Resistance factors, planing hull at low speed

Some quotes from an excellent paper on the topic:

"From Model Scale to Full Size. Investigation on turbulence stimulation in Resistance Model Tests of High Speed Craft, Bertorello et al (FAST 2003).

"The following conclusions can be deduced from the results of the research conducted on three geosims:
- with the naked small models (L < 2.0m) the full turbulence of the flow can be achieved only at the highest speeds (Fn > 0.60)..."

That means for their smallest model Series 64 (L=1.805m) Rn is about 4 million. They also tested geosims with L=2.347m and 4.694m with three different size wire trips and also with sand grit.

"Disturbance of flow should be necessary for the considered hull from to ensure complete turbulence... but it is very difficult to establish the type, the dimension, the longitudinal position and the evaluation of the additional resistance".

They conclude with:
"A small model (L < 2.0m) seems inadequate for the model-ship correlation in the range Fn < 0.60, because of the appreciable laminar flow on the total resistance and of the uncertainty about the value of the additional parasitic drag..."

Personally, I would be very wary of data for Rn < 4 million with or without turbulence stimulators.

All the best,
Leo.

 

Thanks Leo, I used to think Rn=>10^7 is appropriate number.

Besides You mentioned series 64 but we are discussing series 62 where models were 4-8'.

I believe friction is only one side of problem, the other one is transom flow separation that is not modelled.

As to small model tests, would like to mention that Delft Series was studied on 1.6 and 2.0m models.

 

For the yachtsman when fuel was $1 a gallon there were lots and lots of planing yachts made. Now because of fuel cost most are semi disp but there are millions of older boats out there w skippers running them dead slow. There was a discussion on Trawler Forum about whether it was more economical to run a twin on one engine or two. There is great interest in running fast boats slow and in the near future I think the interest is going to increase as fuel will surely become more dear as India and China compete w the rest of us for oil.
One flaw in your discussion is that there is a wide range of resistance in planing hulls, semi-disp hulls, sailboat hulls and full displ power boats. We once compared a 40' Willard FD hull w a GB SD boat. Could a better comparison be made? If specific boats are used and these boats are common then we can fudge up and down to have meaningful thoughts about how another specific boat could fit into the question.

 

This is good point.

As sample, I can show range vs. speed graph for powerboat. It can be seen that there is 'economical planing speed' as well. Semi-displacement mode is likely to suffer in terms of fuel efficiency. Of course, this is data for one particular boat, but...

 

Still no comment about "Cross section of all Series Done by hull type (FL 0pt4).pdf". Or it is still not scientific enough to indicate or prove anything?

 

I only mentioned 4 million as a lower limit for turbulence stimulation.

Below Rn=10^7 I have reservations about the ITTC skin-friction formula, but that is a separate argument altogether.

Sadly there are few (Telfer-like) experiments for any hull series. That is why the paper I cited is so valuable.

Yes, that is a very difficult flow problem.
Even semi-empirical techniques (like those used by Patrick Couser or Doctors and Day) are not much help when the transom is partly-wet. Differences of over 100% between experiments and predictions are still apparent for Fn < 0.5

We really do have a strange situation...
Applied mathematicians are suspicious of using inaccurate experiments to validate their computations; naval architects are suspicious of our predictions because they don't accord with experiments.

I'm not sure if this is due to Prof. Martin Renilson, or some other savvy cynic:
"Everyone believes experimental results except the person who conducted the experiments; nobody believes CFD except the person running the code".

Leo.

 

OK, I just found this file now. What does it show? Draw trend line for hard-chine and round bilge hulls, and see the difference between them - 2 or less.

What You are doing is comparing the extreme points (looking like drop-out points). Methodologically this is completely wrong because we do not know other parameters of compared points: say, L/B, etc. can be very different. These tests were done in different tanks and processed by different methods, turbulence issues, etc. No serious analyst will make conclusions based on drop-out points, once we are looking at set of statistical data, not particular boats. Say, how one can compare wide prismatic chine hull with B/T=8 with round-bilge displacement hull with L/B=2 having much less wetted surface, even if they have same DLR? Again, to evaluate the effect of hull section shape, other parameters should be comparable.

There is good paper by J.Robinson 'PERFORMANCE PREDICTION OF CHINE AND ROUND BILGE HULL FORMS' on such analysis of round-bilge and sharp chine hulls. Results are quite similar to scattered data You posted and those we use for performance predictions.

What I see in posts of PermStress is lack of culture of using engineering/research data: first comparing hulls with different DLR; then using series 62 beyond it range, now taking drop-out points of scattered statistics. These tricks are good for salesmen but not for researchers

 

I only trust calculation methods validated by sea trials of boats Say, in our practice for some systematic series we used to apply additional resistance factors to get reliable results.

 

If we now look back at the FnL 0.3-0.4 measurements of the Series 62, what would be the result of not using stimulators. Re was something like 4e6, thus the BL may have been laminar even over half way of LWL. That would mean, that the friction drag would have been much less (~10-40%) than the ITTC assumption, which was reduced from the measured drag. Thus the "measured" residual drag would have been much less than the real one.

Unless there is some other mechanism counteracting the reduced friction resistance like in the case of smooth vs. dimpled golf ball. That would be different separation properties in an adverse pressure situation, which in a prismatic hull would likely be limited to only the wet transom. Is this an issue for a prismatic hull at FnL 0.3-0.4? Certainly there is flow separation at the sharp transom edge at almost any speed (which does not mean the transom would be dry). I guess the length of the separation bubble would determine how much of the transom is dry. At Re=4e6 the BL is certainly turbulent well before the transom, but the BL can be much thinner than in real scale due to transition far from the bow. How big is the effect at those speeds?

