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

Continuous Improvement

A Kiwiprop update as at Feb 2021:

It is now some years since an update was made to this website and in the intervening period Kiwiprops has continue to prosper and now has an installed base of some 8,000 units in virtually every country dating back to 1998.

 We maintain a database with a build record of every installation and thus able to monitor ongoing performance and functionality issues.

 Today we have over 65,000  propeller years of units in service and consequently over 200,000  blade / reverse screw  years of service upon which to analyze and generate feedback and modifications.

 Rather than dramatic design changes to the unit we have operated a continuous improvement program with successive small changes that have been that have been rigorously tested to the extent that is possible.

 We have also adopted a design constraint making all components backward compatible with all previous units.

 Marine engines come with a whole host of power ranges, maximum engine rpm capability and in addition a wide range of reduction options which are not always the same in ahead and astern. In fact this variation is the norm and one needs to recognize that a Yanmar shaft installation for example with say 2.2:1 reduction in ahead, like all small Yanmar’s will have a reduction ratio of 3.0:1 in reverse.

Another variable on marine engines is the type of clutch they employ which leads to very differing engagement speeds with consequent differences in the force involved with the reverse rollers contacting the blade root surface. At one extreme we have the dog clutch of the Yanmar SD 20 Saildrive which is not actually a clutch at all as it is either fully in a fully out and leads to huge shock loads on the Propeller during Reverse’s engagement.

Many smaller gearboxes today will have what is termed a cone clutch, consisting of a bronze cone into a metal cup and energized by mounting on a spiral spline so that as torque increases, the force on the bronze cone into the cup also increases. These can present difficulties getting them back into neutral with high idle speed, or any glazing of the surfaces of the cone.

Virtually all manufacturers today are phasing out cone clutches and reverting to the normal multipack clutch which has a much smoother engagement and no difficulty engaging neutral. The loading forces on a blade route with these clutches are much lower and lead to significantly lower wear rates.

Today  - Saildrive’s comprise in excess of half the market under about 80 hp and more so in new build production vessels. Yet all Saildrive’s driven by the nature of their drivetrain will have exactly the same reduction ratio in ahead and a stern.

While it is not possible to optimize a particular propeller design for all these very varying constraints, it is important to recognize they exist and make appropriate trade-offs including economic to provide what one considers as an optimal solution.

An optimal solution for sailing vessel will generally attach equal weight to motor and capability and the reduce drag from the feathering function when sailing.

The improvements undertaken over the years can be summarized as follows:

In every instance these changes have been very well documented on our extensive website:

A full database search function on the top right hand corner of our homepage will bring up the information on any keyword entered.


A switch to 50 % glass content blades:

Post mid 2009 we became aware of the existence of a new blade material:     DuPont™ Zytel® HTN53G50HSLR NC010

In simple chemistry terms this constituted a long chain molecule, rather than the previous 35 % glass product we were using which was an aromatic or ring molecule.

The great advantages of this new product were contained 50% glass fibre by weight, was impervious to both hydrocarbons and water meaning it was stable over a very long time frames when immersed. It also of course had much higher strength and stiffness but retaining the obvious zero corrosion potential of the previous product.

The trade-off was a much higher moulding temperature which influenced production and die cooling and of course came at a significantly higher price. It also required a switch to carbide tipped tooling due to the highly abrasive nature with the high glass content in the material.


The development of ogival foiled blades:


The increased strength of the new material allowed for thinner blades which has two benefits – they will generally be more efficient and can in principle be made quieter. This was a comment from above but caution is appropriate as in any aperture situation, particularly on this type of vessel with a very broad keel, it is always going to be a challenge when the propeller blades simply are not going to see continuous smooth streamlines entering the unit.

To retain design flexibility and minimize stockholdings as well as very expensive die costs,

we elected to maintain the existing symmetric foil shape which allows for both left and right hand rotation and then mill off either side of the blade in jig  so that in simple terms it more closely resembles a traditional Propeller with a flat face aft and an ogival foil on the forward face. This has been determined over many years as optimal for motoring functionality – but of course we were restricted as more removal would have a negative affect on feathering stability.

