From: "Perry Mick" <pjmick@mail.viclink.com>
To: "Rotary motors in aircraft" <flyrotary@lancaironline.net>
Subject: Re: [FlyRotary] Re: Performance Props
Date: Wed, 21 Apr 2004 17:45:14 -0800
Message-Id: <20040422013655.M8616@mail.viclink.com>
In-Reply-To: <list-3183508@logan.com>
References: <list-3183508@logan.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1

> > 
> >>Perry, if you could share the formulae you're using I'd be happy to 
> >>make a spreadsheet calculator for this. So far I've found very 
> >>complex prop calculators but nothing straightforward.
> >>
> >>By the way, why are you asking desired cruise speed? 400 knots! =)
> >>
> >>Do your calculations focus on pitch/length or do they also provide 
> >>some means of calculating dimensions at various stations from the 
> >>hub, so forming templates can be made?
> >>
> >>Regards,
> >>Chad
> >>
> > 
> > It's not simple enough to put in a spreadsheet.
> > I think Al Wick uses the Bates Prop program, I don't know anything about 
> > that program and I don't know why he couldn't develop numbers for Bulent.
> > The programs I use are all in the book "Modern Propeller and Duct 
Design" by 
> > Martin Hollman. They are BASIC programs that I have improved on and now 
run 
> > as visual basic macros. They have proven to be very good at predicting 
> > reality.
> > Data entry is actual blade parameters along the length of the blade, so 
the 
> > process is "trial-and-error". Enter a blade design and examine the 
output to 
> > see what the performance is. The Bates program might be the opposite, 
enter 
> > performance numbers and it spits out a blade design (?).
> > I want to know desired cruise speed and cruise RPM as a first estimate 
for 
> > pitch. Actual cruise speed is limited by thrust produced and the drag of 
the 
> > airplane, so equivalent flat plate area of the airplane is also a 
desirable 
> > thing to know.
> > 
> > Perry
> > 
> 
> Perry, would you like to share your algorithms.  I'll turn them into 
> C code ('cause I like C), and put a genetic algorithm on the front 
> end. I'm assuming you have X number of stations that you enter 
> numbers for. Make each station a genetic marker, start with a couple 
> strains and let them mate through a few thousand generations till 
> you get something you really like.
> 
> -- 
Ernest:

I would recommend you purchase that book. They are not really my algorithms 
and I don't want to be accused of copyright infringement. The reason it 
can't easily be put in a spreadsheet is it is really numerical calculus and 
the program runs multiple loops (finite steps whereas in calculus the step 
size is taken to the limit of zero) to develop the performance of a complete 
blade.

Perry