250HP outboard Speed prediction for heavy boat

[Joakim]

Here is a rough measurement of my propeller. The last column is the profile thickness %. The pressure side is completely flat and I measured alfa along it. Both the leading and the trailing edge are equally blunt (about 6 mm thick 8 mm from the edge at r=0.1 m) and the profile looks symmetric.

I tried using MH 126 at two first section and the MH 114 at two outer ones. This reflects the thickness rather well. At 6.5 kW I get ~1300 rpm, ~1100 N and 52% at 6 kn. All else is fine, but thrust is way off compared to VPP. The sections have a bit concave pressure side and clearly asymmetric profile, thus not what my propeller is.

Then I tried flat plate. Now 6.5 kW results to over 1700 rpm, almost 1000 N and 46% at 6 kn. At 1344 rpm the figures would be 2.3 kW, 400 N and 52%.

Then I tried Clark Y, which is flat on the pressure side and got at 6.5 kW a bit over 1500 rpm, over 1000 N and 52%. Using 1344 rpm I got almost 4 kW and a bit over 700 N and 56%.

Thus now I got 6 different values to choose from. Flat plate is obviously wrong. I have very hard to believe in drag of 1100N, thus maybe the Clark Y option is the closest one. Would the symmetric profile cause the power need to jump from 4 to 6.5 kW? Or is the motor outputting only 4 kW. Or a bit of both? Maybe I could play around with xfloil to find the difference.

And that one is also very close to what I already had from Wageningen B-series.

Joakim

[Joakim]

I did some tests with xfoil. I started from naca 4312, which looks rather similar to Clark Y. Then I made it almost symmetrical (trailing and leading edges identical). Xfoil doesn't survive from totally blunt trailing edges, thus only almost.

At zero AoA naca 4312 had Cl = 0.43, Cd=0.0077 and the symmetrical 0.63 and 0.011. The symmetrical one had much flatter pressure distribution and despite higher Cl Cp minimum was at 0.7 compared to 0.9 of naca 4312.

According to JavaProp Cl decreases from 0.5 to 0.15 at the tip and Cd decreases from 0.019 to 0.014. Is there induced drag included or why are the Cd's so high in JavaProp?

Setting Cl to 0.2 puts Cd of naca 4312 to 0.0094 and the symmetrical one to 0.012. Again the symmetrical one has clearly "better" Cp with a maximum of ~0.5 compared to ~0.7 on suction side. On very tip of the pressure side the Cp's are 0.8 for the symmetrical and 1.4 for naca 4312.

Thus according to xfoil the symmetrical profile would have about 30% more drag at the same lift. How would that transfer to propeller efficiency?

But it seems the symmetrical profile would be clearly less prone to cavitation.

Joakim

[Guest625101138]

Joakim
You have provided a lot to comment on.

The one comment I will make now is that you are assuming the power demand for the prop is the rated power. This is unlikely to be the case. I expect it will be designed to have reserve power to cope with high wind. You may have experience that supports or counters this.

It is not unusual to overprop a diesel to get its cruising operation at lower revs so you get maximum efficiency. I expect the motor will go close to achieving rated power less the gear train losses. It should easily achieve your stated speed if the drag is correct.

I will look at the blade information you have provided.

The intention with providing the text file is so you are aware of how to load your own data. The other thing that many people do not realise is that the pitch angle does not need to be perfect. In fact the optimum will have different AoA at different radial points so the best blade will not have perfect pitch angle. So where do you actually measure the pitch? I determine it at the 75% radial position for the given foil usually in its standard configuration. Obviously a cambered foil will still provide thrust under these circumstances. Most of my blades work with slightly negative AoA when optimised.

Generally I find the standard foil sections available within JavaProp are thicker than what ends up optimum. That is why I made my own program so I can choose any foil section. I have never bothered to see if JavaProp can accept other foils. I know there is a version that has more sections than others. Also the software has changed a little since I have been using it. However it is worth pointing out that for the low angles of attack typically seen on water props the foil shape does not make a huge difference to efficiency. A flat plate will be significantly lower like 10% or so but most foils will be within 5% of each other.

JavaProp certainly accounts for induced drag.

