Rudder Design

This is a quick sketch of my situation. I researched rudder info on this site and then followed a link to proboat.com (digital issues #98, 99, 100) where I found an outstanding 3 part series on most everything about designing a rudder written by the highly accomplished Dave Gerr. I found the intermediate shape and speed range to be best suited for my application. Yes, maybe hard-to-find but I can machine a one-piece design of what I need from billet. All the info below is in the ranges that are acceptable according to Mr. Gerr's article.

75 hp diesel, 5,000lb, 18" prop,

LATERAL AREA = 5100 sq in
RUDDER AREA req. = 205 sq. in.
(according to Mr. Gerr's supplied formula)

CHORD = 10 inches
SPAN = 20.5 inches
AREA = 205 sq. in.
ASPECT RATIO = 2.05
BALANCE = 17%-20%(I am using 20%)

RUDDERSTOCK DIA = 1.00 inches
(according to Mr. Gerr's supplied formula)

When I use the supplied offset table for the intermediate shape, I cannot fit the 1.00 dia rudderstock into the profile (center profile in my sketch). It does fit into the NACA 0010 shape - but this is not optimal shape for my speed. I have modified the percent offsets for thickness for the intermediate shape (lower profile in my sketch) to fit the 1.00 stock into that shape by 37%. This is done regularly to the NACA shapes (such as modifying the NACA 0010 thickness values by, say, 11% making it NACA 0010-11%). My NACA 0010 foil sketch is as-designed (unmodified).

The question is (given that the 1/2 perimeters of the lower and upper shapes in my sketch are near identical) that iss this "modified 37% shape too wide? Will the modified shape perform better than the standard NACA 0010 shape?

I do not want to adjust my chord or span due to draft or prop clearance.
I prefer to use either the "modified" intermediate or the NACA 0010.
Any insight will be usefull and helpfull.

 

John
Some Javafoil data attached. The section for your proposed change was not modelled exactly. I have shown what I used.

I was surprised by the result. I expected the drag to be much higher with the chopped tail. It is at low Re# but for the speeds you should get it does not make much difference.

If you want do an exact comparison then JavaProp can accept your coordinates but they need to be normalised.

Your proposed prop should be more progressive than either of the other two but drag throughout the range and best L/D is not much different.

Rick W

 

How will the rudder be attached to the shaft? If the rudder is machined from billet, it will be much stronger than the shaft. If you hit something, the shaft will bend above the rudder. Ideally, there is a weak point where the shaft bends or breaks outside of the boat and not bind or damage the boat, itself. In other words, I suggest using the 1" shaft AND the profile you desire - just taper the shaft to match.
Funny, I am using a different approach but presently confronting almost the identical thing. I'm going with NACA13 tapering to NACA9 at the bottom. I understand the stated hard-over efficiency of the design Rick put forth (Gerr likes that, too) and since my rudder is composite, I figure I can add the "fish tail" later if I stall. Right now, slippin' through the water easily and with minimal cavitation and flutter is my goal. Rick, I really didn't follow - Is it right that the "chopped off" tail is not inefficient (I cruise at 8 and almost eighteen knots)? I start lay-up tomorrow and may as well do it that way if so.

 

I modelled the shape a little more accurately. The cut-off trailing edge does not show any great increase in drag at any speed. It tends to make the rudder more progressive.

The higher drag at low Re# I saw earlier was due to the rough modelling of the shape. There was flow separation at the concave into the flat.


Rick W

 

Hello,
I am sorry that I have to repeat this thing once again, but unfortunately some things need to be repeated a necessary number of times...

Javafoil cannot handle anything that resembles the airfoil shown in the post #4, for at least two reasons:
1) a pointed leading edge leads to the creation of a separation bubble at any angle of attack different from zero.
2) the methods used for boundary-layer transition and detachment estmate is valid for conventionally-shaped airfoils, smooth and with single cusped trailing edge. A panel method will give a singularity for the velocity field at the corners of the cut-off trailing edge like the one you have shown, and since the boundary-layer equations are based on the outer inviscid velocity field, all the subsequent boundary layer estimates will necessarily be faulty.

Please find some time to read carefully the following two pages of Javafoil where the author of the software has explained the limits of the code very clearly:
http://www.mh-aerotools.de/airfoils/jf_about.htm
http://www.mh-aerotools.de/airfoils/jf_analysis_boundarylayer.htm

Both the lift and the drag of a truncated airfoil will generally increase, (the drag will increase more than the lift - so the efficiency will decrease), as can be seen in the first paper I'm enclosing here (pdf attachment n.1 - figures 2.2, 3.2, 4.2). It is also shown by the experimental curve taken from the "Symmetry of Sailing" by Ross Garret (first thumbnail - sorry for the bad quality of the pic). In that graph, "t/h" is the ratio of trailing-edge thickness to the maximum foil thickness, the lower curve is for a smooth airfoil, the upper curve is for NACA standard-roughness airfoil.
I also attach an excerpt form Hoerner's "Fluid-Dynamic drag" on that issue (second thumbnail).

