tooth fails, the sun gear would jump ahead and impact on the next sun tooth, and the remaining teeth would shear off subsequently. It is not unusual for a sun gear to have all its teeth shear off, whereas only one or a few teeth to fail. In fact, it is a likely scenario for a sun to planet mesh where the planet rotation is prevented or retarded by a jammed bearing.

With the limited data, the following tentative conclusions are offered contingent on further investigation:

  1. The primary failure mode is ductile fracture.
  2. The root cause is overload due to a jammed planet bearing.

Recommendations:

  1. Inspect the debris to determine the source of the debris that jammed the bearing.
  2. Inspect all failed teeth on the sun and planet gears to determine whether there is any evidence of fatigue. The failure modes pages on gearboxfailure.com can be used to help identify any fatigue failures. The gear failure atlas and gear failure analysis handbook on the resources page could help as well.
  3. Disassemble and inspect the jammed bearing to look for evidence of the debris jamming and evidence of outer or inner ring spinning.
  4. Inspect the planet gear bearing journal and the housing bore to look for evidence of spinning of inner ring or outer ring spinning.

Damaged Marine Gear
Damaged Marine Gear

Answer from Bob Errichello of GEARTECH: The images that you submitted show severe micropitting in the dedenda (area of the gear teeth below the pitch line) that is escalating into PSO (point source origin) macropitting at the edge of severe polishing in the addenda (area above the pitch line). There is a band along the tips of teeth where contact is light except for contact at the peaks of the hob scallops. This area shows that the original surface finish was relatively rough. The probable root causes of failure are excessively rough surfaces, inadequate lubricant viscosity, or both. This damage is severe, and is likely to progress, and result in failure of the gearbox. Please let me know if you would like assistance in resolving this issue.

Question from J.R.: I have attached for your review a section from a failed spur gear from a torque converter showing a fractured tooth. I believe the failure mode to be low cycle bending fatigue emanating from tooth root fillet. There seems to be evidence of faint progression lines. Although am not 100% sure. Your assistance would be much appreciated.

Fractured-Tooth-Image-1
Fractured Tooth Image 1
Fractured Tooth, Image 2
Fractured Tooth, Image 2

Answer from Bob Errichello of GEARTECH: This is a classic example of bending fatigue failure. Image 1 shows shows the classic concave shape of the fracture surface that starts at the critical point of maximum bending stress in the tensile fillet (left side), progresses downward toward the compression fillet, and then turns upward to exit on the compression fillet (right side). This image also shows macropitting on the load flank of tooth 1 that indicates the gear teeth were misaligned. Since the macropitting proves the highest load was at the end of the tooth where the bending fatigue occurred, the misalignment probably contributed to the failure. Image 2 shows the origin of the fatigue crack is near the left end of the fracture just beyond the first hob feed mark that is evident in the root, A small elliptical zone surrounds the origin. This zone is polished from rubbing of the crack faces. Faint elliptical beach marks are evident radiating outward from the origin and covering perhaps 70% of the fracture surface. The beach marks prove the primary failure mode was bending fatigue (progressive fracture) and not brittle fracture (instantaneous fracture).

Question from Jeff P: When examining a gear contact pattern for ‘apparent’ micropitting, what magnification is recommended (e.g. simple reading magnifying glass or a higher power loop or ?). Possible you can email me what the pitting would look like? I need to inspect a gear and customer states it ‘seems’ to be micropitting. Very thin lines across the teeth but looking frosted like the beginning of micropitting. A 44 MW steam turbine drives the double helix gears to turn a generator.

 Micropitting on Gear
Micropitting on Gear

Answer from Bob Errichello of GEARTECH: It is definitely micropitting along peaks of grind furrows and is especially severe along the pitchline and in the dedenda. I note that the grind stock removal at the ends of the bull gear teeth might be excessive where it has removed the protuberance. Image P1010020 shows what might be macropitting on the turbine end of the bull gear in areas of heavy grind stock removal. The pinion and gears were probably form ground using radial feed to obtain leads with end relief. This increases stock removal near ends of teeth.

