How do propellers propell a boat.

Petros,

Since I am the only one to use the term slip, I assume you are referring to me. You are correct that slip does not do a good job of explaining what happens but it is the common term used in the industry to define the difference between propeller travel and theoretical pitch travel. And, yes, I understand that slip is necessary to produce propeller thrust in a fluid. Its also necessary to produce power in an electric motor , but that is another story. Its also a similar concept to the difference in no load speed and loaded speed in any rotating machinery and that is yet another story.

If you have a widely accepted and more descriptive term to describe this phenomena, let us know and I will use it. Otherwise, I will use slip while knowing that its just a word.

 

Just one person's understanding of propellers...

I argued with myself (always a losing proposition) whether I wanted to dive into this fray, and it probably is an exercise in futility but here goes...

First, the debate about different propeller physics is a lot like asking how a light bulb makes light. It is different for incandescent, CFL, or LED. All make light, but in different ways. There are lots of similarities, but also some critical differences. So, for the purposes of my post here, there will be some simplifications and I apologize in advance if anyone feels this is overly pedantic.

A boat is in motion because thrust is applied to the boat. There are many ways that propulsor thrust can be “connected” to the boat – through an OB motor housing, up the main shaft through a thrust bearing, or even through a nozzle head box in the case of a ducted propeller system. For now, let me limit my comments to a shaft-driven propeller where the thrust is axial to the propeller.

Shaft-line thrust is an accumulation of forces on the blades that are transmitted through the hub and up the shaft. The component of these forces in the axial direction provides the thrust. This is an important point; it is not the sum of all forces, just the proportional vector in the axial direction. The forces are also not equal on all blades at any one time. Just consider a surface-piercing propeller.

So how are these forces created? This seems to be the crux of the debate. Ultimately, it is due to pressure. The total force vector is the integration of all pressures on the blades at any one instant. Thrust is the component in the axial direction. (There is no point debating absolute versus relative pressure because it doesn’t matter – the integration of all pressures will resolve into a single force magnitude and direction, regardless of the type of pressure considered.)

The pressure on the blades can be generated in different ways (i.e., the light bulb analogy). All, however, have a helical frame of reference. Hence the term “screw propeller”, where some terms are similar to a screw thread (e.g., pitch as the axial distance advanced for one rotation). The geometric frame of reference, however, is where the analogy ends.

A propeller does not screw itself through the water. The best description that I can find after 30 years is that a propeller is a set of wings flying along a helical path. Actually, each blade is a short wing of many foils flying along many individual helical paths. Propellers of modest loading “fly”, and will generate forces, using all of the same physics we know about airfoils. (These forces are commonly devolved into lift and drag as useful parameters for airfoil performance.) Pressure differential creates force, from which the useful axial thrust component can be acquired.

So what creates this pressure differential? We know from airfoil physics that lift-wise pressure differential is most significantly affected by angle of attack (leading to a pitch angle against the reference helix) and camber (the nose-tail curvature in the blade). This is common for all propeller types. What is different between types is what happens when water is the fluid and the “lifting” or “pulling” pressures are large.

If the “pulling” pressures are large enough, some water on some of the blade will change state from liquid to gas – cavitation. The change of state alters the macro fluid properties of density and viscosity and the local shape of the cavity changes the flow lines. Beneficial “pulling” pressure actually can increase just after cavitation begins due to additional flow line curvature and then rapidly decline when the cavity significantly increases in size. At some point, the cavity on the “pulling” side (and I am intentionally not calling it the “suction face”) gets so large that it is contributing a substantially smaller part of the whole than the “pushing” side (“pressure face”). This mode is called super-cavitating. It is also the inspiration for surface-piercing propellers where the upper half of the propeller is intentionally pulled above the waterline as a system compromise, allowing the “pushing” pressures on the lower half of immersion to do (virtually) all of the work.

