Canoe with square bottom hull vs rounded hull

Discussion in 'Hydrodynamics and Aerodynamics' started by JosephT, Feb 11, 2019.

1. Joined: Feb 2007
Posts: 429
Likes: 116, Points: 53, Legacy Rep: 160
Location: California, USA

Doug HalseySenior Member

This is a very interesting study that could be very helpful. My compliments, Dolfiman!

I do have some comments & questions though.

Here's the summary of your results from the 1st pdf:

If I am interpreting correctly, you are comparing the drags of the 4 hulls at a constant value for the propulsive power. This means that the drag values correspond to different speeds (in fact smaller speeds for the flat sections.) It would be interesting to see the drag values compared at the same speeds for all the hulls. Do you have plots like that easily available?

Although this is an interesting design study for these particular cases, it doesn't necessarily mean that flat sections would be worse than round ones in other cases. I would point out that the flat sections are wider & shallower than the square or 1/2 square sections referred to earlier in this thread, so they should have higher skin-friction & residual drags than optimal.

Also, according to the figure, the prismatic coefficients of the round hulls are significantly smaller than those for the flat hulls, a difference that I think should give the round shapes a further advantage at the Froude numbers involved.

2. Joined: Aug 2004
Posts: 2,695
Likes: 319, Points: 83, Legacy Rep: 2040
Location: Port Orchard, Washington, USA

jehardimanSenior Member

Sorry I'm late to this, but based on the assumptions: Paddled, identical waterplane/submerged profile (meaning the round bottom is actually a U in the ends), absolute square section; the rounded bottom has a better thrust to speed requirement due to alpha flow losses induced by the paddler and should be faster. Note that just rounding the bilges of the square hull 5% will render any advantage moot based upon water conditions. FWIW I can see how one could set up an analysis that showed they were the same, but that would be flawed. Many people, including engineers, get hung up on "minimum drag' hulls without considering the proposed use. Slack bilges for ocean going sidewheelers and fast sailing ships, hard bilges for screw steamers, sternwheelers, shallow sailers. The hull form has more to do with intended use than minimum drag. As I have said before, there is no "best" hull shape, but there are hull shapes better suited for the intended purpose.

<edited for grammar and thick fngerd>

3. Joined: Jun 2009
Posts: 856
Likes: 106, Points: 43, Legacy Rep: 218
Location: Roaring Forties

JosephTSenior Member

Thank you for chiming in jehardiman. I will upload my results hopefully in the next couple of days so our resident hydrodynamic gurus can chine in (ha!). Totally agree on the intended purpose for each hull. That definitely matters. The use case for these canoe hulls is typical flatwater canoe paddling in the 5-8mph range. I'm using the John Winter's Kaper resistance method offered with FreeShip/Delft and will post the results in detail. I can also export Leo's Michlet files as well if anyone wants to examine. I would definitely encourage others to loft a basic canoe and post their own tests as well. This specific use case will be a good reference point for future discussions.

4. Joined: Aug 2004
Posts: 2,695
Likes: 319, Points: 83, Legacy Rep: 2040
Location: Port Orchard, Washington, USA

jehardimanSenior Member

Resistance methods do not predict propulsive efficiency and the effects of propulsion, i.e. the far field resistance and wetted surface only accounts for ~80-90% of the total ehp required to drive a hull. The other is buried in near field effects (which Michlet specifically does not handle), propulsion geometry ( which John Winter addresses with "With each stroke, the canoe is propelled forward, but, because power is applied off-center and at an angle to the centerline, the canoe does not track perfectly straight"), and seaway effects. Propulsive thrust, and its effect, must be resolved through the center of effective mass of the vessel in all 6 dimensions (x, y, z, roll, pitch, yaw). It is the secondary near field geometry effects that render this question moot for actual physical measurements, singlely they will be meaningless. Statistically, inferences can be drawn; and history points to slack bilge canoes.

<edit to clarify resistance predictors should include wetted surface.>

5. Joined: Jun 2009
Posts: 856
Likes: 106, Points: 43, Legacy Rep: 218
Location: Roaring Forties

JosephTSenior Member

Jehardiman, while this info may not matter to a recreational paddler, it means a lot for competitive racers. I’ll be paying close attention to small differences between both hull types. We also install small trim rudders to keep our vessels precisely on course. Additionally, crew use carbon blades with a marathon stroke that is very efficient...very little side-to-side motion. The smallest reduction in drag from either design (while offering suitable stability) translates into victory over hundreds of miles. These days we’re splitting hairs on hull speed.