We conducted extensive testing on this new foil shape using a friends catamaran fitted with both and old and new foil on each side this negating any variation from hull condition, currently loading and sea state.

Results for this trial are on our website under Ogival Foils.

In addition - we had a very helpful engineer who had had a Kiwiprop and was motoring the inland waterway from New York to Florida and return.  His considered feedback and analysis with a popular 3GM30 on 2.61:1 - which we do value  - was that the new ogival foils delivered between 0.3 and 0.5 knots additional motoring speed over the course of that voyage at the same engine rpm.

This information has been replicated on many situations now and we are very confident that the foil shapes we are using are both optimal for motoring, yet retain adequate strength with a high margin of safety and the shape change has not affected feathering functionality.

There is a very extensive analysis of the testing and design undertaken using computational fluid design (CFD) beginning with the profile of the actual blade die shape and progressing to illustrating the effect of this particular shape and ogival modification on power and thrust. Actual vessel speed versus derived agreed to within 5%. This is all available on our website under Ogival Foils. CFD will deliver the results from a particular foil shape under analysis – it will not tell you what is optimal which has to be carried out using hydro dynamics and trial and error to an extent.


The development of V foils on the lower trailing blade edge:


To ensure feathering functionality, and recognizing that in the real world propellers are subject to fouling, we added two small the foils extension to the lower trailing edge of each blade. This then allowed us to easily grind off during assembly the appropriate side leaving a small extension that when Sailing had the effect of biasing the blade such that the tip favoured movement in the head direction that’s preventing any winding up of the internal torsion spring which could lead to reverse engagement.

We needed to deal with growth on the blades, such as barnacles, oysters and also the rarer situation of for example seaweed or some other flotsam, such as a plastic bag fouling the blade while sailing.

Having this for an extension only on the base of the blade had no effect on motoring performance as the speed of advance at this lower section of the blade was really only matching the forward speed of the vessel so generating no forward thrust.

The analysis of these small foil extensions was undertaken by Flettner - a very early and highly respected German helicopter design engineer - who added a very small piece of metal sheet to the trailing edge of the rudder of an ME109 World War II fighter that could be easily bent with a simple spanner so biasing the rudder to remove any imbalance on the control stick.


Reverse screw switched from UNC ¼” to M8:

From approximately mid 2008 we increased the thread size to M8 which was a heavier screw more appropriate to the higher powered engines and larger blades e.g. 19.50” that were coming into service.

 All Reverse rollers, either conical or Tri-roller design will fit over either screw, as the bearing dimensions have not altered.

 The hexagonally head remains unchanged on both designs. It is a simple task to bore the existing thread with a 7.3 mm drill and re-tap using an  M8 x 1.25 or standard M8 taper tap and stainless lubricant for those wishing to upgrade.


Reverse screw attachment:


It is important to ensure that each of the three M8 threaded Reverse screws ex SS 316  that hold the Tri-Rollers are retained securely in the boss of the unit. The screws are machined with a landing above the thread consisting of the 9.0 Ø Tri-roller bearing and when tightened pull down flush onto this flat.

Any side force on the screw thus generates a tension in the M8 screw as it attempts to roll up about the axis of the flat on the Tri-roller and the flat on the boss.

Each screw thread(s)  is coated with a red high strength grade  MIL  spec Loctite™  277 and torqued down using a torque wrench.

We then had two options to provide a margin of safety with a second level of security to ensure these do not come loose.

One obvious option is a spot of weld from the inside to the boss, but this excludes any potential removal for any reason at a future date. In addition this would introduce a different grade of SS 316 with the inevitable possibility of generating an electro potential across the joint and consequent corrosion.

The approach we use is simply to pin punch the underside of the mushroom headed boss near where the screw exits. This provides a slight distortion and tightens the boss down onto the thread of the screw making any removal very difficult - as all the normal tolerances between the thread of the screw and the thread tapped on the boss have been removed.

This still allows for removal of the screw, normally requiring the addition of heat to soften the Loctite™, but does require a much increased torque to undo the M8 screws.

Our experience from the over 200,000 screw years of service is that unless we have an environment experiencing extreme and abnormal corrosion with electron flows from an external  source to the sharp thread edges, this mounting method has proved to be 100% reliable and excludes any possibility of an electro potential being generated from an additional grade of SS 316.