Unlike a simple airplane airfoil, the helical framework of a propeller blade points the foils toward the far field source of water inflow. A simple analogy, wind is created when the fluid (air) is pulled toward a lower pressure, or pushed away from a higher pressure. A propeller creates additional motion in the incoming flow, called induced velocity, from similar pressure differences. Therefore, the speed of fluid that a propeller “sees” will be larger than the advance velocity of the vehicle. (Just think about a bollard pull condition where the boat speed is nominally zero.) The fluid particles are pulled toward the propeller from ahead, carried through the blades, and pushed aft. Successful propeller design and analysis – regardless of the calculation methodology – requires knowledge of the flow conditions upstream (the wake field) and the ability to model the induced velocities and ultimately the forces.

Still, it is all about pressure. The “momentum” approach is a worthwhile model for macro level understanding of propeller performance. Momentum, velocity and pressure are all part of a particle vector understanding of any fluid. However, it does not help us understand the effects of cavitation or the implications of angle of attack and camber on pressure.

A single lift value on any foil can be achieved with an unlimited combination of angle of attack and camber, where an increase in both increases lift. However, while you may have a common lift magnitude, you will not have like distributions of pressure. Certain combinations of angle of attack (pitch) and camber may have greater “pulling” peaks at the nose or near maximum thickness (which can contribute to different levels and severity of cavitation). They will also have different drag (a necessary part of the total force and pressure picture). This is the behavior that makes cupping work. The extra curvature at the trailing edge – where cavitation can do no harm – allows a reduction in angle of attack and pressure peaks near the leading edge (where cavitation is most harmful).

It is very possible to introduce so much camber into a foil that the angle of attack can becomes negative for a given desired force vector. This brings me to slip. This is one of two terms, in my humble opinion, that have probably done more harm than good (“hull speed” being the other). The calculation of slip requires a formal definition for pitch – but which pitch? Many flat-faced propellers use “face pitch”, to which can be added cup; thereby making the relationship invalid. Some propellers with radial variations in pitch are stamped with a pitch from a given radius; others with a calculated “mean effective pitch”. The point being that any technical relationship where one of the variables can be arbitrarily defined is largely meaningless. Back in the day when all propellers were similar and lightly loaded, and where cupping was not applied, slip was a useful tool. Now, however, the domain has to be very narrowly focused to use the concept of slip in any meaningful way (e.g., sub-cavitating propellers with flat-face and no cup).

Finally, the momentum model tends to hide the effects of viscosity. Drag on the blade (the principal contributor to a propeller’s torque requirements) is not only a force vector from the integration of pressures, but it also includes the frictional effects of surface texture and fluid viscosity. Friction carries fluid with the foil, causing a rotation behind (and also somewhat in front of) the propeller. Without a “swirl corrector” such as a stator or contra-rotating blade set, all propellers will impart rotation to the fluid leaving the propeller.

Finally finally, the tension that exists between empiricists (those who build a model of understanding from observation) and scientists (those who use a model of understanding based on laws and principles) is costly to any technical advancement. One must never dispute the value of the observation out of hand, as it is the observation that gives us the inspiration for the physical laws. However, observations in isolation are risky as there is no guarantee that the behavior is being observed and communicated faithfully. You cannot project one data point into a broad conclusion. That is why we have massive data sets and peer review processes. By appreciating the physical laws – as they are known at this point – one has the opportunity to fit the observation into the broader understanding of the topic and make really valuable discoveries. And that is where real advancement comes from.

That’s my story and I’m sticking to it...

Don MacPherson
HydroComp

 

Nice summary Don!

Yes, I've been lurking. I was sorely tempted to feed the troll by saying something like "Propellers don't push the boat, the thrust bearing does"...

Your summary of the similarities and differences is a good one.

 

How do propellers propel a boat

There is a lot of information in those posts and thank`s for effort which will take me a lot of time to go through and most of it is what is usually accepted explanation. Some are giving a different version which I agree with such as one clearly wrong theory is saying that lift is due to longer path on the suction side of a profile.

To me some of the inbuilt features such as lift for a propeller which is intended to push a boat forward is not needed.

