hollow waterlines, rowboat

[wolle]

Hello,

i read several threads in this forum, and after reading this one:

designing a fast rowboat

i got inspired by these finnish rowboats designed by terhohalme, so i did a redesign of my "project", this came out:
http://www.boatdesign.net/forums/att...1&d=1327908277
http://www.boatdesign.net/forums/att...1&d=1327905780
http://www.boatdesign.net/forums/att...1&d=1327905780

Because of the plump bow, the waterlines are now hollow, wich should be avoided, as i thought. I didnt found something about the negatives of hollow waterlines on the forum with the search, and because the thread of nordvindcrew is more about open water rowing, i dont want to place my questions there.

Main question:
Would the medium speed performance be significaly increased by avoiding these hollow lines, shortening the ends and/or giving the bow and transom more rake?

( I did a test, this change would reduce lwl (10%) and wetted surface (8%), increas entrance angle (from 7 to 10°), the resistance calculation in delftship shows not a markable change, the Rtot for lower speeds are the same, for higher speeds even inreased (that is clear), is the software calculating wrong? )

Second Question:
Am i wrong with this?: As long, as the diagonals stay convex or straight at the end and this is the same with the curve of the sectional areas, there is
nothing to worry about hollow waterlines in the sense of negative effect on speed performance (on a rowboat like this).

Any suggestions would be welcome!
(dimensions are metric, but i can ad a linesplan in feet/inches if you like)

These are the basic specifications
- rowboat for sheltered waters, (optional small sail)
- 2 persons, but should row with only 1 easily
- not for racing, but speed is very important
- medium good initial stability, as needed for pleasure use, thats why beamwl is not so small than terhohalmes designs
- good reserve stability, that is why beamoa is so big, sides are flared, what leads in conjunction with the plump bow to the hollow lines.
- dont like rowing with oar overlap
- should have good reserve buoyancy for additional weight
- construction could be stripplanked wood, glassed outside/inside



wo

[gonzo]

You can have hollow waterlines with a plumb and even a reverse bow. That is a rather burdensome boat, so it will carry a heavy load. If you want fast, make the ends finer and narrow the beam. There is no scale to the drawing. If it is metric, the boat must be 4.5, 9 or 18 meters long.

[wolle]

Sorry for forgetting the dimensions.

LOA = 5.7m = 18.7´(stations are 0.3m = 11.8")
BWL = 0.9m = 35.4"
BOA = 1.3m = 51.2 "

displacement (at designdraft 0.1m) 180kg = 396lbs

[Leo Lazauskas]

Originally Posted by wolle

Because of the plump bow, the waterlines are now hollow, wich should be avoided, as i thought. I didnt found something about the negatives of hollow waterlines on the forum with the search, and because the thread of nordvindcrew is more about open water rowing, i dont want to place my questions there.

Main question:
Would the medium speed performance be significaly increased by avoiding these hollow lines, shortening the ends and/or giving the bow and transom more rake?

Hulls with hollow lines will tend to have greater surface area for
the same displacement which will affect the resistance and hence
speed of the boat. (They are forbidden on Olympic shells, but that
shouldn't be an issue for you.)

You can estimate the resistance of the boat using Michlet. It
should give better estimates than Kaper, but it would be a worthwhile
exercise to compare the predictions.

I'm not sure which version of Freeship/Defltship you are using,
but some versions can export a file in Michlet format.

Good luck!
Leo.

[wolle]

Originally Posted by Leo Lazauskas

.....You can estimate the resistance of the boat using Michlet. It should give better estimates than Kaper, but it would be a worthwhile
exercise to compare the predictions.
...