It might be possible to arrest the micropitting by hand polishing the gear teeth to remove the surface aspertites. However the gears would have to be removed from the gearbox to avoid contaminating the gearbox with debris. Higher viscosity oil, cooler oil, or both might help mitigate the micropitting.

Recommendations:

  1. Determine case depth from heat treat coupons.
  2. Determine stock removal from grinding records.
  3. Calculate effective case depth after grinding and compare to specifications.
  4. Inspect the gear shown in image P1010020 and document macropitting with macro-photos and graphite tapes.
  5. Document the micropitting on several teeth of the pinion and gear with graphite tapes.
  6. Measure surface roughness with a portable surface analyzer on several teeth on the pinion and gear.

Question from RM: The photo attached shows a broken tooth on a high speed shaft of a 2 MW wind turbine gearbox. The crack origin seems to be on a line parallel to the root, but a few mm above the edge of the root. A first hypothesis is that the tooth fracture is caused by bending fatigue. Is this kind of fracture typical for bending fatigue or do you believe that the root cause will be something different?

Broken Tooth
Broken Tooth

Answer from Bob Errichello of GEARTECH:

The liberated tooth might be in the bottom of the oil sump below the pinion. Try to fish it out with a magnet on a stick. We have often found the origin of the failure on the liberated tooth. When a single tooth fails in bending fatigue, the root cause is often an inclusion.

Classic bending fatigue occurs in the root fillet; not on the active flank.
Protuberance is used to avoid grind notches in the root fillet. If the protuberance is inadequate, and a grind notch occurs, it can be the root cause of bending fatigue.

Question from Victor M: We have seen heavy wear on the bearing seats one one of our hot bar rolling mill pinion gearboxes; like the bearing is rotating in the bore.

Wear in Bearing Bore
Wear in Bearing Bore

Answer from Bob Errichello of GEARTECH

Victor, this is a classic case of bearing Outer Ring spinning. This is due to loose fits and is worse in steel housing because steel on steel tends to scuff. Repair would be to sleeve the housing and rebore it. The only known cure is to pin the OR with an anti-rotation pin. However, this depends on the bearing type (SRB, CRB, etc.) and bearing arrangement.

Question from Rajeev S: What are the possible causes of the failed pinion shaft shown in the attached photo?

Failed Pinion
Failed Pinion

Application details are as follows:

Application – Ball Mill
Motor rpm – 1440
Gearbox output rpm – 147
Motor power – 90 kw
Rated torque of gearbox – 5.67 kNm
Pinion shaft fluid coupling mounted end shaft broken from shoulder within 2 month of operation.

Answer from Rob Budny of RBB Engineering (with assistance from Bob Errichello of GEARTECH)

Rajeev, the pictures that you sent made a definitive root cause assignment very difficult, but there were some clues to the failure. We suggest that you investigate the following:

  1. Fit between the key and the shaft: It appears from the photos that the key may have been loose, and thus subjected to impact loads
  2. Geometry of the keyway: The crack may have started at a stress concentration in the bottom of the keyway.
  3. Coupling alignment: The appearance of the fracture surface indicates the presence of large bending loads. These can be caused by misalignment between the two halves of the coupling.

What are the detrimental effects of gear and bearing wear on gearbox reliability?

Answer from Rob Budny of RBB Engineering

Gear and bearing wear has two detrimental effects on gearbox reliability. The first effect of wear is that it changes the shape of gear and bearing contact surfaces. This always results in higher contact stresses, and reduces the life of gears and bearings. The second effect is that the gearbox lubricant is contaminated with the wear debris. The hard particles in the lubricant pass through the gear and bearing contacts, resulting in debris denting. This damage can also lead to premature gearbox failure.