Two images show the flow of air through a propeller (fan).

 

How do propellers propel a boat

Nothing wrong with trying there are a lot of questions waiting for answers from me. I do not think I have ever bought something that I did not have to alter to make it work better or suit my needs.

 

I am still doing my peer review on the previous posts.
If everyone is using the software to design propellers that has not all of the information or correct info required to do the work we must be in trouble.
If you think the software has the right info try reverse building a design by giving the results you want then asking for a model.

 

Actually Tom, that _is_ how the best propeller design software works: We input a desired blade pressure distribution and the software calculates a blade geometry to yield that.

 

How do propellers propel a boat

ckesson

Thank`s cmckesson. you are still finishing up with a very disturbed inefficient flow so there must be errors in the best of propeller design software.

 

There are errors in the best of human propeller designers as well.

Now, no offense intended but I am still trying to understand if there is some point you would like to arrive to, or is this thread just about some leisure general chatting about propellers?

 

The question still is who has got the propeller theory right. There are infinite software for propeller design all finish up with less than satisfactory results. And previous posts have tried to explain their version.

It is amazing the technology and cost,time used to design some simple objects that a child could and do whittle up with a pocket knife.

 

Are you certain that the results are less than satisfactory or are the results just less than perfect, whatever that might be. If you are hoping for a lossless system with 100% efficiency, that will be a first in a mechanical system.

Perhaps you seek a design method that is 100% predictable. That is a much more reasonable quest. The thing that prevents most success in that wish is the lack of knowledge of all the input variables. Tests made by a team at VMI of the Wright brothers propeller showed that they had achieved an efficiency of 85%. Modern propellers don't do much better than that. What that says is that observational research backed by some simple equations ain't all that bad. Maybe if you can control all the possible input variables, some software might be "satisfactory" or even very near perfect. In the real world you cannot know all the variables with precision, much less quantify them..

 

I can not disagree with that.
How about someone rating the prop in post 55 for air medium use. How does it stack up with a sc.. propeller other propellers. Does it comply with accepted propeller design. Maybe even a boat prop use. There are similar ones in marine use.Nice and simple.

 

The hydrodynamics of propellers is extremely complex. More than hydrodynamics, we can talk about a mix of hydrodynamics, thermodynamics and chemical physics.
The most comprehensive theory goes through Navier-Stokes equations, integrated with turbulence and cavitation/ventilation models. However, these equations cannot be solved by hand. They require powerful computers and a lots of memory even for simplest of 3-D cases. Hence, we have to use simplified models for ordinary design tasks.
Being simplified, they cannot cover all the cases, so the designer has to be aware of the limits of their validity in order to make an informed choice of the most appropriate model for a given design job.

Hence, I hope that you will understand that asking about the ultimate propeller theory and who's got it right has very little sense. Everyone's got it right when they chose the right tool for a given job.


That said, when talking about engineering, the mechanical or energy efficiency is only one part of the equation. Pretty often other variables like cost, weight, ease of manufacture, structural robustness and simplicity of maintenance get rated higher than a mere mechanical efficiency.
So, from that point of view, the prop you have shown in the post #55 is perfectly ok if the main geometrical features (diameter and distribution of pitch) have been calculated correctly for the intended use and if you are able to manufacture it. For sure it will be much easier to construct than a propeller with a more complex distribution of chords, cambers, rake, skew, cupped tips etc.

Although, I would have at least explored a possibility to introduce an amount of linear variation of chord from root to tip. That is a very simple modification which can considerably improve efficiency without increasing the complexity of construction.
And would then make a step further - if you are able to manufacture a prop with a linear variation of chord and a linear variation of section angle along the blade, then you will obtain a very simple propeller with characteristics not too far from much more complex ones.

A more detailed design would probably introduce some other features in order to improve the efficiency at the design speed. But as a general rule, the more optimized is a propeller for a certain speed, the less efficient it will be in the off-design conditions. So that issue has to be pondered too, before getting too deep into design.