Four bladed K4 unit for  larger 50 – 75 hp installations:

With the advent and increasing popularity of higher horsepower installations, particularly units such as the Volvo D2–75 and similar Yanmar units required for the ever larger vessels becoming more popular we undertook a development program utilizing as many of the standard components as we could to address this market.

Due to the larger shaft sizes required for these higher power levels a new larger boss was required to accommodate up to 40 mm ISO shaft mountings or 1.500” shaft in SAE mounting.

Blade area to displacement is a critical design ratio for any propeller and the higher displacement typical of these large vessels required a full bladed unit. These are typically smooth running and meant that stress levels per blade were at the 20 hp level typical of the existing K3 3 bladed unit where 60 horsepower over three blades produced the same stress levels per blade.

Developments undertaken on the Tri-roller concept and mounting of the reverse screw was able to be completely duplicated on these larger units as was the Titanium blade mounting pins. The same blades were trimmed to a larger radius at the base to fit the larger boss. Thus a large portion of the components were able to be used on this K4 unit providing positive commonality and economic benefits and reduced component stockholdings.

The first unit was installed for trial in 2011 - there are now  some 200 units installed since 2012.


Threaded Titanium Blade Attachment Pins:


For the initial years our units were produced using simple quarter inch pins and nickel silver that were pressed/tapped into a hole in the blade that had been drilled 0.004” under size.

We were not able to use SS316, as it is prone to crevice corrosion which was likely to be experienced in this application.

We have seen many units over the years where these pins have been 100% successful with no design issues emerging.

However if for some reason, which we did not recommend, the pins had been removed - each time this tended to drag material from the hole and they would become progressively less tight in the blade.

To offer a solution where we could eliminate corrosion with confidence and also ensure that the blade mounting pin was secure under any circumstance we designed a new blade retention pin ex 8 mm Titanium rod stock whilst retaining the quarter inch undersize hole used previously.

These pins turned from titanium rod stock have a slotted head on one end and a female thread on the other which will accept a small male threaded and slotted cap. Both the headed end and the capped end require a standard 45° countersink leaving 25 mm in the blade.

Mounted with a blue medium grade Loctite™ on the thread we have yet to see a scenario in many tens of thousands of operating years of a single failure of this mounting system.

This is extensively documented on our webpage under:  Blade Mounting


TRI Roller – Reversing roller modification(s):


Coupled with the advent of the stiffer and stronger blade material and where rates on the blade roots which contacted the reverse rollers, we undertook an extensive research program to offer an approved solution to the simple conical roller that we had progressed to.

In addition we had found that despite extensive instructions to the contrary,  the fact that the antifouling was often carried out in a yard and not by the owner, we had to assume that the whole unit would be antifouled and this would invariably see the reverse rollers, whose function was to rotate upon contact with the blade root during a reverse function seized up with antifouling paint or Prop-Speed.

In addition using a sliding motion, rather than a rolling motion, would very dramatically reduce the point pressure on the blade root and consequently reduce the wear rate.

After extensive trial and error we found that what we term a Tri-Roller, which was basically a conical roller with three flats machined on it would generate a sliding motion with low contact pressure per unit area during a reverse function.

However to ensure it did not seize up from antifoul application, these three flats would allow the reverse roller to be rotated through 120° for each reverse function engagement using the mechanical force of engagement.

Any addition we designed a small press fit polypropylene cap that could be simply tapped into the upper surface of this or the previous version conical roller - as an additional insurance to keep antifouling and any growth deposits away from the bearing area of the roller and mounting screw.

We have been using these for many years now and have yet to see a Tri Roller  that has frozen and regard this as the optimal approach to the design requirements involved.

In addition - to further reduce anywhere on the blade root we have machined a small tapered cylindrical surface between the conical surface and the flat. We also linish this transition to ensure that the leading edge of the sliding surface does not dig in or scrape the composite material during reverse engagement.

Given the multitude of clutch types that exist in the market we have also added what we term and  “ Impact Screw “  to the blade root at the point of maximum pressure experienced during a reverse engagement function, which occurs approximately when the blade is in a 45° pitch position, on its way to the normal 24° maximum pitch of the Kiwiprop design.