I’ve raced nothing but rounded hulls to date with success. Why switch if they work? I’ve never considered this flat hull concept until now. So I must have a look and see what they can do. Rick W’s boats are impressive, but they aren’t stable without outriggers. They also rely on very smooth water, remaining level and continuous pedal (or electrical) power. Take away any of those and they simply don’t keep up with a human powered boat and a fit crew.

Ok after work nap over...back to work the other hull.

6. Joined: Aug 2017
Posts: 1,123
Likes: 480, Points: 83
Location: France

DolfimanSenior Member

Yes, you are right, this Cp differenciation introduces a biais in the comparison. This discrepancy on Cp is due to the comparison principle done "at same keel line" which prevent to adjust the sections aeras distribution towards fore and aft and so the resulting Cp. So I revised the approach :
At first, I tested the range of Cp leading to the minimum drag at the objective Fn 0,325 within the frame of the used formulation for the residuary drag. Cp = 0,52 leads to a minimum, it is also the lower limit of the range of validity for this formulation (so may be that lower Cp is even better but no data are available to underpin that).
Secondly, with Cp = 0,52 as additional objective, the 3 cases are so re design at same Lwl, Displacement, Cp, GM, Loa, Sheer line and free boards , the adjustements concerning Boa, Bwl, Tc (hull draft) and the rocker of the keel line. Here attached the revised document V1 (where I also add the drag curves comparison crossed the propulsion ones) and the updated application.

The main results are :
*** In terms of drag at Fn 0,325 : the Square version has ~ 11% more drag than the two others quite identical (Round and V Rounded Square), resulting both from more Sw (+12%) and more residuary drag.
*** In terms of speed at 80 W net propulsion (so around Fn 0,325) : the square version leads to ~ 3% less speed, 4,41 Knots instead of 4,54 knots.

As a preliminary conclusion, I would propose :
*** the difference in wetted surface (+12 %), all other things equal, seems a reliable result anyway.
*** the difference in total drag is also dependant of the accuracy of the residuary drag estimation, the above approach based on geometrical parameters and the Delft series data leads also to more drag for the Square version all other things equal.
*** due to the drag slope at Fn around 0,325, such difference leads to only ~ 3% less speed.

Attached Files:

File size:
300.9 KB
Views:
81
• Test Round or Square by Gene-Hull VE Canoe 2.3_2019 02 12.ods
File size:
1.7 MB
Views:
72
Doug Halsey likes this.
7. Joined: Aug 2002
Posts: 14,352
Likes: 619, Points: 123, Legacy Rep: 2031
Location: Milwaukee, WI

gonzoSenior Member

If the plan view of the profile is the same for both, then the round bottom will have a smaller waterline beam. You are comparing apples to bananas. However, if we consider the same displacement on the same waterline length, it would be useful data.

8. Joined: Aug 2017
Posts: 1,123
Likes: 480, Points: 83
Location: France

DolfimanSenior Member

You are right and I must be more precise : the comparison is "at same profile view of the sheer line" so inc. same free boards. As regard the plan view of the sheer line, as Boa is adjusted in each case for the other constraints (Cp, GM and displacement) , of course the sheer lines are different but within the same general shape.

9. Joined: Aug 2002
Posts: 14,352
Likes: 619, Points: 123, Legacy Rep: 2031
Location: Milwaukee, WI

gonzoSenior Member

OK, so do you compare equal prismatic coefficients?

10. Joined: Jun 2009
Posts: 856
Likes: 106, Points: 43, Legacy Rep: 218
Location: Roaring Forties

JosephTSenior Member

Following up on this thread. As it turned out the racing team involved opted for a chined hull with a proven track record so we dropped the square vs. round detailed analysis. Very good points gathered here though and many thanks for all the contributions. If I were to factor in all the input the main hull specs would be the same: They are of equal LWL
• They are of equal BWL
• They have an identical top view
• They have an identical profile view (both have flat bottom with no rocker).
• The only difference is one has a square bottom and the other has your typical rounded bottom.
AND
• They have equal prismatic coefficients
• They have equal displacement
With the lofting tools I use, forcing the target prismatic coefficients & displacements to be the same typically puts in slight hull adjustments, but they're within reason. I have the square hull modeled up per my previous post #30 referenced below. My initial loft with a round hull blew it out of the water on performance so I'm not going to pursue it further. If I get the time at a future date I will match up the prismatic coefficients & displacements. That is the beauty of a round hull. You can tweak it to make it faster. Not so many options with a square hull. With all the input either gathered so far I expect either the round or chined hull should be faster than the square hull. That has been my experience over the years with racing canoes & kayaks. At this point in time we have no desire for a square bottom hull! Speed aside, a square bottom would also not dance waves as well so. Thus, it's just no longer a candidate.