If you are also looking at different foils you should try JavaFoil. I found the interface much easier to work with than Xfoil. There is a lot of similarity with JavaProp. I have compared JavaFoil on a number of foils with test data from Selig and it aligns well. Both these programs can be downloaded and run directly from your computer without internet connected.

I commend you with the precision taken with measurements. With a little more perseverance you should be able to close the gap between what JavaProp is producing and what you are getting with the boat.

Rick W.

[12553Jack]

Hi First post this, it may or may not be useful. We fit fishing Cats with Honda Outboards largest so far 2 x 150hp on 10 mtr x 4 mts cica 6 ton displacment, for performancre I have used the Castle Marine prop calc program ( free from their website ) for the prop size I use Honda technical to get somewhere near and then fine tune with another prop if necessary. Some customers require absolute top speed others best fuel economy for a given speed,because of all the variables trial and error is the only way I have to go on. I think manufactures are very reluctant to make staements on performance just in case it does not perform as predicted.We have not conducted sea trials yet on the 10 mtr mentioned but if anyone is interested I will post the Data.

Jack

[Guest625101138]

Originally Posted by 12553Jack

Hi First post this, it may or may not be useful. We fit fishing Cats with Honda Outboards largest so far 2 x 150hp on 10 mtr x 4 mts cica 6 ton displacment, for performancre I have used the Castle Marine prop calc program ( free from their website ) for the prop size I use Honda technical to get somewhere near and then fine tune with another prop if necessary. Some customers require absolute top speed others best fuel economy for a given speed,because of all the variables trial and error is the only way I have to go on. I think manufactures are very reluctant to make staements on performance just in case it does not perform as predicted.We have not conducted sea trials yet on the 10 mtr mentioned but if anyone is interested I will post the Data.

Jack

Jack
I will certainly be interested in the data. Verification of weight at test and a bit more detail on hull shape would be of interest as well.

Also do you know if outboard manufacturers offer a heavy duty leg for these heavier loads? Is there anything special with the mounting and gearing?

Rick W

[Guest625101138]

Originally Posted by Joakim

Here is a rough measurement of my propeller. The last column is the profile thickness %. The pressure side is completely flat and I measured alfa along it. Both the leading and the trailing edge are equally blunt (about 6 mm thick 8 mm from the edge at r=0.1 m) and the profile looks symmetric.

I tried using MH 126 at two first section and the MH 114 at two outer ones. This reflects the thickness rather well. At 6.5 kW I get ~1300 rpm, ~1100 N and 52% at 6 kn. All else is fine, but thrust is way off compared to VPP. The sections have a bit concave pressure side and clearly asymmetric profile, thus not what my propeller is.

Then I tried flat plate. Now 6.5 kW results to over 1700 rpm, almost 1000 N and 46% at 6 kn. At 1344 rpm the figures would be 2.3 kW, 400 N and 52%.

Then I tried Clark Y, which is flat on the pressure side and got at 6.5 kW a bit over 1500 rpm, over 1000 N and 52%. Using 1344 rpm I got almost 4 kW and a bit over 700 N and 56%.

Thus now I got 6 different values to choose from. Flat plate is obviously wrong. I have very hard to believe in drag of 1100N, thus maybe the Clark Y option is the closest one. Would the symmetric profile cause the power need to jump from 4 to 6.5 kW? Or is the motor outputting only 4 kW. Or a bit of both? Maybe I could play around with xfloil to find the difference.

And that one is also very close to what I already had from Wageningen B-series.

Joakim

Joakim

I have now looked at your data. The attached spreadsheet gives your data in the form required to import into JavaProp. You copy the cells shown in yellow and paste into the import window available on the geometry page. JavaProp will extrapolate.

I have selected the Clark Y foil because it has a 12% section and likely to be the closest section available within JavaProp to your prop given it is around 12% thickness.

With this set up JavaProp gives me 3.19m/s, 4.75kW and 800N at 1344rpm. The efficiency is 58%. I believe you will find this will be quite accurate unless the blade is nothing like a foil. The Clark Y foil is flat on one side and curved on the lifting face.

I also tried a flat plate and modified it by increasing the angle by 3 degrees. The efficiency drops to 53% and power goes up to 5kW. I expect this would be the lowest possible efficiency.

You could spend more time to accurately measure the angle of the blades rather than assuming the pitch is accurately set. Then replace the beta I have calculated for import to JavaProp.