This provides a metal on metal contact from the Tri-roller to the blade root and has virtually eliminated wear at this contact point.

This is well documented on our website under the heading: Impact Screws


Blade root  V Seals:

The very early units we produced did not have a seal in the blade root, but depended upon the low tolerances and shape of the blade extending over the spherical blade carrier which prevented high-pressure water forcing into the blade / blade carrier  and removing grease over time.

To minimize the grease removal we then added an O-ring to the base of the blade which provided improved sealing. These readily available and low cost seals did provide an improved level of ceiling and a small amount of flexibility to accommodate the inevitable tolerances which can change over time between the blade route and its mounting.

To provide a further improved level of sealing, we designed a carbide cutter to machine a stepped recess in the blade root and then made a matching die to produce a softer V – Seal with the ability to accommodate the inevitable wider range of tolerances from assembly variations and wear over time between the blade root and blade carrier casting and leg.

We are confident that these seals do provide a higher level of grease retention, and by the very small quantities required when greasing the blades post haul out. Care must be taken at this stage, as carefully described in our video and manual, not to over pressure when greasing as the seals are so effective they can be distorted from the very high pressures that can be generated with a normal grease gun.


Material changes - Glass reinforced Poly-propylene Nose Cones:

The first units we produced in both Shaft and Saildrive configuration used a white Acetyl Nose Cone for some years. We needed a material available in rod format for machining purposes.

Acetyl has many attributes for this role, it is widely available, is very tough and not prone to cracking so accepting of the four cap screws that hold the two halves of the Nose Cone(s) together. It does however expand very slightly over time when continually immersed as this component inevitably is in service. This could be accommodated by simply providing slightly greater tolerances when new – but being an un-necessary variable - in a perfect world it would not be present.

On the advice of our plastics engineer suppliers we switched to a much harder PETP which is stable underwater but more prone to cracking - particularly if overstressed. Low temperatures for example which shrink the length of the cap screws holding the two halves provides additional stress. Overtightening without a torque wrench also could lead to overstressing. A small percentage of these displayed cracking after some years of service, but continued to deliver the required functionality of transferring forward thrust to the boss and accepting the tail of the internal torsion spring to pre-tension required for the feathering function.

In 2008 we found that we could obtain in rod stock format - a glass reinforced polypropylene product from the US which was not available in New Zealand that met all our material design requirements, very tough, very strong and totally impervious to and dimensionally stable under water. We have used this product exclusively now for nearly 13 years and have yet to experience a failure.

Internal sleeve and aft washer switch to Vesconite:

The first units we produced were from nickel aluminium bronze castings which provide an excellent bearing surface between similar metals. Over time - to cater for increased volume production we switched initially the blade carrier casting, followed by the boss that fits to the shaft taper or spline in the case of the Saildrive to a lost wax investment casting in SS 316.

SS 316 has many admirable properties for continuous immersion in salt water but is prone to what is termed  “ galling “ whereby two moving soft surfaces  “ gall “ or catch and freeze when used in a moving bearing type situation as we had with the 100º of movement between the boss and blade carrier during a reverse engagement function.

To address this we needed the material that was impervious to both saltwater and hydrocarbons as we would be lubricating the bearing and retaining grease inside the unit.

 A fibre reinforced composite product from South Africa used extensively in marine and heavy industry labelled Vesconite was selected.

 We inserted a sleeve between these two components on the bearing service, and a washer with an L-shaped profile to assist in grease retention between the boss and blade carrier aft joint contact surfaces.

 These have proved very resilient over long periods of time and have delivered the functionality required. The larger K4 four bladed unit also has a washer at the forward end as the Nose Cone for this unit is also from SS 316.

Web site development:

Over the years we have developed a very extensive website containing many hundreds of pages of what we believe to be relevant information that is useful to a Kiwiprops or potential Kiwiprops user.

The website has been maintained on a very regular basis to always be current and provide an authoritative source of information relating to the unit.

To assist visitors to the website, on the upper right hand side of the homepage there is a keyword search function that covers the entire database.

Simply entering a keyword will bring up every reference in our database containing that keyword – a very useful function.