[Attachment(s) from Dennis Cox included below]
From: Rick Willoughby <rickwill@bigpond.net.au>
To: harryproa@yahoogroups.com.au
Sent: Sat, August 21, 2010 7:09:35 PM
Subject: Re: [harryproa] My Little Mule - JavaFoil
Rick,
Backing off because I'm reaching a designed number in my head! Had to laugh at that one. Nah... no way... if MLM will do 20, 25+ I'll find it out. I'll run her to destruction or top speed one!... wear a life vest and jump clear and swim home if I have to. Note to self... strap portable radio on good.
I get the jist on stall. That eases the mind somewhat.
A.R. I've set it to 4 per your suggestion.
I would certainly appreciate a second set of eyes and brains... especially with your experience!
I've attached my trade study spread sheet. To summarize if you don't want to mess with the spread sheet. I've...
- I'm using a lower Reynolds number: 3E6 equates to 30 knots. I think 10e6 (equates to 100 knots) is a little ambitious.
- I've bracketed using the NACA 07 with: 010, 012, 015, 310, 312*, 315, 510, 512, 515, 1010, 1012, 1015, 2010, 2012, 2015 forms. I wanted to see what trends that would occur with the symmetric versions all the way to deeply cambered.
- For some strange reason, NACA 07-312 SYM wouldn't analyze in JavaFoil. It showed a flat line Cl for all Reynolds numbers at all AOA.
- In the end the highest L/D also seemed reasonable both from fabrication and having plenty of AOA in both directions to allow for steerage while at max speed.
The end result was...
NACA 07-512 SYM
L/D = 36.5
Rudder height = 32.5" (82.6 cm)
Rudder chord = 8.1" (20.6 cm)
Rudder thickness = 1.0" (2.5 cm)
Drag at 20 knots under full lateral load of 600 pounds = 8.2 pounds (16.4 for both)
It was interesting that at this condition, the rudders are actually at a AOA = -1 degrees.
Cheers,
Dennis
Dennis
From: Rick Willoughby <rickwill@bigpond.net.au>
To: harryproa@yahoogroups.com.au
Sent: Sat, August 21, 2010 7:09:35 PM
Subject: Re: [harryproa] My Little Mule - JavaFoil
Dennis
The maximum lift depends on the stall method used - MH probably wants to be conservative because initial use for JavaFoil was aeroplane design.. You have to be careful here that you are not "polishing the turd" again. Don't focus on the differences with this basic 2D data have a look at what it means to a realistic rudder situation. The difference is not large when you look at safety factors.
From a rudder perspective the only reason you are interested in maximum lift is for the strength based on maximum force:
1. What is the Cl once you set the AR (remember it will be double if it is tight against the hull or properly fenced in other way - something less if not fenced - how much?)
2. From a force perspective at high angle you now have to consider the drag. Where do you place the stops - this is going to limit the force. If you let it go to 90 degrees then the drag will be greater than the lift at any point.
3. What speed are you designing for - 15kts? What happens if you exceed 15kts and get to 20kts? Force has now almost doubled - I cannot imagine you thinking "hang on I am exceeding design speed of the rudders I should back off"!!
4. What AR correction are you going to apply to get 3D data results for the A&VD data?
5. What is your allowance for fatigue?
From a performance perspective the rudder should be sized so it is working around the point of highest L/D. This will give it a good margin to the maximum lift. I can give you the polars for your boat based on various rudder size if you want.
From a drag perspective the detail on the shape and surface finish are likely to be most important.
You will not be far out if you rely on the JavaFoil data. If you build and find a difference to prediction I doubt you will get to the point where you find the JavaFoil data unreliable. It will be because the rudder has ventilated or cavitated or something much simpler like wobbling from loose linkage. It may have a plastic bag on it.
You are focusing on something that is minute in terms of the whole boat. Your term "AND what's is really bad ..." is a myopic focus on these basic numbers rather than anything important.
On a slightly different tack the biggest error I have had with Michlet was explained by water temperature. Leo adjusted the allowable range so I could use for cold mountain lake water. What water temperature are you using for with your rudders? Again will you be thinking "Hang on the water is 5 degrees cooler here than I designed for I have to limit top speed to 14.5kts"!!!
Point is use Javafoil for relative comparisons and tuning of the design. There are a huge number of factors that influence the end result. Many you will not even contemplate, or need to, unless they become an issue and are within your measuring error.
Rick
On 22/08/2010, at 1:45 AM, Dennis Cox wrote:
OK, I've regrouped... yes the page I was referencing had flaps as options, but I was reading off the correct curves.I've un-installed and re-downloaded and re-installed JavaFoil... same results.At RE = 9E6Maximum ClA&VD JavaFoilNACA 0012 1.6 1.209 25% LowNACA 4415 1.64 1.882 15% HighI'm a little concerned that I get different numbers than you and here I'm getting different numbers relative to a well established bible in the industry. AND what's really bad... the estimates are conservative in one case and optimistic in another.Basically, this tells me...(1) if this is the state of the art for approximate versus real world... its not a very good approximate!(2) ... and this is just a sampling out of two foils (of the same class of foils)(3) What if I go to something relatively exotic... like a standard 07 foil. Is it going to be off by 50%... and which way?(4) Then going way out into left field with the symmetric version... which appears to violate several of the dependencies of the analysis. Namely 3D flow starts (guessing) somewhere around 75% of the chord.With all these things I'm ticking off, I'm convincing myself that I would need to oversize at least 50% just to account for the fog.Dennis
From: Dennis Cox <dec720@att.net>
To: harryproa@yahoogroups.com.au
Sent: Sat, August 21, 2010 10:47:36 AM
Subject: Re: [harryproa] My Little Mule - JavaFoil
Hi RickJust starting JavaFoil up, changing nothing which defaults to NACA 0012.Going to the Polar page,Changing only RE and checking Cd at zero degrees, I get:RE = 1E6, Cd = 0.00923RE = 10E6, Cd = 0.01114What's even stranger, I pulled out my Abbot and Von Doenhoff, p. 462. Put inRE = 3E6, 6E6 & 9E6The largest Cl @ (AOA=16) is Cl = 1.6JavaFoil at the same point is Cl = 1.136Dennis
From: Rick Willoughby <rickwill@bigpond.net.au>
To: harryproa@yahoogroups.com.au
Sent: Fri, August 20, 2010 11:18:04 PM
Subject: Re: [harryproa] My Little Mule - JavaFoil
Dennis
I checked your section and it was almost identical to what I got despite the fact I used 61 points. You can get variation with the number of points but not so much if the section has gentle curves.I used smooth finish. I can get what you got by using painted fabric.Check a NACA0012. It should be around 0.009.The Javafoil data I got aligns well with the old NACA report I referenced earlier.RickOn 21/08/2010, at 11:46 AM, Dennis Cox wrote:
Rick Willoughby
Rick Willoughby
03 9796 2415
0419 104 821
__._,_.___
Attachment(s) from Dennis Cox
1 of 1 File(s)
.