Michael Schmiechen, Berlin
News on ship powering trials
Last update and additions: 2016.08.19
Happy end of a very long story
The advanced analysis of the 'model' trial further advanced!
Continuing discussions with Daniel Wiens have lead to further corrections
and improvements of the final files
- Thrust deduction fraction identified
- Partial efficiencies identified
- Traditional results compared
Last update and revisions: 2016.07.29.
In the meantime I have spread the 'gospel' as documented in the
- Correspondence concerning the revision,
alerting the Members of the pertinent 28thITTC Committees and informing
colleagues and friends of the rational theory of trials and monitoring of ship
powering performance.
For ready reference I have added
- An executive abstract,
drafted in response to further critical questions by Dr. Klaus Wagner,
and the derivation and approximation of
- The thrust deduction theorem,
taking advantage of the features of symbolic Maple implemented in Mathcad,
first published as a lecture note in 1997.
Last update and revisions: 2016.07.03.
The publication of my revised worksheets has triggered a detailed correspondence
with Daniel Wiens, a master's student of mathematics at Ulm University, currently
studying my approach to ship powering trials and monitoring at VOITH Marine
Technology, particularly in view of uncertainties due to enviromental conditions
usually only crudely estimated.
By the way Daniel Wiens noticed fundamental problems in my much too simple-minded
approach, trying to identify all basic parameters jointly (!) by solving a system of
non-linear equations. In particular he observed a strong dependence of the results on the
necessary start values, 'educated guesses', due to inherent under- and/or ill-definition
of the problem stated, resulting in multiple, though equivalent (!) solutions.
This 'singularity' is of course unacceptable for applications in the analysis of
quasi-steady model and full scale powering trials and monitoring. And consequently
I have undertaken a second, the following drastic(!) revision of the procedure, trying
to avoid the problems mentioned altogether by a step-wise approach.
Following an explicit definition of the problem to be solved, I have shown in the first
step, that the data of my quasi-steady model test do not only permit to identify the
propulsive efficiency, but the jet, alias 'ideal' efficiency of the propeller as well.
Subsequently I have discovered that my wake convention is not necessary, in fact even
inadequate, thus to be replaced by another convention.
And in the second step I have shown, that my analysis has to correspond as closely as
possible to the traditional procedure in order to produce results as close as possible
to the results obtained by the traditional procedure followed since Froude's days.
The traditional model test procedure is based on the naive convention, that the frictional
deduction, applied as an external towing force, is equivalent to a reduction of the hull
resistance, although this is definitely not the case.
Despite its evident fundamental deficiency I have adopted this convention as well,
in accordance with my overall goal since 1980: to rationalise the traditional procedure
in such a way, that it is not 'only' applicable on model scale and full scale as well,
but acceptable to the community, namely as far as possible in line with traditional
concepts, past experience and current standards and teaching.
As before the whole programme has conveniently been broken down into a series of
Mathcad worksheets, each devoted to a fundamental step of the whole straightforward
procedure of powering prediction based on quasi-steady model tests.
- Coherent symbols developed
- Model data, test conditions
- Numerical routines developed
- Data of quasi-steady model trial
- Quasi-stationary states identified
- Propulsive efficiency identified
- Thrust deduction fraction identified
- Partial efficiencies identified
- Traditional results compared
The first edition of the revised worksheets is published here for open discussion.
Any observation concerning necessary corrections and further explanations will be
gratefully accepted, carefully considered and duly acknowledged. Further, assistance
will be provided concerning evaluations of pertinent data sets model and/or full scale.
Last update and revisions: 2016.06.20.
The endless trouble with the incredibly unprofessional customer service of PTC has of
course not stopped my work concerning the trustworthy identification of the powering
performance of ships. Quite to the contrary, it has forced me to look elsewhere for
the crucial optimum finding routines meeting the state of the art. And I have found
them in the very powerful SciLab environment, freely available in the internet.
The subsequent revision of the advanced analysis of my 'model' trial of 1986,
published in Volume 3 of my METEOR-Festschrift on my website and as a DIN A5
brochure, accounted for all the insights gained during my related numerical studies.
And triggered by the continued scrutiny of my extremely careful reviewer, Dr.-Ing.
Klaus Wagner at Rostock, I detected conceptual 'dirt', not to say unprofessional
fumbling, and programming mistakes, only apparently small.
But due to the intricacy of the whole procedure further conceptual, numerical and
'stilistical' corrections may of course become necessary before general acceptance
of the conventions and individual procedures proposed. Any pertinent indication
of remaining mistakes will be followed up and gratefully acknowledged.
