Michell v CFD and EFD: Round 2007

Sorry, Rick, but I don't know what you mean by "one of these fish tail hulls". Can you attach a picture?

Cheers,
Leo.

 

Leo
I have attached the off1.mlt file from a run for a length constrained hull. The stern is better described as whale-tail.

Rick

 

That looks like Godzilla is trying to cheat something by manipulating the transom hydrostatic resistance and the wave resistance.

The best way to try to stop that happening is to add contraints on the transom beam, draft and/or area. Note, however, that Godzilla might compensate for that by producing strange bulges and hollows elsewhere. Sometimes it's like trying to keep jelly in a fishing net.


Cheers,
Leo.

 

Leo
The height of the flat transom ends up the same height as the depth of the wave trough at the stern. So it makes sense.

I suppose in practice the sinkage would cause the wide stern to drag heavily so the result may be different if sinkage was included.

Rick W

 

That is probably just pure coincidence. Wave elevations calculated by Michlet are not accurate close to the hull because they do not include the local field component. The local field decays fairly rapidly, and becomes insignificant about one ship length behind the hull.

Leo.

 

Michlet Viscosity Limits

Leo
I have found your Michlet limits on viscosity do not suit some of the cool fresh water locations I have been designing for. I have been using the limit but would like to get up to say 1.6E-6 for some cold water applications.

I think this is likely to be a significant factor in the gap I have with some of the performance prediction on the slender hulls.

Rick W.

 

Hey Rick, I'm playing around with that a bit too. Have you tried changing the viscous resistance factor to simulate a changed viscosity? I'm not sure if that really works but it's an idea. I'm looking at how roughness might affect the optimum shape.
John

 

John
I know roughness does not make much difference. If you have ever handled a shark you realise that there must be merit in a somewhat rough surface if natural selection is sound. So something like 40 grit paper could be better than mirror smooth. I have seen power data on a roughly finished hull compared with the same hull mirror finished and almost impossible to pick the difference. I have actually contemplated placing strips of fine hair or fur every meter or so down the hull to dampen the energy in the boundary layer.

I do not know how Michlet applies the viscosity but the top limit of 1.3E-6 is inside what I would like to use for some applications. How ever it applies, it does make a noticeable difference as I have tried different values ranging from 1E-6 to 1.3E-6. I do not know if this is a published correction to the ITTC line or if there is some other factor applying. I expect Leo has made a sound basis for the way it is applied. I know he considers the Grigson method to give more accurate results. I have not compared different viscosity using the Grigson method.

I expect you would find water depth and water temperature more significant drag factors than surface finish for a canoe. If you can manage the perfect roughness it will have less drag.

Rick W

 

Roughness does matter. If it didn't, there'd be no need to remove barnacles. Some guys from the Australian Navy use a form factor of 0.1% per day that the ship is in the water. That translates to an increase in the skin-friction of 36.5% over a year. Of course they don't leave ships in the water that long before scraping them clean.

Michlet uses the viscosity to calculate the Reynolds number that is used in the ITTC line and Grigson's line. It is also used in estimates of the boundary layer displacement thickness, however that should not have a huge effect except for small hulls.

I'll try to increase the allowable limits when I get some time. The dole office are making me attend a course on how to write a CV for a few days next week

All the best,
Leo.

 

Leo
From the point of view of roughness, my meaning was there is likely to be an optimum somewhere between 40 grit type sandpaper and mirror smooth. I have established with accurate measurement that you should not expect a large improvement in performance by achieving mirror finish over roughness perceptible to hand. NASA demonstrated the benefit of riblets for example.

John had used a factor of two for viscosity although primary to see how the optimum hull would change with added viscous drag. No doubt something the size of barnacles add significant drag.

I will look forward to wider viscosity limits in Michlet when you can get around to it. With the unconstrained human powered hulls I expect the water temperature is more significant than sinkage between actual and measured data. The testing was done in water running off snow in Calgary Canada.