I tried to work me in this program, after smal starting problems (one line missing in the in.mlt file from delftship, maybe of my version 3.20) i managed to get some results.
Nice program!

i have to exercise further, but i want to say two small things

- when you google michlet, vou land here:
http://www.cyberiad.net/michlet.htm
the download link leads you to version 807, not the 920

- when playing with michlet, i realised that the LOA in the in.mlt file is signifikant for the results in resistance and wave-making, but the file delftship produces, the value of the project setting is taken, so i had to write the value of LWL from deftship in the in.mlt file for LOA. Is this correct procedure, because michlet only sees underwaterbody, lOA michlet has to be LWL delftship? (maybe others have noted the same).

the graph shows the resistance of hull1 as shown in the first post in pale-color, a shorter design with less wetted surface (hull 2, modified in delftship) with LOA value manually adjusted in full color, and the green curves from hull 2 without adjustment.

as very preliminary conclusion, im still not familiar enough with michlet, i would say that while the frictional component is lower for hull 2, even at lower speeds the increase in waveresistance is substancialy higher.

wo

http://www.boatdesign.net/forums/att...1&d=1328003709
compare_51_57.png

[Leo Lazauskas]

Originally Posted by wolle

I tried to work me in this program, after smal starting problems (one line missing in the in.mlt file from delftship, maybe of my version 3.20) i managed to get some results.
Nice program!

i have to exercise further, but i want to say two small things

- when you google michlet, vou land here:
http://www.cyberiad.net/michlet.htm
the download link leads you to version 807, not the 920

- when playing with michlet, i realised that the LOA in the in.mlt file is signifikant for the results in resistance and wave-making, but the file delftship produces, the value of the project setting is taken, so i had to write the value of LWL from deftship in the in.mlt file for LOA. Is this correct procedure, because michlet only sees underwaterbody, lOA michlet has to be LWL delftship? (maybe others have noted the same).

the graph shows the resistance of hull1 as shown in the first post in pale-color, a shorter design with less wetted surface (hull 2, modified in delftship) with LOA value manually adjusted in full color, and the green curves from hull 2 without adjustment.

I updated the Michlet page to the latest version. For some reason the ISP reverted the page!

Yes, Michlet only uses the underwater portion of the hull. You can input
lines above the water and use the sinkage and trim options to raise and
lower the hull but Defltship only outputs the underwater offsets and
dimensions when it creates a Michlet in.mlt file.

If you post the .fbm file (or email it to me) I can have a look at the Michlet
output to see if anything can be improved.

Leo.

[wolle]

Originally Posted by Leo Lazauskas

.....
If you post the .fbm file ......

The .fbm file should be attached to the my first post, after the .jpg and the .pdf

Does michlet take for the resistance calculation the value of LOA as written in the in.mlt file or is it extracting the length out of the geometry defined in the offset table in the same file?

It looks to me, that the first is the case, as in the graph where curves for

a, full color:
LOA in the in.mlt file corrected to the value lwl=5.17m
b, greenish:
LOA as delftship puts in the in.mlt file during export. delftship writes in the value defined in the project settings, it isnt related to the geometrie, in this case the value from the original hull, 5.7m. (maybe newer versions of delftship behave different, i use 3.20)

shows a lot of more resistance for b. again, everything else except the LOA value in the in.mlt file are the same in both cases.

i know, have to read the michlet manual completely, i should not argue before that, but anyway, thanks for your efforts.

wo

[Eric Sponberg]

I'll be interested to see if you find anything significant in Michlet regarding detectable differences due to the hollowness of waterlines. I am going to go out on a limb and say no, there won't be much difference, but I'll be prepared to stand corrected.

Two thoughts: First, hollow waterlines are often an unintended by-product of the hull-shaping software. The way surface modelling codes are set up, you have to really work hard at manipulating the surface control tools in order to NOT have hollow waterlines. There is no reason why if you have a plumb bow, then you get hollow waterlines. It is possible to have straight or full waterlines with a plumb bow, you just have to work at it harder.

Second, boat with hollow waterlines will exhibit its characteristics when it moves through waves. Hollow waterlines will affect the pitching motion (hobbyhorsing) of the boat in waves. Boats with hollow waterlines will hobbyhorse more than boats with full waterlines. Boats that hobbyhorse are typically slower than those that don't because energy is consumed in the hobbyhorsing rather than in allowing the boat to go faster.

Does Michlet take into account resistance due to waves and boat motions such as pitching, heaving, rolling and yawing?

