On the hydro dynamics of flat bottomed kayaks

Thanks sailhand for your comment.

The model used take into account all the parameters given in this table. They are combined into factors, used in polynomial expressions for the calculations of the total resistance, which is seen as the sum of the viscous drag, the wave drag, and the induced drag ( induced separation ). Regarding the model and the remark that has been made concerning the result I produced, I have the feeling that the coefficients used in the polynomial expressions must be corrected, as the contribution of the different parameters to the various drags are highly depend on the hull shape on which they are applied. This is because these coefficients has been made by mathematical regression of towing tank measurements on hulls, whose shape's parameters were bound to a particular domain. When the model is used on other shapes, these regression are no more applicable.

Initially made for IRC monohulls boats in 1997, I have updated the coefficient of the model to take into account more modern monohull shapes, including chines and more flat bottoms. I have also determined a set of coefficients to be used for the determination of the drag of slender hulls, like catamaran and trimaran. These coefficients have been obtained by regression on CFD and tank test performed on sailing and motor multihulls, from 56' to 140'. These hull shapes are quite different from those of kayaks. This may be the reason why my model would be inadequate for drag prediction of kayak shapes. It lacks some adaptation.



Still, in my experience, the obtained ranking of the different kayak shapes makes sense, acccording to basic fluid mechanics. Even if the absolut values are too high, the same ranking would be obtained with a better hydrodynamic model. I don't think that the planning effect would play a role in the drag of an elongated kayak made for speed, as those I test in my small study. This effect would be more prominent for whitewater kayaks. But I think that your observation is correct, Sailhand, because the rocker / no rocker factor is indeed very important regarding the wave drag and the induced drag, even if this is more an "induced lift" that would cause a drag reduction of kayak with rocker vs kayak without rocker. Again, thanks for your observation.

 

Im not sure I have offered much of anything other than a question born from personal experience. I have used many different models in my design work however as l am in my sixties and a late comer to the computing world other than some simple programming in basic at university in the seventies l have not even the slightest comprehension of the complexities and algorithms involved in these models I use. My hat is off to the developers of Computer based development tools for boat design. These tools have proved remarkably accurate in my experience and most modelling software lve used have been very consistent. Many thanks to all involved in this software over many decades. The interfaces are simple enough for even someone like me to fumble around with them. Cheers

 

This is way less of a difference in resistance than I would have expected. At speed the sharp corners would create pretty significant vortices/turbulance as the water tries to go from the side to the low pressure region under the hull and back again at the stern.

 

Qualitatively, the phenomenon you describe here is correct. Quantitatively, I'm not so sure that, given the flow conditions, theses vortices could have a "significant" contribution to the total drag. I've extracted the graph from this publication https://www.researchgate.net/public...lysis_of_Bodies_with_Different_Cross-Sections; Trying to figure out, generally speaking, how strong would be this contribution.



At low angle of attack, i see a very little drag difference between the two bodies, so I conclude that, at low angle of attack, the section shape has a little influence for elongated bodies. The difference between the upright drag between rec and rec_smooth could even be less than what I've calculated. At angle, a squared section kayak may even be more efficient than a squared section kayak.

As the drag difference is going greater with the angle of attack, we can also say that the vortices/turbulence at the wedge stabilize the flow in the upright direction. As a result, the manoeuvrability of a kayak with squared sections will be lower than a kayak with rounded sections. On the other hand, average skippers would take advantages in sections with variable chine, helping them going straight without increasing the paddling effort. I guess that racing kayaks on flat waters do not have squared sections because of the ability of their skipper to paddle in straight line. Whitewaters kayaks should have no chine at all. This is consistent with what is observed in real world.

Vortices are not always a burden.

 

Alan C: How is your calculation procedure validated, and within what boundaries?

Edit:...and what conditions make the wetted surface less than the waterplane area??

 

I second Dustman, I think the angle at the midpoint will make more drag difference than a flat bottom.

 

I would imagine that long slender bodies(high length to displacement) would be less effected overall by the turbulence induced by the square cross section because there is much less water to move out of the way to begin with, so the proportion of this drag to overall drag would be quite low.

On the other hand, for a given displacement, with the longer body you have more surface area per unit of displacement so the extra wetted surface area could really add up for long skinny hulls.

 

I'd say that long slender bodies have a high length to displacement ratio!

 

Not sure exactly what you are getting at. Are you agreeing with me or throwing out a quip?

 

Reread your statement carefully then!

 

The regression coefficients for slender hulls have been validated by adjustments to datas obtained during 3 campaigns :

- systematic study of a 140' catamaran for the Jules Verne Trophy, using the CFD code ICARE, establishement of an analytical model for catamaran hull forms, realized by ECN ( Ecole Central Nantes, FR) 2005
- optimisation of the hull forms of a 70' footer motor trimaran using 2 CFD codes, comparison with analytical model : Dualsphysics, (University of Manchester) & Openfoam (Streamline company, FR), 2020
- optimisation of the hull forms of a 56' footer motor trimaran using 1 CFD codes, comparison with analytical model & sea tests measures : Dualsphysics, (University of Manchester), 2021

The range of shape ratios that have been validated VS the kayak hull forms ratios. In red, the actual ratios that falls outside the validated boundaries. This is where, to my opinion, the discussion should take place.

Oups....I must correct this error... Thanks for the remark. Let's correct it... The formula for the wetted surface calculation of TRI is quite complicated and I made a mistake. I review the calculations.

 

Thank you; if I understand correctly you are using results from multihull studies to validate the results for monohull? Then I must ask if, and how you are compensating for the influence of the hull-to-hull interference (wavemaking as well as induced) that is "hidden" in the multihull resistance results?

 

I don't compensate. Calculations of the catamaran hull shapes have been performed on bare hull, not in her catamaran configuration. CFD calculations of the trimarans have been done on bare central hull, bare float and central hull + bare float. The hull to hull interference has been estimated in only on case : in order to estimate the sole contributions of each hull's drag, when exploiting the engine rpm measurements, at sea, on the 56' tri.

So, except in the case estimation from sea test, the compensation of the hull to hull interference was not required, and the comparison with CFD calculations was direct.

 

It reads as intended, notice the use of parentheses. "a word, clause, or sentence inserted as an explanation or afterthought into a passage that is grammatically complete without it, in writing usually marked off by curved brackets, dashes, or commas."

 

Dustman, you have written "long slender bodies(high displacement to length)", which is wrong per definition. Long slender bodies are typically characterized by LOW displacement to length or HIGH length to displacement ratios, i.e. the opposite to the sentence you have inserted!