By the way my Canadian friend has now set a new distance record over 24 hours on water using a Godzilla generated hull. The record was previously held by a surf ski paddled by a much younger athlete. Greg has written a report with photos - see link below. In calm conditions he could get 11kph with 120W.
http://www.adventuresofgreg.com/HPB/HPBmain.html

I truly hope your CV writing bears results - good luck.

Rick W.

 

1. No doubt that riblets have some beneficial effects, but I'm not so sure about random grit. Maybe some roughness near the bow would promote transition to turbulent flow, but I don't think it would be wise to roughen the entire hull.

2. Great news about the record attempt. Is there a page that mentions that Godzilla was used in the design? I could put that in my CV to help me to get a job in the public service.

3. If you are tinkering around your shed this long weekend, see if you can knock out a quick cure for cancer so my supervisor can finish the process of submitting my PhD thesis.

All the best,
Leo.

 

I am gradually running out of family, friends and colleagues through the ravages of cancer. Lost a good friend who happened to be my brother-in-law earlier this year. He kept a blog to avoid having to say things over to all his friends and acquaintances:
http://tonydarcy.blogspot.com/
It is a story of ups and down. Sadly my garage does not hold any magic cure. All I could offer was time with him and a few comments. His passing left a deep scar because he was a little younger than I am and we had spent a lot of time together over the years.

As it happens my eldest son does cancer research at the Peter Mac Institute here in Melbourne. What I glean from my conversations with him is that it is a slow process developing even better targeted chemo drugs. He is also waiting on his PhD to get through the review process. It has been 18 months since it was submitted.

So I have some empathy with your situation and hope things work out for you.

I do not recall any mention of Michlet or Godzilla on Greg's web site. I provided him with an iges file for CNC production of the original plug it was laid up on. The Godzilla aspects were transparent to him.

You may not be aware of it but I think you have caused the rules for the US university concrete canoe competition to be changed. The students are now given hull offsets rather than hull dimensions. The leading teams were all using Godzilla for hull design so had an unfair advantage over those not using it. The new hull was probably generated using Godzilla. So it is now more a concrete project than hydrodynamics of hull design.

Rick

 

I've looked at that before and It seems there's more factors than just viscous drag that determine the roughness of a living skin. Experimental results vary but well controlled ones tend to favor smooth and shiny surfaces for shapes that have fully attached flow. Some direct measurements are given in the attached PDF. Cf<0.004 for smooth surfaces and Cf>0.005 for 80grit sandpaper. The values are even higher for 40 and 24 grit sandpaper. Admittedly though, most canoests around here say that a wet-sanded hull is faster than a shiny new one.

If your reynolds numbers are above 10e6 the effect of turbulators is usually not beneficial. The exceptions are when the flow is detaching from the surface with out the extra rougness and a properly placed turbulator causes the flow to remain attached.

Leo describes the method here in section 1.2 and 1.3:
http://www.cyberiad.net/library/kayaks/skmag/skmag.htm

If Michlet simplly takes the calculated skin friction coeficient and multiplies it by the viscous drag form factor (VDFF) of the in.mlt file, you could create an effect equivalent to increasing viscosity by changing VDFF.

The method described in section 2.4 of another article by Leo indicates the VDFF is just a multiplier for Cf:
http://www.cyberiad.net/library/rowing/misbond/misbasic.htm
We surely will get a yeah or nay on this next time Leo looks in.

Using the VDFF to represent a change in viscosity is not straightforward though because the calculated Cf is not a linear function of viscosity.

For example if a 10m boat is going 3m/s and the viscosity is 1.18838 * 10e-6 the Re=25.24 * 10e6 and the ITTC calculated Cf = 0.075/(log10 (R)-2)2 = 0.00257.

If the viscosity were increased to 1.6 * 10e-6 the new Reynolds number would be 18.75 * 10e6 and the ITTC Cf would be .00270

But if the VDFF is just a multiplier of the ITTC Cf you can get the same numerical result as if you changed the viscosity to 1.6 * 10e6 by setting VDFF = 0.00270/0.00257 = 1.05 instead of the default 1.0

However, I could be just all dazed and confused about this.......

John H.