Canoe with square bottom hull vs rounded hull https://www.boatdesign.net/threads/canoe-with-square-bottom-hull-vs-rounded-hull.61808/page-2#post-848098

11. Joined: Aug 2017
Posts: 1,123
Likes: 480, Points: 83
Location: France

DolfimanSenior Member

Dear Joseph, interesting thread indeed, on core aspects of a Canoe optimisation. Some complementary thoughts :
On which basis you assume that a "flat Bottom with no rocker" gives an optimal hull ?
What is your geometrical definition of "rocker" exactly ?
Is it not linked with the speed for which you want to optimise your design ? Various speed leads to various optimal hulls.
What is your objective in this aspect : either a speed in absolute value (if you have no length restriction), or a Froude (if you have a length restriction) ?
Rocker, Cp and wetted surface Sw are naturally linked when all other things are equal + when keeping an isomorph shape of the hull, I tried to show this interaction with Gene-Hull here attached.

Attached Files:

• Canoe_Rocker investigation.pdf
File size:
132 KB
Views:
85
12. Joined: Jun 2009
Posts: 856
Likes: 106, Points: 43, Legacy Rep: 218
Location: Roaring Forties

JosephTSenior Member

Hi Dolphiman, for this comparison a flat bottom (no rocker was used). The proposed flat hull was perfectly flat (no rocker) so we stuck to that for the comparison. Sprint & flatwater racing boats typically have very little rocker. I normally race surfski's & flatwater racing canoes. Both typically have slight rocker to ease maneuvering. Yet another reason to avoid a flat, square hull!

13. Joined: Aug 2017
Posts: 1,123
Likes: 480, Points: 83
Location: France

DolfimanSenior Member

I tried to address with Gene-Hull most of your specifications in post#30 and your approach described above with a round hull inc. a flat bottom line, a displacement of 421 kg and a Cp of 61,94. For that, I proceed in 2 steps :
*** A first serie where the hull draft (Tc) is a variable input data and Bwl is an output data, by paying attention to use as much as possible rounded sections compatible with the objectives, in order to minimise the wetted surface (Sw). That leads to a comprehensive serie from quite squared sections (when Tc is small) to very rounded sections (when Tc is high).
*** The second serie is a refinement, crossing the first one : from one solution of the first serie, the Bwl value is taken fixed and I explore again various hull draft Tc, still at constant Displacement and Cp.
All the results are given attached, with the linesplan and data of each cases in annexes.
In the last figure, speed potential versus stability GM, all the points from the 2 series appear to be roughly on a same curve. Meaning that the search of an optimum by varying such or such parameter is not evident, does not lead necessarly to a better solution if the increase of the speed is at the cost of a decrease of the GM. At the end, one cannot avoid the trade-off GM / speed.

Attached Files:

• Canoe hull investigation.pdf
File size:
632.3 KB
Views:
95
14. Joined: Aug 2004
Posts: 2,695
Likes: 319, Points: 83, Legacy Rep: 2040
Location: Port Orchard, Washington, USA

jehardimanSenior Member

https://www.boatdesign.net/members/dolfiman.62547/;
Thank you for spending the time to pull this tread to the end. It humbles me to find that someone can prove the old builders rules of "this works better" are correct. While today there are many CFD programs, there are those of us that learned BPC from an older NA who said "do it this way".

15. Joined: Jan 2007
Posts: 32
Likes: 3, Points: 8, Legacy Rep: 28
Location: Washington N.C.

otsegJunior Member

Around 1999 I made a 12.5' model to test at the Stevens tank. It was a 1/10th size motor yacht hull. The model was tested with round bilges then tested with some add on chines typical of a planing hull. Round bilge less resistance to 26-28 kts. Chine bilge less resistance above those speeds. We could have read a boat design book by William Garden and saved some money.

Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.