At this stage apparently small, but substantial changes concern the correct identifi-
cation of the quasi-stationary states and the correct identification of the propulsive
efficiency, jointly identified with the very small drift observed in the residua of the
energy balance and accounted for from the 'beginning', but more or less unprofessional.
In the subsequent evaluations the very small residual inertial forces due to the fact,
that data are available only at given instances in time, are accounted for only
in the mean as before.
According to the SciLab DataSheet on Optimization:
'The method used to find the physical parameters of the axiomatic model
is based on the simplex algorithm of Nelder and Mead. This algorithm is known
to be able to manage noisy functions. The neldermead component of the
- SciLab fminsearch function
provides three simplex-based algorithms, which allow to solve unconstrained
and nonlinearly constrained optimization problems. It provides an object
oriented access to the options. While the falgship of Scilab is certainly
the optim function, which provides a set of five algorithms for non-linear
unconstrained (bound constrained) optimization problems.'
In order to prevent any misconceptions, I had explicitly stated, that these
revisions and, in particular, the comparisons with the results of traditional
steady model propulsion tests in the final section 9, requiring additional
nacked hull towing and propeller open water tests are not intended to
provide 'calibrations'.
Their purpose is to provide evidence, that quasi-steady tests of only two minutes
duration permit the reliable, detailed analysis of the powering performance of ship
models, without requiring additional nacked model hull towing and model propeller
open water tests. And, although not necessary, the numerical results 'happen' even
to be very close to those of the traditional model test procedure, of course only as
far as possible. In this context it is sufficient to note, that in the rational approach
promoted the concept of rotative efficiency, a rubbish bin of the traditional approach,
does of course not 'occur'.
But even more interesting than future routine applications on model scale are future
routine applications on full scale, where the traditional approach is not possible, hull
towing and propeller open water tests being impossible. This fact, as a matter of fact,
has been the incentive of my work on trials, stating with the inaugural Schiffs-
technik paper in 1980. And as already shown during the METEOR tests in 1988
quasi-steady tests of only twenty minutes duration performed at any condition,
un-noticed by anybody on board, provide sufficient information with sensors
standardly available, requiring no 'independent', incoherent (!) measurements
of hull speed through water of thrust (!), for the trustworthy analysis of the
powering performance.
In this context I also referred readers to Goethe's dictum:
"Das Höchste wäre: zu begreifen, dass alles Faktische schon Theorie ist."
and to five papers on the history of rationality by Lorraine Daston, director at the
Max-Planck-Institut für Wissenschaftsgeschichte at Berlin. Translations into German
are collected in a slim volume under the title 'Wunder, Beweise und Tatsachen',
Fischer Taschenbuch, Band 14763. Frankdurt/M.: Fischer, 2001. References to the
enlightning original papers to be provided on request.
Last update and revisions: 2016.05.05.
Warning: Minerr and Minimize functions in PTC Mathcad 15 Pro faulty!
Serious malfunctions in PTC's Mathcad 15 Pro fundamental Minimize function,
I had detected in the meantime, have [euphemistically, not to say ignorantly]
been called 'new 'functionality' by PTC 'experts', as the malfunctions
of the Minerr function had been called before.
Using the corresponding functions in SciLab and a simple-minded second order
routine in
- further numerical exercises
I have demonstrated, that finding the global minimum in a non-linear optimum
problem is 'of course' state of the art, as shown in a pertinent
- SciLab script
and in the present case even rather straightforward.
Last update and additions: 2016.04.12.
Due to an update of Windows 8.1 my old Mathcad 8.3 version with the Numerical
Recipes could no longer be installed. Trying PTC Mathcad 15 in late August 2015
I already noticed, that something was wrong with the Minerr function, fundamental
not only for my method of ship powering analysis, but for any other search for
an optimum solution of a non-linear problem.
.
After getting across some minor programme changes I found out, that the new routine
suffered from the fundamental defect, that it does not find the global minimum,
the final proof documented in the following
- detailed error analysis.
In the course of explaining the usage of the PTC Minerr function it has become
subject of a number of
- further crucial tests.
The original programme concerning the identification of the partial efficiencies
has been updated and stream-lined according to the insights gained during work
on the problems encountered, but is still waiting for PTC to provide the correct
Minerr function successfully passing all the tests.
Last update and additions: 2016.02.07.
Ship speed through the water
A crucial magnitude in evaluating trials and monitoring of a ship's
powering performance is the speed of the ship through the water.
Although demonstrated to be error prone the traditional procedure
to determine the speed through the water has been adopted by ITTC,
ISO and IMO.
This subject has among others been addressed in written contributions
- by Klaus Wagner and
- by Michael Schmiechen
to the 9. SVA Research Forum on January 28, 2016.