Eric

[Leo Lazauskas]

Originally Posted by Eric Sponberg

I'll be interested to see if you find anything significant in Michlet regarding detectable differences due to the hollowness of waterlines. I am going to go out on a limb and say no, there won't be much difference, but I'll be prepared to stand corrected.

Two thoughts: First, hollow waterlines are often an unintended by-product of the hull-shaping software. The way surface modelling codes are set up, you have to really work hard at manipulating the surface control tools in order to NOT have hollow waterlines. There is no reason why if you have a plumb bow, then you get hollow waterlines. It is possible to have straight or full waterlines with a plumb bow, you just have to work at it harder.

Second, boat with hollow waterlines will exhibit its characteristics when it moves through waves. Hollow waterlines will affect the pitching motion (hobbyhorsing) of the boat in waves. Boats with hollow waterlines will hobbyhorse more than boats with full waterlines. Boats that hobbyhorse are typically slower than those that don't because energy is consumed in the hobbyhorsing rather than in allowing the boat to go faster.

Does Michlet take into account resistance due to waves and boat motions such as pitching, heaving, rolling and yawing?

Eric

I agree that hollow lines are not necessarily due to the plumb bow. The
classic Wigley hull is one simple example which is convex in all views.
I also agree that for the current boat, the difference with and without
hollows will be quite small.

The main objection to hollows is that they usually mean there is often
(but not always) a shoulder somewhere further along the hull, and this can
kick up a larger than wanted wave.
In my experience, hulls with slight hollows can have lower resistance but
only for a very narrow range of Froude numbers.
I have never really understood why they are banned in Olympic rowing and
kayak events. Extreme hull shapes with bumps and hollows (e.g. Ward's
Optimal Symmetric Ship) have low wave resistance for a narrow range of
Fn, but it's not likely that they would be successful in competitions.

And no, Michlet does not take into account ambient waves or ship motions.
(I occasionally use a combination of Michlet/Flotilla and SMP to estimate
added resistance in waves, but not for rowing shells and kayaks.)

[Leo Lazauskas]

Originally Posted by wolle

Does michlet take for the resistance calculation the value of LOA as written in the in.mlt file or is it extracting the length out of the geometry defined in the offset table in the same file?

It looks to me, that the first is the case, as in the graph where curves for

a, full color:
LOA in the in.mlt file corrected to the value lwl=5.17m
b, greenish:
LOA as delftship puts in the in.mlt file during export. delftship writes in the value defined in the project settings, it isnt related to the geometrie, in this case the value from the original hull, 5.7m. (maybe newer versions of delftship behave different, i use 3.20)

shows a lot of more resistance for b. again, everything else except the LOA value in the in.mlt file are the same in both cases.

In Michlet, the actual offsets (i.e half-beams) are calculated using the
displacement volume, length and draft. So, yes, the length is important.
(Sorry, but I have no control over how Delftship exports lengths into the Michlet file.)
Delftship does create a very smooth, fair hull for Michlet, as can be seen in the attached.

I found the following for Volume = 0.181m^3, and Draft = 0.100m:
With LWL = 5.70m, BWL = 0.8948m and S=3.30m^2.
With LWL = 5.17m, BWL = 0.9865m and S=3.28m^2.

It is not surprising that the wave resistance (in particular) will be significantly
greater for the shorter hull. The wetted surface area, S, is only about 1.3%
less for the shorter LWL, so skin-friction will be about the same for both
lengths. How form drag is affected is anyone's guess

[wolle]

Originally Posted by Eric Sponberg

I'll be interested to see if you find anything significant in Michlet regarding detectable differences due to the hollowness of waterlines. I am going to go out on a limb and say no, there won't be much difference, but I'll be prepared to stand corrected.
.....
Eric

To prove that, i did a second modification to the originall hull. The original one is called 1 (or ps2_1) in the attached files, the modification 2 (or ps2_2).

