# axial vs mixed flow pumps

Discussion in 'Jet Drives' started by Waverunner Eagl, Jan 7, 2013.

1. Joined: Dec 2012
Posts: 11
Likes: 0, Points: 0, Legacy Rep: 10
Location: California

### Waverunner EaglJunior Member

I see axial and mixed flow pumps mentioned a lot, but i have been unable to find any information on the actual differences. Any suggestions of where to start searching?

thanks

2. Joined: Jun 2009
Posts: 1,294
Likes: 262, Points: 83, Legacy Rep: 1165
Location: Sweden

### baeckmoHydrodynamics

Well, I guess your question stems from the never ending debate on jet pump types?? Here’s the explanation:

Pumps are designed for specific combinations of pressure and flow; the operating point. Pump performance is presented in graphic form as a diagram, showing pressure as a function of flow. The pressure that an impeller can produce is a function of its tip (or peripheral) speed, which per se is a function of rotational speed times diameter.

This means that for a pumping operation where the pressure is high and the flow low (in relative terms), you will use a radial impeller; its tip speed is high, but the flow channels are relatively narrow. In the other extreme, you may want a high flow at a low pressure, and alas, now you use an axial flow impeller. Here the important dimension is flow area, while the demands for tip speed is lower. As the driving speed often is restricted by the motor, you will find all combinations of the spectre, from radial to pure axial. A ships propeller is just a pump impeller of extremely high specific speed, operating without a housing. The product of pressure (N/m2) and flow (m3/sec.) is the hydraulic power.

In order to describe, or classify the typical proportions of the various shapes, we use a non-dimensional type number, called the specific speed, Nd. In the SI-system of dimensions it is the product of shaft angular velocity (in radians/second) times the square root of flow (in m3/second), divided by the pump enthalpy (in m2/s2) to the power of 0,75. In the technical system we use rpm*Q^0.5/Head^0.75; which obviously is not dimensionless.

This type number corresponds to the main proportions of an impeller and thus what it is designed for. In the picture below you find various impeller types connected with their specific speeds. Now, the losses in the impellers are mainly depending on two phenomena; friction and turbulence. Those factors combine to make an optimum in a certain range of specific speeds, see the second picture. Consequently we try to combine shaft speed and dimensions so that the design will end up with the highest possible efficiency.

For ship propulsion we strive for high flows at low flow velocities (a requirement for high propulsive efficiency). At the higher specific speeds, the pump efficiency tends to fall. But the losses in the jet wake are reduced (the “impulse efficiency” is still rising) with increasing flow volume, which means that there is a propulsion optimum slightly to the right of the Nd giving max pump efficiency. Recent research has produced improved efficiencies for axial pumps, compared to the previous bench-mark designs, which you will notice as an increased number of high-specific speed (read axial…) pumps for jet propulsion to be seen in the future.

#### Attached Files:

• ###### Nsbild1.pdf
File size:
102.2 KB
Views:
1,412
2 people like this.
3. Joined: Nov 2011
Posts: 247
Likes: 14, Points: 18, Legacy Rep: 117
Location: New England

### johneckSenior Member

Very thorough explanation Baeckmo. From a practical standpoint for a given application an axial pump can be made with a smaller overall diameter than a mixed flow pump and therefore weighs less and takes up less room.

Most (all?) jetskis have axial pumps, but they have also begun to be used on larger vessels. Some of the makers of very large waterjets have recently begun selling axial jets for naval vessels.

4. Joined: Dec 2012
Posts: 11
Likes: 0, Points: 0, Legacy Rep: 10
Location: California

### Waverunner EaglJunior Member

Wow thank you very much for typing that out for me!

My question actually stems from jetskis! I have a few of the older stand up type jet skis, for the model i have there are two pumps available for it that are readily available. I have heard one referred to as an axial pump and one as a mixed flow pump. From reading through jet ski forums, which is a lot of here-say and very little information is soundly mathematically backed, the axial pump is better suited to skis that can achieve higher rpm's. On the chart axial pumps are labeled for the higher Nd, but are a lower efficiency than the mixed flow pumps. I dont know if this is an accurate assumption, but does the efficiency of a mixed flow pump fall off quickly at the higher Nd?

So as far as this applies to skis, if I am going to use a motor that can reach high rpm's i do want the axial pump, otherwise I'll be better off with the mixed flow? I guess there is a lot of assumptions about the pumps in this question that would have to be made.