Last update and additions: 2016.02.12
The fact that since 1998 a simple procedure is available, reliable
under any service condition without requiring any extra instrumentation,
no calibration and no extra measurements, has triggered a correspondence
- zur Fahrt durchs Wasser.
Last update and additions: 2015.12.15
On trustworthy results of trials
On occasion of the 69th Meeting of the STG Committee on Ship-Hydrodynamics
at Kiel on 02.10.2015 a second leaflet winding up the METEOR-Festschrift
has been contributed and published under the title of the leading paper
- 'On trustworthy results
of ship powering trials and monitoring'.
Among others the leaflet contains an executive summary of the evaluation
of the quasi-steady 'model' trial, documented in detail in Volume 3 of the
Festschrift, and and further explanatory letters and discussions.
In order
- to spread the gospel
the leaflet has been brought to the attention of the Members of pertinent
ITTC Committees and of colleagues supposedly interested in my results.
As a reminder and for ready reference the pertinent chapter
- on the emperor's new clothes,
taken from Volume 1 of the METEOR-Festschrift, is published here again.
Last update and additions: 2015.10.09
Publication of a 'Festschrift' and further
Following corrections, improvements and additions of its draft
the 'final'version of
- the third volume
of the METEOR-Festschrift is now being published and open for discussion.
For ready reference links to the two volumes published earlier, in 2014
and 2013, respectively, are provided here as well:
- the second volume
- the first volume
Last update and additions: 2015.09.30/09.22.
In the meantime publication of the third volume of my METEOR-Festschrift
has been brought to the attention of colleagues and groups worldwide by
- short notes,
to make sure that 'everybody' concerned is at least alerted and, maybe some day,
taking advantage of the advanced state of research.
Last update and additions: 2015.06.17
Following my St. Nicholas Present, the preliminary publication of results
of my re-evaluation of the quasi-steady propulsion test performed in 1986,
before the METEOR tests of 1988, the draft of the third, the final volume
celebrating the anniversary of the METEOR tests, has been distributed as an
Easter Egg and as my invitation to contribute to the discussion.
The central part of the volume, designed to be printed as DIN A5 brochure,
consists of the complete details of the procedures developed and of the results
'derived' by the way. The list of 'Concepts, symbols and conventions',
represents a tabular presentation of the axiomatic model adopted.
This constitutive foundation, the 'Explanatory notes' and various
related correspondences are highlighting different aspects my exercise.
Last update and additions: 2015.04.04
The following file provides a preliminary publication of
- The final results
of my re-evaluation of the quasi-steady 'model' propulsion test
performed in 1986, before the METEOR tests of 1988.
This file has been sent as a St. Nicholas' Present
- To colleagues, friends and fans
and has subsequently been brought to the attention of various groups
of colleagues together with the original cover letter as follows:
- To my German colleagues
and further
- To pertinent ITTC Committees.
In the meantime the original cover letter has been amended
by important clarifying additions.
Last update and additions: 2014.12.16
On occasion of the 27th International Towing Tank Conference
2014, held at Copenhagen August 31 to September 05,
- A second Volume
of the 'Festschrift' celebrating the anniversary
of my quasi-steady propulsion tests with the German
research vessel METEOR has been published, based on
the work mostly originated since the STG Meeting
at Hamburg in 2013 and also documented further down
in the Section 'News on ship powering trials'.
Printed copies of the two volumes are still available
on request.
In the files of the PATEs on my website and in the Festschrift
the labels of the plots of propulsive efficiencies reported,
traditionally identified from model tests as confirmed
by Dr. Hollenbach of HSVA, 'the jokers pulled out of the sleeve',
have already been corrected.
As sort of executive summary of both volumes
- A Leaflet
has been published and widely distributed
not only at the ITTC at Copenhagen.
In particular it is being distributed with an addendum
referring to this 'News flash' and the spectacular results
of the final evaluation of my quasi-steady 'model'
propulsion test of 1986, at the 109th Annual Meeting
of STG, Hamburg, November 19 to 21, 2014.
Further, the following related
- Oral Discussions
have been contributed at the Sessions on Propulsion
and on Performance of Ships in Service.
Triggered by various discussions
- An explanatory letter
'On ship theory and paradigms' has been written and
published here, inviting further discussions.
The latter resulted in
- Further explanatory remarks
and the following addendum
- On 'true' models
by Dott.Giulio Gennaro of Genova.
Following the invitation to contribute to the discussion
on the Future of ITTC I have drafted the following
- Pertinent proposals.
In the meantime the final re-evaluation of the
quasi-steady 'model' propulsion test of 1986
is being finalised.
Last update and additions: 2014.12.01