 

There is no simple answer to these questions, and it is still an active area of research. If you hunt around you will find many references to the work of L.J. Doctors, Kevin Maki and many others.

Some workers (like Doctors or Couser and Molland) add a fictitious extension behind the transom to approximate the hollow. I have difficulties with this model because a cavity vented to atmospheric cannot sustain a pressure, and therefore it cannot create waves. You are very welcome to disagree and formulate your own theory

One way of getting a rough estimate is to estimate the drag behind a fully wet transom (where the flow turns almost 90 degrees around the corner and generates large waves) and the case of a fully dry transom with an infinitely long hollow trailing behind (which makes no waves). The "truth" is somewhere between these two extreme cases. In fact at high Fn the wave resistance is approximately equal to the mean of the two extremes for many hulls.

Several attempts have been made at finding regression equations for a broad variety of hulls with transom sterns over a wide range of Fn.
See, for example:
"The hydrodynamics of high-speed transom-stern vessels"
Author: Robards, Simon William
http://unsworks.unsw.edu.au/vital/ac.../unsworks:3426

You should also realise that the flow behind a partly wet transom can be very unsteady, so talking about a boat travelling at a constant Froude number might not very useful.

Good luck!
Leo.

 

In original paper Rn=(3...5)*10^6 corresponds to FnV=1.0.

 

One of possible approaches is to use wind tunnel tests - those can be used to evaluate resistance at very slow speeds where wave drag can be neglected.

Probably such speeds are of no practical interest for small craft but those are used for offshore structures, etc. Will try to find some results.

 

Yes, it is a well-known technique.
Dr. A. Armstrong (of Austal Ships) used wind tunnel tests with NPL hulls to estimate the viscous interference between the demihulls of a catamaran. In his experiments he used "double body" hulls, which is equivalent to having a completely flat free-surface. He also added fairings behind the transom stern.

I'd be interested in any other papers you know of when you have time.

Leo.

 

It is nice to know your opponent read only every other word/line/sentence of what is posted... very scientific indeed...


Trend lines show difference ~1.75, still in the order of 2. close to what I did post at very start of discussion.

Drop out points, although always suspect, (and I understand it perfectly well, thank you) can indicate possible extremes. In general, no one ever is 100% sure he will NOT get accidentally taken to extreme of this kind by some trivial reason.
BT from 2 to 8... Do you really suspect so large hull-form variations in the "cross section..." presented? Or it is a discussion technique?
When using not drop-out points, difference in the order of 2 is still here. Drop-out points show that in certain cases much more is possible. In case nobody did read it, I'l repeat real world data I have at hand:

Boats of similar length, and displacement:
a) yacht, with ~3-4kW/t and folding prop (inefficient by definition, figures as low as 25% are published) do the same or even more speed as
b) planing-shaped boat with ~8-9kW/t and real propeller (efficiency not less than 40%).
It should have something to do with really big difference (certainly more than ~20-30%) in resistance. Or am I still missing something?

Naming someone "amateur" as soon as he do not readily agree is very high discussion/engineering culture indeed.

Germanisher Lloyd guys and customers I sometimes work with (in my spare time, of course ), have slightly different opinion.

...Already old point in this tread, but might be of some interest to check professionalism and level of engineering of some sources I was referring to at the start of discussion: http://bethwaite.com/. A graph from F. Bethwaites book "High Performance Sailing" i did post here (attached "DSC05651-to post.jpg"). Same data, but shifted/compressed/extended right/left to be at same FnL is also attached "R to Fn.pdf". Yes, DLR is different.
But please check attachment with Series 64 graph "64 graph.JPG". With increasing Length/displacement ratio, it has less and less influence. So at least general trends and order of magnitude can be expected to be correct, if DLR of compared hulls is different by ~20%, IF it is HIGH.

As to Series 62 beyond range, all arguments please address to authors of Series 62, I quote "... accuracy about =/-6 % at low speed...". They considered low speed results good enough to be published then, and still consider them good enough for general conclusions (including 2 times more resistance below hull speed) just now. (sorry, it was a few days ago; may be today their attitude is different, I did not check)

From statistics I take order of magnitude here. It is quite good indicator on what areas/types/etc. to avoid from the start, and what sort of gains/loosses could be expected. This correlate quite well with original posters question on what kind of hull shape to select. Or I am again wrong?

 

I just didn't noticed the file; yes maybe not too scientific but it is weekend and sometimes I stay with family My daughter was sitting with me (and trying to read PNA) when I was typing the reply.

By the way, about 'hull speed' - since I posted quote from Larrson, nobody replied...

But I said is below 2, so who is right?

Why do You reference extreme points then???

B/T 2 to 8 is only a suggestion/sample, from my side. There could be other factors, we do not know because we do not see the data for those boats. FYI, difference in total resistance at FnL=0.3 between wide and narrow boats is about 50%, for same length and displacement.

Yes, I agree - 20...40%, even 70% is some cases. But not "2-3 times", at least if we compare options with similar design parameters.

Why You post graphs from book for amateur sailors then? I would never expect this from professional.

Good for You; I believe hydrodynamics is not Your field anyway.

Sorry, but I don't see 'level of engineering' on that website. I see regattas and gallery

OK, at least now You know how to compare hulls - use same DLR For Bethwaite's graphs, L/D^0.333 of compared options is 7...9 - quite a lot.
Yes, I agree at high L/D^0.333 difference is small; for 7...9 it is still significant.
And this graph You posted for 64 is for residual resistance, not for total resistance, isn't it?

Raw results are published, not included in results of series.
You can see Rn for series 62? Can You see the trend of CT? For me, no further explanations are needed.

To select proper hull, take design speed and calculate (or test) few options (including round bilge and chine). That's it.