5. Joined: Dec 2012
Posts: 33
Likes: 2, Points: 0, Legacy Rep: 57
Location: little Rock, AR

### Jet_LoveJunior Member

Not 100% sure but I was thinking Jetskis all have mixed flow in stock form.
A axial flow is a pricy upgrade know as a "Mag pump" by jetskiers. Also in ski's they are always the same size far as taking up room.

I have run each, and will say once I went mag I never went back. Much more hook up in white water. When riding stand up ski's you can't get near prop wash with out falling over for no hook up. With a mag you never lose hook up.

I don't understand why ski's don't come with axial pump's. Most newer pwc's do come with a prop that will work in a mixed or mag pump. Know as large hub prop's.

One can spend 2g's on a nice name brand mag pump set up for a PWC or just get a large hub prop and this pump cone that changes the flow from mixed to axial like this.

I do run the R&D cone and it's the same as a true mag pump.

6. Joined: Dec 2012
Posts: 33
Likes: 2, Points: 0, Legacy Rep: 57
Location: little Rock, AR

### Jet_LoveJunior Member

7. Joined: Dec 2012
Posts: 33
Likes: 2, Points: 0, Legacy Rep: 57
Location: little Rock, AR

### Jet_LoveJunior Member

The pix are of a 6 vain stock mixed pump with and without poor man cone. The last is a 12 vain true mag.

#### Attached Files:

File size:
7.2 KB
Views:
1,785
File size:
7.5 KB
Views:
2,245
• ###### pump3.jpg
File size:
144.8 KB
Views:
4,920
8. Joined: Nov 2011
Posts: 247
Likes: 14, Points: 18, Legacy Rep: 117
Location: New England

### johneckSenior Member

That is not quite correct. Most use axial pumps. That means that the impeller is in a straight cylinder. The diffuser at the back end of the pump is where the difference is. By changing the shape of the diffuser, one can change the operating point for the pump. These certainly could be tuned for max top end speed or better low end performance. It would be interesting to measure the ID of each diffuser at the outlet to see the difference.
A mixed flow pump will have some radial flow in the impeller casing, meaning that the rotating part will increase in diameter aft of the inlet.
Not sure why they put 12 vanes in the diffuser of the OR pump, maybe just to make it look cool.

9. Joined: Dec 2012
Posts: 33
Likes: 2, Points: 0, Legacy Rep: 57
Location: little Rock, AR

### Jet_LoveJunior Member

Hi John,

I'm mixed up now.

You are right on being able to tune them. In addition to the larger hub their are shims too make the cone longer or shorter on each the poor man and the true mag.
Its funny they sure sell or market them as changing from mixed to axail.

10. Joined: Nov 2011
Posts: 247
Likes: 14, Points: 18, Legacy Rep: 117
Location: New England

### johneckSenior Member

Not sure what to say about the marketing, but it is not truly possible to change from mixed to axial simply by changing the diffuser. Do they also change out the nozzle?

I have attached an image of a mixed flow pump, rotor and diffuser. The flow would be upward in that pump.

#### Attached Files:

• ###### Mixed_flo_pump.pdf
File size:
32.1 KB
Views:
990
11. Joined: Dec 2012
Posts: 33
Likes: 2, Points: 0, Legacy Rep: 57
Location: little Rock, AR

### Jet_LoveJunior Member

No, but some do bore the nozzle.

12. Joined: Jun 2009
Posts: 1,294
Likes: 262, Points: 83, Legacy Rep: 1165
Location: Sweden

### baeckmoHydrodynamics

To understand the matching of jet impeller, stator vanes and nozzle to specific engine and operating characteristics, you have to have a basic knowledge of the functioning of rotodynamic pumps. It is a complex system, and the theme of the day is, as always: it depends…… Here some insight without diving too deeply into pump technology. One problem is that as soon as there is a touch of “performance itch”, your vision is blurred by the amounts of snake oil coming with the magic performance bits and pieces that are offered…….

Now, there are two basic conditions to be kept in mind, when jet performance is considered:

A/ The hydraulic power produced by the jet pump (in Watt) is the product of flow (in m3/s) times pressure (in N/m2). Due to inherent losses, the engine has to deliver a higher power, in the proportion of 1/(pump efficiency).