Question from Jose Luis P: My question is a theoretical one. What is it about the wind industry that causes gearboxes to fail more than in other industries? I have the following aspects I can think of:

a) Industry maturity
b) Understanding of the loads or their predictability
c) Designs forced to cut down costs
d) Not enough NPI time
e) Size & manufacturing implications

Thanks for your time.

Answer from Rob Budny of RBB Engineering:

That is an excellent question. I think there are several reasons that wind turbine gearboxes have failed prematurely, some of which you note in your question. I think the main reasons for the relatively high number of premature wind turbine gearbox failures are as follows:

  1. Misunderstanding of loads: Until recently, the tools for predicting wind turbine gearbox loads often underestimated the magnitude of the applied loads.
  2. Insufficient testing: Full scale testing of wind turbine gearboxes is very expensive, and many OEM’s in the past did not perform enough testing to validate their designs. Recently large test stands have become available that have enabled such testing.
  3. Lack of adequate design tools: In the past, tools that accurately predicted the stress in gears and bearings when the gearbox was under load were not available. More recently, advanced tools such as RomaxWind have corrected this deficiency.
  4. Required service life: A wind turbine gearbox is typically expected to last for 20 years. The gearbox may see 100 million main shaft revolutions during this time. This is an extremely long service life, and it is difficult to design and manufacture a gearbox that can meet this requirement.
  5. Inadequate quality assurance: Two leading gearbox failure modes can be attributed to poor manufacturing practices and insufficient attention to quality assurance, grind temper and material cleanliness. Failures due to these failure modes are preventable, especially from grind temper, and improvements have been made that have reduced the number of failures caused by these failure modes.
  6. White etching cracking: Many wind turbine gearbox failures have been caused by bearing white etching cracking. While the root cause of these failures is not agreed upon, steps that reduce the likelihood of failures of this type are known. See the article on preventing white etching failures in the resources section of our site.

Although it is true that gearbox failures have been an issue in the wind industry, it is also true that in the past few years the gearboxes have become much more reliable, as the lessons learned from earlier failures have been incorporated into the design, test, manufacturing, and operations practice of modern wind turbines.

Question from Richard F: Any comments on how to better address residual stresses within gears? Best that I can come up with is reviewing the heat-treatment process as well as perform a study of the parts received from our supplier as well as post heat-treatment.

Answer from Bob Errichello, owner of GEARTECH:

Residual stress in carburized gears
The compressive stress in the case of carburized gears is highly beneficial because it increases fracture toughness and bending fatigue resistance. However, the compressive residual stress in the case is always balanced by a tensile residual stress in the core, and tensile residual stress is generally detrimental. Therefore, a proper balance between the case compressive residual stress and the core tensile residual stress must be achieved to obtain maximum load capacity.
The residual stress gradient is influenced by:

  • material alloy
  • carbon gradient
  • carburizing temperature
  • quench method
  • tempering temperature

For maximum load capacity, the microstructure must be controlled by:

  • using clean steel with a minimum of nonmetallic inclusions
  • using fine grain steel
  • selecting a steel alloy with sufficient hardenability to obtain a microstructure of primarily tempered martensite
  • maximizing residual compressive stress in the case by using a steel with the lowest possible carbon content
  • using reheat quenching rather than direct quenching
  • using proper process control to achieve the optimum residual stress gradient
  • specifying 60 HRC surface hardness, proper case depth for the module, and 35 HRC core hardness
  • controlling retained austenite; limit RA to 20% for maximum bending strength and limit RA to 35% for maximum Hertzian fatigue resistance
  • eliminating bainite, pearlite, and network carbides from the case microstructure
  • eliminating microcracks especially near the surface of the root fillets
  • eliminating defects on the surfaces of the root fillets
  • shot peening the root fillets

Finally, several prototype gears should be destructive tested to determine the microstructural properties and the residual stress gradient should be determined using X-ray diffraction.