I expect the JavaProp data to be reliable. If you accept this then it would be possible to design a prop that will do better. I believe you could get close to 75% efficiency running at 900rpm on a prop that has sufficient strength. Strength at the root will be the limiting factor on efficiency. Power required will be under 3.5kW and engine torque 14Nm. This is still well below the available torque.

If you were inclined to do this then I expect you will improve fuel economy substantially but you are unlikely to get the engine to rev out. Top speed should be a little better than now.

The attached link shows low EAR 2-bladed props in the right size range:
http://www.australpropeller.com.au/fixedyacht.htm
I have never handled one of their props but the proportions are OK.

If you have access to fabrication facilities it is not too hard to make a prop. I can get small ones milled specifically but they are expensive and your loading is quite high so would need stress analysis. A fixed bladed prop will certainly impair sailing performance even if the blades are narrow.

Will be interested in anything you find that helps close the gap between your test data and the JavaProp derived data.

Rick W.

[Guest625101138]

Something else I should comment on.

Known foils have a defined shape and the pitch angle for a prop is given for the foil at zero angle of attack or the nominated offset. If the foil is cambered then it will have lift at zero AoA even if there is no offset.

Hence it is not unreasonable to adjust the blade angle until you get the required thrust at the given rpm. JavaProp will give you the power to achieve the thrust at the rpm.

I design my own blades and am not too worried about what I define as zero angle as long as the manufacturing matches my reference. If I was using a Clark Y I would use foil data with the back side at zero because the flat surface provided an easy reference. So all my data then relates to this.

If you get around to measuring the actual blade profile on the Yanmar then you can do lift curves for it with JavaFoil and find what matches it best.

Right now I am saying your prop efficiency will be between 53 and 58%. If you do more measurement you may be able to get a better answer.

Finally, symmetric foils do not make for efficient propellers. I know they are poor but I do not think they are as bad as a flat plate but please note I have not validated this. I have never bothered to create data for a symmetric foil to make the comparison.

Rick W.

[12553Jack]

Hi Rick

I will post the data when we launch, as we will crane the boat in the crane will having a weighing device fitted. The boat is made by GRP fabrications of Ashington England they have a web site and you may be able to obtain data from them. However we have also fitted 2 x 130 Hondas to a 8 mtr Cheetah Cat this is a commercial fishing vessel out in all weather and worked HARD, performance about 28 knots fully loaded Disp circa 3.5 tons. We have lots of Honda 90s working commercially with 1 pair I am told having run in excess of 7000 hrs. I would suggest that the modern outboard engineering is robust enough to withstand very hard commercial operation.I have to say we are exclusively Honda and it is the outboard of choice for most commercial fishermen but that is not to say other types are not as good.I would also say I am not theoretical in my marine engineering I fit it if it works I record the data and pass it on more a mechanic than a scientist. On the mounting side they are usually mounted on pods bolted to the transom these have to be robust and the forces calculated for the vessel and power output. The gearing is determined by the factory with no options ( that I know of ).I did consider setting up a data base where users could input what they have and how it performed. Say 23 foot Boston Whaler fitted with 200 hp honda size of prop fitted top speed max revs and fuel consumption it would then allow anybody considering that combination to compare to say a guy who fitted a 150 to the same boat and perhaps anther with a 90. Out of interest we fitted the 200 to the whaler it goes rather quickly !!!

Jack

[Joakim]

Rick
I did already measure the propeller (see message #31). I measured the angles using a 25 cm protractor along the flat pressure side. I guess the accuracy would be about 0.5 degrees with that method.

However, it seems Clark Y does not have horizontal pressure side. It is at about 2 degree angle to the reference plane in these co-ordinates: http://www.ae.uiuc.edu/m-selig/ads/coord/clarkysm.dat

So maybe 2 degrees should be subtracted from the angles measured, but can I trust this is really the reference plane in JavaProp?

Any of this has rather little to do with efficiency, which should be with all different approaches a bit over 50%. But I'm still have not been able to clearly define the operating point of the propeller. Using JavaProp and starting from the measured propeller angles and blade widths I can get many different results regarding thrust and power at the same rpm.

Now the propulsion I should need is that 750 N at 6.4 kn which makes 2.5 kW and thus less than 5 kW at assumed efficiency.