B/ The flow is limited by the hull inlet configuration and the rotational speed. At low vehicle speeds and full power ( acceleration “hole shot”), there is always a risk for cavitation, that limits the max flow possible. Increase possible only by improved shape of vanes and hull intake or size.

Now if we first look at what happens when we change the engine characteristics. We select an engine with the same power as the original, but this power comes at higher rpms. Keeping the flow constant (determined by the inlet geometry), the pressure must remain constant. But the impeller pressure is primarily a function of tip speed (i.e. rotational speed times radius) and outlet swirl angle. So, either the impeller outlet radius or the swirl has to be reduced to keep the pressure constant.

With a diagonal-flow pump (moderate specific speed Ns), like the American Turbine, Berkeley et c. , the outlet diameter can be adjusted by turning it down. The producers have a basic set of impellers, with, say four or five different pitch angles, which can be adjusted to match a wide variety of engines. A higher engine speed is accommodated by a “more axial” outflow in this type of pump. PWC’s in general have axial flow pumps with cylindrical impellers (high specific speed Ns). Here the primary matching is done by pitch change. In both cases, an increase in rotational speed will produce cavitation at a lower flow for a certain vehicle speed; noted as a thrust loss.

If we consider a situation where we increase the engine power at constant rpm´s, we have to stay with the same flow as before, due to inlet configuration and cavitation. This means that the pressure has to be increased. With an axial impeller, the medicine is to increase the outlet pitch in order to produce more swirl, tip speed being held constant. At the same time, the jet nozzle has to be reduced slightly to keep flow constant, in spite of the increased pressure. Problem is, that when you squeeze more pressure (=more intense swirl) out of an axial impeller of given diameter, it reacts by displacing the flow outwards towards the periphery, leaving a wake on the hub close to the outlet.

This is where the various cones and alternated stator vanes come in. In order to regain control over the flow in the hub region, you increase the hub diameter. At the same time the stator vanes have to meet the flow coming with a different swirl angle and redirect it into fully axial direction. To manage this increased load, the vane row has to have more working surface in terms of length and number of vanes. Simultaneously, the stator hub has to be adjusted to meet the bigger impeller hub. In some cases, the impeller may operate ok even with a hub wake, but the flow in the stator has to be controlled by increasing the stator hub. This is what you see in the photos above. Remember: there is no “fits all”-solution and there are no free lunches – it all depends…..

13. Joined: Oct 2010
Posts: 604
Likes: 33, Points: 28, Legacy Rep: 390
Location: N Carolina

The post above, has lots of good practical info, but to understand a little more about axial vs. mixed flow, there should probably be a discussion of what to types of flow are "mixed" for the latter case.

The most common type of water pump in industrial applications is the centrifugal pump. Water enters the eye of the pump and vanes sling the water outward by causing it to rotate. This style of pump can develop a lot of pressure, but is generally not as efficient in applications of high flow and small pressure increases. All of the flow direction changes, the flow acceleration and the flow deceleration tend to involve some waste of the input power.

For true axial flow, the impeller is working more like a boat prop or airplane propeller. This minimizes directions changes for the water stream and gets an efficiency advantage due to reduced friction and other flow losses.

For the simplistic boat prop example, one big efficiency loss is actually the induced rotational motion in the outlet flow. This swirl take energy to create, but does not create any forward thrust. One method of improving axial flow pump efficiency is to try to use vanes downstream of the impeller to convert the rotational flow energy to increased velocity in the direction of desired flow.

Mixed flow pumps just blend the two concepts together. Start with an impeller that has an inlet similar to an axial flow pump. Since a rotating impeller tends to spin water, just go ahead and let it happen and then use vanes to get back to non-swirling outlet flow at the discharge. You tend to get a little more developed pressure across the impeller, but this can be converted back to discharge velocity with a nozzle that reduces flow area.

The mixed flow design is probably more complex, but if you do it right, all of the components tend to work together very smoothly to produce the desired results.

For boating, also consider the mechanical side of the issue. An axial flow pump with variable pitch is one choice. A mixed flow pump with an adjustable geometry nozzle is probably an easier / more reliable way to get the same performance flexibility.

14. ### El_GueroPrevious Member

How would you reverse the turbulence without using more energy?

15. Joined: Oct 2010
Posts: 604
Likes: 33, Points: 28, Legacy Rep: 390
Location: N Carolina