• Increase font size
  • Default font size
  • Decrease font size

Lost Propeller

With many thousands of propellers installed over more than 15 years  - it is inevitable that over time situations willl arise where the whole propeller unit is lost from the shaft - which may be either a parallel spline in the case of Saildrives or a shaft mounted unit with a tapered cone / keyway / nut.

To understand what can cause this it is appropriate to review all possible causes by mounting type.


All Saildrive propellers are mounted on a parallel splined shaft to an SAE 16/32 design with 17 teeth on what was originally a 28 mm shaft Ø. Note that even metric shafts use the same spline type designation as above. This is what is termed a face centered spline meaning that neither the outer nor the inner diameter teeth make contact with the propeller boss or shaft. By their nature face centered splines tend to oscillate with more degrees of freedom than other types.

Contact is confined to the angled faces of the teeth of each.

While the original BUKH Saildrives followed by Volvo all adhered to a 28mm Ø shaft - over time differing tolerances have crept in by manufacturer and even by method of manufacture for the same supplier - Rolling the splined teeth vs broaching as Vovo have now done for example - will deliver quite different tolerances with different lead-in characteristics to the starting teeth on the spline.

Thus by their very nature all Saildrive propellers of every make will what is termed " fret " on the shafts as tolerances will be such that they must be able to  be slid on by hand. When Ahead or Reverse is engaged there will be a small amount of movement in a radial direction until the face centered surfaces on the propeller and the shaft come into contact whereupon the rotation of the unit will commence. This is recognised in all locking systems employed by the various propeller manufacturers.

What is critical in all Saildrive mountings is to check that what is termed a thrust collar - a short pipe like piece of material with an ID of 28+ and an OD of ~ 40 with a length of ~ 35 is fitted over the splined shaft. This is a Saildrive component - not a propeller component - with it's own manufacturer's part # and must be checked to ensure it is present before mounting a Kiwiprop or any other type of unit.  Where a rope cutter is fitted this may / will substitute for the functionality of this component depending upon the type of cutter fitted. Previous loss of a propeller makes loss of this component more likely.

In all cases - because the propeller is free to slide forward on the spline - this thrust collar prevents that and transfers the forward thrust of the propeller unit in Ahead onto the Saildrive proper and thus propels the vessel ahead.

Kiwiprops employ a locking system whereby two M8 set screws in the boss at 90º are tightened down into a recess in the main attachment nut which transfers reverse thrust in Astern to the drive shaft. As there is no way for the boss to rotate around the spline, the only way for the unit to come off is for the nut to rotate relative to the shaft and in effect unscrew off the Saildrive thread on the end of the spline.  This will be  M16 x 2 in the case of Yanmar SD20/25's, All Volvo's and other manfacturers, otherwise it will be M20 x 2 in the case of all Yanmar  SD40/50/60 units.

The 2 M8 set screws in the boss will therefore restrain the main attachment nut from becoming undone as they are fixed relative to the boss and net of tolerances - are therefore fixed relative to the shaft. It is virtually impossible for the main nut to come undone while these two M8 set screws are in place and tightened.

To provide an additional level of security K3 Kiwiprops will - as do Volvo - come with another  M8 x 40 Cap screw ( M8 x 70 in the case of K4 units ) that threads into the threaded section of the Saildrive shaft. All Saildrives  are bored and tapped  M8 into the center of the threaded section. This Cap Screw pulls down onto a flat landing machined into the main attachment nut and will have an M8 Nord-Lock™ washer underneath. ( Very early Kiwiprop units may not have this M8 screw or the M8 Nord-Lock washer )

The difference in pitch between the M8 x 1.25 cap screw and the main nut of M16/M20 x 2.00 ensures that if the nut were to start to come loose - it would quickly bind with the cap screw thus preventing that from happening further.

The presence of the Nord-Lock™ washer thus offers a third level of security by offering an additional method of securing the main attachment nut.

The presence of Loctite™ - medium strength - or similar when applied to the main attachment nut and 3 x M8 locking screws offers a fourth level of security in addition to the above.

Kiwiprops mounted with the thrust collar in place with the all three locking screws and attachment nut  inserted wth Loctite™ will remain mounted securely indefiitely. Our high time units on a cat with SD20 Yanmar Saildrives have reported four years of service with 25,000 sailing miles and 2,000 hours on each engine with no issues apparent.

Common sense normal maintenance would include checking the security of these units at haul and/or lubrication time as the one great danger is always hidden corrosion issues that may appear and clearly where the threads of the locking screws or main attachment nut can corrode for whatever reason - then the failsafe locking system devised above will always be at risk. Any stray curents will be always attracted to the sharpest edges available - invariably the threads of the screws which is basic physics and it is these that will corrode first if corrosion where ever to become an issue.