I should have 6.5 kW at propeller shaft, thus things are not adding up. I would really like to know where is the error. I know rpm and speed are very accurate. Thrust required should be OK, but maybe 10% error is possible. So what is left are that the engine is not able to output that 6.5 kW or the propeller efficiency is only ~40%.

Joakim

[Guest625101138]

Joakim
As you have determined, if you are referencing the angle to the flat side then you need to adjust to get it to align with the standard foil section. The Clark Y data will indeed be out by 2 degrees when using the flat side as reference.

I did not realise your third column was degrees. I have now entered your data adjust for the 2 degrees in my spreadsheet. I have redone the JavaProp data and copied it into the spreadsheet to produce a thrust and power curve for the prop at various boat speed for 1344rpm. It indicates that the thrust is about 850N and power is 5kW. I expect you will eventually determine this to be very close.

The conclusion from this is that you are not absorbing 6.5kW at the prop.

You are not reaching full engine speed but this could be governor setting rather than torque limit. It could be that the governor setting is not very precise. A better indication of the power will be the fuel consumption. The data sheet indicates 2.3lph at full throttle. Have you measured fuel consumption for sustained periods at full throttle. If it is only producing 5kW then fuel consumption should be around 1.8lph with prop doing 1344rpm.

Rick W

[Guest625101138]

An indication that the motor is working hard is that you start to get poor burning and the exhaust will show some black. If you are not seeing this when running at full throttle it could be another indicator that the maximum power is not being achieved.

Also if the engine revs are not dropping in a strong breeze it also indicates that there is more torque available and the engine is being governed at 1344rpm at the prop.

It is a mechanical governor and I could imagine it being imprecise and may need regular reset. Someone more familiar with the Yanmars may be able to comment on this.

Rick W.

[Joakim]

Rick
The file I included earlier was a file to be directly imported to JavaProp. JavaProp ignores the last field, which was the profile thickness.

I had already planned to measure the fuel consumption, but haven't found spare time for that. Now the boat is up for the winter until May. I try to measure the fuel consumption and probably even do a towing test in order to verify the thrust needed next summer.

There are two adjusting screws in the governor. One for maximum rpm and one for maximum fuel output. Someone has probably tampered the governor settings, since the sealing wire connecting the screws is missing. The engine revs a bit over 3800 rpm on idle and it used to rev over 3700 rpm with the 14x6 propeller the boat came with, thus it shouldn't be hitting the rpm limit in governor. The 14x6 propeller was put by the earlier owner and the 14x8 is the original setting.

There is no black smoke. Should there really be black smoke, when the engine reaches the design rpm? I thought black smoke is related to overloading the engine on lower rpm, but I'm not a diesel expert.

Joakim

[Guest625101138]

Joakim
You will see in my spreadsheet I have added 2 degrees to the blade angle you provided in your little table. The curve I have provided should now be very close to what your prop is doing at 1344rpm.

Black smoke is not a good thing as it indicates poor fuel burning. However if the motor is working at maximum torque and not being run lean you should see the exhaust darken up at any point in the rev range. If the motor is easily able to get to full revs under light load then it is clearly not being governed back. It could be the maximum fuel setting is low if it has been fiddled with.

As a matter of interest what speed did the boat do with the smaller pitched prop? You could do a check on operating point for this prop as well . This will help confirm the boat data.

Rick W.

[Joakim]

Rick
I only used the 6" prop for two weeks when I bought the boat (it was the lifted up for winter). I only changed the folding blades, which actually came with the boat. I have already sold the 6" blades, thus I can not measure them, but they looked identical except the pitch.

I didn't do very thorough measurements with 6", but it reached about the same boat speed as the current 8", clearly over 6 kn anyway. The rpm was well over 3700 rpm, but I can't remember the exact value. Calculating with these, the thrust by JavaProp would be only ~700 N, depending on the exact rpm and the way I reduce the pitch of the measured propeller.

These propellers are actually sold by Volvo-Penta and probably manufactured by Radice. It is now "the old 2-blade version" of Volvo-Penta, since there is a new one, which came 2008.

Joakim

[Guest625101138]

Joakim
I posted the attached image on another thread. It belatedly answers your question regarding which angle is used to define pitch. You can see there are three possible.

Rick W