Resuming, the choice of how to model, design and manufacture a prop will depend on the general Design Specifications of the project, which will have to be based on the considerations involving the whole life cycle of the propeller. It is a general rule which is valid for any engineering job.

Cheers

 

Am I missing something, or has this thread taken a fork? The original question was about theories to model the physics of how a propeller propels a boat. It has become a debate on whether calculation tools can appropriately model that physics.

It is first useful to separate propeller design (of the component) versus propeller sizing (of the system). I have always contended in the importance of the latter – and built a career around it. A good propeller that is well matched to the vehicle, engine, and transmission will always outperform a really great propeller that is not well matched. As a simple analogy, the best racing surface-piercing propeller is useless on a tug. Most of the performance problems that we see in our business are related to an ill-matched system, not in the propeller itself. Tom is right in saying that propellers are (or seem) like a simple component. Most are a solid piece with some twist in the blades. Finding the best shape and twist for the given system is part of the challenge. However, it is only after you have a good system (propeller sizing) does it make sense to optimize the component (propeller design).

(One a side note, as component propeller designs have marched toward an ideal efficiency, the available room for improvement gets smaller and smaller. Most of the new design challenges are about corollary design concerns, principally noise and vibration.)

Regarding the computational tools used in propeller design, the best are based on some model of the physics. However, as the hydrodynamics of propulsor operation is extremely complex and messy, none can currently accurately capture the observed “in the field” performance without some calibration (or correlation, depending on your perspective) to empirical trial data. There are tools that are nothing more than a statistical representation of observed empirical performance. And so long as the scope of the problem is within the scope of the data set, these are very useful.

Our business is about the “quantitative”, such as accurately predicting the thrust and torque for a given propeller in a given operating condition. The complexity of the calculation model is directly related to the “wiggle room” available for the task. In other words, we match the tool to the task. Our models are also largely empirical, but they are built upon a foundation of the physical laws I described in my last post. Some of these methods represent a propeller as simply a type (B-Series, for example) and its principal parameters (diameter, nominal pitch, BAR, and blade count). With proper correlation, we can accurately use these parametric models to predict performance. The failures typically come in the model of the system. How does the boat itself affect the propeller? Is the proposed top speed and/or drag prediction reasonable? Are the effects of shaft angle and stern flow part of the model? Is cavitation breakdown included? Without these, differences in theoretical laws are really of no consequence because the system is not being properly modeled.

Our more complicated tools – the ones that use a greater definition of propeller geometry and inflow characteristics – are still quantitatively very strong. However, these tools are not needed until later design stages where the hull and stern characteristics have been established and locked in after the system has been properly matched. It is only then that we bring the tools with greater horsepower (and cost) to bear.

Navier-Stokes equations and similar computational methods are just other (albeit more refined) representations of the physical laws. Sometimes they are useful – particularly for observation of potential flow problems. Other times, not so much for what you get. They often require so much computational juice that the model of the system – the really important model, in my opinion – is simplified by the user to the point where it no longer represents the real condition. That being said, these tools can be exceeding useful for those really complex questions if you can justify the cost, time, and complexity of running the computation.

Daiquiri made one comment in passing that I’d like to revisit. None of these calculation models are worth anything if the manufacturing process has substantially greater slop (tolerance) than the calculation model. We work with nearly 100 international marine propeller builders (who knew the world needed so many!), and I can tell you that the detail we put into the definition of the geometric design is largely lost except for a few builders that serve the navies or the very highly-priced propeller system companies.

As Leo has quoted on numerous occasions (and I believe he was re-quoting another): All models are wrong, but some are useful. That is true with propeller physics, so debating the relative merits of computational approaches or even different (complementary, not competing) physics models becomes something like a Coke-Pepsi or PC-Mac argument. There will always be the acolytes of a particular position. My comments – for what they are worth – are for the rest of us.

Don MacPherson
HydroComp

 

You have made two excellent posts here, Mr. MacPherson. Thank you for sharing these bits of your vast experience.