Remember it may come from an adjacent vessel or faulty power suppply on the marina dock so a vessel without it's own corrosion issues may still be at risk.

Based on over 55,000 propeller years of service with just over half being Saildrive mounted,  we have listed below the most frequent causes of propeller loss on Saildrives in descending order.


# 1 Missing Thrust Collar: Operating any Saildrive mounted propeller including a Kiwiprop with this component missing will lead to eventual loss of the propeller as described above.

# 2 Loose M8 Nut Locking Screws:  If these are not tightened and / or they are not inserted with Loctite™ they will possibly come loose as the propeller frets on the spline with each Forward / Reverse engagement and over time will allow the main attachment nut to work loose resulting in the loss of the propeller unit.

# 3 Mounting Underwater:  While quite possible to mount the unit with a diver by coating the threads on the propeller with Loctite™ or similar - care must be taken over time to monitor the locking screws x 3 and main attachment nut as it is not possible to get the Loctite™ to cover the threads remaining underwater namely the main attachment nut and central M8 Cap Screw.

Loctite™ or similar will harden in the absene of oxygen AND the presence of a metal - both conditions being met underwater, but that does not ensure it is in contact with a clean dry surface initially to ensure it binds the two components together. Used correctly  on two dry clean surfaces these products produce very reliable locking over long time frames but to be fully effective they do require clean dry surfaces to react with. 

# 4 Broken SD20 thread: Yanmar machine an M8 thread into the centre of the spline, then an M16 thread on the outer Ø then cross bore a small hole which cumulatively leave little metal and a very weak point at the front of the thread. The actual threaded section can break at this weakened point allowing the propeller to come off. Going to the larger M20 thread on all subsequent Saildrive models has eliminated this known weakness.



All shaft mounted units will be pulled up by a nut over a threaded section at the end of the shaft, preceded  by a conical section which will be a 1:10 ratio in the case of all metric shafts eg 25 mm  30 mm 35 mm etc and 1:16 in the case of all imperial Ø shafts eg 1.000"  1.125"  1.250" etc.

Metric shafts will have a metric thread - imperial shafts a UNC thread - but the mounting principles remain unchanged irrespective of the shaft diameter or conical ratio or thread type.

All shaft mounted units will have a matching keyway machined parallel to the conical surface into the shaft and into the propeller boss. This is what translates the torque of the shaft from the engine / reduction gear  to the rotary motion of the propeller.

Fixed bladed propellers can be secured to the shaft by simply inserting a split pin through the castleated attachment nut thus securely locking the nut to the shaft. If a key were ever to shear or corrode under torque then the shaft will rotate relative to the propeller boss, but the nut - rather than turn with the propeller and become unscrewed leading to the loss of the propeller, will remain firmly locked to the shaft. In these situations loss of a key will entail loss of drive to the propeller but the propeller will be retained on the shaft.

In the case of all feathering or folding type units, including Kiwiprops,  because of the length of the boss protruding aft over the thread,  it is not possible to access the threaded section to insert a split pin into the threaded section so alternative approaches must be devised.

To analyze the possible reasons for the loss of a propeller off a shaft installation requires consideration of the following possibilities.

In all cases it is assumed the shaft has been machined correctly and has not been subsequently modified so that it is still to standard specifications as defined on our web site. This includes the length of the taper, the angle of the conical section and  the length of the thread. It is very difficult to predict issues that are likely to arise once mounting to a non-standard shaft.

The most likely cause of loss of a Kiwiprop or similar non-fixed unit from a shaft is always the shearing of the key which can occur for a number of reasons outlined below.

Once the key shears - the main attachment nut being locked to the boss - not the shaft because of access issues outined above, means that when torque is applied to the shaft, the resistance of the propeller will then cause the propeller to rotate relative to the shaft, but because the locking screws are in the propeller, they too will turn relative to the shaft and being locked to the main attachment nut will cause the main attachment nut to turn relative to the shaft.

Depending upon the rotation of the shaft in Ahead - in the above scenario - either in Ahead or Astern there will arise a situation where the locking screws holding the main attachment nut will unscrew it off the shaft and thus remove the whole propeller from the shaft.


# 1 Corrosion:  The most frequent cause of the key shearing is corrosion. Keys are generally made from brass not bronze which is too hard to extrude so will suffer from de-zincification over time. It may appear substantial on the outer surface but can be totally corroded internally as has been evidenced when a key shears and the remaining now half key  is left in the shaft for examination. In a perfect world keys would be ex SS 316 - but would need to be machined ex solid stock and will be expensive.