However, there are researches aimed at finding the solution for this inconvenient, because cut-off foils are structurally much more desireable. The second paper (pdf attachment n.2) shows one of the solutions examined. By the way - the fig.7 at the page 4 of that paper will graphically show you another reason why Javafoil cannot analyze similar configurations.
Hope that will help.

Cheers to all.

 

I disagree.

Rather than just saying something won't work take a bit of time to use the references you have offered to give some hard design numbers for John and Mark.

Just take a few minutes to apply the formulas or curves from the references you have provided to arrive at some actual values that demonstrate the large error in the JavaFoil data.

Provide your answer in the extra horsepower the modified rudder will rob at 18kts when in line and the loss in maximum steering force available, both compared with a NACA0010 also at 18kts. Lets see what all the fuss is about.

This is the sort of information the guys need to make decisions on. I doubt that they have any concern about your desire for theoretical purity.

Rick W

 

I am sure that both John and Mark, being intelligent guys, will be able to calculate the numbers from the references I gave them.

As soon as I find some time, I'll do the homework.

 

Thanks for the help!
(Today's goofs)


An hour leaning on a grinder for each of those tapers


Broken fishin' rod, hot glue, some glass I layed-up last night


Dry fit, counting wood shims for NACA13 minus 1/2 inch each side for laminate to come


pucky, shmooze - everything got slippery and wanted to leave (me too)

Before it starts; "There's an easier way to do that!" I play around in shallow, rocky areas alot. One of the reasons I took the rudder shoe off is that I have tweeked it by setting down wrong (The main reason for removing is efficiency and swinging a larger prop than it would allow). This new rudder I have designed to shear if I back over a rock coming off a beach or such. To make it replaceable with just a few bolts, I felt like I needed to do it this way.
If I show a complete rudder soon, I guess it will have worked... It will be a month before I get to try it out.

These two are for Tolly - nothing to do with the rudder...The keel was blunt 3.5" and hung this shoe. This is stage one of the stretch next fall.

 

Mark,
You are a busy boy. Nice shop setup, I hope your boat appreciates her pampering. I look forward to learning more about your future project when the dust settles from your present endeavers. LOL, you're definitely willing to NOT take the easy way out. Carry on

 

Well, it seems the thread died so I'll continue to rob its grave...


I finally figured out that I could maintain a larger glass/resin ratio draping with cloth than any other wet lay-up I could do. Some other laminates (biaxial, 3208 BTI, etc.) seemed quicker but filler was needed and I really wanted more glass. It worked nice laying up with the rudder hanging, as I could simply squeegee excess resin onto cardboard under project. It is not as resin rich as it appears -the flash caused that glare. The sharpie line on the lower trailing edge is where I was supposed to cut it and forgot until now! Oh well, I'll try it and see how it works.


Finished shape. The Sharpie lines are from me trying to massage the shaft placement to get the balance under 21% (the old was twenty point something) This is where I wanted to be - Gerr said the fish tail moves the center of pressure aft... Worse case is I'm a little twitchy and I have ways to tweak or I'll feel what I want and build another (I'm good at it, now)!


Lots of roving, a little putty on 1/2" x 2" flat bar. The little pencil zinc screws into the shaft and is hidden from flow by rudder shape.


Slow cure, chopped glass strand and fumed silica. I first tried medium cure, a little fumed silica and perlite - just try and keep perlite from floating in epoxy! Doing again, I would use slow cure and tons of chopped glass strands - trying to save money on resin by mixing with perlite (or vermiculite, which I have done before and doesn't seem to float as bad) was false economy. If you don't have a slow cure, you have to mix it with something, or do multiple small pours to prevent getting too hot.


It looks lumpy in this pic but that is just the sanding/burnishing of the Baltoplate in progress and is actually Lexus smooth.


I don't have the rudder slid up on its core, as I don't have the driveshaft in. ~ 30" depth, 19" chord NACA13.skoshi w/fish tail. Undo four x 1/2" x 1 1/4" 16TPI, blue LocTited and the rudder slides off for shaft change or to try different rudder shape. Shazzam!

 

When are you going to get all this new gear wet?

Rick W

 

Soon. Ten days maybe but I thought I was moving the engine back six inches to account for the lengthened keel but just decided to get a longer shaft instead. Takes a week to get it here. I am the slowest worker you have ever seen. "think twice, build once (then reflect after)"

 

When you do a nice job you have to savour it. Absolutely no point in rushing it.

Rick W

 

This dude is gonna' be so slippery, I'll need to pump fuel off onto other boats...

 

Looking good, looks like you are on track schedule wise. What other loose ends and mini project additions need to be wrapped up before splash? This can't be the only love the boat is receiving after a winter spent warm and toasty inside the shop.