Message-Id: <3.0.3.32.19990315232243.00a8a100@olsusa.com>
Date: Mon, 15 Mar 1999 23:22:43 -0500
To: lancair.list@olsusa.com
From: Marvin Kaye <marvkaye@olsusa.com>
Subject: LC20 flutter -- new comments
Mime-Version: 1.0

          <<<<<<<<<<<<<<<<--->>>>>>>>>>>>>>>>
          <<  Lancair Builders' Mail List  >>
          <<<<<<<<<<<<<<<<--->>>>>>>>>>>>>>>> 
>>
Hi gang,

I just received a post from someone who has another take on the LC20
flutter situation.  Due to the content of this person's post, I have
decided not to make the info available on my website's flutter page, for
reasons which will probably become obvious to you once you have read what
he has to say.  Since the people who have the most abiding interest in this
subject are probably gathered on the LML, I am reposting his letter to me
here for your information and consideration.  This may help you folks who
are out there trying to decide whether or not to incorporate Martin's mods
into your project about what direction to take.  I hope you find it as
interesting as I did.

>>>>
A friend gave me the internet address of your Lancair 360 flutter "problem".
I read the article with some interest since I used to work at Lancair
(though not on the design of the 360) and had to deal with Martin a few
times as part of my employment at Lancair.  Since I no longer work at
Lancair, and since I did not work on the design of the 360, I feel that my
opinions should be somewhat unbiased but perhaps add some insight into what
was going on at the time.

First, Martin contracted Sam McIntosh to "show" him how to do flutter
analysis on the Lancair IV.  Martin thus decided he was an "expert" and has
subsequently charged many people for his "expert" capabilities and can list
20 KITPLANES to his credit.  None of his analysis was scrutinized in the way
a certification report would be reviewed.  In my humble opinion, the
"experts" at flutter analysis are those who were trained and spent YEARS
doing flutter analysis since it is such a specialized topic.  Enough said.

Second, prior to the 360, Martin had done the flutter analysis on the ES.
He originally submitted a report saying the plane was fine.  Later, while
doing analysis on his Stallion which used the same wing, he "discovered" new
problems and wanted Lance to pay for a new analysis on the ES.  Again in my
opinion, Martin should have done the analysis for free since he was
admitting that he had screwed it up to begin with.  Anyway, Lance contracted
with someone else to do the flutter analysis.  FEA models were generated, a
GVT was conducted (and I was part of both of those), and Sam McIntosh did
the flutter analysis, using codes similar to the ones Martin uses (in fact,
Sam is the one who pointed Martin to those codes to begin with).  I think
this pissed off the always arrogant Martin (again, my opinion) and he took
it upon himself to slam any analysis done by an outsider.

Finally, I get to the 360.  Some builders were modifying the 360 aircraft
(remember that Lancair only recommends a few engines) and contracted Martin
to do the flutter analysis.  Martin found problems and now wanted Lancair to
pay for him to do more analysis.  Lancair instead hired someone else, having
lost much faith in Martin's capabilities.  Now Martin takes it upon himself
to slam the analysis done by the outsider.  He says that Norbert didn't use
modern methods for analysis (which is untrue) and didn't do the GVT
correctly (which is untrue) and that he obviously doesn't understand FEA if
he said you had to use a few "tricks" (which is also untrue).  I'll discuss
each of these:
1. Flutter analysis is dependent upon generating a aerodynamic matrix,
a mass matrix, and a stiffness matrix to determine the interactions between
these three sources of force.  All of the flutter methods are rooted in
these three matrices (or add a fourth for damping, but it is usually handled
a little differently) but use different methods to solve for the answers.
One of the biggest areas of differences is in the formulation of the
aerodynamics matrix, but I diverge.  Anyway, Norbert's method certainly uses
one of these basic flutter solution techniques of solving large systems of
differential equations.
2. Norbert's GVT was done correctly.  It is actually quite simple to
excite the fuselage bending and aft fuselage torsion modes in an airplane
using your hand, let alone an electromagnetic shaker.  He did find those
modes and Martin's assertion that he didn't is wrong.  I've now worked with
three different people doing GVTs and they were all close in approach, only
varying in the details.
3. Norbert does understand FEA, and even though I am an FEA "expert", I
understand that it is not required to use FEA to generate the mass and
stiffness matrix inputs to the flutter analysis.  In fact for some
applications, you can generate the mass matrix better using a spreadsheet
method.  Anyway, if you have ever talked or work with an FEA "expert" you
will hear them often discuss a "trick" used to simulate a hinge, or a
sliding contact, or any one of a number of details that aren't generic to
the FEA software.  FEA is similar to flutter analysis in that it is really
just a big matrix problem combining stiffness, deflection, and force.  If a
program doesn't have an element that you need to model, you can often
"trick" the program to do what you want by using combinations of other
elements to give you your desired net effect.  Anyway, to say that a person
doesn't understand FEA because they use a few "tricks" is crazy. 

In summary, I believe the 360s are safe in regards to flutter if built
according to plan.  I also believe that Lancair has gone to experts they can
believe (three different ones that I know of), rather than a self-proclaimed
"expert" and have thus received the best answers they could to a very
technical subject.  I also know that Lancair did not go looking for someone
to tell them the plane was ok, and that if a problem were found, it would
have been addressed.

Name withheld by request.
<<<

So there you have it.  The decision is still yours, but once again, the LML
delivers the goods to help you make it an informed one.  

Keep those posts coming!!!

    <Marv>