We strongly recommend any brass key should be replaced at no more than a 5 years interval as part of normal maintenance.

# 2 Binding: The next most frequent cause is when a situation arises that the keyway is of insufficent depth in the shaft and/or the propeller boss so that when pulled up the propeller appears to pull up tight on the taper but is in fact bound up on the upper surface of the key. Being loose on the cone means that over time the torque transmitted will shear the key leading the to loss of the propeller or the propeller will oscillate around the cone working the attachment nut loose.

Mounting instructions for Kiwiprops are very specific - the unit must be first mounted on the shaft minus the key and marked at the forward end of the shaft then mounted again with the key and checked to ensure it still travels forward to the same position as before. Filing the upper surface of the key offers a simple fix where this issue to arise.

# 3 Taper Length:  A situation can arise where the aft face of the propeller boss does not protrude what is termed a " pull up " distance aft of the small end of the cone / taper. A situation can then arise where the nut pulls up - but at the end of the thread and the boss has not yet been correctly pulled up onto the shaft taper. It can then become loose over time leading invariably to shearing of the key followed by loss of the propeller.  Kiwiprop mounting instructions require this be checked by mounting the unit and measuring the aft face is the required distance aft of the small end of the cone / taper.

# 4 Nut Tight on Thread:  Kiwiprop mounting instructions require that prior to offering the propeller to the shaft - the main attachment nut is run fully down the thread to ensure it not binding and will run freely with the forward face of the nut nearly touching the aft or small end of the cone / taper. If the nut were to bind on the thread prior to pulling the propeller fully home to the required distance on the cone / taper - when mounting the unit the nut will pull up to the correct torque but it will be because it is binding on the thread and not because it is pulling the propeller boss onto the cone / taper. With the boss not pulled up correctly - over time it will become loose and then shear the key with the inevitable total loss outlined above resulting.

# 5 Key / Keyway Dimensions: If the key was undersize and / or the keyway in either or both the boss and the shaft was oversize then when mounted the key will not be a neat press fit into the keyways.  Under that scenario as torque was applied to the shaft the propelelr would move a little to eliminate the slack in the key fit and over time as Reverse / Ahead were engaged, then the key is likely to shear leading to loss of the propeller as described above. Keys should be a neat press fit into both keyways on the shaft and the boss.

# 6 Impact with flotsam or rope wrap: It is possible and even likely on a shaft installation that an impact or rope wrap scenario will simply shear the key on the shaft. Subsequently and perhaps some time later depending upon how tight the boss is on  the cone - when a Reverse or Ahead torque is applied to the shaft the propeller will unscrew as described above. With an engine and it;s flywheel running  at say 1500 rpm and a typically 2.5:1 reduction on the gear train it is simply not possible to bring the propeller to an immediate halt as can happen in an impact situation. The result will always be some form of shearing or slippage in the drive train.

# 7 Cone non standard: If the tapered cone is not accurately machined and to the correct taper angle the propeller boss will be loose on the shaft cone when mounted and have the ability to oscillate relative to the shaft. Over time it will increase this motion and in doing so also oscillate the nut that is locked to the boss. This will eventually cause the nut to become undone over time and with it the whole propeller unit. If the propeller pulls up neatly onto a cone machined accurately to the correct tolernaces as specifiied in the standards for shaft taper attachment - there will be no movement and the unit will remain securely mounted indefinitely.

# 8 Loose Lock Screws: It is critical that the two M8 set screws that lock the nut to the boss are both inserted with a medium strength Loctite™ or similar and pulled up to lock the main attachment nut securely. Were the main attachment nut to become loose over time it will allow again shearing of the key and inevitable loss of the propeller. In all cases these should be torqued down to the levels described in the manual or on the web site video.

# 9 Attachment Nut Loose:  This will be a consequnce of  one of the above situations which allow the propeller boss to move or oscillate on the shaft taper. While you could argue the loss was caused by the nut coming loose - the primary cause will be from one of the scenarios above that allow or force the nut to come loose. With no movement of the propeller boss relative to the shaft there  will be no force acting on the nut so no reason for it to come loose. 

In all cases this scnario is the result of one of the situations above and is not a primary cause of propeller loss.

SUMMARY:  Propellers that are lost are always lost for a reason - almost certain to be one of the above. 

If the unit is mounted correctly following all specified steps onto a strictly standard shaft taper and any keys possibly subject to corrosion are replaced regularly and both keys and keyways are  in specification - a user can have every confidence built on over 40,000+ propeller years of service that a Kiwiprop unit will stay mounted to any shaft taper securely over time.