Things beeing equal (or nearly equal)
- LWL
- displacement
- wetted surface
- draft

Things changed
- avoid of hollow waterlines
- thus more Volume in the ends, higher Cp, now 0.6, before 0.56
- thus smaler beam, now 0.845m, before 0.900m
- thus higher entrance angle on waterline, now 10.1°, original 7.5°

In the image with the resistances, the values of the original hull 1 are in pale color, the one of hull 2 in full color.
As you can see, the Rw (of hull 2) in the case of high speed is lower, that is what to expect, because the Cp 0.6 is apropriate for a s/L like this. But at lower speeds, the waveresistance of the hull without waterlines (2) is clearly worse in comparison with the original hull (1). Please feel free to disprove my results, files .fbm and .mlt are attached for both hulls.
Because im not interested in racing, means rowing most at the time at max speed, i think 1 fits better to my purposes as listed in my first post.

The arise from hollow waterlines because of a plump bow in this case comes from the big flare at the sides, i think, the software is not the bad guy. Transition of the big flare at midships to a plump bow means either hollow waterlines or substancial increas of curvature of the sheerline towards the bow or something inbetween.

Do you think, that the hobbyhorsing thing is a substancial problem with this hull, if the lines are kept as they are in ps2_1?
I found this:
http://dashewoffshore.com/dashew226.asp
(more promotion than facts) but even the hull there has some hollowness in the bowarea.


wo

[Eric Sponberg]

A couple of thoughts. First, to my eye both models have hollow in the waterlines, hull 2 less so than hull 1. Second, the differences in total resistance are really small. At 1.8 m/s (approx. 3.5 knots) the difference in resistance is about 0.0014kN = 0.31 lbs). Third, this difference could be due as much to differences in L/B ratio, Cp, and entrance angle, not necessarily the hollowness of waterlines. So, I don't think anything is proved here regarding the hollowness of the waterlines. It would be instructive to see two other hullforms of the same characteristics as these with respect to length, breadth, wetted surface, and Cp, but without any hollowness whatsoever in the waterlines, and see what the differences are compared to these models. You might see similar effects, but what would the total resistances be by comparison?

Eric

[wolle]

Originally Posted by Eric Sponberg

A couple of thoughts. First, to my eye both models have hollow in the waterlines... Eric

thats true, and in contradiction to my statement in my former statement, it might be easyer with other design methods than with delftship. i looked this morning on the print of the linesplan on the table and had an idea, how to avoid the hollowness 100%, lets see...

Originally Posted by Eric Sponberg

...... It would be instructive to see two other hullforms of the same characteristics as these with respect to length, breadth, wetted surface, and Cp, but without any hollowness whatsoever in the waterlines, and see what the differences are compared to these models. You might see similar effects, but what would the total resistances be by comparison?
Eric

I am not certain if i understand you correct. lenght, beam, wetted surface and Cp kept constant? draft as the only variable? why then two other hullforms?

wo

[Eric Sponberg]

The original question was which hull form has less hesistance, one with hollow waterlines or not, and then, can you tell the difference using Michlet? So, in the models provided so far, both have hollow waterlines, and while one model has different wave resistance than the other, their Cps are different, and so the difference in resistance is probably due to the difference in Cp, not the hollowness of waterlines. Cp and waterline hollow are related, of course, so the point of two more models is that they should have no hollow waterlines whatsoever, but they should have the same Cps as the first two models so that there can be a direct comparison. What is the resistance of the models with non-hollow waterlines vs. the models wtih hollow waterlines? Which model of the four has the least resistance? Can this be detected using Michlet?

The primary performance factors are length, beam, displacement, wetted surface, and Cp. The differences in draft are likely to be minor for the sake of keeping displacement, wetted surface, and Cp the same. The hullforms will have different sections, waterlines of course, and also buttocks, so it is hoped that the differences in these shapes will be detectable in Michlet. To do that, you have to separate Cp from waterline hollow, and that is the purpose of the two extra models.

Eric

[river runner]

I'm not sure I'm understanding what you mean by hollow waterlines, but if you mean what I think you mean, most high performance canoes and seakayaks have hollow watelines, as do most of the lines I've seen for racing sailboats, so I'm a bit confused by this train of thought. I understand that pitching is much more of a problem with row boats, so are your comments regarding hollow waterlines restricted to rowing craft?